From 40ec4fdc9a74bfdb83f13672acdb88af5c91ab46 Mon Sep 17 00:00:00 2001 From: delphij Date: Fri, 15 Jun 2007 07:06:13 +0000 Subject: Virgin import of diffutils 2.8.7. --- contrib/diff/ABOUT-NLS | 768 ++++ contrib/diff/AUTHORS | 112 + contrib/diff/COPYING | 11 +- contrib/diff/ChangeLog | 2142 +++++++++- contrib/diff/Makefile.am | 23 + contrib/diff/NEWS | 124 +- contrib/diff/README | 70 +- contrib/diff/THANKS | 22 + contrib/diff/TODO | 1 + contrib/diff/bootstrap | 180 + contrib/diff/doc/Makefile.am | 24 + contrib/diff/doc/diagmeet.note | 71 + contrib/diff/doc/diff.texi | 4649 ++++++++++++++++++++++ contrib/diff/doc/fdl.texi | 452 +++ contrib/diff/doc/stamp-vti | 4 + contrib/diff/doc/version.texi | 4 + contrib/diff/exgettext | 115 + contrib/diff/lib/Makefile.am | 35 + contrib/diff/lib/alloca.c | 489 +++ contrib/diff/lib/alloca_.h | 68 + contrib/diff/lib/basename.c | 79 + contrib/diff/lib/c-stack.c | 311 ++ contrib/diff/lib/c-stack.h | 19 + contrib/diff/lib/cmpbuf.c | 147 + contrib/diff/lib/cmpbuf.h | 21 + contrib/diff/lib/dirname.c | 119 + contrib/diff/lib/dirname.h | 43 + contrib/diff/lib/error.c | 310 ++ contrib/diff/lib/error.h | 66 + contrib/diff/lib/exclude.c | 263 ++ contrib/diff/lib/exclude.h | 43 + contrib/diff/lib/exit.h | 32 + contrib/diff/lib/exitfail.c | 27 + contrib/diff/lib/exitfail.h | 20 + contrib/diff/lib/file-type.c | 75 + contrib/diff/lib/file-type.h | 166 + contrib/diff/lib/fnmatch.c | 403 ++ contrib/diff/lib/fnmatch_.h | 63 + contrib/diff/lib/fnmatch_loop.c | 1192 ++++++ contrib/diff/lib/getopt.c | 1254 ++++++ contrib/diff/lib/getopt.h | 176 + contrib/diff/lib/getopt1.c | 191 + contrib/diff/lib/getopt_int.h | 129 + contrib/diff/lib/gettext.h | 68 + contrib/diff/lib/gettimeofday.c | 121 + contrib/diff/lib/gnulib.mk | 90 + contrib/diff/lib/hard-locale.c | 74 + contrib/diff/lib/hard-locale.h | 24 + contrib/diff/lib/imaxtostr.c | 3 + contrib/diff/lib/inttostr.c | 49 + contrib/diff/lib/inttostr.h | 47 + contrib/diff/lib/malloc.c | 36 + contrib/diff/lib/mkstemp.c | 42 + contrib/diff/lib/offtostr.c | 3 + contrib/diff/lib/posixver.c | 59 + contrib/diff/lib/posixver.h | 1 + contrib/diff/lib/prepargs.c | 93 + contrib/diff/lib/prepargs.h | 3 + contrib/diff/lib/quotesys.c | 125 + contrib/diff/lib/quotesys.h | 9 + contrib/diff/lib/realloc.c | 39 + contrib/diff/lib/regex.c | 8298 +++++++++++++++++++++++++++++++++++++++ contrib/diff/lib/regex.h | 556 +++ contrib/diff/lib/setmode.c | 67 + contrib/diff/lib/setmode.h | 27 + contrib/diff/lib/stdbool_.h | 93 + contrib/diff/lib/strcase.h | 35 + contrib/diff/lib/strcasecmp.c | 66 + contrib/diff/lib/strftime.c | 1330 +++++++ contrib/diff/lib/stripslash.c | 39 + contrib/diff/lib/strncasecmp.c | 2 + contrib/diff/lib/strtoimax.c | 80 + contrib/diff/lib/strtol.c | 432 ++ contrib/diff/lib/strtoll.c | 33 + contrib/diff/lib/strtoul.c | 20 + contrib/diff/lib/strtoull.c | 27 + contrib/diff/lib/strtoumax.c | 2 + contrib/diff/lib/tempname.c | 344 ++ contrib/diff/lib/time_r.c | 69 + contrib/diff/lib/time_r.h | 46 + contrib/diff/lib/umaxtostr.c | 3 + contrib/diff/lib/unlocked-io.h | 132 + contrib/diff/lib/version-etc.c | 176 + contrib/diff/lib/version-etc.h | 37 + contrib/diff/lib/waitpid.c | 72 + contrib/diff/lib/xalloc.h | 87 + contrib/diff/lib/xmalloc.c | 255 ++ contrib/diff/lib/xstrdup.c | 33 + contrib/diff/lib/xstrtol.c | 295 ++ contrib/diff/lib/xstrtol.h | 91 + contrib/diff/lib/xstrtoul.c | 6 + contrib/diff/lib/xstrtoumax.c | 37 + contrib/diff/man/Makefile.am | 34 + contrib/diff/man/cmp.1 | 63 + contrib/diff/man/diff.1 | 227 ++ contrib/diff/man/diff3.1 | 78 + contrib/diff/man/sdiff.1 | 90 + contrib/diff/src/Makefile.am | 44 + contrib/diff/src/analyze.c | 1038 +++++ contrib/diff/src/cmp.c | 677 ++++ contrib/diff/src/context.c | 478 +++ contrib/diff/src/diff.c | 1353 +++++++ contrib/diff/src/diff.h | 375 ++ contrib/diff/src/diff3.c | 1744 ++++++++ contrib/diff/src/dir.c | 288 ++ contrib/diff/src/ed.c | 169 + contrib/diff/src/ifdef.c | 430 ++ contrib/diff/src/io.c | 859 ++++ contrib/diff/src/normal.c | 71 + contrib/diff/src/sdiff.c | 1225 ++++++ contrib/diff/src/side.c | 281 ++ contrib/diff/src/system.h | 326 ++ contrib/diff/src/util.c | 775 ++++ 113 files changed, 38786 insertions(+), 28 deletions(-) create mode 100644 contrib/diff/ABOUT-NLS create mode 100644 contrib/diff/AUTHORS create mode 100644 contrib/diff/Makefile.am create mode 100644 contrib/diff/THANKS create mode 100644 contrib/diff/TODO create mode 100755 contrib/diff/bootstrap create mode 100644 contrib/diff/doc/Makefile.am create mode 100644 contrib/diff/doc/diagmeet.note create mode 100644 contrib/diff/doc/diff.texi create mode 100644 contrib/diff/doc/fdl.texi create mode 100644 contrib/diff/doc/stamp-vti create mode 100644 contrib/diff/doc/version.texi create mode 100755 contrib/diff/exgettext create mode 100644 contrib/diff/lib/Makefile.am create mode 100644 contrib/diff/lib/alloca.c create mode 100644 contrib/diff/lib/alloca_.h create mode 100644 contrib/diff/lib/basename.c create mode 100644 contrib/diff/lib/c-stack.c create mode 100644 contrib/diff/lib/c-stack.h create mode 100644 contrib/diff/lib/cmpbuf.c create mode 100644 contrib/diff/lib/cmpbuf.h create mode 100644 contrib/diff/lib/dirname.c create mode 100644 contrib/diff/lib/dirname.h create mode 100644 contrib/diff/lib/error.c create mode 100644 contrib/diff/lib/error.h create mode 100644 contrib/diff/lib/exclude.c create mode 100644 contrib/diff/lib/exclude.h create mode 100644 contrib/diff/lib/exit.h create mode 100644 contrib/diff/lib/exitfail.c create mode 100644 contrib/diff/lib/exitfail.h create mode 100644 contrib/diff/lib/file-type.c create mode 100644 contrib/diff/lib/file-type.h create mode 100644 contrib/diff/lib/fnmatch.c create mode 100644 contrib/diff/lib/fnmatch_.h create mode 100644 contrib/diff/lib/fnmatch_loop.c create mode 100644 contrib/diff/lib/getopt.c create mode 100644 contrib/diff/lib/getopt.h create mode 100644 contrib/diff/lib/getopt1.c create mode 100644 contrib/diff/lib/getopt_int.h create mode 100644 contrib/diff/lib/gettext.h create mode 100644 contrib/diff/lib/gettimeofday.c create mode 100644 contrib/diff/lib/gnulib.mk create mode 100644 contrib/diff/lib/hard-locale.c create mode 100644 contrib/diff/lib/hard-locale.h create mode 100644 contrib/diff/lib/imaxtostr.c create mode 100644 contrib/diff/lib/inttostr.c create mode 100644 contrib/diff/lib/inttostr.h create mode 100644 contrib/diff/lib/malloc.c create mode 100644 contrib/diff/lib/mkstemp.c create mode 100644 contrib/diff/lib/offtostr.c create mode 100644 contrib/diff/lib/posixver.c create mode 100644 contrib/diff/lib/posixver.h create mode 100644 contrib/diff/lib/prepargs.c create mode 100644 contrib/diff/lib/prepargs.h create mode 100644 contrib/diff/lib/quotesys.c create mode 100644 contrib/diff/lib/quotesys.h create mode 100644 contrib/diff/lib/realloc.c create mode 100644 contrib/diff/lib/regex.c create mode 100644 contrib/diff/lib/regex.h create mode 100644 contrib/diff/lib/setmode.c create mode 100644 contrib/diff/lib/setmode.h create mode 100644 contrib/diff/lib/stdbool_.h create mode 100644 contrib/diff/lib/strcase.h create mode 100644 contrib/diff/lib/strcasecmp.c create mode 100644 contrib/diff/lib/strftime.c create mode 100644 contrib/diff/lib/stripslash.c create mode 100644 contrib/diff/lib/strncasecmp.c create mode 100644 contrib/diff/lib/strtoimax.c create mode 100644 contrib/diff/lib/strtol.c create mode 100644 contrib/diff/lib/strtoll.c create mode 100644 contrib/diff/lib/strtoul.c create mode 100644 contrib/diff/lib/strtoull.c create mode 100644 contrib/diff/lib/strtoumax.c create mode 100644 contrib/diff/lib/tempname.c create mode 100644 contrib/diff/lib/time_r.c create mode 100644 contrib/diff/lib/time_r.h create mode 100644 contrib/diff/lib/umaxtostr.c create mode 100644 contrib/diff/lib/unlocked-io.h create mode 100644 contrib/diff/lib/version-etc.c create mode 100644 contrib/diff/lib/version-etc.h create mode 100644 contrib/diff/lib/waitpid.c create mode 100644 contrib/diff/lib/xalloc.h create mode 100644 contrib/diff/lib/xmalloc.c create mode 100644 contrib/diff/lib/xstrdup.c create mode 100644 contrib/diff/lib/xstrtol.c create mode 100644 contrib/diff/lib/xstrtol.h create mode 100644 contrib/diff/lib/xstrtoul.c create mode 100644 contrib/diff/lib/xstrtoumax.c create mode 100644 contrib/diff/man/Makefile.am create mode 100644 contrib/diff/man/cmp.1 create mode 100644 contrib/diff/man/diff.1 create mode 100644 contrib/diff/man/diff3.1 create mode 100644 contrib/diff/man/sdiff.1 create mode 100644 contrib/diff/src/Makefile.am create mode 100644 contrib/diff/src/analyze.c create mode 100644 contrib/diff/src/cmp.c create mode 100644 contrib/diff/src/context.c create mode 100644 contrib/diff/src/diff.c create mode 100644 contrib/diff/src/diff.h create mode 100644 contrib/diff/src/diff3.c create mode 100644 contrib/diff/src/dir.c create mode 100644 contrib/diff/src/ed.c create mode 100644 contrib/diff/src/ifdef.c create mode 100644 contrib/diff/src/io.c create mode 100644 contrib/diff/src/normal.c create mode 100644 contrib/diff/src/sdiff.c create mode 100644 contrib/diff/src/side.c create mode 100644 contrib/diff/src/system.h create mode 100644 contrib/diff/src/util.c (limited to 'contrib') diff --git a/contrib/diff/ABOUT-NLS b/contrib/diff/ABOUT-NLS new file mode 100644 index 0000000..2f50c66 --- /dev/null +++ b/contrib/diff/ABOUT-NLS @@ -0,0 +1,768 @@ +Notes on the Free Translation Project +************************************* + +Free software is going international! The Free Translation Project is +a way to get maintainers of free software, translators, and users all +together, so that will gradually become able to speak many languages. +A few packages already provide translations for their messages. + + If you found this `ABOUT-NLS' file inside a distribution, you may +assume that the distributed package does use GNU `gettext' internally, +itself available at your nearest GNU archive site. But you do _not_ +need to install GNU `gettext' prior to configuring, installing or using +this package with messages translated. + + Installers will find here some useful hints. These notes also +explain how users should proceed for getting the programs to use the +available translations. They tell how people wanting to contribute and +work at translations should contact the appropriate team. + + When reporting bugs in the `intl/' directory or bugs which may be +related to internationalization, you should tell about the version of +`gettext' which is used. The information can be found in the +`intl/VERSION' file, in internationalized packages. + +Quick configuration advice +========================== + +If you want to exploit the full power of internationalization, you +should configure it using + + ./configure --with-included-gettext + +to force usage of internationalizing routines provided within this +package, despite the existence of internationalizing capabilities in the +operating system where this package is being installed. So far, only +the `gettext' implementation in the GNU C library version 2 provides as +many features (such as locale alias, message inheritance, automatic +charset conversion or plural form handling) as the implementation here. +It is also not possible to offer this additional functionality on top +of a `catgets' implementation. Future versions of GNU `gettext' will +very likely convey even more functionality. So it might be a good idea +to change to GNU `gettext' as soon as possible. + + So you need _not_ provide this option if you are using GNU libc 2 or +you have installed a recent copy of the GNU gettext package with the +included `libintl'. + +INSTALL Matters +=============== + +Some packages are "localizable" when properly installed; the programs +they contain can be made to speak your own native language. Most such +packages use GNU `gettext'. Other packages have their own ways to +internationalization, predating GNU `gettext'. + + By default, this package will be installed to allow translation of +messages. It will automatically detect whether the system already +provides the GNU `gettext' functions. If not, the GNU `gettext' own +library will be used. This library is wholly contained within this +package, usually in the `intl/' subdirectory, so prior installation of +the GNU `gettext' package is _not_ required. Installers may use +special options at configuration time for changing the default +behaviour. The commands: + + ./configure --with-included-gettext + ./configure --disable-nls + +will respectively bypass any pre-existing `gettext' to use the +internationalizing routines provided within this package, or else, +_totally_ disable translation of messages. + + When you already have GNU `gettext' installed on your system and run +configure without an option for your new package, `configure' will +probably detect the previously built and installed `libintl.a' file and +will decide to use this. This might be not what is desirable. You +should use the more recent version of the GNU `gettext' library. I.e. +if the file `intl/VERSION' shows that the library which comes with this +package is more recent, you should use + + ./configure --with-included-gettext + +to prevent auto-detection. + + The configuration process will not test for the `catgets' function +and therefore it will not be used. The reason is that even an +emulation of `gettext' on top of `catgets' could not provide all the +extensions of the GNU `gettext' library. + + Internationalized packages have usually many `po/LL.po' files, where +LL gives an ISO 639 two-letter code identifying the language. Unless +translations have been forbidden at `configure' time by using the +`--disable-nls' switch, all available translations are installed +together with the package. However, the environment variable `LINGUAS' +may be set, prior to configuration, to limit the installed set. +`LINGUAS' should then contain a space separated list of two-letter +codes, stating which languages are allowed. + +Using This Package +================== + +As a user, if your language has been installed for this package, you +only have to set the `LANG' environment variable to the appropriate +`LL_CC' combination. Here `LL' is an ISO 639 two-letter language code, +and `CC' is an ISO 3166 two-letter country code. For example, let's +suppose that you speak German and live in Germany. At the shell +prompt, merely execute `setenv LANG de_DE' (in `csh'), +`export LANG; LANG=de_DE' (in `sh') or `export LANG=de_DE' (in `bash'). +This can be done from your `.login' or `.profile' file, once and for +all. + + You might think that the country code specification is redundant. +But in fact, some languages have dialects in different countries. For +example, `de_AT' is used for Austria, and `pt_BR' for Brazil. The +country code serves to distinguish the dialects. + + The locale naming convention of `LL_CC', with `LL' denoting the +language and `CC' denoting the country, is the one use on systems based +on GNU libc. On other systems, some variations of this scheme are +used, such as `LL' or `LL_CC.ENCODING'. You can get the list of +locales supported by your system for your country by running the command +`locale -a | grep '^LL''. + + Not all programs have translations for all languages. By default, an +English message is shown in place of a nonexistent translation. If you +understand other languages, you can set up a priority list of languages. +This is done through a different environment variable, called +`LANGUAGE'. GNU `gettext' gives preference to `LANGUAGE' over `LANG' +for the purpose of message handling, but you still need to have `LANG' +set to the primary language; this is required by other parts of the +system libraries. For example, some Swedish users who would rather +read translations in German than English for when Swedish is not +available, set `LANGUAGE' to `sv:de' while leaving `LANG' to `sv_SE'. + + Special advice for Norwegian users: The language code for Norwegian +bokma*l changed from `no' to `nb' recently (in 2003). During the +transition period, while some message catalogs for this language are +installed under `nb' and some older ones under `no', it's recommended +for Norwegian users to set `LANGUAGE' to `nb:no' so that both newer and +older translations are used. + + In the `LANGUAGE' environment variable, but not in the `LANG' +environment variable, `LL_CC' combinations can be abbreviated as `LL' +to denote the language's main dialect. For example, `de' is equivalent +to `de_DE' (German as spoken in Germany), and `pt' to `pt_PT' +(Portuguese as spoken in Portugal) in this context. + +Translating Teams +================= + +For the Free Translation Project to be a success, we need interested +people who like their own language and write it well, and who are also +able to synergize with other translators speaking the same language. +Each translation team has its own mailing list. The up-to-date list of +teams can be found at the Free Translation Project's homepage, +`http://www.iro.umontreal.ca/contrib/po/HTML/', in the "National teams" +area. + + If you'd like to volunteer to _work_ at translating messages, you +should become a member of the translating team for your own language. +The subscribing address is _not_ the same as the list itself, it has +`-request' appended. For example, speakers of Swedish can send a +message to `sv-request@li.org', having this message body: + + subscribe + + Keep in mind that team members are expected to participate +_actively_ in translations, or at solving translational difficulties, +rather than merely lurking around. If your team does not exist yet and +you want to start one, or if you are unsure about what to do or how to +get started, please write to `translation@iro.umontreal.ca' to reach the +coordinator for all translator teams. + + The English team is special. It works at improving and uniformizing +the terminology in use. Proven linguistic skill are praised more than +programming skill, here. + +Available Packages +================== + +Languages are not equally supported in all packages. The following +matrix shows the current state of internationalization, as of January +2004. The matrix shows, in regard of each package, for which languages +PO files have been submitted to translation coordination, with a +translation percentage of at least 50%. + + Ready PO files af am ar az be bg bs ca cs da de el en en_GB eo es + +----------------------------------------------------+ + a2ps | [] [] [] [] | + aegis | () | + ant-phone | () | + anubis | | + ap-utils | | + aspell | [] | + bash | [] [] [] [] | + batchelor | | + bfd | [] [] | + binutils | [] [] | + bison | [] [] [] | + bluez-pin | [] [] [] | + clisp | | + clisp | [] [] [] | + console-tools | [] [] | + coreutils | [] [] [] [] | + cpio | [] [] [] | + darkstat | [] () [] | + diffutils | [] [] [] [] [] [] [] | + e2fsprogs | [] [] [] | + enscript | [] [] [] [] | + error | [] [] [] [] [] | + fetchmail | [] () [] [] [] [] | + fileutils | [] [] [] | + findutils | [] [] [] [] [] [] [] | + flex | [] [] [] [] | + fslint | | + gas | [] | + gawk | [] [] [] [] | + gbiff | [] | + gcal | [] | + gcc | [] [] | + gettext | [] [] [] [] [] | + gettext-examples | [] [] [] [] | + gettext-runtime | [] [] [] [] [] | + gettext-tools | [] [] [] | + gimp-print | [] [] [] [] [] | + gliv | | + glunarclock | [] [] | + gnubiff | [] | + gnucash | [] () [] [] | + gnucash-glossary | [] () [] | + gnupg | [] () [] [] [] [] | + gpe-aerial | [] | + gpe-beam | [] [] | + gpe-calendar | [] [] | + gpe-clock | [] [] | + gpe-conf | [] [] | + gpe-contacts | [] [] | + gpe-edit | [] | + gpe-go | [] | + gpe-login | [] [] | + gpe-ownerinfo | [] [] | + gpe-sketchbook | [] [] | + gpe-su | [] [] | + gpe-taskmanager | [] [] | + gpe-timesheet | [] | + gpe-today | [] [] | + gpe-todo | [] [] | + gphoto2 | [] [] [] [] | + gprof | [] [] [] | + gpsdrive | () () () | + gramadoir | [] | + grep | [] [] [] [] [] [] | + gretl | [] | + gtick | [] () | + hello | [] [] [] [] [] [] | + id-utils | [] [] | + indent | [] [] [] [] | + iso_3166 | [] [] [] [] [] [] [] [] [] [] | + iso_3166_1 | [] [] [] [] [] [] | + iso_3166_2 | | + iso_3166_3 | [] | + iso_4217 | [] [] [] [] | + iso_639 | | + jpilot | [] [] [] | + jtag | | + jwhois | [] | + kbd | [] [] [] [] [] | + latrine | () | + ld | [] [] | + libc | [] [] [] [] [] [] | + libgpewidget | [] [] | + libiconv | [] [] [] [] [] | + lifelines | [] () | + lilypond | [] | + lingoteach | | + lingoteach_lessons | () () | + lynx | [] [] [] [] | + m4 | [] [] [] [] | + mailutils | [] [] | + make | [] [] [] | + man-db | [] () [] [] () | + minicom | [] [] [] | + mysecretdiary | [] [] [] | + nano | [] () [] [] [] | + nano_1_0 | [] () [] [] [] | + opcodes | [] | + parted | [] [] [] [] [] | + ptx | [] [] [] [] [] | + python | | + radius | [] | + recode | [] [] [] [] [] [] [] | + rpm | [] [] | + screem | | + scrollkeeper | [] [] [] [] [] [] | + sed | [] [] [] [] [] [] | + sh-utils | [] [] [] | + shared-mime-info | | + sharutils | [] [] [] [] [] [] | + silky | () | + skencil | [] () [] | + sketch | [] () [] | + soundtracker | [] [] [] | + sp | [] | + tar | [] [] [] [] | + texinfo | [] [] [] | + textutils | [] [] [] [] | + tin | () () | + tp-robot | | + tuxpaint | [] [] [] [] [] [] [] | + unicode-han-tra... | | + unicode-transla... | | + util-linux | [] [] [] [] [] | + vorbis-tools | [] [] [] [] | + wastesedge | () | + wdiff | [] [] [] [] | + wget | [] [] [] [] [] [] | + xchat | [] [] [] [] | + xfree86_xkb_xml | [] [] | + xpad | [] | + +----------------------------------------------------+ + af am ar az be bg bs ca cs da de el en en_GB eo es + 4 0 0 1 9 4 1 40 41 60 78 17 1 5 13 68 + + et eu fa fi fr ga gl he hr hu id is it ja ko lg + +-------------------------------------------------+ + a2ps | [] [] [] () () | + aegis | | + ant-phone | [] | + anubis | [] | + ap-utils | [] | + aspell | [] [] | + bash | [] [] | + batchelor | [] [] | + bfd | [] | + binutils | [] [] | + bison | [] [] [] [] | + bluez-pin | [] [] [] [] [] | + clisp | | + clisp | [] | + console-tools | | + coreutils | [] [] [] [] [] [] | + cpio | [] [] [] [] | + darkstat | () [] [] [] | + diffutils | [] [] [] [] [] [] [] | + e2fsprogs | | + enscript | [] [] | + error | [] [] [] [] | + fetchmail | [] | + fileutils | [] [] [] [] [] [] | + findutils | [] [] [] [] [] [] [] [] [] [] [] | + flex | [] [] [] | + fslint | [] | + gas | [] | + gawk | [] [] [] | + gbiff | [] | + gcal | [] | + gcc | [] | + gettext | [] [] [] | + gettext-examples | [] [] | + gettext-runtime | [] [] [] [] [] | + gettext-tools | [] [] [] | + gimp-print | [] [] | + gliv | () | + glunarclock | [] [] [] [] | + gnubiff | [] | + gnucash | () [] | + gnucash-glossary | [] | + gnupg | [] [] [] [] [] [] [] | + gpe-aerial | [] | + gpe-beam | [] | + gpe-calendar | [] [] [] | + gpe-clock | [] | + gpe-conf | [] | + gpe-contacts | [] [] | + gpe-edit | [] [] | + gpe-go | [] | + gpe-login | [] [] | + gpe-ownerinfo | [] [] [] | + gpe-sketchbook | [] | + gpe-su | [] | + gpe-taskmanager | [] | + gpe-timesheet | [] [] [] | + gpe-today | [] [] | + gpe-todo | [] [] | + gphoto2 | [] [] [] | + gprof | [] [] | + gpsdrive | () () () | + gramadoir | [] [] | + grep | [] [] [] [] [] [] [] [] [] [] [] | + gretl | [] [] | + gtick | [] [] [] | + hello | [] [] [] [] [] [] [] [] [] [] [] [] [] | + id-utils | [] [] [] [] | + indent | [] [] [] [] [] [] [] [] [] | + iso_3166 | [] [] [] [] [] [] [] | + iso_3166_1 | [] [] [] [] [] | + iso_3166_2 | | + iso_3166_3 | | + iso_4217 | [] [] [] [] [] [] | + iso_639 | | + jpilot | [] () | + jtag | [] | + jwhois | [] [] [] [] | + kbd | [] | + latrine | [] | + ld | [] | + libc | [] [] [] [] [] [] | + libgpewidget | [] [] [] [] | + libiconv | [] [] [] [] [] [] [] [] [] | + lifelines | () | + lilypond | [] | + lingoteach | [] [] | + lingoteach_lessons | | + lynx | [] [] [] [] | + m4 | [] [] [] [] | + mailutils | | + make | [] [] [] [] [] [] | + man-db | () () | + minicom | [] [] [] [] | + mysecretdiary | [] [] | + nano | [] [] [] [] | + nano_1_0 | [] [] [] [] | + opcodes | [] | + parted | [] [] [] | + ptx | [] [] [] [] [] [] [] | + python | | + radius | [] | + recode | [] [] [] [] [] [] | + rpm | [] [] | + screem | | + scrollkeeper | [] | + sed | [] [] [] [] [] [] [] [] [] | + sh-utils | [] [] [] [] [] [] [] | + shared-mime-info | [] [] [] | + sharutils | [] [] [] [] [] | + silky | () [] () () | + skencil | [] | + sketch | [] | + soundtracker | [] [] | + sp | [] () | + tar | [] [] [] [] [] [] [] [] [] | + texinfo | [] [] [] [] | + textutils | [] [] [] [] [] [] | + tin | [] () | + tp-robot | [] | + tuxpaint | [] [] [] [] [] [] [] [] [] | + unicode-han-tra... | | + unicode-transla... | [] [] | + util-linux | [] [] [] [] () [] | + vorbis-tools | [] | + wastesedge | () | + wdiff | [] [] [] [] [] [] | + wget | [] [] [] [] [] [] [] | + xchat | [] [] [] | + xfree86_xkb_xml | [] [] | + xpad | [] [] | + +-------------------------------------------------+ + et eu fa fi fr ga gl he hr hu id is it ja ko lg + 22 2 1 26 106 28 24 8 10 41 33 1 26 33 12 0 + + lt lv mk mn ms mt nb nl nn no nso pl pt pt_BR ro ru + +-----------------------------------------------------+ + a2ps | [] [] () () [] [] [] | + aegis | () () () | + ant-phone | [] [] | + anubis | [] [] [] [] [] [] | + ap-utils | [] () [] | + aspell | [] | + bash | [] [] [] | + batchelor | [] | + bfd | [] | + binutils | [] | + bison | [] [] [] [] [] | + bluez-pin | [] [] [] | + clisp | | + clisp | [] | + console-tools | [] | + coreutils | [] [] | + cpio | [] [] [] [] [] | + darkstat | [] [] [] [] | + diffutils | [] [] [] [] [] [] | + e2fsprogs | [] | + enscript | [] [] [] [] | + error | [] [] [] | + fetchmail | [] [] () [] | + fileutils | [] [] [] | + findutils | [] [] [] [] [] | + flex | [] [] [] [] | + fslint | [] [] | + gas | | + gawk | [] [] [] | + gbiff | [] [] | + gcal | | + gcc | | + gettext | [] [] [] | + gettext-examples | [] [] [] | + gettext-runtime | [] [] [] [] | + gettext-tools | [] [] | + gimp-print | [] | + gliv | [] [] [] | + glunarclock | [] [] [] [] | + gnubiff | [] | + gnucash | [] [] () [] | + gnucash-glossary | [] [] | + gnupg | [] | + gpe-aerial | [] [] [] [] | + gpe-beam | [] [] [] [] | + gpe-calendar | [] [] [] [] | + gpe-clock | [] [] [] [] | + gpe-conf | [] [] [] [] | + gpe-contacts | [] [] [] [] | + gpe-edit | [] [] [] [] | + gpe-go | [] [] [] | + gpe-login | [] [] [] [] | + gpe-ownerinfo | [] [] [] [] | + gpe-sketchbook | [] [] [] [] | + gpe-su | [] [] [] [] | + gpe-taskmanager | [] [] [] [] | + gpe-timesheet | [] [] [] [] | + gpe-today | [] [] [] [] | + gpe-todo | [] [] [] [] | + gphoto2 | [] | + gprof | [] [] | + gpsdrive | () () [] | + gramadoir | () [] | + grep | [] [] [] [] [] | + gretl | | + gtick | [] [] [] | + hello | [] [] [] [] [] [] [] [] [] [] | + id-utils | [] [] [] [] | + indent | [] [] [] [] | + iso_3166 | [] [] [] | + iso_3166_1 | [] [] | + iso_3166_2 | | + iso_3166_3 | [] | + iso_4217 | [] [] [] [] [] [] [] [] | + iso_639 | [] | + jpilot | () () | + jtag | | + jwhois | [] [] [] [] () | + kbd | [] [] [] | + latrine | [] | + ld | | + libc | [] [] [] [] | + libgpewidget | [] [] [] | + libiconv | [] [] [] [] [] | + lifelines | | + lilypond | | + lingoteach | | + lingoteach_lessons | | + lynx | [] [] [] | + m4 | [] [] [] [] [] | + mailutils | [] [] [] | + make | [] [] [] [] | + man-db | [] | + minicom | [] [] [] [] | + mysecretdiary | [] [] [] | + nano | [] [] [] [] [] | + nano_1_0 | [] [] [] [] [] [] | + opcodes | [] [] | + parted | [] [] [] [] | + ptx | [] [] [] [] [] [] [] [] | + python | | + radius | [] [] | + recode | [] [] [] [] | + rpm | [] [] [] | + screem | | + scrollkeeper | [] [] [] [] [] | + sed | [] [] [] | + sh-utils | [] [] | + shared-mime-info | [] [] | + sharutils | [] [] | + silky | () | + skencil | [] [] | + sketch | [] [] | + soundtracker | | + sp | | + tar | [] [] [] [] [] [] | + texinfo | [] [] [] [] | + textutils | [] [] | + tin | | + tp-robot | [] | + tuxpaint | [] [] [] [] [] [] [] [] | + unicode-han-tra... | | + unicode-transla... | | + util-linux | [] [] [] | + vorbis-tools | [] [] [] | + wastesedge | | + wdiff | [] [] [] [] [] | + wget | [] [] [] | + xchat | [] [] [] | + xfree86_xkb_xml | [] [] | + xpad | [] [] | + +-----------------------------------------------------+ + lt lv mk mn ms mt nb nl nn no nso pl pt pt_BR ro ru + 1 2 0 3 12 0 10 69 6 7 1 40 26 36 76 63 + + sk sl sr sv ta th tr uk ven vi wa xh zh_CN zh_TW zu + +-----------------------------------------------------+ + a2ps | [] [] [] [] | 16 + aegis | | 0 + ant-phone | | 3 + anubis | [] [] | 9 + ap-utils | () | 3 + aspell | | 4 + bash | | 9 + batchelor | | 3 + bfd | [] [] | 6 + binutils | [] [] [] | 8 + bison | [] [] | 14 + bluez-pin | [] [] [] | 14 + clisp | | 0 + clisp | | 5 + console-tools | | 3 + coreutils | [] [] [] [] | 16 + cpio | [] [] | 14 + darkstat | [] [] [] () () | 12 + diffutils | [] [] [] | 23 + e2fsprogs | [] [] | 6 + enscript | [] [] | 12 + error | [] [] [] | 15 + fetchmail | [] [] | 11 + fileutils | [] [] [] [] [] | 17 + findutils | [] [] [] [] [] [] | 29 + flex | [] [] | 13 + fslint | | 3 + gas | [] | 3 + gawk | [] [] | 12 + gbiff | | 4 + gcal | [] [] | 4 + gcc | [] | 4 + gettext | [] [] [] [] [] | 16 + gettext-examples | [] [] [] [] [] | 14 + gettext-runtime | [] [] [] [] [] [] [] [] | 22 + gettext-tools | [] [] [] [] [] [] | 14 + gimp-print | [] [] | 10 + gliv | | 3 + glunarclock | [] [] [] | 13 + gnubiff | | 3 + gnucash | [] [] | 9 + gnucash-glossary | [] [] [] | 8 + gnupg | [] [] [] [] | 17 + gpe-aerial | [] | 7 + gpe-beam | [] | 8 + gpe-calendar | [] [] [] [] | 13 + gpe-clock | [] [] [] | 10 + gpe-conf | [] [] | 9 + gpe-contacts | [] [] [] | 11 + gpe-edit | [] [] [] [] [] | 12 + gpe-go | | 5 + gpe-login | [] [] [] [] [] | 13 + gpe-ownerinfo | [] [] [] [] | 13 + gpe-sketchbook | [] [] | 9 + gpe-su | [] [] [] | 10 + gpe-taskmanager | [] [] [] | 10 + gpe-timesheet | [] [] [] [] | 12 + gpe-today | [] [] [] [] [] | 13 + gpe-todo | [] [] [] [] | 12 + gphoto2 | [] [] [] | 11 + gprof | [] [] | 9 + gpsdrive | [] [] | 3 + gramadoir | [] | 5 + grep | [] [] [] [] | 26 + gretl | | 3 + gtick | | 7 + hello | [] [] [] [] [] | 34 + id-utils | [] [] | 12 + indent | [] [] [] [] | 21 + iso_3166 | [] [] [] [] [] [] [] | 27 + iso_3166_1 | [] [] [] | 16 + iso_3166_2 | | 0 + iso_3166_3 | | 2 + iso_4217 | [] [] [] [] [] [] | 24 + iso_639 | | 1 + jpilot | [] [] [] [] [] | 9 + jtag | [] | 2 + jwhois | () [] [] | 11 + kbd | [] [] | 11 + latrine | | 2 + ld | [] [] | 5 + libc | [] [] [] [] | 20 + libgpewidget | [] [] [] [] | 13 + libiconv | [] [] [] [] [] [] [] [] | 27 + lifelines | [] | 2 + lilypond | [] | 3 + lingoteach | | 2 + lingoteach_lessons | () | 0 + lynx | [] [] [] | 14 + m4 | [] [] | 15 + mailutils | | 5 + make | [] [] [] | 16 + man-db | [] | 5 + minicom | | 11 + mysecretdiary | [] [] | 10 + nano | [] [] [] [] | 17 + nano_1_0 | [] [] [] | 17 + opcodes | [] [] | 6 + parted | [] [] [] | 15 + ptx | [] [] | 22 + python | | 0 + radius | | 4 + recode | [] [] [] | 20 + rpm | [] [] | 9 + screem | [] [] | 2 + scrollkeeper | [] [] [] | 15 + sed | [] [] [] [] [] [] | 24 + sh-utils | [] [] | 14 + shared-mime-info | [] [] | 7 + sharutils | [] [] [] [] | 17 + silky | () | 3 + skencil | [] | 6 + sketch | [] | 6 + soundtracker | [] [] | 7 + sp | [] | 3 + tar | [] [] [] [] [] | 24 + texinfo | [] [] [] | 14 + textutils | [] [] [] [] | 16 + tin | | 1 + tp-robot | | 2 + tuxpaint | [] [] [] [] [] | 29 + unicode-han-tra... | | 0 + unicode-transla... | | 2 + util-linux | [] [] | 15 + vorbis-tools | | 8 + wastesedge | | 0 + wdiff | [] [] [] | 18 + wget | [] [] [] [] [] [] [] [] | 24 + xchat | [] [] [] [] [] | 15 + xfree86_xkb_xml | [] [] [] [] [] | 11 + xpad | | 5 + +-----------------------------------------------------+ + 63 teams sk sl sr sv ta th tr uk ven vi wa xh zh_CN zh_TW zu + 131 domains 47 19 28 83 0 0 59 13 1 1 11 0 22 22 0 1373 + + Some counters in the preceding matrix are higher than the number of +visible blocks let us expect. This is because a few extra PO files are +used for implementing regional variants of languages, or language +dialects. + + For a PO file in the matrix above to be effective, the package to +which it applies should also have been internationalized and +distributed as such by its maintainer. There might be an observable +lag between the mere existence a PO file and its wide availability in a +distribution. + + If January 2004 seems to be old, you may fetch a more recent copy of +this `ABOUT-NLS' file on most GNU archive sites. The most up-to-date +matrix with full percentage details can be found at +`http://www.iro.umontreal.ca/contrib/po/HTML/matrix.html'. + +Using `gettext' in new packages +=============================== + +If you are writing a freely available program and want to +internationalize it you are welcome to use GNU `gettext' in your +package. Of course you have to respect the GNU Library General Public +License which covers the use of the GNU `gettext' library. This means +in particular that even non-free programs can use `libintl' as a shared +library, whereas only free software can use `libintl' as a static +library or use modified versions of `libintl'. + + Once the sources are changed appropriately and the setup can handle +the use of `gettext' the only thing missing are the translations. The +Free Translation Project is also available for packages which are not +developed inside the GNU project. Therefore the information given above +applies also for every other Free Software Project. Contact +`translation@iro.umontreal.ca' to make the `.pot' files available to +the translation teams. + diff --git a/contrib/diff/AUTHORS b/contrib/diff/AUTHORS new file mode 100644 index 0000000..f5f0e90 --- /dev/null +++ b/contrib/diff/AUTHORS @@ -0,0 +1,112 @@ +Authors of GNU diffutils. + + Copyright 2001 Free Software Foundation, Inc. + + This file is part of GNU diffutils. + + GNU diffutils is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU diffutils is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with GNU diffutils; see the file COPYING. If not, write to + the Free Software Foundation, Inc., 59 Temple Place - Suite 330, + Boston, MA 02111-1307, USA. + +The following contributions warranted legal paper exchanges with the +Free Software Foundation. Also see files ChangeLog and THANKS. + +DIFFUTILS Leonard H. Tower Jr. US 1949 1987-03-09 +Assigns diff (diff.c, initial version). + +DIFFUTILS Torbjorn Granlund Sweden 1961 1988-01-11 +Assigns cmp. +tege@matematik.su.se + +DIFFUTILS Mike Haertel US 1967 1988-09-16 +Assigns changes to diff. + +DIFFUTILS David S. Hayes US ? 1988-01-12 +Assigns changes to diff. + +DIFFUTILS Randall Smith US 1964 1988-09-21 +Assigns diff3. + +DIFFUTILS Richard Stallman US 1953 1988-01-15 +Assigns changes to GNU Diff. + +DIFFUTILS F. Thomas May US 1965 1989-08-22 +Assigns changes to diff (for -D). + +DIFFUTILS Optimal Solutions, Inc. 1989-08-14 +Disclaims changes by Thomas May to diff. + +DIFFUTILS Wayne Davison 1990-09-10 +Disclaims changes to diff. + +DIFFUTILS Digital Research Inc. 1990-09-13 +Disclaims changes by Wayne Davison to diff. + +DIFFUTILS Paul Eggert 1990-03-16 +Disclaims changes to diff. +eggert@twinsun.com + +DIFFUTILS Paul Eggert 1990-08-14 +Disclaims changes to GNU Diff. +eggert@twinsun.com + +DIFFUTILS Twin Sun Inc. 1990-03-16 +Disclaims changes to GNU Diff by Paul Eggert. + +DIFFUTILS Twin Sun Inc. 1990-08-14 +Disclaims changes to GNU Diff by Paul Eggert. + +DIFFUTILS Chip Rosenthal US 1959 1990-03-06 +Assigns changes to diff. +chip@chinacat.Unicom.COM + +DIFFUTILS Unicom Systems Development 1990-03-06 +Disclaims changes by Chip Rosenthal to diff. + +GCC DIFFUTILS Paul Eggert and Twin Sun Inc. 1992-03-11 +Disclaims changes by Paul Eggert to gcc and diff. +eggert@twinsun.com + +DIFF Wayne Davison 1993-06-20 +Disclaims diffcvt.c. + +DIFFUTILS Francois Pinard Canada 1949 1993-01-15 +Assigns wdiff and future changes submitted to the FSF. +pinard@iro.umontreal.ca + +DIFFUTILS Patrick D'Cruze Australia 1971 1994-11-10 +Assigns changes (makefile.in, analyze.c, cmp.c, error.c, diff.c, +diff3.c, getopt.c, getopt1.c, regex.c, sdiff.c, util.c, xmalloc.c; +new file: language.++) + +DIFFUTILS Paul R. Eggert US 1954 1997-04-07 +Assigns past and future changes. +eggert@twinsun.com + +DIFFUTILS Paul R. Eggert US 1954 1997-04-07 +Assigns past and future changes to manual. +eggert@twinsun.com + +ANY DIFFUTILS GNATS Cyclic Software 1997-11-11 +Assigns past and future works (work for hire by Tim Pierce (diffutils) and +Abe Feldman (GNATS)). +kingdon@cyclic.com + +WEBPAGES Gregory B. Harvey Canada 1976 1998-02-14 +Assigns web pages describing GNU Diffutils and future changes. + +DIFFUTILS Olga Nikulin Russia 1965 2001-01-11 +Assigns changes to diff. (diffutils-2.7.2/analyze.c, context.c, diff.[ch], +ed.c, ifdef.c, io.c, normal.c, side.c, util.c) +onikulin@yahoo.com diff --git a/contrib/diff/COPYING b/contrib/diff/COPYING index a43ea21..d60c31a 100644 --- a/contrib/diff/COPYING +++ b/contrib/diff/COPYING @@ -2,7 +2,7 @@ Version 2, June 1991 Copyright (C) 1989, 1991 Free Software Foundation, Inc. - 675 Mass Ave, Cambridge, MA 02139, USA + 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA Everyone is permitted to copy and distribute verbatim copies of this license document, but changing it is not allowed. @@ -279,7 +279,7 @@ POSSIBILITY OF SUCH DAMAGES. END OF TERMS AND CONDITIONS - Appendix: How to Apply These Terms to Your New Programs + How to Apply These Terms to Your New Programs If you develop a new program, and you want it to be of the greatest possible use to the public, the best way to achieve this is to make it @@ -291,7 +291,7 @@ convey the exclusion of warranty; and each file should have at least the "copyright" line and a pointer to where the full notice is found. - Copyright (C) 19yy + Copyright (C) This program is free software; you can redistribute it and/or modify it under the terms of the GNU General Public License as published by @@ -305,14 +305,15 @@ the "copyright" line and a pointer to where the full notice is found. You should have received a copy of the GNU General Public License along with this program; if not, write to the Free Software - Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. + Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA + Also add information on how to contact you by electronic and paper mail. If the program is interactive, make it output a short notice like this when it starts in an interactive mode: - Gnomovision version 69, Copyright (C) 19yy name of author + Gnomovision version 69, Copyright (C) year name of author Gnomovision comes with ABSOLUTELY NO WARRANTY; for details type `show w'. This is free software, and you are welcome to redistribute it under certain conditions; type `show c' for details. diff --git a/contrib/diff/ChangeLog b/contrib/diff/ChangeLog index 6cfc779..291919d 100644 --- a/contrib/diff/ChangeLog +++ b/contrib/diff/ChangeLog @@ -1,3 +1,2100 @@ +2004-04-13 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.8.7. + * configure.ac (AM_GNU_GETTEXT_VERSION): Add. + (XGETTEXT): Restore from pre-2004-04-12 version. This fixes + a bug that lost many msgids in doc/diffutils.pot. + * bootstrap: New file. + * exgettext: Don't generate a temporary file, as this runs afoul + of "make distcheck" which operates with read-only directories. + * Makefile.am (EXTRA_DIST): Add bootstrap. + Remove config/config.rpath as it is deduced automatically these days. + +2004-04-12 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.8.6. + + * NEWS: Add news for 2.8.4, 2.8.6. + + * README: Move copyright notice to end. Defer to "configure + --help" for special "configure" options. Suggest latest libiconv. + Update version numbers of Autoconf etc. to current. + + * configure.ac: Quote various arguments better. + (AC_CONFIG_MACRO_DIR): Add call, specifying "m4". + (AC_CONFIG_HEADER): Replaces AM_CONFIG_HEADER. + (gl_USE_SYSTEM_EXTENSIONS): Replaces AC_GNU_SOURCE. + (AC_ISC_POSIX): Remove; nobody ports to ancient ISC any more. + (AC_PROG_CPP, AC_PROG_INSTALL, AC_C_INLINE, + AC_HEADER_STDBOOL, AC_HEADER_STDC, AM_GNU_GETTEXT, XGETTEXT, + AC_HEADER_STAT, AC_FUNC_VPRINTF, jm_FUNC_GLIBC_UNLOCKED_IO, + jm_FUNC_GNU_STRFTIME, jm_FUNC_MALLOC, jm_FUNC_REALLOC, + jm_PREREQ_C_STACK, jm_PREREQ_ERROR, jm_PREREQ_HARD_LOCALE, + jm_PREREQ_QUOTEARG, jm_PREREQ_REGEX, AC_FUNC_FNMATCH_GNU, jm_AC_DOS): + Remove; not needed here, as our files don't use them directly + or we rely on gnulib modules. + (AC_C_CONST): Remove; we assume C89 now. + (AC_CHECK_HEADERS): Remove libintl.h, limits.h, stdlib.h, string.h, + time.h. + (AC_CHECK_TYPE): Remove ptrdiff_t, ssize_t. + (AC_CHECK_FUNCS): Remove diraccess, strchr, strerror, tmpnam). + (AC_REPLACE_FUNCS): Remove memchr, mkstemp, strcasecmp. + (GNULIB_AUTOCONF_SNIPPET): Add call. This replaces much of + the above. + (AC_CONFIG_FILES): Remove lib/posix/Makefile. + (AC_CONFIG_COMMANDS): Remove. + + * doc/diff.texi (dircategory): Change to "Text creation and + manipulation" from "GNU packages". + (Translations): New node. + (Overview): Improve quality of algorithm citations. + (Binary): -q doesn't exactly cause diff to assume files are binary. + (Normal): Place after Side by Side, since it's less important. + (Detailed Context, Detailed Unified, Detailed ed, + Detailed if-then-else, diff3 Hunks, Detailed diff3 Normal): + Place at end of menu. + (Detailed Unified): Mention that fractional timestamps are + omitted on hosts that don't support them. + Mention what happens when hunks contain just one line. + (Line Group Formats, Reject Names): Fix duplicate-word typos. + (Comparing Directories): Trailing white space and empty lines are + ignored in the -X file. + (diff Options): Add --strip-trailing-cr. + (Projects): gnu -> gvc@gnu.org. + + * lib/Makefile.am (SUBDIRS): Remove. + (EXTRA_DIST, noinst_HEADERS): Remove most entries. + (libdiffutils_a_SOURCES): Now just lib_SOURCES. + (lib_SOURCES): New macro. + (DISTCLEANFILES, MOSTLYCLEANFILES): Set to empty now. + (gnulib.mk): Include: this does most of the work eliminated + by the above changes. + + * lib/inttostr.c (inttostr): Protect i < 0 by compile-time + test intended to suppress compiler warnings. + * lib/inttostr.h: Include limits.h unilaterally. + (CHAR_BIT): Remove. + (PARAMS): Remove; all uses changed. + * lib/setmode.c (__attribute__): New macro. + (set_binary_mode): Define only if HAVE_SETMODE_DOS. + Otherwise define a dummy static char, as C89 requires + that we define something. + * lib/setmode.h (set_binary_mode): Return true, not 1. + + * src/analyze.c, src/context.c, src/diff.c, src/io.c, src/util.c: + Do not include regex.h, as diff.h does this now. + + * src/cmp.c: Sort includes. Include , . + (specify_comparison_type): Don't report an error if the comparison + type has already been specified the same way as this one. + + * src/cmp.c (usage): Mention exit status. + * src/diff.c (option_help_msgid): Likewise. + * src/diff3.c (usage): Likewise. + * src/sdiff.c (usage): Likewise. + + * src/cmp.c (main): Adjust to latest gnulib c_stack_action + calling conventions. + * src/diff.c (main): Likewise. + * src/diff3.c (main): Likewise. + * src/sdiff.c (main): Likewise. + + * src/cmp.c (main): Adjust to latest version_etc calling conventions. + * src/diff.c (main): Likewise. + * src/diff3.c (main): Likewise. + * src/sdiff.c (main): Likewise. + + * src/diff.c: Include . + (binary): Define to true if not declared. + (longopts): Set tabsize flag to 1. + (main): Don't output nanoseconds if platform lacks them. + Don't treat files as binary if !binary. + (set_mtime_to_now): Use 0, not NULL. + (compare_files): Mark files as nonexistent if it looks like + 'patch' created inaccessible regular empty files to denote + nonexistent backups. Don't compare such files. + Clear st_* members of status of nonexistent file. + Remove now-unnecessary tests. + + * src/diff.h: Include regex.h, unlocked-io.h. + (struct file_data.changed): Now char *, not bool *, to save + space on hosts where bool takes more space than char. + All uses changed. + + * src/diff3.c: Include unlocked-io.h. + (strip_trailing_cr): New var. + (STRIP_TRAILING_CR_OPTION): New enum. + (longopts, main, option_help_msgid, read_diff): + Add --strip-trailing-cr support. + (read_diff): Exit with status 126 (not 127) if errno != ENOENT + after failed execvp in child. Report this in parent. + + * src/dir.c: Include . + (failed_locale_specific_sorting): Renamed from failed_strcoll. + All uses changed. + (compare_names): Don't invoke strcasecmp first thing when + ignore_file_name_case; if locale_specific_sorting, we should + just use that. + + * src/ifdef.c (next_line): Remove; replace with... + (next_line0, next_line1): New vars. + (print_ifdef_script, print_ifdef_hunk): + Use them to fix line-number computation bug. + + * src/io.c (find_and_hash_each_line): Don't convert char * + to unsigned char *; just leave pointers as char *. This + lessens the number of potentially-dangerous casts. + * src/util.c (lines_differ): Likewise. + + * src/sdiff.c: Include , . + (check_child_status): Renamed from ck_editor_status, and + accept a new arg MAX_OK_STATUS. All callers changed. + Handle status 126/127 as per POSIX. + (edit): Likewise. + (main): Likewise. Fix getopt typo: -E wasn't supported. + + * src/system.h (S_IRWXU, S_IRWXG, S_IRWXO): Define if not defined. + (S_IXUSR, S_IXGRP, S_IXOTH): Remove. + Include unconditionally, since we can assume C89 now. + Likewise for , . + (getenv, EXIT_SUCCESS, EXIT_FAILURE, SSIZE_MAX, strchr, strrchr, + memcmp, memcpy): Remove decl; no longer needed. + (strcasecoll, strcasecmp): Define if not built in. + (CTYPE_DOMAIN, ISPRINT, ISSPACE, TOLOWER, _tolower, errno): Remove; + we now assume C89 or better. All uses changed. + Include unconditionally now, since gnulib supports it + if the C compiler doesn't. All boolean uses of 0 and 1 now + changed to false and true. + (lin_is_printable_as_long_int): Renamed from lin_is_printable_as_long. + + * src/util.c (begin_output): Fix bug: 0 wasn't cast to char * arg, + which led to undefined behavior on 64-bit hosts. + Use more-standard test for exit status 126 versus 127. + (finish_output): Likewise. + (analyze_hunk): Do not cast bool to int. + +2004-03-15 Paul Eggert + + * src/cmp.c (main): Don't consider two files with the same name to + be the same, if their initial skip values differ. This fixes a + bug reported by Hideki Iwamoto in + . + +2004-03-11 Paul Eggert + + * src/analyze.c (diag): Return void, not lin, since the return + value wasn't needed. All callers changed. + (diag, diff_2_files): + Use 'true' and 'false' instead of '1' and '0', when appropriate. + (compareseq): Use lin const * local variables instead of lin *. + Don't bother checking diag's return value. + (shift_boundaries, build_reverse_script, build_script, diff_2_files): + Use char arrays, not bool arrays, since + sizeof (bool) might be greater than 1. + +2004-02-09 Paul Eggert + + * m4/setmode.m4 (AC_FUNC_SETMODE_DOS): AC_LIBOBJ(setmode) if + we would return true. + +2002-10-14 Paul Eggert + + * src/Makefile.am (diff3.$(OBJEXT), diff.$(OBJEXT), + sdiff.$(OBJEXT)): Rename from (misspelled) diff3.$(OBJECT), + diff.$(OBJECT), sdiff.$(OBJECT). Patch by Paul D. Smith in + . + Bug reported by Chris Bainbridge. + +2002-10-13 Paul Eggert + + * src/Makefile.am (MOSTLYCLEANFILES): Add paths.ht. + (paths.h): Send output to paths.ht first, and then rename to + paths.h at the end. This avoids problems if the disk is full. + It also works around what appears to be a bug with GNU make -j + (3.79.1); see . + +2002-06-27 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.8.4. + + * config/config.sub: Sync with latest version maintained in other + packages. + + * lib/file-type.h: Protect against double inclusion. Detect + whether has been included. Fix from Jim Meyering. + + * src/analyze.c (briefly_report): Don't say "Binary files differ", + since one of the files may not be a binary file. + Bug reported by Dan Jacobson. + +2002-06-22 Paul Eggert + + * lib/c-stack.c (segv_handler, c_stack_action) [! defined + SA_SIGINFO]: Do not assume SA_SIGINFO behavior. + Bug reported by Jim Meyering on NetBSD 1.5.2. + +2002-06-16 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.8.3. + + * config/depcomp, config/missing, README: Update to automake 1.6.2. + + * po/LINGUAS: Add en_US. + * po/en_US.po: New file. + * po/POTFILES.in: Remove lib/freesoft.c. + Add lib/file-type.c, lib/version-etc.c, lib/xmalloc.c. + +2002-06-15 Paul Eggert + + * doc/diff.texi (Special Files): Document behavior of symlink + loops. + + * lib/Makefile.am (noinst_HEADERS): Remove freesoft.h. + Add version-etc.h. + (libdiffutils_a_SOURCES): Remove freesoft.c. Add version-etc.c. + * lib/freesoft.c, lib/freesoft.h: Remove. + * lib/version-etc.h (PARAMS): Remove; we now assume C89 at least. + + * lib/version-etc.h (version_etc): Remove package and version args. + (version_etc_copyright): Remove. + * lib/version-etc.c: Likewise. + Do not include unlocked-io.h; no longer needed. + Include gettext.h rather than libinto.h. + (_): Define unconditionally. + (version_etc): Adjust wording to match current GNU coding standards. + Translate "(C)" if possible. + + * lib/version-etc.c, lib/version-etc.h: New files, taken from + fileutils. + + * src/Makefile.am (cmp_SOURCES, diff3_SOURCES, sdiff_SOURCES, + diff_SOURCES): Remove version.c. + (MAINTAINERCLEANFILES, $(srcdir)/version.c): Remove. + + * src/cmp.c: Include version-etc.h, not freesoft.h. + (copyright_notice): Remove. + (main): Use version_etc to print version. + * src/diff.c, src/diff3.c, src/sdiff.c: Likewise. + + * src/cmp.c (version_string): Remove decl. + * src/diff.h, src/diff3.c, src/sdiff.c: Likewise. + +2002-06-11 Paul Eggert + + * lib/fnmatch.c, lib/fnmatch_loop.c (WIDE_CHAR_SUPPORT): + New macro. Use it uniformly instead of + (defined HAVE_WCTYPE_H && defined HAVE_WCHAR_H). + It also uses HAVE_BTOWC, to fix a porting bug on Solaris 2.5.1 + reported by Vin Shelton. + * m4/fnmatch.m4 (_AC_LIBOBJ_FNMATCH): Check for btowc. + + * NEWS, configure.ac (AC_INIT): Version 2.8.2. + + * ABOUT-NLS, config/config.guess, config/config.sub, + config/depcomp, config/texinfo.tex, lib/posix/regex.h, + m4/c-bs-a.m4, m4/gettext.m4, m4/gnu-source.m4, m4/lib-link.m4, + m4/malloc.m4: + Update to recent version (maintained in other packages). + * m4/prereq.m4 (jm_PREREQ_EXCLUSIVE): AC_FUNC_FNMATCH_GNU + no longer takes a lib. + + * README: Incorporate contents from INSTALLME. + * INSTALLME: Remove. + * Makefile.am (EXTRA_DIST): Remove INSTALLME. + + * configure.ac (AC_GNU_SOURCE): Move up, so that it affects + later compilations properly. + (DEFAULT_DIFF_PROGRAM, AC_TYPE_SIGNAL): Remove. + (jm_AC_TYPE_INTMAX_T): Add. + (AC_FUNC_FNMATCH_GNU): Use this, instead of AC_FUNC_FNMATCH. + (AC_CONFIG_LINKS): regex.hin renamed from regex_.h. + + * doc/diff.texi: Reword "@option{-f} and @option{--foo}" to + "@option{-f} or @option{--foo}". + Use @acronym instead of @sc where appropriate. + (Specified Lines): Renamed from Specified Folding. + (Comparison, Blank Lines): + Clarify wordings for Blank Lines and Specified Lines nodes. + (Binary): Mention --verbose and --print-bytes. + (Tabs, sdiff Option Summary, diff Options): + New option --tabsize=COLUMNS. + + * lib/Makefile.am (EXTRA_DIST): Add fnmatch_loop.c. + (noinst_HEADERS): fnmatch_.h renamed from fnmatch.hin. + regex_.h renamed from regex.hin. + Add file-type.h. + (libdiffutils_a_SOURCES): Add file-type.c. + (DISTCLEANFILES): Remove fnmatch.hno, regex.hno. + + * lib/c-stack.c (__attribute__): New macro. + (EOVERFLOW): Define if not defined. + (stack_t): Define to struct sigaltstack if not defined or declared. + Include , if available. + Include if DEBUG. + Do not include or . + (c_stack_die): Remove info and context args. All uses changed. + (segv_action): Likewise. + (alternate_signal_stack): Change uintmax_t to long, to ease porting. + (get_stack_location, min_address_from_argv, max_address_from_argv, + null_action): New functions. + (stack_base, stack_size): New vars. + (segv_handler): context arg may not be used. + Use global stack_base, stack_size if + ! HAVE_XSI_STACK_OVERFLOW_HEURISTIC. + Add debug code. + Invoke die (rather than segv_action) to exit. + (c_stack_action): Accept new argv arg, and simpler handler arg. + All uses changed. Move code into new functions above. + Allow null action. + [! (defined SA_ONSTACK && defined _SC_PAGESIZE)]: Assume all segvs + are stack overflows. + (main) [DEBUG]: Describe what output should be like. + + * lib/c-stack.h (siginfo_t, c_stack_die): Remove decl. + + * lib/file-type.c, lib/file-type.h: New files. These contain code + that was in src/diff.c, but is now librarified and spiffed up a + bit. Jim Meyering suggested this. + + * lib/fnmatch.c (alloca, __builtin_expect): Define for non-GCC hosts. + : Include only if HAVE_STRINGS_H. + : Include if we include stdlib.h. + Do not comment out all code if ! HAVE_FNMATCH_GNU. + (getenv): Do not declare if HAVE_DECL_GETENV. + (__strchrnul, __wcschrnul): Remove; not used. + (MEMPCPY): Use mempcpy if not _LIBC; use memcpy if neither _LIBC + nor HAVE_MEMPCPY. + (FOLD) [HANDLE_MULTIBYTE]: Do not pass wide char to ISUPPER. + (STRLEN, STRCAT, MEMPCPY) [HANDLE_MULTIBYTE && !defined _LIBC]: + Use wcslen rather than __wcslen, and likewise for wcscat, wmempcpy. + (MEMPCPY) [HANDLE_MULTIBYTE]: Use wmempcpy if not _LIBC; use wmemcpy + if neither _LIBC nor HAVE_WMEMPCPY. + * lib/fnmatch_.h (__const): Do not define to empty, as this breaks + Sun cc. The code doesn't work with K&R anyway. + * lib/fnmatch_loop.c (struct patternlist.str): Size 1, not 0, + as C89 requires this. + (NEW_PATTERN): Use offsetof, not sizeof, since str now has size 1. + * lib/fnmatch_.h: Import from glibc fnmatch.h. + * lib/fnmatch.c, lib/fnmatch_loop.c: Import from glibc. + + * lib/posixver.c: Include posixver.h. + + * lib/regex_.h: Renamed from lib/regex.hin. + + * m4/c-stack.m4 (jm_PREREQ_C_STACK): Do not AC_REQUIRE + jm_AC_TYPE_UINTMAX_T and do not use uintmax_t. + Check for sys/resource.h, uccontext.h. + Check for decls and existence of getcontext, sigaltstack. + Check for stack_t. + + * m4/codeset.m4, m4/glibc21.m4, m4/lcmessage.m4: Remove. + + * m4/fnmatch.m4: Update to latest Autoconf CVS for AC_FUNC_FNMATCH_GNU. + * m4/gnu-source.m4: Likewise, for AC_GNU_SOURCE (renamed from + AC__GNU_SOURCE). + + * m4/mbstate_t.m4 (AC_TYPE_MBSTATE_T): Renamed from AC_MBSTATE_T. + All uses changed. Upgrade to recent Autoconf CVS. + + * m4/stdbool.m4 (AC_HEADER_STDBOOL): Do not cast pointer to + bool in integer constant expression; C99 does not allow it. + Reported by Bruno Haible. + + * po/LINGUAS: Add hu, pt_BR. + * po/hu.po, po/pt_BR.po: New files. + + * src/Makefile.am (noinst_HEADERS): Remove diff.h. + (DEFS): Remove. + (diff_sources): Add diff.h. + (MOSTLYCLEANFILES): New macro. + (cmp.$(OBJEXT) diff3.$(OBJECT) diff.$(OBJECT) sdiff.$(OBJECT)): Depend + on paths.h. + (paths.h): New rule. + + * src/analyze.c, src/cmp.c, src/diff.c, src/diff3.c, src/io.c, + src/sdiff.c: Include . + + * src/cmp.c: Include paths.h. + (copyright_notice): Renamed from copyright_string. + Now a msgid, so that copyright symbol can be translated. + All uses changed. + * src/diff.c, src/diff3.c, src/sdiff.c: Likewise. + + * src/diff.c: Include posixver.h. + (TABSIZE_OPTION): New constant. + (main): Allow widths up to SIZE_MAX. + (filetype): Move to lib/file-type.c and rename to file_type. + All uses changed. + + * src/diff.c (longopts, main, usage): New option --tabsize=COLUMNS. + * src/io.c (find_and_hash_each_line): Likewise. + + * src/diff.h (TAB_WIDTH): Remove. + (tabsize): New decl. + (sdiff_half_width, sdiff_column2_offset): Now size_t rather than + unsigned int. + + * src/diff3.c (skipwhite, readnum): New functions. + (process_diff_control): Use them. + (SKIPWHITE, READNUM): Remove. + (read_diff): Don't worry about errno == ENOEXEC. + + * src/sdiff.c (catchsig, signal_handler, initial_action): Signal + handlers return void, not RETSIGTYPE, since we no longer support + K&R. + (TABSIZE_OPTION): New constant. + (longopts, usage, main): New option --tabsize=COLUMNS. + (cleanup): New arg signo. All uses changed. + (ck_editor_status, main, edit): Don't worry about ENOEXEC. + + * src/side.c (tab_from_to, print_half_line, print_1sdiff_line): + New option --tabsize=COLUMNS. + + * src/system.h (S_ISBLK, S_ISCHR, S_ISDIR, S_ISFIFO, S_ISREG, + S_ISSOCK): Remove; now in lib/file-type.h. + + * src/util.c (finish_output): Check for ENOEXEC. + (lines_differ, output_1_line): New option --tabsize=COLUMNS. + (analyze_hunk): If -b or -w is also specified, -B now considers + lines to be empty if they contain only white space. + +2002-04-05 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.8.1. + + * configure.ac (AC_HEADER_STDBOOL): Add. + (AC_CHECK_HEADERS): Remove stdbool.h. + * m4/stdbool.m4: New file. + * m4/prereq.m4 (jm_PREREQ_EXCLUDE): + Use AC_HEADER_STDBOOL rather than AC_CHECK_HEADERS(stdbool.h). + (jm_PREREQ_HASH): Likewise. + + * src/system.h (SSIZE_MAX): Define if limits.h doesn't. + + * src/analyze.c (diff_2_files): Assign PTRDIFF_MAX - 1 to a + size_t variable, just in case there's a problem with ptrdiff_t + versus size_t. + + * lib/cmpbuf.c (errno): Remove decl; K&R C is no longer supported. + Include limits.h. + (SIZE_MAX, SSIZE_MAX): Define if standard headers don't. + (MIN): New macro. + (block_read): Do not attempt to read more than SSIZE_MAX bytes, as the + resulting behavior is implementation-defined. Work around bug in + Tru64 5.1, which can't read more than INT_MAX bytes at a time. + * src/cmp.c (cmp): Use block_read instead of read, to work + around Tru64 5.1 bug. + * src/diff3.c (read_diff): Likewise. + * src/diff3.c: Include cmpbuf.h. + + * THANKS: Add Ulrich Drepper. + + * INSTALLME: Mention GNU texinfo. + + * doc/diff.texi: + Use new @copying directive. + Put @contents first, not last, since Texinfo now suggests this. + Fix bug in -w documentation noted by Karl Berry. + Mention links for speedup. + New node "Speedups" for future speedups. + Just say "Index", not "Concept Index". + +2002-03-26 Paul Eggert + + * src/Makefile.am: + (INCLUDES): Remove this obsolete macro, replacing it with: + (AM_CPPFLAGS): New macro. + +2002-03-26 Albert Chin-A-Young + + * src/Makefile.am (datadir): Remove, as it conflicts with --datadir. + +2002-03-26 Paul Eggert + + * doc/diff.texi (dircategory GNU packages): Fix typo: a "* " was + missing before the menu entry. Bug diagnosed by Adam Heath. + Also, put this dircategory after the Individual utilities dircategory, + to work around a compatibility problem with Debian install-info. + +2002-03-24 Eli Zaretskii + + * src/io.c (sip): Do not mishandle buffered count when reverting + to text mode. + +2002-03-23 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.8. + * configure.ac (AC_PREREQ): 2.53. + * INSTALLME: Upgrade to gettext 0.11.1 and help2man 1.27. + + * doc/diff.texi: Upgrade the description of `patch' to GNU patch + 2.5.4, and revamp the documentation accordingly. + + * src/diff.c (main): Fix typo that prevented diff -y from working. + Bug reported by Mitsuru Chinen. + +2002-03-15 Paul Eggert + + * lib/c-stack.c (c_stack_die) [!HAVE_SIGINFO_T]: Don't use info. + Bug reported by Eli Zaretskii. + +2002-03-15 Eli Zaretskii + + * ms/config.sed: Tweak editing of install-info-am target. + +2002-03-12 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.7.10. + + * NEWS: cmp -l -s and cmp -s -l are not allowed. + Deprecate diff -h, -H, -L, -P, --inhibit-hunk-merge. + + * configure.ac (jm_PREREQ_HARD_LOCALE): Add. + (AM_INIT_AUTOMAKE): Do not distribute shar file. + + * doc/diff.texi (Overview): byte != character. + (Detailed Context, Detailed Unified, Alternate Names, diff Options): + Do not document diff -L. + (Comparing Directories, Making Patches, diff Options): + Do not document diff -P. + (diff Performance, sdiff Option Summary, diff Options, sdiff Options): + Do not document diff -H. + (diff Performance, diff Options): Do not document --horizon-lines. + (cmp Options): Prefer -b to -c. + (cmp Options, diff Options, diff3 Options, patch Options, + sdiff Options): Put short options next to the similar long options. + Document --help, and use the same wording for --verbose. + (diff3 Options): Fix typo in description of -E, which used wrongly used + "-e" instead of "-E". + + * lib/hard-locale.c (alloca): Remove. + Include stdlib.h if available, for malloc. + (hard_locale): Use malloc, not alloca, so that we need not worry about + alloca issues. Test for storage allocation failure. + + * m4/prereq.m4 (jm_PREREQ): Add jm_PREREQ_HARD_LOCALE. + (jm_PREREQ_HARD_LOCALE): New macro. + + * src/cmp.c (specify_comparison_type): New function. + (check_stdout): "indices and codes" -> "byte numbers and values" + (main): Detect clashing options. + (cmp): Use "byte" rather than "char" if a translation for "byte" + is available, even when in the POSIX locale. + + * src/diff.c (option_help_msgid): Do not document -L, -P, + --horizon-lines, --inhibit-hunk-merge, -H. + * src/diff.h: -L -> --label + +2002-03-11 Paul Eggert + + * NEWS, configure.ac (AC_INIT): Version 2.7.9. + + * INSTALLME: Update to autoconf 2.53, automake 1.6, help2man + 1.25 with patch. + + * configure.ac (AC_INIT): + Change package name from diff to diffutils. + (AM_INIT_AUTOMAKE): Use new form, with option gnits, + rather than old from that duplicated AC_INIT. + (AM_MISSING_PROG): Add help2man. + (REGEX_MALLOC): Define. + (AC_CONFIG_FILES): Add man/Makefile. + + * Makefile.am (AUTOMAKE_OPTIONS): Remove. + * doc/Makefile.am (AUTOMAKE_OPTIONS): Remove. + * lib/Makefile.am (AUTOMAKE_OPTIONS): Likewise. + * ms/Makefile.am (AUTOMAKE_OPTIONS): Likewise. + * src/Makefile.am (AUTOMAKE_OPTIONS): Likewise. + + * lib/c-stack.c: Include + (ENOTSUP): Define if errno.h doesn't. + (SA_NODEFER, SA_ONSTACK, SA_RESETHAND, SA_SIGINFO, SIGSTKSZ, + _SC_PAGESIZE, ALTERNATE_STACK_SIZE, stack_t, sigaltstack): + Remove; we now assume them all when + HAVE_XSI_STACK_OVERFLOW_HEURISTIC, so we don't need + substitutes. + (): Include only if HAVE_XSI_STACK_OVERFLOW_HEURISTIC. + (alternate_signal_stack): Now of size SIGSTKSZ. + (segv_handler): Simplify, under the assumption that + HAVE_XSI_STACK_OVERFLOW_HEURISTIC is nonzero. + (c_stack_action): Likewise. + (exit_failure) [DEBUG]: Initialize to 0, not 1. + (recurse, main) [DEBUG]: Remove main args. + + * m4/c-stack.m4 (AC_SYS_XSI_STACK_OVERFLOW_HEURISTIC): Change + wording of message. Do not check for stdbool.h or ucontext.h, + or for ucontext_t or sigaction or sigaltstack. + + * po/LINGUAS: Add zh_TW. + + * Makefile.am (SUBDIRS): Add man. + * man/Makefile.am: New file. + * src/cmp.c (usage): Reword for help2man. + * src/diff.c (option_help_msgid): Likewise. + * src/diff3.c (option_help_msgid, usage): Likewise. + * src/sdiff3.c (option_help_msgid, usage): Likewise. + Reword for help2man. + + * THANKS: Add email address for Tower. + + * config/config.guess, config/config.sub, config/depcomp, + config/install-sh, config/mdate-sh, config/missing, + config/mkinstalldirs, config/texinfo.tex: Update + to recent version (maintained in other packages). + +2002-03-04 Bruno Haible + + * m4/gettext.m4 (AM_GNU_GETTEXT): Set LIBINTL and LTLIBINTL to empty if + no preinstalled GNU gettext was found. + +2002-03-02 Eli Zaretskii + + * ms/config.sed: Tweak editing of install-info-am and + uninstall-info-am targets, to include 8+3-butchered names of Info + files. + +2002-02-28 Paul Eggert + + * NEWS, configure.ac (AC_INIT, AM_INIT_AUTOMAKE): Version 2.7.8. + + * doc/diff.texi: Add vr index. + Update copyright to 2002. + (Standards conformance): New chapter. + (Binary): Differing binary files are trouble unless the user asked for + brief output. + (Detailed Context): Prefer ISO time stamp format in discussion. + (Detailed Unified, Pagination): Likewise. + (Less Context): Likewise. Also use short option. + (Alternate Names): Separate option from arg. + (Making Patches): Mention -U 2. + (diff Options): Deprecate -LINES, as POSIX 1003.1-2001 does not + allow it. + + * INSTALLME: Update advice for Solaris installation problems. + We no longer use a test version of gettext. + Autoconf test version updated from 2.52f to 2.52h. + POSIX 1003.1-2001 patch for Automake. + + * configure.ac (AC__GNU_SOURCE): Add this, + replacing AH_VERBATIM of _GNU_SOURCE. + (tempname): Use AC_LIBOBJS, not LIBOBJS=, as now required by autoconf. + (jm_PREREQ_C_STACK): Add. + (AC_CONFIG_FILES): Remove intl/Makefile. + (AM_GNU_GETTEXT): Add external arg, from gettext 0.11. + + * lib/c-stack.c, lib/c-stack.h, lib/exitfail.c, lib/exitfail.h, + lib/posixver.c, lib/posixver.h, m4/c-stack.m4, m4/gnu-source.m4, + po/cs.po, po/ja.po: New files. + + * intl/ChangeLog, intl/Makefile.in, intl/VERSION, + intl/bindtextdom.c, intl/config.charset, intl/dcgettext.c, + intl/dcigettext.c, intl/dcngettext.c, intl/dgettext.c, + intl/dngettext.c, intl/explodename.c, intl/finddomain.c, + intl/gettext.c, intl/gettextP.h, intl/gmo.h, intl/hash-string.h, + intl/intl-compat.c, intl/l10nflist.c, intl/libgnuintl.h, + intl/loadinfo.h, intl/loadmsgcat.c, intl/localcharset.c, + intl/locale.alias, intl/localealias.c, intl/localename.c, + intl/ngettext.c, intl/os2compat.c, intl/os2compat.h, intl/osdep.c, + intl/plural-eval.c, intl/plural-exp.c, intl/plural-exp.h, + intl/plural.c, intl/plural.y, intl/ref-add.sin, intl/ref-del.sin, + intl/textdomain.c, m4/isc-posix.m4, m4/libtool.m4: Remove. + + * ABOUT-NLS: Update to Gettext 0.11. + + * Makefile.am (SUBDIRS): Remove intl. + + * config/config.guess, config/config.rpath, config/config.sub, + config/texinfo.tex, config/depcomp, config/texinfo.tex, + lib/tempname.c: Update to latest version from other packages. + + * lib/xalloc.h (xalloc_exit_failure): Remove; subsumed by exit_failure. + * lib/xmalloc.c: Include exitfail.h. + (xalloc_exit_failure): Remove; subsumed by exit_failure. + All uses changed. + + * lib/Makefile.am (noinst_HEADERS): Add c-stack.h, exitfail.h. + (libdiffutils_a_SOURCES): Add c-stack.c, exitfail.c, quotesys.c. + (INCLUDES): Remove. + + * lib/cmpbuf.h (buffer_lcm): New arg LCM_MAX. + * lib/cmpbuf.c: Include errno.h. + (errno): Declare if !STDC_HEADERS. + Include signal.h. + (SA_RESTART): Define if not already defined. + Include . + (PTRDIFF_MAX): Define if not already defined. + (TYPE_SIGNED, TYPE_MINIMUM, TYPE_MAXIMUM): Likewise. + (block_read): Accommodate ancient AIX hosts that set errno to EINTR + after uncaught SIGCONT. + (buffer_lcm): Return a reasonable size if the multiple is too large. + New arg LCM_MAX. All callers changed. + + * lib/hard-locale.c: Include "hard-locale.h". + (hard_locale): Ignore ENABLE_NLS, since we want to operate on + locales other than LC_MESSAGES. + + * m4/prereq.m4 (jm_PREREQ): Add jm_PREREQ_POSIXVER. + (jm_PREREQ_POSIXVER): New macro. + + * m4/setmode.m4 (AC_FUNC_SETMODE_DOS): + Check for fcntl.h and unistd.h unconditionally. + Suggested by Bruno Haible. + + * po/LINGUAS: Add cs, ja. + * po/POTFILES.in: Add lib/c-stack.c, src/dir.c. + + * src/Makefile.am (datadir): @DATADIRNAME@ -> share. + (INCLUDES): Remove intl. + (LDADD): Change INTLLIBS to LIBINTL. + No longer need to link libdiffutils.a twice. + + * src/analyze.c (diff_2_files): + Avoid arithmetic overflow in buffer size calculation. + + * src/cmp.c: Include c-stack.h, exitfail.h. + (hard_locale_LC_MESSAGES): Depend on ENABLE_NLS. + (try_help, check_stdout, main, cmp): 2 -> EXIT_TROUBLE. + (main): Check for stack overflow. + 0 -> EXIT_SUCCESS. + 1 -> EXIT_FAILURE. + (cmp): Likewise. + Accommodate ancient AIX hosts that set errno to + EINTR after uncaught SIGCONT. + + * src/context.c (pr_context_hunk): + Do not dump core if an enormous context causes an + arithmetic overflow. + (pr_unidiff_hunk): Likewise. + (find_hunk): Likewise. + + * src/diff.h: unsigned -> unsigned int. + * src/diff.c: Include c-stack.h, exitfail.h. + Do not include signal.h. + (specify_style, specify_value): Bring these routines back, as POSIX + requires that the order of options not matter. + (shortopts): New constant. + (group_format_option, line_format_option): New constants. + (main): 0 -> EXIT_SUCCESS, 1 -> EXIT_FAILURE, 2 -> EXIT_TROUBLE. + Ensure that order of options does not matter. + Check for stack overflow. + If contexts overflow, substitute LIN_MAX, as that's good enough. + If multiple contexts are specified, use their maximum. + -c is equivalent to -C 3 now, instead of having an implicit context; + likewise for -u and -U 3. + Use specify_style and specify_value. + (SIGCHLD): Do not define; now done in a header. + Use new style time stamp format for -u / -U. + Reject numeric-string options if operating in POSIX 1003.1-2001 mode. + Avoid overflow problems with tab width. + Simplify from-file and to-file code. + (usage): Do not mention obsolete options. + (filetype): Do not mention whether a file is executable. + Add typed memory objects. + (compare_files): 0 -> EXIT_SUCCESS, 1 -> EXIT_FAILURE, 2 -> + EXIT_TROUBLE. + + * src/diff3.c: Include c-stack.h, exitfail.h. + (ALLOCATE): Remove. All uses changed to xmalloc, or to xmalloc plus + an overflow check. + (myread): Remove. + (main): Check for stack overflow. + 0 -> EXIT_SUCCESS, 1 -> EXIT_FAIULRE, 2 -> EXIT_TROUBLE. + (try_help): Likewise. + (process_diff): Check for integer overflow, to avoid core dumps. + 2 -> EXIT_TROUBLE. + (read_diff): Exit with status 126 if the file is not executable, + for compatibility with POSIX 1003.1-2001. + Accommodate ancient AIX hosts that set errno to EINTR after uncaught + SIGCONT. + Check for integer overflow to avoid core dumps. + (fatal, perror_with_exit): 2 -> EXIT_TROUBLE. + + * src/dir.c (dir_read): + Ignore st_size of directories: POSIX says it's garbage. + Check for integer overflow to avoid core dumps. + (diff_dirs): 0 -> EXIT_SUCCESS, 2 -> EXIT_TROUBLE. + + * src/ifdef.c: Include . + (format_group, print_ifdef_lines): Avoid core dumps with bad formats. + (do_printf_spec): Avoid alloca. + + * src/io.c (sip): + Avoid integer overflow and core dumps if buffer alignments are + preposterously incompatible. + (slurp): Do not dump core if the file is growing as we read it. + If a regular file grows, keep reading until we catch up with its EOF. + (find_and_hash_each_line): Check for integer overflow to avoid cores. + (GUESS_LINES): Remove. + (guess_lines): New function. Avoid integer overflow. + (find_identical_ends): Use it. + Avoid integer overflow and possible core dumps. + + * src/sdiff.c: Include c-stack.h, exitfail.h. Do not include signal.h. + 0 -> EXIT_SUCCESS, 1 -> EXIT_FAILURE, 2 -> EXIT_TROUBLE. + (ck_editor_status): New function. + (main): Check for stack overflow. + Adopt POSIX convention for subsidiary programs not found. + (diffarg): Check for integer overflow to avoid core dumps. + (trapsigs): Remove SA_INTERRUPT special case; now done by header. + (SIGCHLD): Likewise. + (edit): Adopt POSIX convention for subsidiary programs not found. + + * src/side.c: unsigned -> unsigned int. + + * src/system.h: Don't use alloca or include . + unsigned -> unsigned int + (EXIT_SUCCESS, EXIT_FAILURE, EXIT_TROUBLE): Define if not defined. + Include signal.h. + (SA_RESTART): Define if not defined. + (SIGCHLD): Likewise. + + * src/util.c: 2 -> EXIT_TROUBLE. + Adopt POSIX convention for ENOEXEC and exit status 126. + unsigned -> unsigned int + +2002-01-24 Paul Eggert + + * NEWS, configure.ac (AC_INIT, AM_INIT_AUTOMAKE): Version 2.7.7. + + * intl/plural.c: Regenerate with Bison 1.31. + + * ABOUT-NLS, intl/*: Update to Gettext 0.11-pre5++. + * INSTALL: Update to Autoconf 2.52f. + + * INSTALLME: New file. + * Makefile.am (EXTRA_DIST): Add config/config.rpath, INSTALLME. + (DISTCLEANFILES): Remove. + * NEWS: Reformat for imminent 2.8 release. + * README: Mention INSTALLME. + * README-alpha: Move most of contents to INSTALLME. + * THANKS: Add Bruno Haible, Jim Meyering, and Eli Zaretskii. + + * config: New subdirectory, containing the following files from .: + config.guess, config.sub, depcomp, missing, install-sh, mkinstalldirs. + Move the following files here from doc: texinfo.tex, mdate-sh. + * config/config.guess, config/config.sub, config/texinfo.tex: + Update to latest version from FSF. + * config/config.rpath: New file, from Gettext 0.11-pre5++. + + * configure.ac (AC_INIT): Use new 3-arg form. + (AC_CONFIG_SRCDIR): Specify src/diff.c here, not in AC_INIT. + (ALL_LINGUAS): Remove: now in po/LINGUAS as per Gettext 0.11. + (AC_CONFIG_AUX_DIR): New macro invocation. + + * lib/Makefile.am (noinst_HEADERS): Add gettext.h. + * lib/gettext.h: New file, from Gettext 0.11-pre5++. + * lib/prepargs.c: Include . Reported by Bruno Haible. + + * m4/codeset.m4, m4/gettext.m4, glibc21.m4, iconv.m4, isc-posix.m4, + lcmessage.m4, progtest.m4: Upgrade to Gettext 0.11-pre5++. + * m4/lib-ld.m4, m4/lib-link.m4, m4/lib-prefix.m4: New files, from + Gettext 0.11-pre5++. + + * po/LINGUAS: New file. + * po/Makefile.in.in: Upgrade to Gettext 0.11-pre5++. + * po/Makevars, po/Rules-quot, po/boldquot.sed: New files, + from Gettext 0.11-pre5++. + + * src/cmp.c (copyright_string): Update to 2002. + * src/diff.c (copyright_string): Likewise. + * src/diff3.c (copyright_string): Likewise. + * src/sdiff.c (copyright_string): Likewise. + + * src/cmp.c (specify_ignore_initial): Renamed from + parse_ignore_initial, with different signature, to take the + maximum of multiple options rather than the last one. + All uses changed. + + * src/cmp.c (bytes, specify_ignore_initial, cmp): Use UINTMAX_MAX + instead of (uintmax_t) -1, to avoid warnings on some compilers. + * src/io.c (file_block_read): Likewise, for SIZE_MAX. + + * src/cmp.c (usage): Reformat messages to ease translation. + * src/diff3.c (usage): Likewise. + * src/sdiff.c (usage): Likewise. + + * src/cmp.c (main): Two files with the same name are identical + only if the same offset is specified. + (block_compare_and_count): Avoid cast to unsigned char. + + * src/diff3.c (main): Remove unused variable. + + * src/dir.c: Include + (struct dirdata): New member nnames. + (locale_specific_sorting, failed_strcoll): New vars. + (dir_read): Renamed from dir_sort. Don't sort the dir. + Set new nnames member of struct dirdata. All callers changed. + (compare_names): Don't check for errno after strcasecmp. + Use strcoll only if locale_specific_sorting is nonzero. + If strcoll fails, longjmp out rather than returning a value + that might result in an invalid comparison function that might + make qsort dump core. + (diff_dirs): Sort the directory ourselves. Use setjmp to recover + from strcoll failure, falling back on native byte comparison. + Make local variables volatile if they need to preserve their value + after setjmp/longjmp. + + * src/sdiff.c (handler_index_of_SIGINT, handler_index_of_SIGPIPE): + New macros. + (main): Do not confuse signal numbers with their indices. + Bug reported by Bruno Haible. + (edit): Cat lin to long before printing with %ld, since lin might + be narrow than long. + + * src/system.h (UINTMAX_MAX): New macro. + Include gettext.h, not libgettext.h. + (N_): Do not wrap arg in parentheses. Fix from Bruno Haible. + + * src/util.c (finish_output): Ensure that werrno is initialized. + (lines_differ): Have an explicit do-nothing case for + IGNORE_NO_WHITE_SPACE, to pacify gcc -Wall. + +2001-12-29 Eli Zaretskii + + * src/sdiff.c (interact): After extracting rlen from the editor + command, test for a terminating null character, not for a newline. + + * ms/config.bat: Allow longer source directory names without + overflowing the line length limits. Create the cache in the + build directory, not in the source directory + * ms/config.sed: Fix AC_CONFIG_LINKS for when symlinks are + unavailable. + +2001-12-23 Paul Eggert + + * NEWS, configure.ac (AM_INIT_AUTOMAKE): Version 2.7.6. + + * configure.ac (ALL_LINGUAS): Add tr. + + * src/util.c (begin_output): + Have child exit with status 127 rather than reporting + failure on its own. Set errno to 0 before invoking popen. + (finish_output): Report errno on pclose failure. + Distinguish between subsidiary program not found, and failure. + + * src/sdiff.c (not_found, execdiff): Remove. + (DIFF_PROGRAM_OPTION): New constant. + (longopts, option_help_msgid, main): Add --diff-program=PROGRAM. + (check_stdout): New function. + (main): Remove DIFF_PROGRAM. Check stdout after printing version. + Use check_stdout after printing help. Use execvp/perror_fatail rather + than execdiff. Set errno to 0 before invoking popen. + Check for pclose failure properly. + (main, edit): If child exec fails, exit with 127 rather than trying to + print diagnostic. + Distinguish between subsidiary program failing and not being found. + (edit): Handle signals the same way, regardless of whether we're using + system or fork+exec. Check for system returning -1. + + * src/diff3.c (DIFF_PROGRAM_OPTION, HELP_OPTION): New constants. + (longopts, main): Use them. + (longopts, main, option_help_msgid): New option --diff-option=PROGRAM. + (main): Remove DIFF_PROGRAM support. + Check stdout after printing version. + (check_stdout): Report errno info if fclose fails. + (read_diff): Have child exit with status 127 when program is not found, + rather than trying to have the child report failure. Check for + pclose returning -1. + + * src/diff.c (DEFAULT_WIDTH): Remove. + (main): Use 130 instead of DEFAULT_WIDTH, since it's not really + builder-settable. Do not prepend DIFF_OPTIONS. + (check-stdout): If fclose (stdout) fails, print errno info. + (option_help_msgid): Default context is 3, not 2. + (usage): Work even if ptrdiff_t is wider than int. + + * doc/diff.texi (diff Options): Remove DIFF_OPTIONS. + (Invoking diff3, Invoking sdiff): Remove DIFF_PROGRAM. + (diff3 Options, sdiff Options): Add --diff-program. + + * src/cmp.c (valid_suffixes): + Add '0', to support suffixes like "MB" and "MiB". + (check_stdout): Don't assume that the translations of "write failed" + and of "standard output" lack '%'. + (main): Check stdout after printing version. + + * lib/setmode.c: [HAVE_FCNTL_H && HAVE_SETMODE_DOS]: Include . + [!HAVE_SETMODE_DOS]: Do not include . + (set_binary_mode): Return mode (not 1) if fd is a tty. + Do not assume that O_TEXT is zero. + + * doc/diff.texi (cmp Options): + In byte counts, a plain suffix (without any integer) + is assumed to modify the integer 1. Index terms like "kibibyte". + Document plain "k". + + (Reporting Bugs): Mention bug-report archive and test version + location. Ask for "diff --version" in bug reports. + +2001-12-13 Paul Eggert + + * src/diff.c (DEFAULT_WIDTH): Remove; couldn't be changed without + also changing option_help_msgid. All uses replaced with 130. + + * lib/setmode.c: Include fcntl.h and unistd.h only if + HAVE_SETMODE_DOS. + (setmode): Assume a file is binary unless the mode is O_TEXT. + * ms/README: Fix minor typos. + +2001-12-13 Eli Zaretskii + + * ms/README: New file. + + * lib/setmode.c (set_binary_mode) [HAVE_SETMODE_DOS]: Don't assume + O_TEXT has a zero value. If FD is a terminal device, do nothing + and return MODE, thus pretending that it was already in the + requested MODE. + [HAVE_FCNTL_H]: Include fcntl.h (needed for O_BINARY). + + * ms/config.sed: Remove the split prevention of config.status. + Fix Sed commands for converting absolute file names into + top_srcdir-relative ones. + + * ms/config.bat: Fix typos. + +2001-12-12 Neal H Walfield + + * diff.c (option_help_msgid): Correct the default context width + from 2 to 3. + +2001-12-11 Paul Eggert + + * m4/Makefile.am.in: Remove jm-glibc-io.m4 + + * NEWS, configure.ac (AM_INIT_AUTOMAKE): Version 2.7.5. + + * configure.ac (PR_PROGRAM): Use AC_DEFINE_UNQUOTED, so that + $PR_PROGRAM is expanded by sh. + (ptrdiff_t, ssize_t): Use AC_CHECK_TYPE with a default of int, + not AC_CHECK_TYPES. + (jm_AC_DOS, AC_FUNC_SETMODE_DOS): New macros. + (AC_CONFIG_FILES): Add ms/Makefile. + + * doc/diff.texi: Add --no-ignore-file-name-case. + File name case sensitivity now affects file name exclusion. + Fix typos. + + * src/util.c: Include dirname.h. + (dir_file_pathname): Use base_name rather than file_name_lastdirchar. + + * src/system.h (S_IXUSR, S_IXGRP, S_IXOTH): New macros. + Include rather than rolling it ourselves. + (file_name_lastdirchar, HAVE_SETMODE, set_binary_mode): Remove. + + * src/sdiff.c: Include . + (expand_name): Use base_name rather than file_name_lastdirchar, for + portability to DOS. + (main): Initialize xalloc_exit_failure before possibly invoking + any memory allocator. + + * src/io.c: Include setmode.h. + + * src/diff3.c (main): + Initialize xalloc_exit_failure before possibly invoking any memory + allocator. + + * src/diff.c: Include dirname.h, setmode.h. + + (main): Later values and/or styles now silently override earlier. + (specify_value, specify_style): Likewise. All callers changed. + Remove. + (binary, main, option_help_msgid, compare_files): + HAVE_SETMODE -> HAVE_SETMODE_DOS. + (NO_IGNORE_FILE_NAME_CASE_OPTION): New constant. + (longopts, main, option_help_msgid): Support it. + (exclude_options): New function. + (main): Use it. Initialize xalloc_exit_failure before potentially + allocating memory. + + (filetype): Distinguish executable files from others, as POSIX + suggests. + + (compare_files): Use base_name instead of file_name_lastdirchar. + + * src/cmp.c: Include , . + (hard_locale_LC_MESSAGES): New macro. + (sprintc): Remove int width arg; it's now the caller's responsibility + to pad. All callers changed. + (stat_buf): New static var; was formerly a local var in 'main'. + (valid_suffixes): Add 'K', for 'KiB'. + (option_help_msgid): Don't confuse bytes with characters. + (main): Set xalloc_exit_failure before invoking anything that might + allocate memory. Fix bug: -n was incorrectly ignored when optimizing + the case of regular files with different lengths. + (cmp): Use an index column wide enough to store this comparison's + indexes. In locales other than the POSIX locale, say "byte" + rather than "char". + + * ms/config.bat: pc -> ms + + * ms/Makefile.am, m4/setmode.m4, lib/setmode.c, lib/setmode.h: + New file. + + * lib/Makefile.am (noinst_HEADERS): Add dirname.h, setmode.h. + (libdiffutils_a_SOURCES): Add basename.c, setmode.c. + + * Makefile.am (SUBDIRS): Add ms. + +2001-12-10 Paul Eggert + + * m4/fnmatch.m4: Test for FNM_CASEFOLD. + +2001-12-03 Paul Eggert + + * lib/posix/regex.h: Fix copyright notice. + +2001-12-03 Paul Eggert + + * NEWS, configure.ac (AM_INIT_AUTOMAKE): Version 2.7.4. + + * diff.texi (direntry, Overview, Comparison, Binary, Invoking cmp): + Use "byte" rather than "character" when talking about cmp, since + it compares bytes, not character. + (Invoking cmp): New trailing operands FROM-SKIP and TO-SKIP. + -i or --ignore-initial now accepts FROM-SKIP:TO-SKIP. + New option -n or --bytes. + Count operands now may be in octal or hex, and may be followed by a + size multiplier. + + * configure.ac (DEFAULT_DIFF_PROGRAM): + Define to "diff", not "$bindir/diff" (which didn't work anyway). + (AC_CHECK_MEMBERS): Add struct stat.st_blksize, struct stat.st_rdev. + (AC_STRUCT_ST_BLKSIZE, AC_STRUCT_ST_RDEV): Remove; obsolescent. + (AC_FUNC_FORK): Use this, instead of obsolescent AC_FUNC_VFORK. + (AC_CONFIG_FILES, AC_CONFIG_COMMANDS): Add. + (AC_OUTPUT): Remove args; they were obsolescent. + + * util.c (setup_output, begin_output, finish_output): + HAVE_FORK -> HAVE_WORKING_FORK || HAVE_WORKING_VFORK. + * sdiff.c (diffpid, cleanup, main, edit): Likewise. + * diff3.c (read_diff): Likewise. + + * system.h (STAT_BLOCKSIZE): + Use HAVE_STRUCT_STAT_ST_BLKSIZE, not HAVE_ST_BLKSIZE. + (vfork): New macro. + (HAVE_FORK): Remove. + (set_binary_mode): New macro. + + * sdiff.c (main): HAVE_VFORK -> HAVE_WORKING_VFORK. + (edit): Reopen the temporary file after the editor has run, in case + the editor operates by unlinking the old file and linking a new one. + (P_tmpdir): Rename from PVT_tmpdir; this fixes a typo. + All uses changed. + + * io.c (sip, read_files): + Remove tests for HAVE_SETMODE; use set_binary_mode + instead of setmode. + (sip): Fix typo in backward lseek when reverting to text mode. + + * config.site, config.sed, config.bat: New file. + + * Makefile.am (EXTRA_DIST): Add xstrtol.c. + (noinst_HEADERS): Add xstrtol.h. + (libdiffutils_a_SOURCES): Add xstrtoumax.c. + + * cmp.c: : Include. + (ignore_initial): Now an array with 2 elements. All uses changed. + (bytes): New var. + (HELP_OPTION): New constant. + (long_options, main): Use it. + (long_options, option_help_msgid, main, cmp): + Add support for -n or --bytes. + (parse_ignore_initial): New function. + (option_help_msgid, main): Add -i M:N. + (usage, main): Add two optional trailing operands, a la BSD. + (main): setmode -> set_binary_mode. + (cmp): Report byte number of what we've seen, not of the entire file. + This is to be consistent with the line number, which is always relative + with what we've seen. + +2001-12-02 Paul Eggert + + * diff.c (main, compare_files): setmode -> set_binary_mode. + + * xstrtol.c (__xstrtol): Don't accept 'Ki'; require 'KiB'. + + * xstrtol.c (__xstrtol): Add support for IEC 60027-2. + +2001-11-25 Paul Eggert + + * NEWS, configure.ac (AM_INIT_AUTOMAKE): Version 2.7.3. + + * README-alpha: New file. + + * src/Makefile.am (INCLUDES): Add -I../lib, for regex.h. + + * configure.ac: + Don't set LIB_CLOCK_GETTIME to 'none required'; set it to + the empty string instead. + + * lib/Makefile.am (EXTRA_DIST): Add strtoimax.c, strtol.c. + + * Makefile.am (SUBDIRS): Put intl before lib, so that libintl.h exists. + + * lib/Makefile.am (noinst_HEADERS): Add unlocked-io.h. + + * configure.ac (__EXTENSIONS__): New define, for the unlocked macros. + + * README: Add copyright notice. + Remove stuff that doesn't apply any more. + + * doc/diff.texi: offsets -> indices for cmp + + * src/cmp.c (option_help_msgid): offsets -> indices + + * src/diff.c (option_help_msgid): + Don't mention --binary on POSIX hosts. + + * src/sdiff.c (STRIP_TRAILING_CR_OPTION): New constant. + (longopts, option_help_msgid, main): Add -E, --ignore-tab-expansion, + --strip-trailing-cr. + + * doc/diff.texi: Change direcategory from Utilities to GNU Packages. + Add individual utilities. + Switch to Free Documentation License. + @code -> @command + @samp -> @option + GNU -> @sc{gnu} + Expand tabs to spaces, except when in an example that actually + uses tabs. + Prefer @node with just one arg. + Document -E or --ignore-tab-expansion, --strip-trailing-cr, + --ignore-file-name-case. + Regular expressions are now grep style, not Emacs style. + cmp's -c or --print-chars option is now -b or --print-bytes. + Time stamps now depend on LC_TIME. + -p now implies ^[[:alpha:]$_]. + Flags now include ' and 0. + cmp -i is an alias for --ignore-initial + Document --from-file, --to-file. + Document DIFF_OPTIONS. + + * configure.ac (AC_CHECK_FUNCS): Add gettimeofday, clock_gettime. + (LIB_CLOCK_GETTIME): New subst. + + * src/system.h: Assume C89 or better. + (_GNU_SOURCE): Remove; config.h now defines it. + (alloca): Declare like coreutils does it. + (verify, TYPE_SIGNED, TYPE_MINIMUM, TYPE_MAXIMUM, O_RDWR, + S_IRUSR, S_IWUSR): New macros. + (STAT_BLOCKSIZE): Parenthesize definiens. + : Include if HAVE_INTTYPES_H. + (CHAR_MAX, INT_MAX): Remove. + (PTRDIFF_MAX, SIZE_MAX): New macros. + (strtoumax): New decl. + Include stddef.h. + (bzero): Remove. + (bindtextdomain, textdomain, N_): New macros. + (ISPRINT, ISSPACE): Remove ifndef wrappers. + (ISUPPER, ISDIGIT): Remove. + (TOLOWER): New macro. + (MIN): Renamed from min; all callers changed. + (MAX): Likewise, from max. + (lin): New type. + (LIN_MAX): New macro. + (file_name_cmp): Renamed from filename_cmp. All callers changed. + (file_name_lastdirchar): Renamed from file_name_lastdirchar. + All callers changed. + (could_be_mvfs_stat_bug, could_be_nfs_stat_bug, + dev_may_have_duplicate_ino): Remove. + (HAVE_SETMODE, NULL_DEVICE): New macros. + (same_file): Do not check attributes. + (same_file_attributes): New macro. + + * src/util.c: Assume C89 or better. + int -> bool for booleans. + int -> lin for line numbers. + int -> size_t for sizes. + Use angle-brackets when including quotesys.h. + Include error.h, regex.h, xalloc.h. + (message5): sizeof -> offsetof + (begin_output): Invoke pr without -f. + (lines_differ): Renamed from line_cmp, and return bool not 3-way int. + All callers changed. + Add support for IGNORE_TAB_EXPANSION. + (change_letter): Now an array rather than a function. All + callers changed. + (translate_range): Translate line numbers to long, not lin, + for convenience with printf. + (analyze_hunk): Return enum changes instead of a count of + inserts and deletes. All callers changed. + (zalloc): New function. + + * src/side.c: Assume C89 or better. + int -> bool for booleans. + int -> lin for line numbers. + + * src/sdiff.c: Assume C89 or better. + int -> bool for booleans. + int -> lin for line numbers. + Use angle-brackets when including getopt.h, quotesys.h. + Include error.h, freesoft.h, stdio.h, xalloc.h. + (copyright_string): Use only most recent year. + (authorship_msgid, option_help_msgid): Wrap in N_(). + + (tmpname): Now volatile. + (tmpmade): Remove. + (tmp): New var. + (private_tempnam, exists, letters): Remove. + (temporary_file): New function. + (edit): Use it. + (interact): Use strtoumax, not atoi. + + * src/normal.c: Assume C89 or better. + int -> lin for line numbers. + + * src/io.c: Assume C89 or better. + int -> bool for booleans. + int -> lin for line numbers. + int -> size_t for sizes. + Use angle-brackets when including cmpbuf.h. + Include regex.h, xalloc.h. + (word): Remove; now done in system.h. + (hash_value): New type; use it instead of 'unsigned' for hash values. + (file_block_read): New function. + (sip, slurp): Use it. Now static. + (sip): Ensure block size is a multiple of word size. Clear eof flag. + (slurp): Use xalloc_die to report memory exhaustion. + (find_and_hash_each_line): Use TOLOWER instead of _tolower. + Add support for IGNORE_TAB_EXPANSION. + (prepare_text_end): Strip trailing CR if requested. + (find_identical_ends): Prepare the text only once, + if they're duplicates. + Let the compiler take advantage more of the fact that the buffers are + word-aligned. + (primes): Remove. + (prime_offset): New var. + (read_var): Use prime_offset instead of primes. + Use zalloc instead of xmalloc + bzero. + + * src/ifdef.c: Assume C89 or better. + int -> lin for line numbers. + (format_group): Use strtoumax to parse line numbers. + (format_group, print_ifdef_lines): Use do_printf_spec to + handle printf specs. + (groups_letter_value): Don't use _tolower; it's locale-dependent. + (do_printf_spec): Renamed from scan_printf_spec; now does the printing. + + * src/ed.c: Assume C89 or better. + int -> lin for line numbers (or 'long' when that's more convenient). + (print_ed_hunk): Fix bug when handling double-dot inserts. + + * src/dir.c: Assume C89 or better. + int -> bool for booleans. + Include error.h, exclude.h, xalloc.h. + + (dir_sort): Return 0 on error, 1 on success. All callers changed. + compare_names -> compare_names_for_qsort. + + (compare_names): Try strcasecmp if ignore_file_name_case. Then try + strcoll. Use file_name_cmp only as a last resort. Warn about + strcasecmp or strcoll failure. + (compare_names_for_qsort): New function. + + (diff_dirs): Use compare_names rather than filename_cmp. + + * src/diff3.c: Assume C89 or better. + int -> bool for booleans. + int -> lin for line numbers. + Use angle-brackets when including getopt.h, quotesys.h. + Include error.h, freesoft.h, inttostr.h, xalloc.h. + (copyright_string): Use only most recent year. + (authorship_msgid, option_help_msgid): Wrap in N_(). + + Rename the following variables for consistency with user-visible + option spellings. All uses changed. + (text): Renamed from always_text. + (initial_tab): Renamed from tab_align_flag. + + (horizon_lines): Remove. Remove all uses. + + (main): Invoke bindtextdomain and textdomain after setlocale. + Rename "DIFF" to "DIFF_PROGRAM". + + Try to compare file0 to file1, because this is where changes are + expected to come from. Diffing between these pairs of files is more + likely to avoid phantom changes from file0 to file1. + However, use file2 as the common file if this is a 3-way diff, + for backward compatibility. Suggested by Karl Tomlinson. + + (create_diff3_block): Use xcalloc instead of malloc + bzero. + + (INT_STRLEN_BOUND): Remove; now in system.h. + + (read_diff): Always use --horizon-lines=100 rather than trying + to guess it. + Do not pass --inhibit-hunk-merge. + Minimum chunk size is 1, not 8KiB. + Use xalloc_die to report memory exhaustion. + (undotlines): Use long for start, not int. + + * src/diff.h: Assume C89 or better. + int -> bool for booleans. + int -> lin for line numbers. + Don't include regex.h. + (enum changes): New enum. + (enum line_class): Remove; subsumed by enum changes. + (enum output_style): New constant OUTPUT_UNSPECIFIED. + + (ignore_space_change_flag, ignore_all_space_flag): Remove. + (ignore_white_space): New decl, subsuming the above two. All + uses changed. + + Rename the following decls for consistency with user-visible + option spellings. All uses changed. + (text): Renamed from always_text_flag. + (ignore_blank_lines): Renamed from ignore_blank_lines_flag. + (ignore_case): Renamed from ignore_case_flag. + (brief): Renamed from no_details_flag. + (initial_tab): Renamed from tab_align_flag. + (expand_tabs): Renamed from tab_expand_flag. + (starting_file): Renamed from dir_start_file. + (paginate): Renamed from paginate_flag. + (sdiff_merge_assist): Renamed from sdiff_help_sdiff. + (left_column): Renamed from sdiff_left_only. + (suppress_common_lines): Renamed from sdiff_skip_common_lines. + (speed_large_files): Renamed from heuristic. + (minimal): Renamed from no_discards. + + (inhibit_hunk_merge): Remove. + + (strip_trailing_cr, excluded, time_format): New decls. + + (files_can_be_treated_as_binary): Renamed from ignore_some_changes. + + (group_format, line_format): Now char const *[], not char *[]. + + (struct file_data): Buffer is now word*, not char*, as it's always + aligned and this can help the compiler. buffered_chars -> buffered + (since it's a byte count, not a char count). All uses changed. + New member `eof'. + + (FILE_BUFFER): New macro. + + (excluded_filename, error, free_software_msgid): Remove decls; now in + other .h files. + + (sip, slurp): Remove decls. + (file_block_read): New decl. + (change_letter): Now an array, not a function. + (lines_differ): Renamed from line_cmp. + (analyze_hunk): Now returns enum changes rather than two change counts. + + * src/Makefile.am (diff_LDADD): New symbol. + + * src/diff.c: Assume C89 or better. + int -> bool for booleans. + long -> off_t for line numbers. + Use angle-brackets when including getopt.h, fnmatch.h, quotesys.h. + Include error.h, exclude.h, freesoft.h, hard-locale.h, prepargs.h, + regex.h, signal.h, xalloc.h. + (copyright_string): Use only most recent year. + (authorship_msgid, option_help_msgid): Wrap in N_(). + + Rename the following variables for consistency with user-visible + option spellings. All uses changed. + (binary): Renamed from binary_flag. + (new_file): Renamed from entire_new_file_flag. + (unidirectional_new_file): Renamed from unidirectional_new_file_flag. + (report_identical_files): Renamed from print_file_same_flag. + + (numeric_arg): Remove. + + (exclude, exclude_alloc, exclude_count, excluded_filename, add_exclude, + add_exclude_file): + Remove; now done by exclude.h. + + (BINARY_OPTION, FROM_FILE_OPTION, HELP_OPTION, HORIZON_LINES_OPTION, + IGNORE_FILE_NAME_CASE_OPTION, INHIBIT_HUNK_MERGE_OPTION, + LEFT_COLUMN_OPTION, LINE_FORMAT_OPTION, NORMAL_OPTION, + SDIFF_MERGE_ASSIST_OPTION, STRIP_TRAILING_CR_OPTION, + SUPPRESS_COMMON_LINES_OPTION, TO_FILE_OPTION, + UNCHANGED_LINE_FORMAT_OPTION, OLD_LINE_FORMAT_OPTION, + NEW_LINE_FORMAT_OPTION, UNCHANGED_GROUP_FORMAT_OPTION, + OLD_GROUP_FORMAT_OPTION, NEW_GROUP_FORMAT_OPTION, + CHANGED_GROUP_FORMAT_OPTION): New constants. + (longopts, main): Use them. + + (longopts, main, option_help_msgid): Add -E, --from-file, --to-file. + + (main): Invoke bindtextdomain and textdomain after setlocale. + Use grep syntax, not Emacs, for regular expressions. + Use exclude.h, not our own functions. + Use ISO 8601 time format in hard locales. + Prepend DIFF_OPTIONS. + Don't update ignore_some_changes. + Use strtoumax instead of numeric_arg. + Use specify_value when appropriate. + error -> try_help when appropriate. + -p now means ^[[:alpha:]$_], not ^[_a-zA-Z$]. + Ignore --inhibit-hunk-merge. + Prefer changed group formats to unchanged ones. + Remove now-unnecessary casts. + Set files_can_be_treated_as_binary. + + (specify_value): Renamed from specify_format. All uses changed. + + (specify_style): Default is now unspecified, not normal. All + uses changed. + + (set_mtime_to_now): New function. + (compare_files): Use it. Use memset, not bzero. + Set stdin mtime to current time even when stdin is not a regular file. + Check for same file attributes, as well as for same file. + Use files_can_be_treated_as_binary. + "write failed" -> "standard output on output failure. + + * src/context.c: Assume C89 or better. + int -> lin for line numbers. + Include inttostr.h, regex.h. + (TIMESPEC_NS): New macro. + (nstrftime): New decl. + (print_context_label): Use nstrftime and time_format to format times. + Print numeric time stamp value if localtime fails. + (print_context_function): New function. + (pr_context_hunk, pr_unidiff_hunk): Use it. + (find_function): Use size_t for sizes, not int. + + * src/cmp.c: Assume C89 or better. + int -> bool for booleans. + long -> off_t for line numbers. + Use angle-brackets when including cmpbuf.h, getopt.h. + Include error.h, freesoft.h, inttostr.h, xalloc.h. + (copyright_string): Use only most recent year. + (authorship_msgid): Wrap in N_(). + (buffer): Now word*, not char*. All uses changed. + (word): Remove macro; now in system.h. + (long_options, option_help_msgid, main): -c --print-chars -> + -b --print-bytes + (check_stdout): "write failed" -> "standard output" + (option_help_msgid): Wrap in N_(). + (main): Invoke bindtextdomain and textdomain after setlocale. + Use strtoumax instead of doing the work ourselves. + Check for same_file_attributes as well as same_file. + (cmp): Use ssize_t for read returns, not size_t. + Do not assume that size_t is not narrower than int. + Do not assume that line numbers fit in 'long'. + (block_compare_and_count, block_compare): + Compiler now checks that buffers are word-aligned. + (block_compare_and_count): Count sizes with size_t, not long. + (sprintc): byte arg is unsigned char, not unsigned. + + * src/analyze.c: Assume C89 or better. + int -> lin for line numbers. + int -> bool for booleans. + unsigned int -> size_t for sizes. + Use angle-brackets when including cmpbuf.h. + Include error.h, regex.h, xalloc.h. + (discard_confusing_lines, diff_2_files): Use zalloc rather + than xalloc+bzero. + (discard_confusing_lines): unsigned int -> lin for values that + are really line numbers. + (shift_boundaries): Do not inhibit hunk merges. + (build_reverse_script, build_script, diff_2_files): Use |, not ||. + (diff_2_files): no_details_flag & ~ignore_some_changes -> + files_can_be_treated_as_binary. Esure that buffer size is a multiple + of sizeof (word). Use file_block_read to read buffers. + (diff_2_files): Abort if output style is not one of the + expected styles. + +2001-11-23 Paul Eggert + + * src/Makefile.am, m4/vararrays.m4: New file. + + * m4/prereq.m4 (jm_PREREQ_READUTMP): + Remove, as it gives autoheader the willies. + + * m4/README, lib/prepargs.h, lib/prepargs.c, lib/offtostr.c, + lib/umaxtostr.c, lib/inttostr.c, lib/inttostr.h, + lib/imaxtostr.c, lib/freesoft.h: New files. + + * lib/freesoft.c: Include config.h, freesoft.h rather than diff.h. + (free_software_msgid): Wrap contents in N_. + + * lib/cmpbuf.h: Use prototypes instead of old-style functions. + + * lib/cmpbuf.c: + Don't include system.h; instead, include config.h, unistd.h. + Use prototypes instead of old-style functions. + (block_read): Don't assume that int is no wider than size_t. + + * lib/Makefile.am, po/POTFILES.in: New file. + +2001-11-22 Paul Eggert + + * pc/config.h: + Define filename_cmp as an object-like macro, not as a function-like + macro. + + * exgettext: Always operate in the C locale. + Set AWK using a method that works even with broken shells. + + * doc/Makefile.am: New file. + + * configure.ac (AC_INIT): + Use src/diff.c, not diff.h, as the source files got removed. + (AM_CONFIG_HEADER): Switch from AC_CONFIG_HEADER. + (AC_ARG_PROGRAM, AC_MINIX): Remove. + + (AC_PREREQ, AM_INIT_AUTOMAKE, ALL_LINGUAS, AC_PROG_AWK, + AM_PROG_CC_STDC, AC_PROG_RANLIB, AC_C_INLINE, AC_C_VARARRAYS, + DEFAULT_DIFF_PROGRAM, DEFAULT_EDITOR_PROGRAM, + AC_STRUCT_ST_MTIM_NSEC): Add. + + (PR_PROGRAM): AC_DEFINE. + + (AC_SYS_LARGEFILE): Use instead of our homebrew version. + + (_GNU_SOURCE): Define if not defined. + + (AC_CHECK_HEADERS): Add stdbool.h, unistd.h. + (AC_CHECK_TYPES): Add ptrdiff_t, uintmax_t. + (AM_GNU_GETTEXT, XGETTEXT): Add. + + (WITH_MVFS_STAT_BUG, WITH_NFS_STAT_BUG): Remove. + (HAVE_MEMCHR): Remove. + (AC_CHECK_FUNCS): Add diraccess. + (AC_REPLACE_FUNCS): Add memchr, waitpid. + (jm_FUNC_GLIBC_UNLOCKED_IO, jm_FUNC_GNU_STRFTIME, jm_FUNC_MALLOC, + jm_FUNC_REALLOC, jm_PREREQ_ERROR, jm_PREREQ_QUOTEARG, jm_PREREQ_REGEX, + jm_PREREQ_TEMPNAME, jm_AC_PREREQ_XSTRTOUMAX, AC_FUNC_FNMATCH): Add. + (fnmatch.h, regex.h): Do not create these files unless we're using + our own fnmatch and regex. + + (AC_OUTPUT): Add doc/Makefile, intl/Makefile, lib/Makefile, + lib/posix/Makefile, m4/Makefile, po/Makefile.in, src/Makefile. + + * Makefile.am: New file. + + * po/en_GB.po: Don't translate "program" to "programme". + +2001-11-20 Paul Eggert + + * m4/prereq.m4: New file. + +2001-03-16 Paul Eggert + + * lib/tempname.c (uint64_t): + Define if not defined, and if UINT64_MAX is not defined. + +2001-02-26 Paul Eggert + + * lib/tempname.c: glibc 1.32 + +2001-02-17 Paul Eggert + + * m4/Makefile.am.in: GNU fileutils 4.1 + +2001-01-09 Paul Eggert + + * lib/tempname.c (struct_stat64): New macro. + (direxists, __gen_tempname): Use it. This avoids a portability problem + with Solaris 8. + + * lib/tempname.c (): Include if HAVE_CONFIG_H. + (, , ): + Include only if STDC_HEADERS || _LIBC. + (): Include only if HAVE_FCNTL_H || _LIBC. + (): Include only if HAVE_UNISTD_H || _LIBC. + (): Include only if HAVE_SYS_TIME_H || _LIBC. + (__set_errno): Define this macro if doesn't. + (P_tmpdir, TMP_MAX, __GT_FILE, __GT_BIGFILE, __GT_DIR, __GT_NOCREATE): + Define these macros if doesn't. + (S_ISDIR, S_IRUSR, S_IWUSR, S_IXUSR): + Define these macros if + doesn't. Ignore S_ISDIR if STAT_MACROS_BROKEN. + (stat64, __getpid, __gettimeofday, __mkdir, __open, __open64, lxstat64, + __xstat64): Define if not _LIBC. + (__secure_getenv): Define if ! (HAVE___SECURE_GETENV || _LIBC). + (__gen_tempname): Invoke gettimeofday only if HAVE_GETTIMEOFDAY + || _LIBC; otherwise, fall back on plain "time". + Use macros like S_IRUSR | S_IWUSR rather than octal values like 0600. + + * lib/mkstemp.c (__GT_FILE): Define to zero if not defined. + +2000-10-25 Paul Eggert + + * lib/hard-locale.c: New file. + +2000-02-05 Paul Eggert + + * exgettext: From GCC repository + +1999-07-06 Paul Eggert + + * lib/mkstemp.c: glibc 2.2 + +1998-12-11 Paul Eggert + + * src/sdiff.c (lf_snarf): + Fix bug when help line wrapped around the input buffer. + +1998-09-15 Paul Eggert + + * diff.texi: Add @dircategory and @direntry. + +1998-09-14 Paul Eggert + + * Makefile.in (VERSION): Version 2.7.2. + (DEFAULT_DIFF_PROGRAM): Renamed from DIFF_PROGRAM. + (PR_PROGRAM): All `configure' to define it. + (srcs): Add $(diffutils_srcs), freesoft.c, quotearg.c instead of + quote.c, quotearg.h. + (distfiles): Add acconfig.h, message/*. + (all): Depend on $(destfiles), not info. + (version.c): Parenthesize `GNU diffutils'. + (common_o): Add freesoft.o + (diff_o): quote.o -> quotearg.o + (diff3_o, sdiff_o): Likewise. + (diff.dvi): Depend on version.texi. + (diff.o diff3.o quotearg.o sdiff.o util.o): + New dependency on quotearg.h + (diff3.o): DIFF_PROGRAM -> DEFAULT_DIFF_PROGRAM. + (sdiff.o): Likewise. + (messages.po): Remove. + (message/msgid.po, message/template.po): New rules. + (maintainer-clean): Renamed from realclean. + (install): Install from source directory, if applicable. + Invoke install-info if needed. + (install-strip): New rule. + (check): Set DIFF. + (stamp-h.in): Don't put the date into the timestamp. + (D_dirs): Add $D/message. + ($D.tar.gz): Compress with gzip -9. + Don't use ln to create distribution; it doesn't work with symlinks. + (srcs, distfiles, diff_o, diff3_o, sdiff_o): Rename quotearg.c to + quotesys.c and quotearg.h to quotesys.h. + + * configure.in (AC_PATH_PROG): Add PR_PROGRAM. + If available, prefer support for large files unless the user specified + one of the CPPFLAGS, LDFLAGS, or LIBS variables. + (AC_STRUCT_ST_RDEV): Add. + (HAVE_ST_FSTYPE_STRING): Add. + (--with-mvfs-stat-bug, --with-nfs-stat-bug): New options. + (HAVE_MEMCHR): New macro. + (AC_CHECK_FUNCS): Add sicprocmask. + + * diff.h (XTERN): Renamed from EXTERN. + (struct filedata): Remove dir_p arg. + (struct comparison): New type. + (diff_2_files, diff_dirs)" Ise ot/ + (error): Add printf attribute if applicable. + (free_software_msgid): New decl. + (pr_program): New decl. + (fatal): Add noreturn attribute. + (pfatal_with_name): Likewise. + + * system.h (__attribute__): New macro. + (getenv): Don't declare if HAVE_STDLIB_H. + (CHAR_MAX): New macro. + (): New include. + (): Include before . + (could_be_mvfs_stat_bug, could_be_nfs_stat_bug, + dev_may_have_duplicate_ino, same_special_file): New macros. + (same_file): Use them. + + * cmp.c (authorship_msgid): New var. + (free_software_msgid): New decl. + (error): Now has printf attribute. + (try_help): Likewise. + (long_options): Don't assume ASCII. + (try_help): Now accepts operand arg. + (main): Check for -1, not EOF, when calling getopt_long. + Report --ignore-initial value when complaining about it. + Output copyright and free software info with -v. + Don't assume ASCII. + Report last operand when one is missing. + Report text of extra operand. + Move block_read into cmpbuf.c. + + * diff.c (authorship_msgid): New var. + (quotesys.h): Include. + (ck_atoi): Remove. + (function_regexp_list, ignore_regexp_list): Now static. + (binary_flag): Renamed from binary_I_O. + (entire_new_file_flag, unidirectional_new_file_flag, + print_file_same_flag): Now static. + (numeric_arg): Renamed from ck_atoi. + New argument specifying the argument type. + (longopts, main): Don't assume ASCII. + (longopts): Remove old aliases --file-label, --entire-new-file, + --ascii, --print. + (main): Check for -1, not EOF, when calling getopt_long. + Use numeric_arg to report errors. + Report error if -l specified but pagination is not supported. + Report error if -S is specified twice with conflicting values. + Have --version conform to the new GNU standards. + Add new --from-file, --to-file, --inhibit-hun,-merge options. + Make the horizon at least as large as the context. + Add casts to pacify gcc -Wall. + (try_help): Add operand arg. + (option_help_msgid): Doc fix to match above. + (usage): Indent option_help_msgid. + (compare_files): Now takes struct comparison + instead of two directory names and a depth. + (NONEXISTENT, UNOPENED, ERRNO_ENCODE, ERRNO_DECODE): + New macros. + (DIR_P): New macro. + Report error if fflush does. + + * cmpbuf.c (block_read): Moved here from cmp.c. + + * cmpbuf.h (block_read): New decl. + + * io.c (cmpbuf.h): Include. + (slurp): Check for arithmetic overflow when computing buffer size. + + * dir.c (diff_dirs): Check for recursive directory loop. + Arg is now struct comparison const *. + (dir_loop): New function + + * analyze.c (no_discards): Remove. + (inhibit): Remove. + (shift_boundaries): Don't inhibit. If inhibit_hunk_merge is nonzero, + don't merge hunks. + (briefly_report): Now returns 2 if trouble, CHANGES otherwise. + (diff_2_files): Now takes struct comparison. If briefly_report reports + trouble, pass it on to caller. + + * side.c (print_half_line): Add brackets to pacify GCC -Wall. + + * sdiff.c (quotesys.h): Include. + (DIFF_PROGRAM, DEFAULT_EDITOR_PROGRAM): Remove. + (free_software_msgid, editor_program, not_found): New vars. + (diffbin, edbin): Remove. + (editor_program): Renamed from edbin. + (edit, interact): Now take extra string arg. + (exiterr, fatal, perror_fatal, try_help): Add noreturn attribute. + (sigset_t, sigemptyset, sigmask, sigaddset, SIG_BLOCK, SIG_SETMASK): + (sigprocmask): New macros, if !HAVE_SIGPROCMASK. + (error): Now has printf attribute. + (longopts, main): Don't assume ASCII. + (try_help): New operand arg. + (usage): Conform to new GNU standards. + (main): Set static vars for editor and diff program. + Compare getopt_long result to -1, not EOF. + -v conforms to new GNU standard. + Complain better about extra and missing operands. + If HAVE_VFORK, block SIGINT and SIGPIPE in the parent, since when + the child munges its handlers it may somp on the parent. + Pass rname to intract. + Translate not-found message before forking. + (give_help): Just output it all at once. + (edit): New args lname, lline, rname, rline. + (edit): New command 'd'. + (interact): New args lname, rname. + + * util.c (quotesys.h): Include. + (PR_PROGRAM): New macro. + (pfatal_with_name): Abort if error returns. + (fatal): Likewise. + (print_message_queue): Free message chain after printing. + (currently_recursive): Renamed from current_depth, and now a boolean. + (begin_output): Report error if fflush does. + Avoid stdio and gettext in child. + + * diff3.c (quotesys.h): Include. + (free_software_msgid): New decl. + (RANGE_START, RANGE_END): Renamed from START and END. + (fatal, perror_with_exit, try_help): Add noreturn attribute. + (error): Add printf attribute. + (diff_program): Now a ptr, not an array. + Initialize to DEFAULT_DIFF_PROGRAM instead of DIFF_PROGRAM. + (longopts, main): Don't assume ASCII. + (main): Use DIFF environment var to specify name of diff program. + Compare getopt_long result to -1, not EOF. + -v now reports version according to new GNU standard. + Report spelling of extra operand, or last operand before missing one. + (try_help): Now takes operand arg. + (option_help_ms): Fix typo: missing comma. + (usage): Update as per current GNU standards. + (environ): Remove decl. + (read_diff): Invoke diff with --inhibit-hunk-merge. + Translate `not found' message before forking. + Quote name of diff program. + Pass horizon lines. + `memory exhausted' -> `Memory exhausted' + + * pc/makefile (%.exe): Remove. + (pc-clean): Remove *.exe + * pc/makefile.sed (DEFAULT_DIFF_PROGRAM): Renamed from DIFF_PROGRAM. + When editing mkinstalldirs rule, look for exec_prefix and prefix. + Add .exe when installing files. + * pc/emx/config.h (same_file): Add. + * pc/config.h (same_file): Remove. + * pc/djgpp/config.h: Adjust to latest patch from eliz. + * pc/djgpp/makefile.sed: Don't alter PROGRAMS. + * pc/pc.c: Update FSF address. + (quote_system_arg): Renamed from system_quote_arg. + + * README: Add --with-mvfs-stat-bug, --with-nfs-stat-bug. + + * getmsgids: Add copyright date and update FSF address. + + * diff.texi: Document recent changes. + The patch doc still corresponds to patch 2.2, unfortunately. + Update GNU bug reporting address. Omit Larry Wall's address; + it's obsolete and he's busy with perl. + + * context.c: Fix spacing. + + * NEWS: Mention --from-file=FILE, --to-file=FILE, ed. + + * acconfig.h, freesoft.c, message/de.po, message/en_UK.po, + message/es.po, message/fr.po, message/pl.po, message/sv.po: + New files. + + * ed.c: Remove `#if 0'ed code. + + * normal.c, waitpid.c: Update FSF address. + +1998-03-15 Paul Eggert + + * quotesys.c: Renamed from quotearg.c. + + * quotesys.h: Renamed from quotearg.h + (__QUOTESYS_P): Renamed from __QUOTEARG_P. + +1997-05-05 Paul Eggert + + * quotesys.c, quotesys.h: New file. + +Mon Nov 14 05:10:56 1994 Paul Eggert + + Add internationalization support. + Several messages have been changed slightly, + to make them more consistent and easier to translate. + All strings that are messages are passed through gettext once before + being used, so that they can be localized. + Each function and macro whose first parameter is a gettext msgid + has had its first parameter's name changed so it ends in `msgid'. + All arrays of msgids have had their names changed to end in `msgid'. + `getmsgids' uses this to determine which strings are msgids. + + * pc/COPYING, pc/INSTALL, pc/config.h, + pc/djgpp/config.h, pc/djgpp/makefile.sed, + pc/emx/config.h, pc/emx/diff.def, pc/emx/gnuregex.def, + pc/emx/makefile.sed, + pc/makefile, pc/makefile.sed, pc/pc.c: New files, for PC support. + + * getmsgids: New file. + + * Makefile.in (PACKAGE, VERSION, diffutils_srcs, D): New vars. + (version.c, version.texi, messages.po): New files. + messages.po is built automatically from source files and `getmsgids'. + (distfiles): Add them, pc/*, and getmsgids. + (diff.info): Now depends on version.texi. + (realclean): Clean messages.po, version.*. + (dist): Just build $D.tar.gz. + ($D.tar.gz): New file, takes over old `dist' function. + Don't assume $(distfiles) are all in same directory. + + * configure.in (AC_CHECK_HEADERS): Add libintl.h, locale.h. + (AC_CHECK_LIB): Check for -lintl. + + * analyze.c (briefly_report): Rewrite `message (A?"B":"C")' as + `if (A) message ("B") : message ("C")'; this is for getmsgids. + (briefly_report, diff_2_files): For label, use file_label if set. + * diff.c (compare_files): Likewise. + + * system.h (gettext): Declare; use a stub if ! HAVE_LIBINTL_H. + (setlocale): Declare; use a stub if ! HAVE_LOCALE_H. + + * cmp.c, diff.c, diff3.c, sdiff.c (main): + Invoke setlocale first thing, to tell library we're internationalized. + (option_help_msgid): New constant. + (usage): Use it, so message is translated one option at a time. + * sdiff (help_msgid, give_help): Likewise. + + * cmp.c (sprintc): Renamed from `printc'. + Now outputs to a buffer instead of stdout. + (cmp): Use new sprintc; it's easier to internationalize. + + * diff.c (main): -D FOO now outputs `/* ! FOO */ instead of + `/* not FOO */'. + + * sdiff.c (version_string): Fix decl typo: `const' was missing. + (trapsigs): Ignore sigaction failure, to be compatible with `signal'. + + * util.c (struct msg, message5, print_message_queue): + Allocate just one block of memory to save a message. + +Wed Nov 9 17:42:44 1994 Paul Eggert + + * sdiff.c (trapsigs): Don't check signal return value, since it's + bogus under djgpp. + +Mon Oct 31 07:27:27 1994 Paul Eggert + + * Makefile.in (srcs, diff_o, diff3_o, sdiff_o): + New files quote.c, quote.o. + + * diff.h (function_regexp, ignore_regexp): Replace lists of compiled + regexps with these single compiled regexps. All users changed. + (regexp_list,function_regexp_list,ignore_regexp_list): Move to diff.c. + * diff.c (add_regexp): Build one big regexp instead of a regexp list. + (summarize_regexp_list): New function. + (regexp_list): Redesigned struct; moved here from diff.h. + (function_regexp_list, ignore_regexp_list): Likewise, for vars. + + * context.c (find_function): Simplify interface: + don't return size of function line. All callers changed. + (print_context_script, find_function): INT_MAX now denotes no + previous match; this is simpler than `- file->prefix_lines - 1'. + + * diff3.c (read_diff): Quote arguments with system_quote_arg. + * sdiff.c (main): Use system_quote_arg to compute command. + * diff.c (option_list): Quote options with system_quote_arg. + * util.c (begin_output): Use system_quote_arg to compute command. + + * util.c (pr_program): New var. + (analyze_hunk): Fix off-by-1 line length bug. + Match with one big regexp instead of a list of regexps. + Use new `trivial_length' local instead of comparing first byte to `\n'. + Help the compiler with linbuf local vars. + + * system.h (system_quote_arg): + New function; replaces SYSTEM_QUOTE_ARG macro. + +Sat Oct 15 20:09:12 1994 Paul Eggert + + * system.h (_tolower): Define if not already defined. + * io.c (find_and_hash_each_line): Change tolower to _tolower; this + speeds up diff -i considerably on some hosts (e.g. Solaris 2.3). + * util.c (line_cmp): Likewise. + * ifdef.c (groups_letter_value): Likewise. + + * diff.h (ignore_some_line_changes): Remove. All users changed. + * io.c (find_and_hash_each_line): Don't invoke line_cmp if the length + differs and -i is in force. Don't assume ISSPACE ('\n') is nonzero. + + * diff.h (xmalloc_exit_failure): New variable. + All `main' programs set this variable at the start. + xmalloc and xrealloc are now taken from GNU library. + * cmp.c (main): Align buffer size to word size; some mallocs care. + * io.c (slurp): Likewise. + * diff.c (add_exclude): Can now assume xrealloc (0, ...) works. + (add_regexp): Free storage on failure. Allocate storage all at one go. + * system.h (malloc, realloc): Remove unused declarations. + * diff3.c, sdiff.c, util.c (xmalloc, xrealloc): Remove. + * sdiff.c (diffarg): Take advantage of cleaner xrealloc semantics. + + * io.c (ROL): Use sizeof to make it more generic. + + * Makefile.in (common_o): New variable. + Link error.o and xmalloc.o into all programs. + (check): Depend on $(PROGRAMS). + + * diff.h (error): Change to GNU library standard. All callers changed. + * diff3.c (main): Use strerror (EISDIR) instead of "Is a directory". + (fatal, perror_with_exit): Use `error'. + * util.c (perror_with_name, fatal): Use GNU `error'. + (error): Remove. + +Wed Oct 12 17:04:40 1994 David J. MacKenzie (djm@duality.gnu.ai.mit.edu) + + * cmp.c (main): Set xmalloc_exit_failure. + Sat Oct 1 05:24:19 1994 Paul Eggert * Version 2.7 released. @@ -190,7 +2287,7 @@ Fri Sep 2 16:01:49 1994 Paul Eggert * system.h (STAT_BLOCKSIZE): Don't define if already defined. (min, max): Undef if already defined. (filename_cmp, filename_lastdirchar, HAVE_FORK, HAVE_SETMODE, - initialize_main O_BINARY, same_file): New macros. + initialize_main, O_BINARY, same_file): New macros. Fri Jun 17 11:23:53 1994 David J. MacKenzie (djm@geech.gnu.ai.mit.edu) @@ -580,7 +2677,7 @@ Sat Jun 5 23:10:40 1993 Paul Eggert (eggert@twinsun.com) Wed May 26 17:16:02 1993 Paul Eggert (eggert@twinsun.com) * diff.c (main): Cast args to compare_files, for traditional C. - * side.c (print_sdiff_common_lines_print_sdiff_hunk): Likewise. + * side.c (print_sdiff_common_lines, print_sdiff_hunk): Likewise. * analyze.c, diff3.c, sdiff.c, util.c: Don't assume NULL is defined properly. @@ -1407,9 +3504,9 @@ Sun Mar 25 15:58:42 1990 Richard Stallman (rms at sugar-bombs.ai.mit.edu) * analyze.c (discard_confusing_lines): `many' wasn't being used; use it. - Cancelling provisionals near start of run must handle already - cancelled provisionals. - Cancelling subruns of provisionals was cancelling last nonprovisional. + Canceling provisionals near start of run must handle already + canceled provisionals. + Canceling subruns of provisionals was canceling last nonprovisional. Sat Mar 24 14:02:51 1990 Richard Stallman (rms at sugar-bombs.ai.mit.edu) @@ -1417,7 +3514,7 @@ Sat Mar 24 14:02:51 1990 Richard Stallman (rms at sugar-bombs.ai.mit.edu) Threshold for line occurring many times scales by square root of total lines. Within each run, cancel any long subrun of provisionals. - Don't update `provisional' while cancelling provisionals. + Don't update `provisional' while canceling provisionals. In big outer loop, handle provisional and nonprovisional separately. Thu Mar 22 16:35:33 1990 Richard Stallman (rms at sugar-bombs.ai.mit.edu) @@ -1477,7 +3574,7 @@ Thu Feb 8 02:43:16 1990 Jim Kingdon (kingdon at pogo.ai.mit.edu) Fri Feb 2 23:21:38 1990 Richard Stallman (rms at sugar-bombs.ai.mit.edu) - * analyze.c (diif_2_files): If -B or -I, don't return 1 + * analyze.c (diff_2_files): If -B or -I, don't return 1 if all changes were ignored. Wed Jan 24 20:43:57 1990 Richard Stallman (rms at albert.ai.mit.edu) @@ -1496,7 +3593,7 @@ Wed Jan 10 16:06:38 1990 Richard Stallman (rms at sugar-bombs.ai.mit.edu) Sun Dec 24 10:29:20 1989 Richard Stallman (rms at sugar-bombs.ai.mit.edu) * io.c (find_equiv_class): Fix typo that came from changing init of B - to an assigment. + to an assignment. * version.c: New file. * diff.c (main): -v prints version number. @@ -1523,7 +3620,7 @@ Thu Nov 16 13:51:10 1989 Richard Stallman (rms at sugar-bombs.ai.mit.edu) * analyze.c (shift_boundaries): Test for END at end of range before indexing by it. - Fix typo `preceeding' in var names. + Fix misspelling of `preceding' in var names. Sat Nov 11 14:04:16 1989 Richard Stallman (rms at sugar-bombs.ai.mit.edu) @@ -1758,9 +3855,24 @@ Thu Nov 3 16:30:24 1988 Randall Smith (randy at gluteus.ai.mit.edu) * ed.c, analyze.c, context.c: Shortened names to 12 characters for the sake of System V (too simple not to do). - -Local Variables: -mode: indented-text -left-margin: 8 -version-control: never -End: + + +Copyright (C) 1988, 1989, 1990, 1991, 1992, 1993, 1994, 1997, 1998, 1999, +2000, 2001, 2002 Free Software Foundation, Inc. + +This file is part of GNU Diffutils. + +This program is free software; you can redistribute it and/or modify +it under the terms of the GNU General Public License as published by +the Free Software Foundation; either version 2, or (at your option) +any later version. + +This program is distributed in the hope that they will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING. If not, write to +the Free Software Foundation, Inc., 59 Temple Place - Suite 330, +Boston, MA 02111-1307, USA. diff --git a/contrib/diff/Makefile.am b/contrib/diff/Makefile.am new file mode 100644 index 0000000..87f7c3d --- /dev/null +++ b/contrib/diff/Makefile.am @@ -0,0 +1,23 @@ +# Main Automakefile for GNU diffutils. + +# Copyright (C) 2001, 2002, 2004 Free Software Foundation, Inc. + +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; either version 2, or (at your option) +## any later version. + +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. + +## You should have received a copy of the GNU General Public License +## along with this program; if not, write to the Free Software +## Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA +## 02111-1307, USA. + +EXTRA_DIST = bootstrap exgettext +SUBDIRS = doc lib m4 ms src man po + +ACLOCAL_AMFLAGS = -I m4 diff --git a/contrib/diff/NEWS b/contrib/diff/NEWS index dcde122..3854f8b 100644 --- a/contrib/diff/NEWS +++ b/contrib/diff/NEWS @@ -1,6 +1,102 @@ +Version 2.8.7 contains no user-visible changes. + +User-visible changes in version 2.8.6: + +* New diff3 option --strip-trailing-cr. + +* With -N and -P, inaccessible empty regular files (the kind of files + that 'patch' creates to indicate nonexistent backups) are now + treated as nonexistent when they are in the 'backup' file position. + +* If multiple SKIP values are given to cmp, e.g., `cmp -i 10 -i 20', + cmp now uses the maximal value instead of the last one. + +* diff now omits the ".000000000" on hosts that do not support + fractional time stamps. + +Version 2.8.5 was not publicly released. + +User-visible changes in version 2.8.4: + +* Diff now simply prints "Files A and B differ" instead of "Binary + files A and B differ". The message is output if either A or B + appears to be a binary file, and the old wording was misleading + because it implied that both files are binary, which is not + necessarily the case. + +User-visible changes in version 2.8.3: + +* New locale: en_US. + +User-visible changes in version 2.8.2: + +* New diff and sdiff option: + --tabsize=COLUMNS +* If --ignore-space-change or --ignore-all-space is also specified, + --ignore-blank-lines now considers lines to be empty if they contain + only white space. +* More platforms now handle multibyte characters correctly when + excluding files by name (diff -x and -X). +* New locales: hu, pt_BR. + +User-visible changes in version 2.8.1: + +* Documentation fixes. + +User-visible changes in version 2.8: + +* cmp and diff now conform to POSIX 1003.1-2001 (IEEE Std 1003.1-2001) + if the underlying system conforms to POSIX and if the _POSIX2_VERSION + environment variable is set to 200112. Conformance removes support + for `diff -NUM', where NUM is a number. Use -C NUM or -U NUM instead. +* cmp now supports trailing operands SKIP1 and SKIP2, like BSD cmp. +* cmp -i or --ignore-initial now accepts SKIP1:SKIP2 option value. +* New cmp option: -n or --bytes. +* cmp's old -c or --print-chars option has been renamed; + use -b or --print-bytes instead. +* cmp now outputs "byte" rather than "char" outside the POSIX locale. +* cmp -l's index column width now adjusts to fit larger (or smaller) files. +* cmp -l -s and cmp -s -l are not allowed. Use cmp -s or cmp -l instead. +* diff uses ISO 8601 style time stamps for output times (e.g. "2001-11-23 + 16:44:36.875702460 -0800") unless in the C or POSIX locale and the + -c style is specified. +* diff's -I and -F options use the regexp syntax of grep, not of Emacs. +* diff now accepts multiple context arguments, and uses their maximum value. +* New diff and sdiff options: + -E --ignore-tab-expansion + --strip-trailing-cr +* New diff options: + --from-file=FILE, --to-file=FILE + --ignore-file-name-case + --no-ignore-file-name-case +* New diff3 and sdiff option: + --diff-program=PROGRAM +* The following diff options are still accepted, but are no longer documented. + They may be withdrawn in future releases. + -h (omit; it has no effect) + -H (use --speed-large-files instead) + -L (use --label instead) + -P (use --unidirectional-new-file instead) + --inhibit-hunk-merge (omit; it has no effect) +* Recursive diffs now sort file names according to the LC_COLLATE locale + category if possible, instead of using native byte comparison. +* Recursive diffs now detect and report directory loops. +* Diff printf specs can now use the "0" and "'" flags. +* The new sdiff interactive command `ed' precedes each version with a header. +* On 64-bit hosts, files larger than 2 GB can be compared. +* Some internationalization support has been added, but multibyte locales + are still not completely supported yet. +* Some diagnostics have been reworded slightly for consistency. + Also, `diff -D FOO' now outputs `/* ! FOO */' instead of `/* not FOO */'. +* The `patch' part of the manual now describes `patch' version 2.5.4. +* Man pages are now distributed and installed. +* There is support for DJGPP; see the 'ms' subdirectory and the files + m4/dos.m4 and */setmode.*. + + User-visible changes in version 2.7: -* New diff option: --binary (useful only on non-Posix hosts) +* New diff option: --binary (useful only on non-POSIX hosts) * diff -b and -w now ignore line incompleteness; -B no longer does this. * cmp -c now uses locale to decide which output characters to quote. * Help and version messages are reorganized. @@ -64,8 +160,8 @@ User-visible changes in version 2.4: * The format spec %0 introduced in version 2.1 has been removed, since it is incompatible with printf specs like %02d. To represent a null char, use %c'\0' instead. -* cmp and diff now conform to Posix.2 (ISO/IEC 9945-2:1993) - if the underlying system conforms to Posix: +* cmp and diff now conform to POSIX 1003.2-1992 (ISO/IEC 9945-2:1993) + if the underlying system conforms to POSIX: - Some messages' wordings are changed in minor ways. - ``White space'' is now whatever C's `isspace' says it is. - When comparing directories, if `diff' finds a file that is not a regular @@ -124,3 +220,25 @@ User-visible changes in version 2.0: * Add long-named equivalents for other diff3 options. * diff options -F (--show-function-line) and -I (--ignore-matching-lines) can now be given more than once. + + + +Copyright (C) 1993, 1994, 1998, 2001, 2002, 2004 Free Software +Foundation, Inc. + +This file is part of GNU Diffutils. + +This program is free software; you can redistribute it and/or modify +it under the terms of the GNU General Public License as published by +the Free Software Foundation; either version 2, or (at your option) +any later version. + +This program is distributed in the hope that they will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. + +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING. If not, write to +the Free Software Foundation, Inc., 59 Temple Place - Suite 330, +Boston, MA 02111-1307, USA. diff --git a/contrib/diff/README b/contrib/diff/README index bfb9965..0c37230 100644 --- a/contrib/diff/README +++ b/contrib/diff/README @@ -1,9 +1,69 @@ +README for GNU DIFF + This directory contains the GNU diff, diff3, sdiff, and cmp utilities. Their features are a superset of the Unix features and they are -significantly faster. cmp has been moved here from the GNU textutils. +significantly faster. + +Please see the file COPYING for copying conditions. + +Please see the file doc/version.texi for version information. + +Please see the file doc/diff.texi (or doc/diff.info) for documentation +that can be printed with TeX, or read with the `info' program or with +Emacs's `M-x info'. Brief man pages are in man/*, but they are no +substitute for the documentation. + +Please see the file ABOUT-NLS for notes about translations. + +Please see the file INSTALL for generic compilation and installation +instructions. Briefly, you can run "./configure; make install". The +command "./configure --help" lists the supported --enable and --with +options. + +If you have a problem with internationalization, you might be able to +work around it as described in ABOUT-NLS by invoking `./configure +--disable-nls'. Many of the problems arise from dynamic linking +issues on non-GNU platforms (e.g. with the iconv library). Such +problems tend to be shared by other GNU applications on these +platforms, and can usually be fixed by carefully tweaking your non-GNU +installation. If you have an older version of libiconv, please +upgrade to the latest one; see . If +the problem seems isolated to diffutils, though, please report a bug. + +This program requires a Standard C compiler (C89 or later). If you +have a nonstandard compiler, please install GCC first. + +If you make changes to the source code, you may need appropriate +versions of GNU build tools to regenerate the intermediate files. The +following versions were used to generate the intermediate files in +this distribution: + +* Autoconf 2.59 +* Automake 1.8.3 +* gettext 0.14.1 +* help2man 1.33 +* Texinfo 4.7 + +Please report bugs to . + +----- + +Copyright (C) 1992, 1998, 2001, 2002, 2004 Free Software Foundation, +Inc. + +This file is part of GNU Diffutils. + +This program is free software; you can redistribute it and/or modify +it under the terms of the GNU General Public License as published by +the Free Software Foundation; either version 2, or (at your option) +any later version. -See the file COPYING for copying conditions. -See the file diff.texi (or diff.info*) for documentation. -See the file INSTALL for compilation and installation instructions. +This program is distributed in the hope that it will be useful, +but WITHOUT ANY WARRANTY; without even the implied warranty of +MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +GNU General Public License for more details. -Report bugs to bug-gnu-utils@prep.ai.mit.edu +You should have received a copy of the GNU General Public License +along with this program; see the file COPYING. If not, write to +the Free Software Foundation, Inc., 59 Temple Place - Suite 330, +Boston, MA 02111-1307, USA. diff --git a/contrib/diff/THANKS b/contrib/diff/THANKS new file mode 100644 index 0000000..4fc388c --- /dev/null +++ b/contrib/diff/THANKS @@ -0,0 +1,22 @@ +Thanks to all the following for their contributions to GNU diffutils: + +Thomas Bushnell +Wayne Davison +Ulrich Drepper +Paul Eggert +Jay Fenlason +John Gilmore +Torbjorn Granlund +Mike Haertel +Bruno Haible +Chris Hanson +Jim Kingdon +Tom Lord +David J. MacKenzie +Roland McGrath +Jim Meyering +Eugene W. Myers +Randy Smith +Richard Stallman +Leonard H. Tower Jr. +Eli Zaretskii diff --git a/contrib/diff/TODO b/contrib/diff/TODO new file mode 100644 index 0000000..98211ed --- /dev/null +++ b/contrib/diff/TODO @@ -0,0 +1 @@ +Add --include option (opposite of --exclude). diff --git a/contrib/diff/bootstrap b/contrib/diff/bootstrap new file mode 100755 index 0000000..f72f57d --- /dev/null +++ b/contrib/diff/bootstrap @@ -0,0 +1,180 @@ +#! /bin/sh + +# Bootstrap this package from CVS. + +# Copyright (C) 2003, 2004 Free Software Foundation, Inc. + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. + +# You should have received a copy of the GNU General Public License +# along with this program; if not, write to the Free Software +# Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA +# 02111-1307, USA. + +# Written by Paul Eggert. + +package=diffutils + +# Parse options. + +for option +do + case $option in + --help) + echo "$0: usage: $0 [--gnulib-srcdir=DIR] [--cvs-auth=AUTH-METHOD] [--cvs-user=USERNAME] [--skip-po]" + exit;; + --gnulib-srcdir=*) + GNULIB_SRCDIR=`expr "$1" : '--gnulib-srcdir=\(.*\)'`;; + --cvs-auth=*) + CVS_AUTH=`expr "$1" : '--cvs-auth=\(.*\)'`;; + --cvs-user=*) + CVS_USER=`expr "$1" : '--cvs-user=\(.*\)'`;; + --skip-po) + SKIP_PO=t;; + *) + echo >&2 "$0: $option: unknown option" + exit 1;; + esac +done + +echo "$0: Bootstrapping CVS $package..." + +build_cvs_prefix() { + CVS_PREFIX=:${1}: + if [ "${2}" != - ]; then + CVS_PREFIX=${CVS_PREFIX}${2}@ + fi +} + +# Get gnulib files. + +case ${GNULIB_SRCDIR--} in +-) + if [ ! -d gnulib ]; then + echo "$0: getting gnulib files..." + + trap exit 1 2 13 15 + trap 'rm -fr gnulib; exit 1' 0 + + case "${CVS_AUTH--}" in + -) : ${CVS_RSH:?} + CVS_PREFIX="";; + pserver) build_cvs_prefix $CVS_AUTH ${CVS_USER:-anoncvs};; + gserver|server) + build_cvs_prefix $CVS_AUTH ${CVS_USER--};; + ext) : ${CVS_RSH:?} + build_cvs_prefix $CVS_AUTH ${CVS_USER--};; + *) echo "$0: Unknown CVS access method" >&2 + exit 1;; + esac + if [ "${CVS_AUTH--}" = "pserver" ]; then + cvs -d ${CVS_PREFIX}subversions.gnu.org:/cvsroot/gnulib login || exit + fi + cvs -q -d ${CVS_PREFIX}subversions.gnu.org:/cvsroot/gnulib co gnulib || exit + + trap 0 + fi + GNULIB_SRCDIR=gnulib +esac + +<$GNULIB_SRCDIR/gnulib-tool || exit + +gnulib_modules=' +c-stack +dirname +error +exclude +exit +exitfail +file-type +fnmatch-gnu +getopt +hard-locale +inttostr +mkstemp +posixver +regex +strcase +strftime +strtoumax +unlocked-io +version-etc +xalloc +xstrtoumax +' + +previous_gnulib_modules= +while [ "$gnulib_modules" != "$previous_gnulib_modules" ]; do + previous_gnulib_modules=$gnulib_modules + gnulib_modules=` + (echo "$gnulib_modules" + for gnulib_module in $gnulib_modules; do + $GNULIB_SRCDIR/gnulib-tool --extract-dependencies $gnulib_module + done) | sort -u + ` +done + +gnulib_files=` + (for gnulib_module in $gnulib_modules; do + $GNULIB_SRCDIR/gnulib-tool --extract-filelist $gnulib_module + done) | sort -u +` + +gnulib_dirs=`echo "$gnulib_files" | sed 's,/[^/]*$,,' | sort -u` +mkdir -p $gnulib_dirs || exit + +for gnulib_file in $gnulib_files; do + dest=$gnulib_file + + case $gnulib_file in + m4/onceonly_2_57.m4) dest=m4/onceonly.m4;; + esac + + rm -f $dest && + echo "$0: Copying file $GNULIB_SRCDIR/$gnulib_file" && + cp -p $GNULIB_SRCDIR/$gnulib_file $dest || exit +done + + +# Get translations. + +case $SKIP_PO in +'') + echo "$0: getting translations into po..." + (cd po && + rm -f dummy `ls | sed -n '/\.gmo$/p; /\.po/p'` && + wget -nv -nd -r -l 1 -A .po -C off \ + http://www2.iro.umontreal.ca/~gnutra/po/maint/$package/ && + ls *.po | sed 's/\.po$//' >LINGUAS + ) || exit;; +esac + + +# Reconfigure, getting other files. + +echo "$0: autoreconf --verbose --install --force ..." +autoreconf --verbose --install --force || exit + + +# Generate autoconf and automake snippets. + +(echo '# This file is generated automatically by "bootstrap".' && + echo 'AC_DEFUN([GNULIB_AUTOCONF_SNIPPET],[' && + $GNULIB_SRCDIR/gnulib-tool --extract-autoconf-snippet $gnulib_modules && + echo '])' +) >m4/gnulib.m4 || exit + +(echo '# This file is generated automatically by "bootstrap".' && + $GNULIB_SRCDIR/gnulib-tool --extract-automake-snippet $gnulib_modules +) >lib/gnulib.mk || exit + + +echo "$0: done. Now you can run './configure'." diff --git a/contrib/diff/doc/Makefile.am b/contrib/diff/doc/Makefile.am new file mode 100644 index 0000000..574fdab --- /dev/null +++ b/contrib/diff/doc/Makefile.am @@ -0,0 +1,24 @@ +# Makefile for GNU diffutils documentation. + +# Copyright (C) 2001, 2002 Free Software Foundation, Inc. + +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; either version 2, or (at your option) +## any later version. + +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. + +## You should have received a copy of the GNU General Public License +## along with this program; if not, write to the Free Software Foundation, +## Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. + +AM_MAKEINFOFLAGS = --no-split + +info_TEXINFOS = diff.texi +diff_TEXINFOS = fdl.texi + +EXTRA_DIST = diagmeet.note diff --git a/contrib/diff/doc/diagmeet.note b/contrib/diff/doc/diagmeet.note new file mode 100644 index 0000000..8f7242c --- /dev/null +++ b/contrib/diff/doc/diagmeet.note @@ -0,0 +1,71 @@ +Here is a comparison matrix which shows a case in which +it is possible for the forward and backward scan in `diag' +to meet along a nonzero length of diagonal simultaneous +(so that bdiag[d] and fdiag[d] are not equal) +even though there is no snake on that diagonal at the meeting point. + + + 85 1 1 1 159 1 1 17 + 1 2 3 4 +60 + 1 2 +1 + 2 2 3 4 +71 + 3 3 4 5 +85 + 4 3 4 5 +17 + 5 4 5 +1 + 6 4 5 6 +183 + 7 5 6 7 +10 + 8 6 7 +1 + 9 6 7 8 +12 + 7 8 9 10 +13 + 10 8 9 10 +14 + 10 9 10 +17 + 10 10 +1 + 10 9 10 +1 + 8 10 10 10 +183 + 8 7 9 9 9 +10 + 7 6 8 9 8 8 +1 + 6 5 7 7 +1 + 5 6 6 +1 + 5 5 5 +50 + 5 4 4 4 +1 + 4 3 3 +85 + 5 4 3 2 2 +1 + 2 1 +17 + 5 4 3 2 1 1 +1 + 1 0 + 85 1 1 1 159 1 1 17 + + + + + + + + + diff --git a/contrib/diff/doc/diff.texi b/contrib/diff/doc/diff.texi new file mode 100644 index 0000000..a8724f4 --- /dev/null +++ b/contrib/diff/doc/diff.texi @@ -0,0 +1,4649 @@ +\input texinfo @c -*-texinfo-*- +@comment $Id: diff.texi,v 1.25 2004/04/12 07:44:35 eggert Exp $ +@comment %**start of header +@setfilename diff.info +@include version.texi +@settitle Comparing and Merging Files +@syncodeindex vr cp +@setchapternewpage odd +@comment %**end of header +@copying +This manual is for GNU Diffutils +(version @value{VERSION}, @value{UPDATED}), +and documents the @acronym{GNU} @command{diff}, @command{diff3}, +@command{sdiff}, and @command{cmp} commands for showing the +differences between files and the @acronym{GNU} @command{patch} command for +using their output to update files. + +Copyright @copyright{} 1992, 1993, 1994, 1998, 2001, 2002, 2004 Free +Software Foundation, Inc. + +@quotation +Permission is granted to copy, distribute and/or modify this document +under the terms of the GNU Free Documentation License, Version 1.1 or +any later version published by the Free Software Foundation; with no +Invariant Sections, with the Front-Cover texts being ``A GNU Manual,'' +and with the Back-Cover Texts as in (a) below. A copy of the +license is included in the section entitled ``GNU Free Documentation +License.'' + +(a) The FSF's Back-Cover Text is: ``You have freedom to copy and modify +this GNU Manual, like GNU software. Copies published by the Free +Software Foundation raise funds for GNU development.'' +@end quotation +@end copying + +@c Debian install-info (up through at least version 1.9.20) uses only the +@c first dircategory. Put this one first, as it is more useful in practice. +@dircategory Individual utilities +@direntry +* cmp: (diff)Invoking cmp. Compare 2 files byte by byte. +* diff: (diff)Invoking diff. Compare 2 files line by line. +* diff3: (diff)Invoking diff3. Compare 3 files line by line. +* patch: (diff)Invoking patch. Apply a patch to a file. +* sdiff: (diff)Invoking sdiff. Merge 2 files side-by-side. +@end direntry + +@dircategory Text creation and manipulation +@direntry +* Diff: (diff). Comparing and merging files. +@end direntry + +@titlepage +@title Comparing and Merging Files +@subtitle for Diffutils @value{VERSION} and @code{patch} 2.5.4 +@subtitle @value{UPDATED} +@author David MacKenzie, Paul Eggert, and Richard Stallman +@page +@vskip 0pt plus 1filll +@insertcopying +@end titlepage + +@shortcontents +@contents + +@ifnottex +@node Top +@top Comparing and Merging Files + +@insertcopying +@end ifnottex + +@menu +* Overview:: Preliminary information. +* Comparison:: What file comparison means. + +* Output Formats:: Formats for two-way difference reports. +* Incomplete Lines:: Lines that lack trailing newlines. +* Comparing Directories:: Comparing files and directories. +* Adjusting Output:: Making @command{diff} output prettier. +* diff Performance:: Making @command{diff} smarter or faster. + +* Comparing Three Files:: Formats for three-way difference reports. +* diff3 Merging:: Merging from a common ancestor. + +* Interactive Merging:: Interactive merging with @command{sdiff}. + +* Merging with patch:: Using @command{patch} to change old files into new ones. +* Making Patches:: Tips for making and using patch distributions. + +* Invoking cmp:: Compare two files byte by byte. +* Invoking diff:: Compare two files line by line. +* Invoking diff3:: Compare three files line by line. +* Invoking patch:: Apply a diff file to an original. +* Invoking sdiff:: Side-by-side merge of file differences. + +* Standards conformance:: Conformance to the @acronym{POSIX} standard. +* Projects:: If you've found a bug or other shortcoming. + +* Copying This Manual:: How to make copies of this manual. +* Translations:: Available translations of this manual. +* Index:: Index. +@end menu + +@node Overview +@unnumbered Overview +@cindex overview of @command{diff} and @command{patch} + +Computer users often find occasion to ask how two files differ. Perhaps +one file is a newer version of the other file. Or maybe the two files +started out as identical copies but were changed by different people. + +You can use the @command{diff} command to show differences between two +files, or each corresponding file in two directories. @command{diff} +outputs differences between files line by line in any of several +formats, selectable by command line options. This set of differences is +often called a @dfn{diff} or @dfn{patch}. For files that are identical, +@command{diff} normally produces no output; for binary (non-text) files, +@command{diff} normally reports only that they are different. + +You can use the @command{cmp} command to show the byte and line numbers +where two files differ. @command{cmp} can also show all the bytes +that differ between the two files, side by side. A way to compare +two files character by character is the Emacs command @kbd{M-x +compare-windows}. @xref{Other Window, , Other Window, emacs, The @acronym{GNU} +Emacs Manual}, for more information on that command. + +You can use the @command{diff3} command to show differences among three +files. When two people have made independent changes to a common +original, @command{diff3} can report the differences between the original +and the two changed versions, and can produce a merged file that +contains both persons' changes together with warnings about conflicts. + +You can use the @command{sdiff} command to merge two files interactively. + +You can use the set of differences produced by @command{diff} to distribute +updates to text files (such as program source code) to other people. +This method is especially useful when the differences are small compared +to the complete files. Given @command{diff} output, you can use the +@command{patch} program to update, or @dfn{patch}, a copy of the file. If you +think of @command{diff} as subtracting one file from another to produce +their difference, you can think of @command{patch} as adding the difference +to one file to reproduce the other. + +This manual first concentrates on making diffs, and later shows how to +use diffs to update files. + +@acronym{GNU} @command{diff} was written by Paul Eggert, Mike Haertel, +David Hayes, Richard Stallman, and Len Tower. Wayne Davison designed and +implemented the unified output format. The basic algorithm is described +by Eugene W. Myers in ``An O(ND) Difference Algorithm and its Variations'', +@cite{Algorithmica} Vol.@: 1 No.@: 2, 1986, pp.@: 251--266; and in ``A File +Comparison Program'', Webb Miller and Eugene W. Myers, +@cite{Software---Practice and Experience} Vol.@: 15 No.@: 11, 1985, +pp.@: 1025--1040. +@c From: "Gene Myers" +@c They are about the same basic algorithm; the Algorithmica +@c paper gives a rigorous treatment and the sub-algorithm for +@c delivering scripts and should be the primary reference, but +@c both should be mentioned. +The algorithm was independently discovered as described by E. Ukkonen in +``Algorithms for Approximate String Matching'', +@cite{Information and Control} Vol.@: 64, 1985, pp.@: 100--118. +@c From: "Gene Myers" +@c Date: Wed, 29 Sep 1993 08:27:55 MST +@c Ukkonen should be given credit for also discovering the algorithm used +@c in GNU diff. +Unless the @option{--minimal} option is used, @command{diff} uses a +heuristic by Paul Eggert that limits the cost to @math{O(N^1.5 log N)} +at the price of producing suboptimal output for large inputs with many +differences. Related algorithms are surveyed by Alfred V. Aho in +section 6.3 of ``Algorithms for Finding Patterns in Strings'', +@cite{Handbook of Theoretical Computer Science} (Jan Van Leeuwen, +ed.), Vol.@: A, @cite{Algorithms and Complexity}, Elsevier/MIT Press, +1990, pp.@: 255--300. + +@acronym{GNU} @command{diff3} was written by Randy Smith. @acronym{GNU} +@command{sdiff} was written by Thomas Lord. @acronym{GNU} @command{cmp} +was written by Torbj@"orn Granlund and David MacKenzie. + +@acronym{GNU} @command{patch} was written mainly by Larry Wall and Paul Eggert; +several @acronym{GNU} enhancements were contributed by Wayne Davison and +David MacKenzie. Parts of this manual are adapted from a manual page +written by Larry Wall, with his permission. + +@node Comparison +@chapter What Comparison Means +@cindex introduction + +There are several ways to think about the differences between two files. +One way to think of the differences is as a series of lines that were +deleted from, inserted in, or changed in one file to produce the other +file. @command{diff} compares two files line by line, finds groups of +lines that differ, and reports each group of differing lines. It can +report the differing lines in several formats, which have different +purposes. + +@acronym{GNU} @command{diff} can show whether files are different +without detailing the differences. It also provides ways to suppress +certain kinds of differences that are not important to you. Most +commonly, such differences are changes in the amount of white space +between words or lines. @command{diff} also provides ways to suppress +differences in alphabetic case or in lines that match a regular +expression that you provide. These options can accumulate; for +example, you can ignore changes in both white space and alphabetic +case. + +Another way to think of the differences between two files is as a +sequence of pairs of bytes that can be either identical or +different. @command{cmp} reports the differences between two files +byte by byte, instead of line by line. As a result, it is often +more useful than @command{diff} for comparing binary files. For text +files, @command{cmp} is useful mainly when you want to know only whether +two files are identical, or whether one file is a prefix of the other. + +To illustrate the effect that considering changes byte by byte +can have compared with considering them line by line, think of what +happens if a single newline character is added to the beginning of a +file. If that file is then compared with an otherwise identical file +that lacks the newline at the beginning, @command{diff} will report that a +blank line has been added to the file, while @command{cmp} will report that +almost every byte of the two files differs. + +@command{diff3} normally compares three input files line by line, finds +groups of lines that differ, and reports each group of differing lines. +Its output is designed to make it easy to inspect two different sets of +changes to the same file. + +@menu +* Hunks:: Groups of differing lines. +* White Space:: Suppressing differences in white space. +* Blank Lines:: Suppressing differences whose lines are all blank. +* Specified Lines:: Suppressing differences whose lines all match a pattern. +* Case Folding:: Suppressing differences in alphabetic case. +* Brief:: Summarizing which files are different. +* Binary:: Comparing binary files or forcing text comparisons. +@end menu + +@node Hunks +@section Hunks +@cindex hunks + +When comparing two files, @command{diff} finds sequences of lines common to +both files, interspersed with groups of differing lines called +@dfn{hunks}. Comparing two identical files yields one sequence of +common lines and no hunks, because no lines differ. Comparing two +entirely different files yields no common lines and one large hunk that +contains all lines of both files. In general, there are many ways to +match up lines between two given files. @command{diff} tries to minimize +the total hunk size by finding large sequences of common lines +interspersed with small hunks of differing lines. + +For example, suppose the file @file{F} contains the three lines +@samp{a}, @samp{b}, @samp{c}, and the file @file{G} contains the same +three lines in reverse order @samp{c}, @samp{b}, @samp{a}. If +@command{diff} finds the line @samp{c} as common, then the command +@samp{diff F G} produces this output: + +@example +1,2d0 +< a +< b +3a2,3 +> b +> a +@end example + +@noindent +But if @command{diff} notices the common line @samp{b} instead, it produces +this output: + +@example +1c1 +< a +--- +> c +3c3 +< c +--- +> a +@end example + +@noindent +It is also possible to find @samp{a} as the common line. @command{diff} +does not always find an optimal matching between the files; it takes +shortcuts to run faster. But its output is usually close to the +shortest possible. You can adjust this tradeoff with the +@option{-d} or @option{--minimal} option (@pxref{diff Performance}). + +@node White Space +@section Suppressing Differences in Blank and Tab Spacing +@cindex blank and tab difference suppression +@cindex tab and blank difference suppression + +The @option{-E} or @option{--ignore-tab-expansion} option ignores the +distinction between tabs and spaces on input. A tab is considered to be +equivalent to the number of spaces to the next tab stop (@pxref{Tabs}). + +The @option{-b} or @option{--ignore-space-change} option is stronger. +It ignores white space at line end, and considers all other sequences of +one or more white space characters within a line to be equivalent. With this +option, @command{diff} considers the following two lines to be equivalent, +where @samp{$} denotes the line end: + +@example +Here lyeth muche rychnesse in lytell space. -- John Heywood$ +Here lyeth muche rychnesse in lytell space. -- John Heywood $ +@end example + +The @option{-w} or @option{--ignore-all-space} option is stronger still. +It ignores differences even if one line has white space where +the other line has none. @dfn{White space} characters include +tab, newline, vertical tab, form feed, carriage return, and space; +some locales may define additional characters to be white space. +With this option, @command{diff} considers the +following two lines to be equivalent, where @samp{$} denotes the line +end and @samp{^M} denotes a carriage return: + +@example +Here lyeth muche rychnesse in lytell space.-- John Heywood$ + He relyeth much erychnes seinly tells pace. --John Heywood ^M$ +@end example + +@node Blank Lines +@section Suppressing Differences Whose Lines Are All Blank +@cindex blank line difference suppression + +The @option{-B} or @option{--ignore-blank-lines} option ignores changes +that consist entirely of blank lines. With this option, for example, a +file containing +@example +1. A point is that which has no part. + +2. A line is breadthless length. +-- Euclid, The Elements, I +@end example +@noindent +is considered identical to a file containing +@example +1. A point is that which has no part. +2. A line is breadthless length. + + +-- Euclid, The Elements, I +@end example + +Normally this option affects only lines that are completely empty, but +if you also specify the @option{-b} or @option{--ignore-space-change} +option, or the @option{-w} or @option{--ignore-all-space} option, +lines are also affected if they look empty but contain white space. +In other words, @option{-B} is equivalent to @samp{-I '^$'} by +default, but it is equivalent to @option{-I '^[[:space:]]*$'} if +@option{-b} or @option{-w} is also specified. + +@node Specified Lines +@section Suppressing Differences Whose Lines All Match a Regular Expression +@cindex regular expression suppression + +To ignore insertions and deletions of lines that match a +@command{grep}-style regular expression, use the @option{-I +@var{regexp}} or @option{--ignore-matching-lines=@var{regexp}} option. +You should escape +regular expressions that contain shell metacharacters to prevent the +shell from expanding them. For example, @samp{diff -I '^[[:digit:]]'} ignores +all changes to lines beginning with a digit. + +However, @option{-I} only ignores the insertion or deletion of lines that +contain the regular expression if every changed line in the hunk---every +insertion and every deletion---matches the regular expression. In other +words, for each nonignorable change, @command{diff} prints the complete set +of changes in its vicinity, including the ignorable ones. + +You can specify more than one regular expression for lines to ignore by +using more than one @option{-I} option. @command{diff} tries to match each +line against each regular expression. + +@node Case Folding +@section Suppressing Case Differences +@cindex case difference suppression + +@acronym{GNU} @command{diff} can treat lower case letters as +equivalent to their upper case counterparts, so that, for example, it +considers @samp{Funky Stuff}, @samp{funky STUFF}, and @samp{fUNKy +stuFf} to all be the same. To request this, use the @option{-i} or +@option{--ignore-case} option. + +@node Brief +@section Summarizing Which Files Differ +@cindex summarizing which files differ +@cindex brief difference reports + +When you only want to find out whether files are different, and you +don't care what the differences are, you can use the summary output +format. In this format, instead of showing the differences between the +files, @command{diff} simply reports whether files differ. The @option{-q} +or @option{--brief} option selects this output format. + +This format is especially useful when comparing the contents of two +directories. It is also much faster than doing the normal line by line +comparisons, because @command{diff} can stop analyzing the files as soon as +it knows that there are any differences. + +You can also get a brief indication of whether two files differ by using +@command{cmp}. For files that are identical, @command{cmp} produces no +output. When the files differ, by default, @command{cmp} outputs the byte +and line number where the first difference occurs, or reports that one +file is a prefix of the other. You can use +the @option{-s}, @option{--quiet}, or @option{--silent} option to +suppress that information, so that @command{cmp} +produces no output and reports whether the files differ using only its +exit status (@pxref{Invoking cmp}). + +@c Fix this. +Unlike @command{diff}, @command{cmp} cannot compare directories; it can only +compare two files. + +@node Binary +@section Binary Files and Forcing Text Comparisons +@cindex binary file diff +@cindex text versus binary diff + +If @command{diff} thinks that either of the two files it is comparing is +binary (a non-text file), it normally treats that pair of files much as +if the summary output format had been selected (@pxref{Brief}), and +reports only that the binary files are different. This is because line +by line comparisons are usually not meaningful for binary files. + +@command{diff} determines whether a file is text or binary by checking the +first few bytes in the file; the exact number of bytes is system +dependent, but it is typically several thousand. If every byte in +that part of the file is non-null, @command{diff} considers the file to be +text; otherwise it considers the file to be binary. + +Sometimes you might want to force @command{diff} to consider files to be +text. For example, you might be comparing text files that contain +null characters; @command{diff} would erroneously decide that those are +non-text files. Or you might be comparing documents that are in a +format used by a word processing system that uses null characters to +indicate special formatting. You can force @command{diff} to consider all +files to be text files, and compare them line by line, by using the +@option{-a} or @option{--text} option. If the files you compare using this +option do not in fact contain text, they will probably contain few +newline characters, and the @command{diff} output will consist of hunks +showing differences between long lines of whatever characters the files +contain. + +You can also force @command{diff} to report only whether files differ +(but not how). Use the @option{-q} or @option{--brief} option for +this. + +Normally, differing binary files count as trouble because the +resulting @command{diff} output does not capture all the differences. +This trouble causes @command{diff} to exit with status 2. However, +this trouble cannot occur with the @option{-a} or @option{--text} +option, or with the @option{-q} or @option{--brief} option, as these +options both cause @command{diff} to generate a form of output that +represents differences as requested. + +In operating systems that distinguish between text and binary files, +@command{diff} normally reads and writes all data as text. Use the +@option{--binary} option to force @command{diff} to read and write binary +data instead. This option has no effect on a @acronym{POSIX}-compliant system +like @acronym{GNU} or traditional Unix. However, many personal computer +operating systems represent the end of a line with a carriage return +followed by a newline. On such systems, @command{diff} normally ignores +these carriage returns on input and generates them at the end of each +output line, but with the @option{--binary} option @command{diff} treats +each carriage return as just another input character, and does not +generate a carriage return at the end of each output line. This can be +useful when dealing with non-text files that are meant to be +interchanged with @acronym{POSIX}-compliant systems. + +The @option{--strip-trailing-cr} causes @command{diff} to treat input +lines that end in carriage return followed by newline as if they end +in plain newline. This can be useful when comparing text that is +imperfectly imported from many personal computer operating systems. +This option affects how lines are read, which in turn affects how they +are compared and output. + +If you want to compare two files byte by byte, you can use the +@command{cmp} program with the @option{-l} or @option{--verbose} +option to show the values of each differing byte in the two files. +With @acronym{GNU} @command{cmp}, you can also use the @option{-b} or +@option{--print-bytes} option to show the @acronym{ASCII} representation of +those bytes. @xref{Invoking cmp}, for more information. + +If @command{diff3} thinks that any of the files it is comparing is binary +(a non-text file), it normally reports an error, because such +comparisons are usually not useful. @command{diff3} uses the same test as +@command{diff} to decide whether a file is binary. As with @command{diff}, if +the input files contain a few non-text bytes but otherwise are like +text files, you can force @command{diff3} to consider all files to be text +files and compare them line by line by using the @option{-a} or +@option{--text} option. + +@node Output Formats +@chapter @command{diff} Output Formats +@cindex output formats +@cindex format of @command{diff} output + +@command{diff} has several mutually exclusive options for output format. +The following sections describe each format, illustrating how +@command{diff} reports the differences between two sample input files. + +@menu +* Sample diff Input:: Sample @command{diff} input files for examples. +* Context:: Showing differences with the surrounding text. +* Side by Side:: Showing differences in two columns. +* Normal:: Showing differences without surrounding text. +* Scripts:: Generating scripts for other programs. +* If-then-else:: Merging files with if-then-else. +@end menu + +@node Sample diff Input +@section Two Sample Input Files +@cindex @command{diff} sample input +@cindex sample input for @command{diff} + +Here are two sample files that we will use in numerous examples to +illustrate the output of @command{diff} and how various options can change +it. + +This is the file @file{lao}: + +@example +The Way that can be told of is not the eternal Way; +The name that can be named is not the eternal name. +The Nameless is the origin of Heaven and Earth; +The Named is the mother of all things. +Therefore let there always be non-being, + so we may see their subtlety, +And let there always be being, + so we may see their outcome. +The two are the same, +But after they are produced, + they have different names. +@end example + +This is the file @file{tzu}: + +@example +The Nameless is the origin of Heaven and Earth; +The named is the mother of all things. + +Therefore let there always be non-being, + so we may see their subtlety, +And let there always be being, + so we may see their outcome. +The two are the same, +But after they are produced, + they have different names. +They both may be called deep and profound. +Deeper and more profound, +The door of all subtleties! +@end example + +In this example, the first hunk contains just the first two lines of +@file{lao}, the second hunk contains the fourth line of @file{lao} +opposing the second and third lines of @file{tzu}, and the last hunk +contains just the last three lines of @file{tzu}. + +@node Context +@section Showing Differences in Their Context +@cindex context output format +@cindex @samp{!} output format + +Usually, when you are looking at the differences between files, you will +also want to see the parts of the files near the lines that differ, to +help you understand exactly what has changed. These nearby parts of the +files are called the @dfn{context}. + +@acronym{GNU} @command{diff} provides two output formats that show context +around the differing lines: @dfn{context format} and @dfn{unified +format}. It can optionally show in which function or section of the +file the differing lines are found. + +If you are distributing new versions of files to other people in the +form of @command{diff} output, you should use one of the output formats +that show context so that they can apply the diffs even if they have +made small changes of their own to the files. @command{patch} can apply +the diffs in this case by searching in the files for the lines of +context around the differing lines; if those lines are actually a few +lines away from where the diff says they are, @command{patch} can adjust +the line numbers accordingly and still apply the diff correctly. +@xref{Imperfect}, for more information on using @command{patch} to apply +imperfect diffs. + +@menu +* Context Format:: An output format that shows surrounding lines. +* Unified Format:: A more compact output format that shows context. +* Sections:: Showing which sections of the files differences are in. +* Alternate Names:: Showing alternate file names in context headers. +@end menu + +@node Context Format +@subsection Context Format + +The context output format shows several lines of context around the +lines that differ. It is the standard format for distributing updates +to source code. + +To select this output format, use the @option{-C @var{lines}}, +@option{--context@r{[}=@var{lines}@r{]}}, or @option{-c} option. The +argument @var{lines} that some of these options take is the number of +lines of context to show. If you do not specify @var{lines}, it +defaults to three. For proper operation, @command{patch} typically needs +at least two lines of context. + +@menu +* Example Context:: Sample output in context format. +* Less Context:: Another sample with less context. +* Detailed Context:: A detailed description of the context output format. +@end menu + +@node Example Context +@subsubsection An Example of Context Format + +Here is the output of @samp{diff -c lao tzu} (@pxref{Sample diff Input}, +for the complete contents of the two files). Notice that up to three +lines that are not different are shown around each line that is +different; they are the context lines. Also notice that the first two +hunks have run together, because their contents overlap. + +@example +*** lao 2002-02-21 23:30:39.942229878 -0800 +--- tzu 2002-02-21 23:30:50.442260588 -0800 +*************** +*** 1,7 **** +- The Way that can be told of is not the eternal Way; +- The name that can be named is not the eternal name. + The Nameless is the origin of Heaven and Earth; +! The Named is the mother of all things. + Therefore let there always be non-being, + so we may see their subtlety, + And let there always be being, +--- 1,6 ---- + The Nameless is the origin of Heaven and Earth; +! The named is the mother of all things. +! + Therefore let there always be non-being, + so we may see their subtlety, + And let there always be being, +*************** +*** 9,11 **** +--- 8,13 ---- + The two are the same, + But after they are produced, + they have different names. ++ They both may be called deep and profound. ++ Deeper and more profound, ++ The door of all subtleties! +@end example + +@node Less Context +@subsubsection An Example of Context Format with Less Context + +Here is the output of @samp{diff -C 1 lao tzu} (@pxref{Sample diff +Input}, for the complete contents of the two files). Notice that at +most one context line is reported here. + +@example +*** lao 2002-02-21 23:30:39.942229878 -0800 +--- tzu 2002-02-21 23:30:50.442260588 -0800 +*************** +*** 1,5 **** +- The Way that can be told of is not the eternal Way; +- The name that can be named is not the eternal name. + The Nameless is the origin of Heaven and Earth; +! The Named is the mother of all things. + Therefore let there always be non-being, +--- 1,4 ---- + The Nameless is the origin of Heaven and Earth; +! The named is the mother of all things. +! + Therefore let there always be non-being, +*************** +*** 11 **** +--- 10,13 ---- + they have different names. ++ They both may be called deep and profound. ++ Deeper and more profound, ++ The door of all subtleties! +@end example + +@node Detailed Context +@subsubsection Detailed Description of Context Format + +The context output format starts with a two-line header, which looks +like this: + +@example +*** @var{from-file} @var{from-file-modification-time} +--- @var{to-file} @var{to-file-modification time} +@end example + +@noindent +@vindex LC_TIME +@cindex time stamp format, context diffs +The time stamp normally looks like @samp{2002-02-21 23:30:39.942229878 +-0800} to indicate the date, time with fractional seconds, and time +zone in @uref{ftp://ftp.isi.edu/in-notes/rfc2822.txt, Internet RFC +2822 format}. (The fractional seconds are omitted on hosts that do +not support fractional time stamps.) However, a traditional time +stamp like @samp{Thu Feb 21 23:30:39 2002} is used if the +@env{LC_TIME} locale category is either @samp{C} or @samp{POSIX}. + +You can change the header's content with the +@option{--label=@var{label}} option; see @ref{Alternate Names}. + +Next come one or more hunks of differences; each hunk shows one area +where the files differ. Context format hunks look like this: + +@example +*************** +*** @var{from-file-line-numbers} **** + @var{from-file-line} + @var{from-file-line}@dots{} +--- @var{to-file-line-numbers} ---- + @var{to-file-line} + @var{to-file-line}@dots{} +@end example + +If a hunk contains two or more lines, its line numbers look like +@samp{@var{start},@var{end}}. Otherwise only its end line number +appears. An empty hunk is considered to end at the line that precedes +the hunk. + +The lines of context around the lines that differ start with two space +characters. The lines that differ between the two files start with one +of the following indicator characters, followed by a space character: + +@table @samp +@item ! +A line that is part of a group of one or more lines that changed between +the two files. There is a corresponding group of lines marked with +@samp{!} in the part of this hunk for the other file. + +@item + +An ``inserted'' line in the second file that corresponds to nothing in +the first file. + +@item - +A ``deleted'' line in the first file that corresponds to nothing in the +second file. +@end table + +If all of the changes in a hunk are insertions, the lines of +@var{from-file} are omitted. If all of the changes are deletions, the +lines of @var{to-file} are omitted. + +@node Unified Format +@subsection Unified Format +@cindex unified output format +@cindex @samp{+-} output format + +The unified output format is a variation on the context format that is +more compact because it omits redundant context lines. To select this +output format, use the @option{-U @var{lines}}, +@option{--unified@r{[}=@var{lines}@r{]}}, or @option{-u} +option. The argument @var{lines} is the number of lines of context to +show. When it is not given, it defaults to three. + +At present, only @acronym{GNU} @command{diff} can produce this format and +only @acronym{GNU} @command{patch} can automatically apply diffs in this +format. For proper operation, @command{patch} typically needs at +least three lines of context. + +@menu +* Example Unified:: Sample output in unified format. +* Detailed Unified:: A detailed description of unified format. +@end menu + +@node Example Unified +@subsubsection An Example of Unified Format + +Here is the output of the command @samp{diff -u lao tzu} +(@pxref{Sample diff Input}, for the complete contents of the two files): + +@example +--- lao 2002-02-21 23:30:39.942229878 -0800 ++++ tzu 2002-02-21 23:30:50.442260588 -0800 +@@@@ -1,7 +1,6 @@@@ +-The Way that can be told of is not the eternal Way; +-The name that can be named is not the eternal name. + The Nameless is the origin of Heaven and Earth; +-The Named is the mother of all things. ++The named is the mother of all things. ++ + Therefore let there always be non-being, + so we may see their subtlety, + And let there always be being, +@@@@ -9,3 +8,6 @@@@ + The two are the same, + But after they are produced, + they have different names. ++They both may be called deep and profound. ++Deeper and more profound, ++The door of all subtleties! +@end example + +@node Detailed Unified +@subsubsection Detailed Description of Unified Format + +The unified output format starts with a two-line header, which looks +like this: + +@example +--- @var{from-file} @var{from-file-modification-time} ++++ @var{to-file} @var{to-file-modification-time} +@end example + +@noindent +@cindex time stamp format, unified diffs +The time stamp looks like @samp{2002-02-21 23:30:39.942229878 -0800} +to indicate the date, time with fractional seconds, and time zone. +The fractional seconds are omitted on hosts that do not support +fractional time stamps. + +You can change the header's content with the +@option{--label=@var{label}} option; see @xref{Alternate Names}. + +Next come one or more hunks of differences; each hunk shows one area +where the files differ. Unified format hunks look like this: + +@example +@@@@ @var{from-file-line-numbers} @var{to-file-line-numbers} @@@@ + @var{line-from-either-file} + @var{line-from-either-file}@dots{} +@end example + +If a hunk contains just one line, only its start line number appears. +Otherwise its line numbers look like @samp{@var{start},@var{count}}. +An empty hunk is considered to start at the line that follows the hunk. + +If a hunk and its context contain two or more lines, its +line numbers look like @samp{@var{start},@var{count}}. Otherwise only +its end line number appears. An empty hunk is considered to end at +the line that precedes the hunk. + +The lines common to both files begin with a space character. The lines +that actually differ between the two files have one of the following +indicator characters in the left print column: + +@table @samp +@item + +A line was added here to the first file. + +@item - +A line was removed here from the first file. +@end table + +@node Sections +@subsection Showing Which Sections Differences Are in +@cindex headings +@cindex section headings + +Sometimes you might want to know which part of the files each change +falls in. If the files are source code, this could mean which +function was changed. If the files are documents, it could mean which +chapter or appendix was changed. @acronym{GNU} @command{diff} can +show this by displaying the nearest section heading line that precedes +the differing lines. Which lines are ``section headings'' is +determined by a regular expression. + +@menu +* Specified Headings:: Showing headings that match regular expressions. +* C Function Headings:: Showing headings of C functions. +@end menu + +@node Specified Headings +@subsubsection Showing Lines That Match Regular Expressions +@cindex specified headings +@cindex regular expression matching headings + +To show in which sections differences occur for files that are not +source code for C or similar languages, use the @option{-F @var{regexp}} +or @option{--show-function-line=@var{regexp}} option. @command{diff} +considers lines that match the @command{grep}-style regular expression +@var{regexp} to be the beginning +of a section of the file. Here are suggested regular expressions for +some common languages: + +@c Please add to this list, e.g. Fortran, Pascal, Perl, Python. +@table @samp +@item ^[[:alpha:]$_] +C, C++, Prolog +@item ^( +Lisp +@item ^@@node +Texinfo +@end table + +This option does not automatically select an output format; in order to +use it, you must select the context format (@pxref{Context Format}) or +unified format (@pxref{Unified Format}). In other output formats it +has no effect. + +The @option{-F} or @option{--show-function-line} option finds the nearest +unchanged line that precedes each hunk of differences and matches the +given regular expression. Then it adds that line to the end of the +line of asterisks in the context format, or to the @samp{@@@@} line in +unified format. If no matching line exists, this option leaves the output for +that hunk unchanged. If that line is more than 40 characters long, it +outputs only the first 40 characters. You can specify more than one +regular expression for such lines; @command{diff} tries to match each line +against each regular expression, starting with the last one given. This +means that you can use @option{-p} and @option{-F} together, if you wish. + +@node C Function Headings +@subsubsection Showing C Function Headings +@cindex C function headings +@cindex function headings, C + +To show in which functions differences occur for C and similar +languages, you can use the @option{-p} or @option{--show-c-function} option. +This option automatically defaults to the context output format +(@pxref{Context Format}), with the default number of lines of context. +You can override that number with @option{-C @var{lines}} elsewhere in the +command line. You can override both the format and the number with +@option{-U @var{lines}} elsewhere in the command line. + +The @option{-p} or @option{--show-c-function} option is equivalent to +@option{-F '^[[:alpha:]$_]'} if the unified format is specified, otherwise +@option{-c -F '^[[:alpha:]$_]'} (@pxref{Specified Headings}). @acronym{GNU} +@command{diff} provides this option for the sake of convenience. + +@node Alternate Names +@subsection Showing Alternate File Names +@cindex alternate file names +@cindex file name alternates + +If you are comparing two files that have meaningless or uninformative +names, you might want @command{diff} to show alternate names in the header +of the context and unified output formats. To do this, use the +@option{--label=@var{label}} option. The first time +you give this option, its argument replaces the name and date of the +first file in the header; the second time, its argument replaces the +name and date of the second file. If you give this option more than +twice, @command{diff} reports an error. The @option{--label} option does not +affect the file names in the @command{pr} header when the @option{-l} or +@option{--paginate} option is used (@pxref{Pagination}). + +Here are the first two lines of the output from @samp{diff -C 2 +--label=original --label=modified lao tzu}: + +@example +*** original +--- modified +@end example + +@node Side by Side +@section Showing Differences Side by Side +@cindex side by side +@cindex two-column output +@cindex columnar output + +@command{diff} can produce a side by side difference listing of two files. +The files are listed in two columns with a gutter between them. The +gutter contains one of the following markers: + +@table @asis +@item white space +The corresponding lines are in common. That is, either the lines are +identical, or the difference is ignored because of one of the +@option{--ignore} options (@pxref{White Space}). + +@item @samp{|} +The corresponding lines differ, and they are either both complete +or both incomplete. + +@item @samp{<} +The files differ and only the first file contains the line. + +@item @samp{>} +The files differ and only the second file contains the line. + +@item @samp{(} +Only the first file contains the line, but the difference is ignored. + +@item @samp{)} +Only the second file contains the line, but the difference is ignored. + +@item @samp{\} +The corresponding lines differ, and only the first line is incomplete. + +@item @samp{/} +The corresponding lines differ, and only the second line is incomplete. +@end table + +Normally, an output line is incomplete if and only if the lines that it +contains are incomplete; @xref{Incomplete Lines}. However, when an +output line represents two differing lines, one might be incomplete +while the other is not. In this case, the output line is complete, +but its the gutter is marked @samp{\} if the first line is incomplete, +@samp{/} if the second line is. + +Side by side format is sometimes easiest to read, but it has limitations. +It generates much wider output than usual, and truncates lines that are +too long to fit. Also, it relies on lining up output more heavily than +usual, so its output looks particularly bad if you use varying +width fonts, nonstandard tab stops, or nonprinting characters. + +You can use the @command{sdiff} command to interactively merge side by side +differences. @xref{Interactive Merging}, for more information on merging files. + +@menu +* Side by Side Format:: Controlling side by side output format. +* Example Side by Side:: Sample side by side output. +@end menu + +@node Side by Side Format +@subsection Controlling Side by Side Format +@cindex side by side format + +The @option{-y} or @option{--side-by-side} option selects side by side +format. Because side by side output lines contain two input lines, the +output is wider than usual: normally 130 print columns, which can fit +onto a traditional printer line. You can set the width of the output +with the @option{-W @var{columns}} or @option{--width=@var{columns}} +option. The output is split into two halves of equal width, separated by a +small gutter to mark differences; the right half is aligned to a tab +stop so that tabs line up. Input lines that are too long to fit in half +of an output line are truncated for output. + +The @option{--left-column} option prints only the left column of two +common lines. The @option{--suppress-common-lines} option suppresses +common lines entirely. + +@node Example Side by Side +@subsection An Example of Side by Side Format + +Here is the output of the command @samp{diff -y -W 72 lao tzu} +(@pxref{Sample diff Input}, for the complete contents of the two files). + +@example +The Way that can be told of is n < +The name that can be named is no < +The Nameless is the origin of He The Nameless is the origin of He +The Named is the mother of all t | The named is the mother of all t + > +Therefore let there always be no Therefore let there always be no + so we may see their subtlety, so we may see their subtlety, +And let there always be being, And let there always be being, + so we may see their outcome. so we may see their outcome. +The two are the same, The two are the same, +But after they are produced, But after they are produced, + they have different names. they have different names. + > They both may be called deep and + > Deeper and more profound, + > The door of all subtleties! +@end example + +@node Normal +@section Showing Differences Without Context +@cindex normal output format +@cindex @samp{<} output format + +The ``normal'' @command{diff} output format shows each hunk of differences +without any surrounding context. Sometimes such output is the clearest +way to see how lines have changed, without the clutter of nearby +unchanged lines (although you can get similar results with the context +or unified formats by using 0 lines of context). However, this format +is no longer widely used for sending out patches; for that purpose, the +context format (@pxref{Context Format}) and the unified format +(@pxref{Unified Format}) are superior. Normal format is the default for +compatibility with older versions of @command{diff} and the @acronym{POSIX} +standard. Use the @option{--normal} option to select this output +format explicitly. + +@menu +* Example Normal:: Sample output in the normal format. +* Detailed Normal:: A detailed description of normal output format. +@end menu + +@node Example Normal +@subsection An Example of Normal Format + +Here is the output of the command @samp{diff lao tzu} +(@pxref{Sample diff Input}, for the complete contents of the two files). +Notice that it shows only the lines that are different between the two +files. + +@example +1,2d0 +< The Way that can be told of is not the eternal Way; +< The name that can be named is not the eternal name. +4c2,3 +< The Named is the mother of all things. +--- +> The named is the mother of all things. +> +11a11,13 +> They both may be called deep and profound. +> Deeper and more profound, +> The door of all subtleties! +@end example + +@node Detailed Normal +@subsection Detailed Description of Normal Format + +The normal output format consists of one or more hunks of differences; +each hunk shows one area where the files differ. Normal format hunks +look like this: + +@example +@var{change-command} +< @var{from-file-line} +< @var{from-file-line}@dots{} +--- +> @var{to-file-line} +> @var{to-file-line}@dots{} +@end example + +There are three types of change commands. Each consists of a line +number or comma-separated range of lines in the first file, a single +character indicating the kind of change to make, and a line number or +comma-separated range of lines in the second file. All line numbers are +the original line numbers in each file. The types of change commands +are: + +@table @samp +@item @var{l}a@var{r} +Add the lines in range @var{r} of the second file after line @var{l} of +the first file. For example, @samp{8a12,15} means append lines 12--15 +of file 2 after line 8 of file 1; or, if changing file 2 into file 1, +delete lines 12--15 of file 2. + +@item @var{f}c@var{t} +Replace the lines in range @var{f} of the first file with lines in range +@var{t} of the second file. This is like a combined add and delete, but +more compact. For example, @samp{5,7c8,10} means change lines 5--7 of +file 1 to read as lines 8--10 of file 2; or, if changing file 2 into +file 1, change lines 8--10 of file 2 to read as lines 5--7 of file 1. + +@item @var{r}d@var{l} +Delete the lines in range @var{r} from the first file; line @var{l} is where +they would have appeared in the second file had they not been deleted. +For example, @samp{5,7d3} means delete lines 5--7 of file 1; or, if +changing file 2 into file 1, append lines 5--7 of file 1 after line 3 of +file 2. +@end table + +@node Scripts +@section Making Edit Scripts +@cindex script output formats + +Several output modes produce command scripts for editing @var{from-file} +to produce @var{to-file}. + +@menu +* ed Scripts:: Using @command{diff} to produce commands for @command{ed}. +* Forward ed:: Making forward @command{ed} scripts. +* RCS:: A special @command{diff} output format used by @acronym{RCS}. +@end menu + +@node ed Scripts +@subsection @command{ed} Scripts +@cindex @command{ed} script output format + +@command{diff} can produce commands that direct the @command{ed} text editor +to change the first file into the second file. Long ago, this was the +only output mode that was suitable for editing one file into another +automatically; today, with @command{patch}, it is almost obsolete. Use the +@option{-e} or @option{--ed} option to select this output format. + +Like the normal format (@pxref{Normal}), this output format does not +show any context; unlike the normal format, it does not include the +information necessary to apply the diff in reverse (to produce the first +file if all you have is the second file and the diff). + +If the file @file{d} contains the output of @samp{diff -e old new}, then +the command @samp{(cat d && echo w) | ed - old} edits @file{old} to make +it a copy of @file{new}. More generally, if @file{d1}, @file{d2}, +@dots{}, @file{dN} contain the outputs of @samp{diff -e old new1}, +@samp{diff -e new1 new2}, @dots{}, @samp{diff -e newN-1 newN}, +respectively, then the command @samp{(cat d1 d2 @dots{} dN && echo w) | +ed - old} edits @file{old} to make it a copy of @file{newN}. + +@menu +* Example ed:: A sample @command{ed} script. +* Detailed ed:: A detailed description of @command{ed} format. +@end menu + +@node Example ed +@subsubsection Example @command{ed} Script + +Here is the output of @samp{diff -e lao tzu} (@pxref{Sample +diff Input}, for the complete contents of the two files): + +@example +11a +They both may be called deep and profound. +Deeper and more profound, +The door of all subtleties! +. +4c +The named is the mother of all things. + +. +1,2d +@end example + +@node Detailed ed +@subsubsection Detailed Description of @command{ed} Format + +The @command{ed} output format consists of one or more hunks of +differences. The changes closest to the ends of the files come first so +that commands that change the number of lines do not affect how +@command{ed} interprets line numbers in succeeding commands. @command{ed} +format hunks look like this: + +@example +@var{change-command} +@var{to-file-line} +@var{to-file-line}@dots{} +. +@end example + +Because @command{ed} uses a single period on a line to indicate the +end of input, @acronym{GNU} @command{diff} protects lines of changes +that contain a single period on a line by writing two periods instead, +then writing a subsequent @command{ed} command to change the two +periods into one. The @command{ed} format cannot represent an +incomplete line, so if the second file ends in a changed incomplete +line, @command{diff} reports an error and then pretends that a newline +was appended. + +There are three types of change commands. Each consists of a line +number or comma-separated range of lines in the first file and a single +character indicating the kind of change to make. All line numbers are +the original line numbers in the file. The types of change commands +are: + +@table @samp +@item @var{l}a +Add text from the second file after line @var{l} in the first file. For +example, @samp{8a} means to add the following lines after line 8 of file +1. + +@item @var{r}c +Replace the lines in range @var{r} in the first file with the following +lines. Like a combined add and delete, but more compact. For example, +@samp{5,7c} means change lines 5--7 of file 1 to read as the text file +2. + +@item @var{r}d +Delete the lines in range @var{r} from the first file. For example, +@samp{5,7d} means delete lines 5--7 of file 1. +@end table + +@node Forward ed +@subsection Forward @command{ed} Scripts +@cindex forward @command{ed} script output format + +@command{diff} can produce output that is like an @command{ed} script, but +with hunks in forward (front to back) order. The format of the commands +is also changed slightly: command characters precede the lines they +modify, spaces separate line numbers in ranges, and no attempt is made +to disambiguate hunk lines consisting of a single period. Like +@command{ed} format, forward @command{ed} format cannot represent incomplete +lines. + +Forward @command{ed} format is not very useful, because neither @command{ed} +nor @command{patch} can apply diffs in this format. It exists mainly for +compatibility with older versions of @command{diff}. Use the @option{-f} or +@option{--forward-ed} option to select it. + +@node RCS +@subsection @acronym{RCS} Scripts +@cindex @acronym{RCS} script output format + +The @acronym{RCS} output format is designed specifically for use by +the Revision Control System, which is a set of free programs used for +organizing different versions and systems of files. Use the +@option{-n} or @option{--rcs} option to select this output format. It +is like the forward @command{ed} format (@pxref{Forward ed}), but it +can represent arbitrary changes to the contents of a file because it +avoids the forward @command{ed} format's problems with lines +consisting of a single period and with incomplete lines. Instead of +ending text sections with a line consisting of a single period, each +command specifies the number of lines it affects; a combination of the +@samp{a} and @samp{d} commands are used instead of @samp{c}. Also, if +the second file ends in a changed incomplete line, then the output +also ends in an incomplete line. + +Here is the output of @samp{diff -n lao tzu} (@pxref{Sample +diff Input}, for the complete contents of the two files): + +@example +d1 2 +d4 1 +a4 2 +The named is the mother of all things. + +a11 3 +They both may be called deep and profound. +Deeper and more profound, +The door of all subtleties! +@end example + +@node If-then-else +@section Merging Files with If-then-else +@cindex merged output format +@cindex if-then-else output format +@cindex C if-then-else output format +@cindex @command{ifdef} output format + +You can use @command{diff} to merge two files of C source code. The output +of @command{diff} in this format contains all the lines of both files. +Lines common to both files are output just once; the differing parts are +separated by the C preprocessor directives @code{#ifdef @var{name}} or +@code{#ifndef @var{name}}, @code{#else}, and @code{#endif}. When +compiling the output, you select which version to use by either defining +or leaving undefined the macro @var{name}. + +To merge two files, use @command{diff} with the @option{-D @var{name}} or +@option{--ifdef=@var{name}} option. The argument @var{name} is the C +preprocessor identifier to use in the @code{#ifdef} and @code{#ifndef} +directives. + +For example, if you change an instance of @code{wait (&s)} to +@code{waitpid (-1, &s, 0)} and then merge the old and new files with +the @option{--ifdef=HAVE_WAITPID} option, then the affected part of your code +might look like this: + +@example + do @{ +#ifndef HAVE_WAITPID + if ((w = wait (&s)) < 0 && errno != EINTR) +#else /* HAVE_WAITPID */ + if ((w = waitpid (-1, &s, 0)) < 0 && errno != EINTR) +#endif /* HAVE_WAITPID */ + return w; + @} while (w != child); +@end example + +You can specify formats for languages other than C by using line group +formats and line formats, as described in the next sections. + +@menu +* Line Group Formats:: Formats for general if-then-else line groups. +* Line Formats:: Formats for each line in a line group. +* Example If-then-else:: Sample if-then-else format output. +* Detailed If-then-else:: A detailed description of if-then-else format. +@end menu + +@node Line Group Formats +@subsection Line Group Formats +@cindex line group formats +@cindex formats for if-then-else line groups + +Line group formats let you specify formats suitable for many +applications that allow if-then-else input, including programming +languages and text formatting languages. A line group format specifies +the output format for a contiguous group of similar lines. + +For example, the following command compares the TeX files @file{old} +and @file{new}, and outputs a merged file in which old regions are +surrounded by @samp{\begin@{em@}}-@samp{\end@{em@}} lines, and new +regions are surrounded by @samp{\begin@{bf@}}-@samp{\end@{bf@}} lines. + +@example +diff \ + --old-group-format='\begin@{em@} +%<\end@{em@} +' \ + --new-group-format='\begin@{bf@} +%>\end@{bf@} +' \ + old new +@end example + +The following command is equivalent to the above example, but it is a +little more verbose, because it spells out the default line group formats. + +@example +diff \ + --old-group-format='\begin@{em@} +%<\end@{em@} +' \ + --new-group-format='\begin@{bf@} +%>\end@{bf@} +' \ + --unchanged-group-format='%=' \ + --changed-group-format='\begin@{em@} +%<\end@{em@} +\begin@{bf@} +%>\end@{bf@} +' \ + old new +@end example + +Here is a more advanced example, which outputs a diff listing with +headers containing line numbers in a ``plain English'' style. + +@example +diff \ + --unchanged-group-format='' \ + --old-group-format='-------- %dn line%(n=1?:s) deleted at %df: +%<' \ + --new-group-format='-------- %dN line%(N=1?:s) added after %de: +%>' \ + --changed-group-format='-------- %dn line%(n=1?:s) changed at %df: +%<-------- to: +%>' \ + old new +@end example + +To specify a line group format, use @command{diff} with one of the options +listed below. You can specify up to four line group formats, one for +each kind of line group. You should quote @var{format}, because it +typically contains shell metacharacters. + +@table @option +@item --old-group-format=@var{format} +These line groups are hunks containing only lines from the first file. +The default old group format is the same as the changed group format if +it is specified; otherwise it is a format that outputs the line group as-is. + +@item --new-group-format=@var{format} +These line groups are hunks containing only lines from the second +file. The default new group format is same as the changed group +format if it is specified; otherwise it is a format that outputs the +line group as-is. + +@item --changed-group-format=@var{format} +These line groups are hunks containing lines from both files. The +default changed group format is the concatenation of the old and new +group formats. + +@item --unchanged-group-format=@var{format} +These line groups contain lines common to both files. The default +unchanged group format is a format that outputs the line group as-is. +@end table + +In a line group format, ordinary characters represent themselves; +conversion specifications start with @samp{%} and have one of the +following forms. + +@table @samp +@item %< +stands for the lines from the first file, including the trailing newline. +Each line is formatted according to the old line format (@pxref{Line Formats}). + +@item %> +stands for the lines from the second file, including the trailing newline. +Each line is formatted according to the new line format. + +@item %= +stands for the lines common to both files, including the trailing newline. +Each line is formatted according to the unchanged line format. + +@item %% +stands for @samp{%}. + +@item %c'@var{C}' +where @var{C} is a single character, stands for @var{C}. +@var{C} may not be a backslash or an apostrophe. +For example, @samp{%c':'} stands for a colon, even inside +the then-part of an if-then-else format, which a colon would +normally terminate. + +@item %c'\@var{O}' +where @var{O} is a string of 1, 2, or 3 octal digits, +stands for the character with octal code @var{O}. +For example, @samp{%c'\0'} stands for a null character. + +@item @var{F}@var{n} +where @var{F} is a @code{printf} conversion specification and @var{n} is one +of the following letters, stands for @var{n}'s value formatted with @var{F}. + +@table @samp +@item e +The line number of the line just before the group in the old file. + +@item f +The line number of the first line in the group in the old file; +equals @var{e} + 1. + +@item l +The line number of the last line in the group in the old file. + +@item m +The line number of the line just after the group in the old file; +equals @var{l} + 1. + +@item n +The number of lines in the group in the old file; equals @var{l} - @var{f} + 1. + +@item E, F, L, M, N +Likewise, for lines in the new file. + +@end table + +@vindex LC_NUMERIC +The @code{printf} conversion specification can be @samp{%d}, +@samp{%o}, @samp{%x}, or @samp{%X}, specifying decimal, octal, +lower case hexadecimal, or upper case hexadecimal output +respectively. After the @samp{%} the following options can appear in +sequence: a series of zero or more flags; an integer +specifying the minimum field width; and a period followed by an +optional integer specifying the minimum number of digits. +The flags are @samp{-} for left-justification, @samp{'} for separating +the digit into groups as specified by the @env{LC_NUMERIC} locale category, +and @samp{0} for padding with zeros instead of spaces. +For example, @samp{%5dN} prints the number of new lines in the group +in a field of width 5 characters, using the @code{printf} format @code{"%5d"}. + +@item (@var{A}=@var{B}?@var{T}:@var{E}) +If @var{A} equals @var{B} then @var{T} else @var{E}. +@var{A} and @var{B} are each either a decimal constant +or a single letter interpreted as above. +This format spec is equivalent to @var{T} if +@var{A}'s value equals @var{B}'s; otherwise it is equivalent to @var{E}. + +For example, @samp{%(N=0?no:%dN) line%(N=1?:s)} is equivalent to +@samp{no lines} if @var{N} (the number of lines in the group in the +new file) is 0, to @samp{1 line} if @var{N} is 1, and to @samp{%dN lines} +otherwise. +@end table + +@node Line Formats +@subsection Line Formats +@cindex line formats + +Line formats control how each line taken from an input file is +output as part of a line group in if-then-else format. + +For example, the following command outputs text with a one-character +change indicator to the left of the text. The first character of output +is @samp{-} for deleted lines, @samp{|} for added lines, and a space for +unchanged lines. The formats contain newline characters where newlines +are desired on output. + +@example +diff \ + --old-line-format='-%l +' \ + --new-line-format='|%l +' \ + --unchanged-line-format=' %l +' \ + old new +@end example + +To specify a line format, use one of the following options. You should +quote @var{format}, since it often contains shell metacharacters. + +@table @option +@item --old-line-format=@var{format} +formats lines just from the first file. + +@item --new-line-format=@var{format} +formats lines just from the second file. + +@item --unchanged-line-format=@var{format} +formats lines common to both files. + +@item --line-format=@var{format} +formats all lines; in effect, it sets all three above options simultaneously. +@end table + +In a line format, ordinary characters represent themselves; +conversion specifications start with @samp{%} and have one of the +following forms. + +@table @samp +@item %l +stands for the contents of the line, not counting its trailing +newline (if any). This format ignores whether the line is incomplete; +@xref{Incomplete Lines}. + +@item %L +stands for the contents of the line, including its trailing newline +(if any). If a line is incomplete, this format preserves its +incompleteness. + +@item %% +stands for @samp{%}. + +@item %c'@var{C}' +where @var{C} is a single character, stands for @var{C}. +@var{C} may not be a backslash or an apostrophe. +For example, @samp{%c':'} stands for a colon. + +@item %c'\@var{O}' +where @var{O} is a string of 1, 2, or 3 octal digits, +stands for the character with octal code @var{O}. +For example, @samp{%c'\0'} stands for a null character. + +@item @var{F}n +where @var{F} is a @code{printf} conversion specification, +stands for the line number formatted with @var{F}. +For example, @samp{%.5dn} prints the line number using the +@code{printf} format @code{"%.5d"}. @xref{Line Group Formats}, for +more about printf conversion specifications. + +@end table + +The default line format is @samp{%l} followed by a newline character. + +If the input contains tab characters and it is important that they line +up on output, you should ensure that @samp{%l} or @samp{%L} in a line +format is just after a tab stop (e.g.@: by preceding @samp{%l} or +@samp{%L} with a tab character), or you should use the @option{-t} or +@option{--expand-tabs} option. + +Taken together, the line and line group formats let you specify many +different formats. For example, the following command uses a format +similar to normal @command{diff} format. You can tailor this command +to get fine control over @command{diff} output. + +@example +diff \ + --old-line-format='< %l +' \ + --new-line-format='> %l +' \ + --old-group-format='%df%(f=l?:,%dl)d%dE +%<' \ + --new-group-format='%dea%dF%(F=L?:,%dL) +%>' \ + --changed-group-format='%df%(f=l?:,%dl)c%dF%(F=L?:,%dL) +%<--- +%>' \ + --unchanged-group-format='' \ + old new +@end example + +@node Example If-then-else +@subsection An Example of If-then-else Format + +Here is the output of @samp{diff -DTWO lao tzu} (@pxref{Sample +diff Input}, for the complete contents of the two files): + +@example +#ifndef TWO +The Way that can be told of is not the eternal Way; +The name that can be named is not the eternal name. +#endif /* ! TWO */ +The Nameless is the origin of Heaven and Earth; +#ifndef TWO +The Named is the mother of all things. +#else /* TWO */ +The named is the mother of all things. + +#endif /* TWO */ +Therefore let there always be non-being, + so we may see their subtlety, +And let there always be being, + so we may see their outcome. +The two are the same, +But after they are produced, + they have different names. +#ifdef TWO +They both may be called deep and profound. +Deeper and more profound, +The door of all subtleties! +#endif /* TWO */ +@end example + +@node Detailed If-then-else +@subsection Detailed Description of If-then-else Format + +For lines common to both files, @command{diff} uses the unchanged line +group format. For each hunk of differences in the merged output +format, if the hunk contains only lines from the first file, +@command{diff} uses the old line group format; if the hunk contains only +lines from the second file, @command{diff} uses the new group format; +otherwise, @command{diff} uses the changed group format. + +The old, new, and unchanged line formats specify the output format of +lines from the first file, lines from the second file, and lines common +to both files, respectively. + +The option @option{--ifdef=@var{name}} is equivalent to +the following sequence of options using shell syntax: + +@example +--old-group-format='#ifndef @var{name} +%<#endif /* ! @var{name} */ +' \ +--new-group-format='#ifdef @var{name} +%>#endif /* @var{name} */ +' \ +--unchanged-group-format='%=' \ +--changed-group-format='#ifndef @var{name} +%<#else /* @var{name} */ +%>#endif /* @var{name} */ +' +@end example + +You should carefully check the @command{diff} output for proper nesting. +For example, when using the @option{-D @var{name}} or +@option{--ifdef=@var{name}} option, you should check that if the +differing lines contain any of the C preprocessor directives +@samp{#ifdef}, @samp{#ifndef}, @samp{#else}, @samp{#elif}, or +@samp{#endif}, they are nested properly and match. If they don't, you +must make corrections manually. It is a good idea to carefully check +the resulting code anyway to make sure that it really does what you +want it to; depending on how the input files were produced, the output +might contain duplicate or otherwise incorrect code. + +The @command{patch} @option{-D @var{name}} option behaves like +the @command{diff} @option{-D @var{name}} option, except it operates on +a file and a diff to produce a merged file; @xref{patch Options}. + +@node Incomplete Lines +@chapter Incomplete Lines +@cindex incomplete lines +@cindex full lines +@cindex newline treatment by @command{diff} + +When an input file ends in a non-newline character, its last line is +called an @dfn{incomplete line} because its last character is not a +newline. All other lines are called @dfn{full lines} and end in a +newline character. Incomplete lines do not match full lines unless +differences in white space are ignored (@pxref{White Space}). + +An incomplete line is normally distinguished on output from a full +line by a following line that starts with @samp{\}. However, the +@acronym{RCS} format (@pxref{RCS}) outputs the incomplete line as-is, +without any trailing newline or following line. The side by side +format normally represents incomplete lines as-is, but in some cases +uses a @samp{\} or @samp{/} gutter marker; @xref{Side by Side}. The +if-then-else line format preserves a line's incompleteness with +@samp{%L}, and discards the newline with @samp{%l}; @xref{Line +Formats}. Finally, with the @command{ed} and forward @command{ed} +output formats (@pxref{Output Formats}) @command{diff} cannot +represent an incomplete line, so it pretends there was a newline and +reports an error. + +For example, suppose @file{F} and @file{G} are one-byte files that +contain just @samp{f} and @samp{g}, respectively. Then @samp{diff F G} +outputs + +@example +1c1 +< f +\ No newline at end of file +--- +> g +\ No newline at end of file +@end example + +@noindent +(The exact message may differ in non-English locales.) +@samp{diff -n F G} outputs the following without a trailing newline: + +@example +d1 1 +a1 1 +g +@end example + +@noindent +@samp{diff -e F G} reports two errors and outputs the following: + +@example +1c +g +. +@end example + +@node Comparing Directories +@chapter Comparing Directories + +@vindex LC_COLLATE +You can use @command{diff} to compare some or all of the files in two +directory trees. When both file name arguments to @command{diff} are +directories, it compares each file that is contained in both +directories, examining file names in alphabetical order as specified by +the @env{LC_COLLATE} locale category. Normally +@command{diff} is silent about pairs of files that contain no differences, +but if you use the @option{-s} or @option{--report-identical-files} option, +it reports pairs of identical files. Normally @command{diff} reports +subdirectories common to both directories without comparing +subdirectories' files, but if you use the @option{-r} or +@option{--recursive} option, it compares every corresponding pair of files +in the directory trees, as many levels deep as they go. + +For file names that are in only one of the directories, @command{diff} +normally does not show the contents of the file that exists; it reports +only that the file exists in that directory and not in the other. You +can make @command{diff} act as though the file existed but was empty in the +other directory, so that it outputs the entire contents of the file that +actually exists. (It is output as either an insertion or a +deletion, depending on whether it is in the first or the second +directory given.) To do this, use the @option{-N} or @option{--new-file} +option. + +If the older directory contains one or more large files that are not in +the newer directory, you can make the patch smaller by using the +@option{--unidirectional-new-file} option instead of @option{-N}. +This option is like @option{-N} except that it only inserts the contents +of files that appear in the second directory but not the first (that is, +files that were added). At the top of the patch, write instructions for +the user applying the patch to remove the files that were deleted before +applying the patch. @xref{Making Patches}, for more discussion of +making patches for distribution. + +To ignore some files while comparing directories, use the @option{-x +@var{pattern}} or @option{--exclude=@var{pattern}} option. This option +ignores any files or subdirectories whose base names match the shell +pattern @var{pattern}. Unlike in the shell, a period at the start of +the base of a file name matches a wildcard at the start of a pattern. +You should enclose @var{pattern} in quotes so that the shell does not +expand it. For example, the option @option{-x '*.[ao]'} ignores any file +whose name ends with @samp{.a} or @samp{.o}. + +This option accumulates if you specify it more than once. For example, +using the options @option{-x 'RCS' -x '*,v'} ignores any file or +subdirectory whose base name is @samp{RCS} or ends with @samp{,v}. + +If you need to give this option many times, you can instead put the +patterns in a file, one pattern per line, and use the @option{-X +@var{file}} or @option{--exclude-from=@var{file}} option. Trailing +white space and empty lines are ignored in the pattern file. + +If you have been comparing two directories and stopped partway through, +later you might want to continue where you left off. You can do this by +using the @option{-S @var{file}} or @option{--starting-file=@var{file}} +option. This compares only the file @var{file} and all alphabetically +later files in the topmost directory level. + +If two directories differ only in that file names are lower case in +one directory and upper case in the upper, @command{diff} normally +reports many differences because it compares file names in a +case sensitive way. With the @option{--ignore-file-name-case} option, +@command{diff} ignores case differences in file names, so that for example +the contents of the file @file{Tao} in one directory are compared to +the contents of the file @file{TAO} in the other. The +@option{--no-ignore-file-name-case} option cancels the effect of the +@option{--ignore-file-name-case} option, reverting to the default +behavior. + +If an @option{-x @var{pattern}} or @option{--exclude=@var{pattern}} +option, or an @option{-X @var{file}} or +@option{--exclude-from=@var{file}} option, +is specified while the @option{--ignore-file-name-case} option is in +effect, case is ignored when excluding file names matching the +specified patterns. + +@node Adjusting Output +@chapter Making @command{diff} Output Prettier + +@command{diff} provides several ways to adjust the appearance of its output. +These adjustments can be applied to any output format. + +@menu +* Tabs:: Preserving the alignment of tab stops. +* Pagination:: Page numbering and time-stamping @command{diff} output. +@end menu + +@node Tabs +@section Preserving Tab Stop Alignment +@cindex tab stop alignment +@cindex aligning tab stops + +The lines of text in some of the @command{diff} output formats are +preceded by one or two characters that indicate whether the text is +inserted, deleted, or changed. The addition of those characters can +cause tabs to move to the next tab stop, throwing off the alignment of +columns in the line. @acronym{GNU} @command{diff} provides two ways +to make tab-aligned columns line up correctly. + +The first way is to have @command{diff} convert all tabs into the correct +number of spaces before outputting them; select this method with the +@option{-t} or @option{--expand-tabs} option. To use this form of output with +@command{patch}, you must give @command{patch} the @option{-l} or +@option{--ignore-white-space} option (@pxref{Changed White Space}, for more +information). @command{diff} normally assumes that tab stops are set +every 8 print columns, but this can be altered by the +@option{--tabsize=@var{columns}} option. + +The other method for making tabs line up correctly is to add a tab +character instead of a space after the indicator character at the +beginning of the line. This ensures that all following tab characters +are in the same position relative to tab stops that they were in the +original files, so that the output is aligned correctly. Its +disadvantage is that it can make long lines too long to fit on one line +of the screen or the paper. It also does not work with the unified +output format, which does not have a space character after the change +type indicator character. Select this method with the @option{-T} or +@option{--initial-tab} option. + +@node Pagination +@section Paginating @command{diff} Output +@cindex paginating @command{diff} output + +It can be convenient to have long output page-numbered and time-stamped. +The @option{-l} or @option{--paginate} option does this by sending the +@command{diff} output through the @command{pr} program. Here is what the page +header might look like for @samp{diff -lc lao tzu}: + +@example +2002-02-22 14:20 diff -lc lao tzu Page 1 +@end example + +@node diff Performance +@chapter @command{diff} Performance Tradeoffs +@cindex performance of @command{diff} + +@acronym{GNU} @command{diff} runs quite efficiently; however, in some +circumstances you can cause it to run faster or produce a more compact +set of changes. + +One way to improve @command{diff} performance is to use hard or +symbolic links to files instead of copies. This improves performance +because @command{diff} normally does not need to read two hard or +symbolic links to the same file, since their contents must be +identical. For example, suppose you copy a large directory hierarchy, +make a few changes to the copy, and then often use @samp{diff -r} to +compare the original to the copy. If the original files are +read-only, you can greatly improve performance by creating the copy +using hard or symbolic links (e.g., with @acronym{GNU} @samp{cp -lR} or +@samp{cp -sR}). Before editing a file in the copy for the first time, +you should break the link and replace it with a regular copy. + +You can also affect the performance of @acronym{GNU} @command{diff} by +giving it options that change the way it compares files. +Performance has more than one dimension. These options improve one +aspect of performance at the cost of another, or they improve +performance in some cases while hurting it in others. + +The way that @acronym{GNU} @command{diff} determines which lines have +changed always comes up with a near-minimal set of differences. +Usually it is good enough for practical purposes. If the +@command{diff} output is large, you might want @command{diff} to use a +modified algorithm that sometimes produces a smaller set of +differences. The @option{-d} or @option{--minimal} option does this; +however, it can also cause @command{diff} to run more slowly than +usual, so it is not the default behavior. + +When the files you are comparing are large and have small groups of +changes scattered throughout them, you can use the +@option{--speed-large-files} option to make a different modification to +the algorithm that @command{diff} uses. If the input files have a constant +small density of changes, this option speeds up the comparisons without +changing the output. If not, @command{diff} might produce a larger set of +differences; however, the output will still be correct. + +Normally @command{diff} discards the prefix and suffix that is common to +both files before it attempts to find a minimal set of differences. +This makes @command{diff} run faster, but occasionally it may produce +non-minimal output. The @option{--horizon-lines=@var{lines}} option +prevents @command{diff} from discarding the last @var{lines} lines of the +prefix and the first @var{lines} lines of the suffix. This gives +@command{diff} further opportunities to find a minimal output. + +Suppose a run of changed lines includes a sequence of lines at one end +and there is an identical sequence of lines just outside the other end. +The @command{diff} command is free to choose which identical sequence is +included in the hunk. In this case, @command{diff} normally shifts the +hunk's boundaries when this merges adjacent hunks, or shifts a hunk's +lines towards the end of the file. Merging hunks can make the output +look nicer in some cases. + +@node Comparing Three Files +@chapter Comparing Three Files +@cindex comparing three files +@cindex format of @command{diff3} output + +Use the program @command{diff3} to compare three files and show any +differences among them. (@command{diff3} can also merge files; see +@ref{diff3 Merging}). + +The ``normal'' @command{diff3} output format shows each hunk of +differences without surrounding context. Hunks are labeled depending +on whether they are two-way or three-way, and lines are annotated by +their location in the input files. + +@xref{Invoking diff3}, for more information on how to run @command{diff3}. + +@menu +* Sample diff3 Input:: Sample @command{diff3} input for examples. +* Example diff3 Normal:: Sample output in the normal format. +* diff3 Hunks:: The format of normal output format. +* Detailed diff3 Normal:: A detailed description of normal output format. +@end menu + +@node Sample diff3 Input +@section A Third Sample Input File +@cindex @command{diff3} sample input +@cindex sample input for @command{diff3} + +Here is a third sample file that will be used in examples to illustrate +the output of @command{diff3} and how various options can change it. The +first two files are the same that we used for @command{diff} (@pxref{Sample +diff Input}). This is the third sample file, called @file{tao}: + +@example +The Way that can be told of is not the eternal Way; +The name that can be named is not the eternal name. +The Nameless is the origin of Heaven and Earth; +The named is the mother of all things. + +Therefore let there always be non-being, + so we may see their subtlety, +And let there always be being, + so we may see their result. +The two are the same, +But after they are produced, + they have different names. + + -- The Way of Lao-Tzu, tr. Wing-tsit Chan +@end example + +@node Example diff3 Normal +@section An Example of @command{diff3} Normal Format + +Here is the output of the command @samp{diff3 lao tzu tao} +(@pxref{Sample diff3 Input}, for the complete contents of the files). +Notice that it shows only the lines that are different among the three +files. + +@example +====2 +1:1,2c +3:1,2c + The Way that can be told of is not the eternal Way; + The name that can be named is not the eternal name. +2:0a +====1 +1:4c + The Named is the mother of all things. +2:2,3c +3:4,5c + The named is the mother of all things. + +====3 +1:8c +2:7c + so we may see their outcome. +3:9c + so we may see their result. +==== +1:11a +2:11,13c + They both may be called deep and profound. + Deeper and more profound, + The door of all subtleties! +3:13,14c + + -- The Way of Lao-Tzu, tr. Wing-tsit Chan +@end example + +@node Detailed diff3 Normal +@section Detailed Description of @command{diff3} Normal Format + +Each hunk begins with a line marked @samp{====}. Three-way hunks have +plain @samp{====} lines, and two-way hunks have @samp{1}, @samp{2}, or +@samp{3} appended to specify which of the three input files differ in +that hunk. The hunks contain copies of two or three sets of input +lines each preceded by one or two commands identifying where the lines +came from. + +Normally, two spaces precede each copy of an input line to distinguish +it from the commands. But with the @option{-T} or @option{--initial-tab} +option, @command{diff3} uses a tab instead of two spaces; this lines up +tabs correctly. @xref{Tabs}, for more information. + +Commands take the following forms: + +@table @samp +@item @var{file}:@var{l}a +This hunk appears after line @var{l} of file @var{file}, and +contains no lines in that file. To edit this file to yield the other +files, one must append hunk lines taken from the other files. For +example, @samp{1:11a} means that the hunk follows line 11 in the first +file and contains no lines from that file. + +@item @var{file}:@var{r}c +This hunk contains the lines in the range @var{r} of file @var{file}. +The range @var{r} is a comma-separated pair of line numbers, or just one +number if the range is a singleton. To edit this file to yield the +other files, one must change the specified lines to be the lines taken +from the other files. For example, @samp{2:11,13c} means that the hunk +contains lines 11 through 13 from the second file. +@end table + +If the last line in a set of input lines is incomplete +(@pxref{Incomplete Lines}), it is distinguished on output from a full +line by a following line that starts with @samp{\}. + +@node diff3 Hunks +@section @command{diff3} Hunks +@cindex hunks for @command{diff3} +@cindex @command{diff3} hunks + +Groups of lines that differ in two or three of the input files are +called @dfn{diff3 hunks}, by analogy with @command{diff} hunks +(@pxref{Hunks}). If all three input files differ in a @command{diff3} +hunk, the hunk is called a @dfn{three-way hunk}; if just two input files +differ, it is a @dfn{two-way hunk}. + +As with @command{diff}, several solutions are possible. When comparing the +files @samp{A}, @samp{B}, and @samp{C}, @command{diff3} normally finds +@command{diff3} hunks by merging the two-way hunks output by the two +commands @samp{diff A B} and @samp{diff A C}. This does not necessarily +minimize the size of the output, but exceptions should be rare. + +For example, suppose @file{F} contains the three lines @samp{a}, +@samp{b}, @samp{f}, @file{G} contains the lines @samp{g}, @samp{b}, +@samp{g}, and @file{H} contains the lines @samp{a}, @samp{b}, +@samp{h}. @samp{diff3 F G H} might output the following: + +@example +====2 +1:1c +3:1c + a +2:1c + g +==== +1:3c + f +2:3c + g +3:3c + h +@end example + +@noindent +because it found a two-way hunk containing @samp{a} in the first and +third files and @samp{g} in the second file, then the single line +@samp{b} common to all three files, then a three-way hunk containing +the last line of each file. + +@node diff3 Merging +@chapter Merging From a Common Ancestor +@cindex merging from a common ancestor + +When two people have made changes to copies of the same file, +@command{diff3} can produce a merged output that contains both sets of +changes together with warnings about conflicts. + +One might imagine programs with names like @command{diff4} and @command{diff5} +to compare more than three files simultaneously, but in practice the +need rarely arises. You can use @command{diff3} to merge three or more +sets of changes to a file by merging two change sets at a time. + +@command{diff3} can incorporate changes from two modified versions into a +common preceding version. This lets you merge the sets of changes +represented by the two newer files. Specify the common ancestor version +as the second argument and the two newer versions as the first and third +arguments, like this: + +@example +diff3 @var{mine} @var{older} @var{yours} +@end example + +@noindent +You can remember the order of the arguments by noting that they are in +alphabetical order. + +@cindex conflict +@cindex overlap +You can think of this as subtracting @var{older} from @var{yours} and +adding the result to @var{mine}, or as merging into @var{mine} the +changes that would turn @var{older} into @var{yours}. This merging is +well-defined as long as @var{mine} and @var{older} match in the +neighborhood of each such change. This fails to be true when all three +input files differ or when only @var{older} differs; we call this +a @dfn{conflict}. When all three input files differ, we call the +conflict an @dfn{overlap}. + +@command{diff3} gives you several ways to handle overlaps and conflicts. +You can omit overlaps or conflicts, or select only overlaps, +or mark conflicts with special @samp{<<<<<<<} and @samp{>>>>>>>} lines. + +@command{diff3} can output the merge results as an @command{ed} script that +that can be applied to the first file to yield the merged output. +However, it is usually better to have @command{diff3} generate the merged +output directly; this bypasses some problems with @command{ed}. + +@menu +* Which Changes:: Selecting changes to incorporate. +* Marking Conflicts:: Marking conflicts. +* Bypassing ed:: Generating merged output directly. +* Merging Incomplete Lines:: How @command{diff3} merges incomplete lines. +* Saving the Changed File:: Emulating System V behavior. +@end menu + +@node Which Changes +@section Selecting Which Changes to Incorporate +@cindex overlapping change, selection of +@cindex unmerged change + +You can select all unmerged changes from @var{older} to @var{yours} for merging +into @var{mine} with the @option{-e} or @option{--ed} option. You can +select only the nonoverlapping unmerged changes with @option{-3} or +@option{--easy-only}, and you can select only the overlapping changes with +@option{-x} or @option{--overlap-only}. + +The @option{-e}, @option{-3} and @option{-x} options select only +@dfn{unmerged changes}, i.e.@: changes where @var{mine} and @var{yours} +differ; they ignore changes from @var{older} to @var{yours} where +@var{mine} and @var{yours} are identical, because they assume that such +changes have already been merged. If this assumption is not a safe +one, you can use the @option{-A} or @option{--show-all} option +(@pxref{Marking Conflicts}). + +Here is the output of the command @command{diff3} with each of these three +options (@pxref{Sample diff3 Input}, for the complete contents of the files). +Notice that @option{-e} outputs the union of the disjoint sets of changes +output by @option{-3} and @option{-x}. + +Output of @samp{diff3 -e lao tzu tao}: +@example +11a + + -- The Way of Lao-Tzu, tr. Wing-tsit Chan +. +8c + so we may see their result. +. +@end example + +Output of @samp{diff3 -3 lao tzu tao}: +@example +8c + so we may see their result. +. +@end example + +Output of @samp{diff3 -x lao tzu tao}: +@example +11a + + -- The Way of Lao-Tzu, tr. Wing-tsit Chan +. +@end example + +@node Marking Conflicts +@section Marking Conflicts +@cindex conflict marking +@cindex @samp{<<<<<<<} for marking conflicts + +@command{diff3} can mark conflicts in the merged output by +bracketing them with special marker lines. A conflict +that comes from two files @var{A} and @var{B} is marked as follows: + +@example +<<<<<<< @var{A} +@r{lines from @var{A}} +======= +@r{lines from @var{B}} +>>>>>>> @var{B} +@end example + +A conflict that comes from three files @var{A}, @var{B} and @var{C} is +marked as follows: + +@example +<<<<<<< @var{A} +@r{lines from @var{A}} +||||||| @var{B} +@r{lines from @var{B}} +======= +@r{lines from @var{C}} +>>>>>>> @var{C} +@end example + +The @option{-A} or @option{--show-all} option acts like the @option{-e} +option, except that it brackets conflicts, and it outputs all changes +from @var{older} to @var{yours}, not just the unmerged changes. Thus, +given the sample input files (@pxref{Sample diff3 Input}), @samp{diff3 +-A lao tzu tao} puts brackets around the conflict where only @file{tzu} +differs: + +@example +<<<<<<< tzu +======= +The Way that can be told of is not the eternal Way; +The name that can be named is not the eternal name. +>>>>>>> tao +@end example + +And it outputs the three-way conflict as follows: + +@example +<<<<<<< lao +||||||| tzu +They both may be called deep and profound. +Deeper and more profound, +The door of all subtleties! +======= + + -- The Way of Lao-Tzu, tr. Wing-tsit Chan +>>>>>>> tao +@end example + +The @option{-E} or @option{--show-overlap} option outputs less information +than the @option{-A} or @option{--show-all} option, because it outputs only +unmerged changes, and it never outputs the contents of the second +file. Thus the @option{-E} option acts like the @option{-e} option, +except that it brackets the first and third files from three-way +overlapping changes. Similarly, @option{-X} acts like @option{-x}, except +it brackets all its (necessarily overlapping) changes. For example, +for the three-way overlapping change above, the @option{-E} and @option{-X} +options output the following: + +@example +<<<<<<< lao +======= + + -- The Way of Lao-Tzu, tr. Wing-tsit Chan +>>>>>>> tao +@end example + +If you are comparing files that have meaningless or uninformative names, +you can use the @option{--label=@var{label}} +option to show alternate names in the @samp{<<<<<<<}, @samp{|||||||} +and @samp{>>>>>>>} brackets. This option can be given up to three +times, once for each input file. Thus @samp{diff3 -A --label X +--label Y --label Z A +B C} acts like @samp{diff3 -A A B C}, except that the output looks like +it came from files named @samp{X}, @samp{Y} and @samp{Z} rather than +from files named @samp{A}, @samp{B} and @samp{C}. + +@node Bypassing ed +@section Generating the Merged Output Directly +@cindex merged @command{diff3} format + +With the @option{-m} or @option{--merge} option, @command{diff3} outputs the +merged file directly. This is more efficient than using @command{ed} to +generate it, and works even with non-text files that @command{ed} would +reject. If you specify @option{-m} without an @command{ed} script option, +@option{-A} is assumed. + +For example, the command @samp{diff3 -m lao tzu tao} +(@pxref{Sample diff3 Input} for a copy of the input files) would output +the following: + +@example +<<<<<<< tzu +======= +The Way that can be told of is not the eternal Way; +The name that can be named is not the eternal name. +>>>>>>> tao +The Nameless is the origin of Heaven and Earth; +The Named is the mother of all things. +Therefore let there always be non-being, + so we may see their subtlety, +And let there always be being, + so we may see their result. +The two are the same, +But after they are produced, + they have different names. +<<<<<<< lao +||||||| tzu +They both may be called deep and profound. +Deeper and more profound, +The door of all subtleties! +======= + + -- The Way of Lao-Tzu, tr. Wing-tsit Chan +>>>>>>> tao +@end example + +@node Merging Incomplete Lines +@section How @command{diff3} Merges Incomplete Lines +@cindex incomplete line merging + +With @option{-m}, incomplete lines (@pxref{Incomplete Lines}) are simply +copied to the output as they are found; if the merged output ends in an +conflict and one of the input files ends in an incomplete +line, succeeding @samp{|||||||}, @samp{=======} or @samp{>>>>>>>} +brackets appear somewhere other than the start of a line because +they are appended to the incomplete line. + +Without @option{-m}, if an @command{ed} script option is specified and an +incomplete line is found, @command{diff3} generates a warning and acts as +if a newline had been present. + +@node Saving the Changed File +@section Saving the Changed File +@cindex System V @command{diff3} compatibility + +Traditional Unix @command{diff3} generates an @command{ed} script without the +trailing @samp{w} and @samp{q} commands that save the changes. +System V @command{diff3} generates these extra commands. @acronym{GNU} +@command{diff3} normally behaves like traditional Unix +@command{diff3}, but with the @option{-i} option it behaves like +System V @command{diff3} and appends the @samp{w} and @samp{q} +commands. + +The @option{-i} option requires one of the @command{ed} script options +@option{-AeExX3}, and is incompatible with the merged output option +@option{-m}. + +@node Interactive Merging +@chapter Interactive Merging with @command{sdiff} +@cindex diff merging +@cindex interactive merging + +With @command{sdiff}, you can merge two files interactively based on a +side-by-side @option{-y} format comparison (@pxref{Side by Side}). Use +@option{-o @var{file}} or @option{--output=@var{file}} to specify where to +put the merged text. @xref{Invoking sdiff}, for more details on the +options to @command{sdiff}. + +Another way to merge files interactively is to use the Emacs Lisp +package @command{emerge}. @xref{emerge, , emerge, emacs, The +@acronym{GNU} Emacs Manual}, for more information. + +@menu +* sdiff Option Summary:: Summary of @command{sdiff} options. +* Merge Commands:: Merging two files interactively. +@end menu + +@node sdiff Option Summary +@section Specifying @command{diff} Options to @command{sdiff} +@cindex @command{sdiff} output format + +The following @command{sdiff} options have the same meaning as for +@command{diff}. @xref{diff Options}, for the use of these options. + +@example +-a -b -d -i -t -v +-B -E -I @var{regexp} + +--expand-tabs +--ignore-blank-lines --ignore-case +--ignore-matching-lines=@var{regexp} --ignore-space-change +--ignore-tab-expansion +--left-column --minimal --speed-large-files +--strip-trailing-cr --suppress-common-lines +--tabsize=@var{columns} --text --version --width=@var{columns} +@end example + +For historical reasons, @command{sdiff} has alternate names for some +options. The @option{-l} option is equivalent to the +@option{--left-column} option, and similarly @option{-s} is equivalent +to @option{--suppress-common-lines}. The meaning of the @command{sdiff} +@option{-w} and @option{-W} options is interchanged from that of +@command{diff}: with @command{sdiff}, @option{-w @var{columns}} is +equivalent to @option{--width=@var{columns}}, and @option{-W} is +equivalent to @option{--ignore-all-space}. @command{sdiff} without the +@option{-o} option is equivalent to @command{diff} with the @option{-y} +or @option{--side-by-side} option (@pxref{Side by Side}). + +@node Merge Commands +@section Merge Commands +@cindex merge commands +@cindex merging interactively + +Groups of common lines, with a blank gutter, are copied from the first +file to the output. After each group of differing lines, @command{sdiff} +prompts with @samp{%} and pauses, waiting for one of the following +commands. Follow each command with @key{RET}. + +@table @samp +@item e +Discard both versions. +Invoke a text editor on an empty temporary file, +then copy the resulting file to the output. + +@item eb +Concatenate the two versions, edit the result in a temporary file, +then copy the edited result to the output. + +@item ed +Like @samp{eb}, except precede each version with a header that +shows what file and lines the version came from. + +@item el +Edit a copy of the left version, then copy the result to the output. + +@item er +Edit a copy of the right version, then copy the result to the output. + +@item l +Copy the left version to the output. + +@item q +Quit. + +@item r +Copy the right version to the output. + +@item s +Silently copy common lines. + +@item v +Verbosely copy common lines. This is the default. +@end table + +@vindex EDITOR +The text editor invoked is specified by the @env{EDITOR} environment +variable if it is set. The default is system-dependent. + +@node Merging with patch +@chapter Merging with @command{patch} + +@command{patch} takes comparison output produced by @command{diff} and applies +the differences to a copy of the original file, producing a patched +version. With @command{patch}, you can distribute just the changes to a +set of files instead of distributing the entire file set; your +correspondents can apply @command{patch} to update their copy of the files +with your changes. @command{patch} automatically determines the diff +format, skips any leading or trailing headers, and uses the headers to +determine which file to patch. This lets your correspondents feed a +mail message containing a difference listing directly to +@command{patch}. + +@command{patch} detects and warns about common problems like forward +patches. It saves any patches that it could not apply. It can also maintain a +@code{patchlevel.h} file to ensure that your correspondents apply +diffs in the proper order. + +@command{patch} accepts a series of diffs in its standard input, usually +separated by headers that specify which file to patch. It applies +@command{diff} hunks (@pxref{Hunks}) one by one. If a hunk does not +exactly match the original file, @command{patch} uses heuristics to try to +patch the file as well as it can. If no approximate match can be found, +@command{patch} rejects the hunk and skips to the next hunk. @command{patch} +normally replaces each file @var{f} with its new version, putting reject +hunks (if any) into @samp{@var{f}.rej}. + +@xref{Invoking patch}, for detailed information on the options to +@command{patch}. + +@menu +* patch Input:: Selecting the type of @command{patch} input. +* Revision Control:: Getting files from @acronym{RCS}, @acronym{SCCS}, etc. +* Imperfect:: Dealing with imperfect patches. +* Creating and Removing:: Creating and removing files with a patch. +* Patching Time Stamps:: Updating time stamps on patched files. +* Multiple Patches:: Handling multiple patches in a file. +* patch Directories:: Changing directory and stripping directories. +* Backups:: Whether backup files are made. +* Backup Names:: Backup file names. +* Reject Names:: Reject file names. +* patch Messages:: Messages and questions @command{patch} can produce. +* patch and POSIX:: Conformance to the @acronym{POSIX} standard. +* patch and Tradition:: @acronym{GNU} versus traditional @command{patch}. +@end menu + +@node patch Input +@section Selecting the @command{patch} Input Format +@cindex @command{patch} input format + +@command{patch} normally determines which @command{diff} format the patch +file uses by examining its contents. For patch files that contain +particularly confusing leading text, you might need to use one of the +following options to force @command{patch} to interpret the patch file as a +certain format of diff. The output formats listed here are the only +ones that @command{patch} can understand. + +@table @option +@item -c +@itemx --context +context diff. + +@item -e +@itemx --ed +@command{ed} script. + +@item -n +@itemx --normal +normal diff. + +@item -u +@itemx --unified +unified diff. +@end table + +@node Revision Control +@section Revision Control +@cindex revision control +@cindex version control +@cindex @acronym{RCS} +@cindex ClearCase +@cindex @acronym{SCCS} + +If a nonexistent input file is under a revision control system +supported by @command{patch}, @command{patch} normally asks the user +whether to get (or check out) the file from the revision control +system. Patch currently supports @acronym{RCS}, ClearCase and +@acronym{SCCS}. Under @acronym{RCS} and @acronym{SCCS}, +@command{patch} also asks when the input file is read-only and matches +the default version in the revision control system. + +@vindex PATCH_GET +The @option{-g @var{num}} or @option{--get=@var{num}} option affects access +to files under supported revision control systems. If @var{num} is +positive, @command{patch} gets the file without asking the user; if +zero, @command{patch} neither asks the user nor gets the file; and if +negative, @command{patch} asks the user before getting the file. The +default value of @var{num} is given by the value of the +@env{PATCH_GET} environment variable if it is set; if not, the default +value is zero if @command{patch} is conforming to @acronym{POSIX}, negative +otherwise. @xref{patch and POSIX}. + +@vindex VERSION_CONTROL +The choice of revision control system is unaffected by the +@env{VERSION_CONTROL} environment variable (@pxref{Backup Names}). + +@node Imperfect +@section Applying Imperfect Patches +@cindex imperfect patch application + +@command{patch} tries to skip any leading text in the patch file, +apply the diff, and then skip any trailing text. Thus you can feed a +mail message directly to @command{patch}, and it should work. If the +entire diff is indented by a constant amount of white space, +@command{patch} automatically ignores the indentation. If a context +diff contains trailing carriage return on each line, @command{patch} +automatically ignores the carriage return. If a context diff has been +encapsulated by prepending @w{@samp{- }} to lines beginning with @samp{-} +as per @uref{ftp://ftp.isi.edu/in-notes/rfc934.txt, Internet RFC 934}, +@command{patch} automatically unencapsulates the input. + +However, certain other types of imperfect input require user +intervention or testing. + +@menu +* Changed White Space:: When tabs and spaces don't match exactly. +* Reversed Patches:: Applying reversed patches correctly. +* Inexact:: Helping @command{patch} find close matches. +* Dry Runs:: Predicting what @command{patch} will do. +@end menu + +@node Changed White Space +@subsection Applying Patches with Changed White Space +@cindex white space in patches + +Sometimes mailers, editors, or other programs change spaces into tabs, +or vice versa. If this happens to a patch file or an input file, the +files might look the same, but @command{patch} will not be able to match +them properly. If this problem occurs, use the @option{-l} or +@option{--ignore-white-space} option, which makes @command{patch} compare +blank characters (i.e.@: spaces and tabs) loosely so that any nonempty +sequence of blanks in the patch file matches any nonempty sequence of +blanks in the input files. Non-blank +characters must still match exactly. Each line of the context must +still match a line in the input file. + +@node Reversed Patches +@subsection Applying Reversed Patches +@cindex reversed patches + +Sometimes people run @command{diff} with the new file first instead of +second. This creates a diff that is ``reversed''. To apply such +patches, give @command{patch} the @option{-R} or @option{--reverse} option. +@command{patch} then attempts to swap each hunk around before applying it. +Rejects come out in the swapped format. + +Often @command{patch} can guess that the patch is reversed. If the first +hunk of a patch fails, @command{patch} reverses the hunk to see if it can +apply it that way. If it can, @command{patch} asks you if you want to have +the @option{-R} option set; if it can't, @command{patch} continues to apply +the patch normally. This method cannot detect a reversed patch if it is +a normal diff and the first command is an append (which should have been +a delete) since appends always succeed, because a null context matches +anywhere. But most patches add or change lines rather than delete them, +so most reversed normal diffs begin with a delete, which fails, and +@command{patch} notices. + +If you apply a patch that you have already applied, @command{patch} thinks +it is a reversed patch and offers to un-apply the patch. This could be +construed as a feature. If you did this inadvertently and you don't +want to un-apply the patch, just answer @samp{n} to this offer and to +the subsequent ``apply anyway'' question---or type @kbd{C-c} to kill the +@command{patch} process. + +@node Inexact +@subsection Helping @command{patch} Find Inexact Matches +@cindex inexact patches +@cindex fuzz factor when patching + +For context diffs, and to a lesser extent normal diffs, @command{patch} can +detect when the line numbers mentioned in the patch are incorrect, and +it attempts to find the correct place to apply each hunk of the patch. +As a first guess, it takes the line number mentioned in the hunk, plus +or minus any offset used in applying the previous hunk. If that is not +the correct place, @command{patch} scans both forward and backward for a +set of lines matching the context given in the hunk. + +First @command{patch} looks for a place where all lines of the context +match. If it cannot find such a place, and it is reading a context or +unified diff, and the maximum fuzz factor is set to 1 or more, then +@command{patch} makes another scan, ignoring the first and last line of +context. If that fails, and the maximum fuzz factor is set to 2 or +more, it makes another scan, ignoring the first two and last two lines +of context are ignored. It continues similarly if the maximum fuzz +factor is larger. + +The @option{-F @var{lines}} or @option{--fuzz=@var{lines}} option sets the +maximum fuzz factor to @var{lines}. This option only applies to context +and unified diffs; it ignores up to @var{lines} lines while looking for +the place to install a hunk. Note that a larger fuzz factor increases +the odds of making a faulty patch. The default fuzz factor is 2; there +is no point to setting it to more than the number of lines of context +in the diff, ordinarily 3. + +If @command{patch} cannot find a place to install a hunk of the patch, it +writes the hunk out to a reject file (@pxref{Reject Names}, for information +on how reject files are named). It writes out rejected hunks in context +format no matter what form the input patch is in. If the input is a +normal or @command{ed} diff, many of the contexts are simply null. The +line numbers on the hunks in the reject file may be different from those +in the patch file: they show the approximate location where @command{patch} +thinks the failed hunks belong in the new file rather than in the old +one. + +If the @option{--verbose} option is given, then +as it completes each hunk @command{patch} tells you whether the hunk +succeeded or failed, and if it failed, on which line (in the new file) +@command{patch} thinks the hunk should go. If this is different from the +line number specified in the diff, it tells you the offset. A single +large offset @emph{may} indicate that @command{patch} installed a hunk in +the wrong place. @command{patch} also tells you if it used a fuzz factor +to make the match, in which case you should also be slightly suspicious. + +@command{patch} cannot tell if the line numbers are off in an @command{ed} +script, and can only detect wrong line numbers in a normal diff when it +finds a change or delete command. It may have the same problem with a +context diff using a fuzz factor equal to or greater than the number of +lines of context shown in the diff (typically 3). In these cases, you +should probably look at a context diff between your original and patched +input files to see if the changes make sense. Compiling without errors +is a pretty good indication that the patch worked, but not a guarantee. + +A patch against an empty file applies to a nonexistent file, and vice +versa. @xref{Creating and Removing}. + +@command{patch} usually produces the correct results, even when it must +make many guesses. However, the results are guaranteed only when +the patch is applied to an exact copy of the file that the patch was +generated from. + +@node Dry Runs +@subsection Predicting what @command{patch} will do +@cindex testing @command{patch} +@cindex dry runs for @command{patch} + +It may not be obvious in advance what @command{patch} will do with a +complicated or poorly formatted patch. If you are concerned that the +input might cause @command{patch} to modify the wrong files, you can +use the @option{--dry-run} option, which causes @command{patch} to +print the results of applying patches without actually changing any +files. You can then inspect the diagnostics generated by the dry run +to see whether @command{patch} will modify the files that you expect. +If the patch does not do what you want, you can modify the patch (or +the other options to @command{patch}) and try another dry run. Once +you are satisfied with the proposed patch you can apply it by invoking +@command{patch} as before, but this time without the +@option{--dry-run} option. + +@node Creating and Removing +@section Creating and Removing Files +@cindex creating files +@cindex empty files, removing +@cindex removing empty files + +Sometimes when comparing two directories, a file may exist in one +directory but not the other. If you give @command{diff} the +@option{-N} or @option{--new-file} option, or if you supply an old or +new file that is named @file{/dev/null} or is empty and is dated the +Epoch (1970-01-01 00:00:00 UTC), @command{diff} outputs a patch that +adds or deletes the contents of this file. When given such a patch, +@command{patch} normally creates a new file or removes the old file. +However, when conforming to @acronym{POSIX} (@pxref{patch and POSIX}), +@command{patch} does not remove the old file, but leaves it empty. +The @option{-E} or @option{--remove-empty-files} option causes +@command{patch} to remove output files that are empty after applying a +patch, even if the patch does not appear to be one that removed the +file. + +If the patch appears to create a file that already exists, +@command{patch} asks for confirmation before applying the patch. + +@node Patching Time Stamps +@section Updating Time Stamps on Patched Files +@cindex time stamps on patched files + +When @command{patch} updates a file, it normally sets the file's +last-modified time stamp to the current time of day. If you are using +@command{patch} to track a software distribution, this can cause +@command{make} to incorrectly conclude that a patched file is out of +date. For example, if @file{syntax.c} depends on @file{syntax.y}, and +@command{patch} updates @file{syntax.c} and then @file{syntax.y}, then +@file{syntax.c} will normally appear to be out of date with respect to +@file{syntax.y} even though its contents are actually up to date. + +The @option{-Z} or @option{--set-utc} option causes @command{patch} to +set a patched file's modification and access times to the time stamps +given in context diff headers. If the context diff headers do not +specify a time zone, they are assumed to use Coordinated Universal +Time (@acronym{UTC}, often known as @acronym{GMT}). + +The @option{-T} or @option{--set-time} option acts like @option{-Z} or +@option{--set-utc}, except that it assumes that the context diff +headers' time stamps use local time instead of @acronym{UTC}. This option +is not recommended, because patches using local time cannot easily be +used by people in other time zones, and because local time stamps are +ambiguous when local clocks move backwards during daylight-saving time +adjustments. If the context diff headers specify a time zone, this +option is equivalent to @option{-Z} or @option{--set-utc}. + +@command{patch} normally refrains from setting a file's time stamps if +the file's original last-modified time stamp does not match the time +given in the diff header, of if the file's contents do not exactly +match the patch. However, if the @option{-f} or @option{--force} +option is given, the file's time stamps are set regardless. + +Due to the limitations of the current @command{diff} format, +@command{patch} cannot update the times of files whose contents have +not changed. Also, if you set file time stamps to values other than +the current time of day, you should also remove (e.g., with @samp{make +clean}) all files that depend on the patched files, so that later +invocations of @command{make} do not get confused by the patched +files' times. + +@node Multiple Patches +@section Multiple Patches in a File +@cindex multiple patches +@cindex intuiting file names from patches + +If the patch file contains more than one patch, and if you do not +specify an input file on the command line, @command{patch} tries to +apply each patch as if they came from separate patch files. This +means that it determines the name of the file to patch for each patch, +and that it examines the leading text before each patch for file names +and prerequisite revision level (@pxref{Making Patches}, for more on +that topic). + +@command{patch} uses the following rules to intuit a file name from +the leading text before a patch. First, @command{patch} takes an +ordered list of candidate file names as follows: + +@itemize @bullet +@item +If the header is that of a context diff, @command{patch} takes the old +and new file names in the header. A name is ignored if it does not +have enough slashes to satisfy the @option{-p@var{num}} or +@option{--strip=@var{num}} option. The name @file{/dev/null} is also +ignored. + +@item +If there is an @samp{Index:} line in the leading garbage and if either +the old and new names are both absent or if @command{patch} is +conforming to @acronym{POSIX}, @command{patch} takes the name in the +@samp{Index:} line. + +@item +For the purpose of the following rules, the candidate file names are +considered to be in the order (old, new, index), regardless of the +order that they appear in the header. +@end itemize + +@noindent +Then @command{patch} selects a file name from the candidate list as +follows: + +@itemize @bullet +@item +If some of the named files exist, @command{patch} selects the first +name if conforming to @acronym{POSIX}, and the best name otherwise. + +@item +If @command{patch} is not ignoring @acronym{RCS}, ClearCase, and @acronym{SCCS} +(@pxref{Revision Control}), and no named files exist but an @acronym{RCS}, +ClearCase, or @acronym{SCCS} master is found, @command{patch} selects the +first named file with an @acronym{RCS}, ClearCase, or @acronym{SCCS} master. + +@item +If no named files exist, no @acronym{RCS}, ClearCase, or @acronym{SCCS} master +was found, some names are given, @command{patch} is not conforming to +@acronym{POSIX}, and the patch appears to create a file, @command{patch} +selects the best name requiring the creation of the fewest +directories. + +@item +If no file name results from the above heuristics, you are asked for +the name of the file to patch, and @command{patch} selects that name. +@end itemize + +To determine the @dfn{best} of a nonempty list of file names, +@command{patch} first takes all the names with the fewest path name +components; of those, it then takes all the names with the shortest +basename; of those, it then takes all the shortest names; finally, it +takes the first remaining name. + +@xref{patch and POSIX}, to see whether @command{patch} is conforming +to @acronym{POSIX}. + +@node patch Directories +@section Applying Patches in Other Directories +@cindex directories and patch +@cindex patching directories + +The @option{-d @var{directory}} or @option{--directory=@var{directory}} +option to @command{patch} makes directory @var{directory} the current +directory for interpreting both file names in the patch file, and file +names given as arguments to other options (such as @option{-B} and +@option{-o}). For example, while in a mail reading program, you can patch +a file in the @file{/usr/src/emacs} directory directly from a message +containing the patch like this: + +@example +| patch -d /usr/src/emacs +@end example + +Sometimes the file names given in a patch contain leading directories, +but you keep your files in a directory different from the one given in +the patch. In those cases, you can use the +@option{-p@var{number}} or @option{--strip=@var{number}} +option to set the file name strip count to @var{number}. The strip +count tells @command{patch} how many slashes, along with the directory +names between them, to strip from the front of file names. A sequence +of one or more adjacent slashes is counted as a single slash. By +default, @command{patch} strips off all leading directories, leaving +just the base file names. + +For example, suppose the file name in the patch file is +@file{/gnu/src/emacs/etc/NEWS}. Using @option{-p0} gives the +entire file name unmodified, @option{-p1} gives +@file{gnu/src/emacs/etc/NEWS} (no leading slash), @option{-p4} gives +@file{etc/NEWS}, and not specifying @option{-p} at all gives @file{NEWS}. + +@command{patch} looks for each file (after any slashes have been stripped) +in the current directory, or if you used the @option{-d @var{directory}} +option, in that directory. + +@node Backups +@section Backup Files +@cindex backup file strategy + +Normally, @command{patch} creates a backup file if the patch does not +exactly match the original input file, because in that case the +original data might not be recovered if you undo the patch with +@samp{patch -R} (@pxref{Reversed Patches}). However, when conforming +to @acronym{POSIX}, @command{patch} does not create backup files by +default. @xref{patch and POSIX}. + +The @option{-b} or @option{--backup} option causes @command{patch} to +make a backup file regardless of whether the patch matches the +original input. The @option{--backup-if-mismatch} option causes +@command{patch} to create backup files for mismatches files; this is +the default when not conforming to @acronym{POSIX}. The +@option{--no-backup-if-mismatch} option causes @command{patch} to not +create backup files, even for mismatched patches; this is the default +when conforming to @acronym{POSIX}. + +When backing up a file that does not exist, an empty, unreadable +backup file is created as a placeholder to represent the nonexistent +file. + +@node Backup Names +@section Backup File Names +@cindex backup file names + +Normally, @command{patch} renames an original input file into a backup +file by appending to its name the extension @samp{.orig}, or @samp{~} +if using @samp{.orig} would make the backup file name too +long.@footnote{A coding error in @acronym{GNU} @command{patch} version +2.5.4 causes it to always use @samp{~}, but this should be fixed in +the next release.} The @option{-z @var{backup-suffix}} or +@option{--suffix=@var{backup-suffix}} option causes @command{patch} to +use @var{backup-suffix} as the backup extension instead. + +@vindex SIMPLE_BACKUP_SUFFIX +Alternately, you can specify the extension for backup files with the +@env{SIMPLE_BACKUP_SUFFIX} environment variable, which the options +override. + +@command{patch} can also create numbered backup files the way +@acronym{GNU} Emacs does. With this method, instead of having a +single backup of each file, @command{patch} makes a new backup file +name each time it patches a file. For example, the backups of a file +named @file{sink} would be called, successively, @file{sink.~1~}, +@file{sink.~2~}, @file{sink.~3~}, etc. + +@vindex PATCH_VERSION_CONTROL +@vindex VERSION_CONTROL +The @option{-V @var{backup-style}} or +@option{--version-control=@var{backup-style}} option takes as an +argument a method for creating backup file names. You can alternately +control the type of backups that @command{patch} makes with the +@env{PATCH_VERSION_CONTROL} environment variable, which the +@option{-V} option overrides. If @env{PATCH_VERSION_CONTROL} is not +set, the @env{VERSION_CONTROL} environment variable is used instead. +Please note that these options and variables control backup file +names; they do not affect the choice of revision control system +(@pxref{Revision Control}). + +The values of these environment variables and the argument to the +@option{-V} option are like the @acronym{GNU} Emacs @code{version-control} +variable (@pxref{Backup Names, , , emacs, The @acronym{GNU} Emacs Manual}, +for more information on backup versions in Emacs). They also +recognize synonyms that are more descriptive. The valid values are +listed below; unique abbreviations are acceptable. + +@table @option +@item t +@itemx numbered +Always make numbered backups. + +@item nil +@itemx existing +Make numbered backups of files that already have them, simple backups of +the others. This is the default. + +@item never +@itemx simple +Always make simple backups. +@end table + +You can also tell @command{patch} to prepend a prefix, such as a +directory name, to produce backup file names. The @option{-B +@var{prefix}} or @option{--prefix=@var{prefix}} option makes backup +files by prepending @var{prefix} to them. The @option{-Y +@var{prefix}} or @option{--basename-prefix=@var{prefix}} prepends +@var{prefix} to the last file name component of backup file names +instead; for example, @option{-Y ~} causes the backup name for +@file{dir/file.c} to be @file{dir/~file.c}. If you use either of +these prefix options, the suffix-based options are ignored. + +If you specify the output file with the @option{-o} option, that file is +the one that is backed up, not the input file. + +Options that affect the names of backup files do not affect whether +backups are made. For example, if you specify the +@option{--no-backup-if-mismatch} option, none of the options described +in this section have any affect, because no backups are made. + +@node Reject Names +@section Reject File Names +@cindex reject file names + +The names for reject files (files containing patches that +@command{patch} could not find a place to apply) are normally the name +of the output file with @samp{.rej} appended (or @samp{#} if using +@samp{.rej} would make the backup file name too long). + +Alternatively, you can tell @command{patch} to place all of the rejected +patches in a single file. The @option{-r @var{reject-file}} or +@option{--reject-file=@var{reject-file}} option uses @var{reject-file} as +the reject file name. + +@node patch Messages +@section Messages and Questions from @command{patch} +@cindex @command{patch} messages and questions +@cindex diagnostics from @command{patch} +@cindex messages from @command{patch} + +@command{patch} can produce a variety of messages, especially if it +has trouble decoding its input. In a few situations where it's not +sure how to proceed, @command{patch} normally prompts you for more +information from the keyboard. There are options to produce more or +fewer messages, to have it not ask for keyboard input, and to +affect the way that file names are quoted in messages. + +@menu +* More or Fewer Messages:: Controlling the verbosity of @command{patch}. +* patch and Keyboard Input:: Inhibiting keyboard input. +* patch Quoting Style:: Quoting file names in diagnostics. +@end menu + +@command{patch} exits with status 0 if all hunks are applied successfully, +1 if some hunks cannot be applied, and 2 if there is more serious trouble. +When applying a set of patches in a loop, you should check the +exit status, so you don't apply a later patch to a partially patched +file. + +@node More or Fewer Messages +@subsection Controlling the Verbosity of @command{patch} +@cindex verbose messages from @command{patch} +@cindex inhibit messages from @command{patch} + +You can cause @command{patch} to produce more messages by using the +@option{--verbose} option. For example, when you give this option, +the message @samp{Hmm...} indicates that @command{patch} is reading text in +the patch file, attempting to determine whether there is a patch in that +text, and if so, what kind of patch it is. + +You can inhibit all terminal output from @command{patch}, unless an error +occurs, by using the @option{-s}, @option{--quiet}, or @option{--silent} +option. + +@node patch and Keyboard Input +@subsection Inhibiting Keyboard Input +@cindex keyboard input to @command{patch} + +There are two ways you can prevent @command{patch} from asking you any +questions. The @option{-f} or @option{--force} option assumes that you know +what you are doing. It causes @command{patch} to do the following: + +@itemize @bullet +@item +Skip patches that do not contain file names in their headers. + +@item +Patch files even though they have the wrong version for the +@samp{Prereq:} line in the patch; + +@item +Assume that patches are not reversed even if they look like they are. +@end itemize + +@noindent +The @option{-t} or @option{--batch} option is similar to @option{-f}, in that +it suppresses questions, but it makes somewhat different assumptions: + +@itemize @bullet +@item +Skip patches that do not contain file names in their headers +(the same as @option{-f}). + +@item +Skip patches for which the file has the wrong version for the +@samp{Prereq:} line in the patch; + +@item +Assume that patches are reversed if they look like they are. +@end itemize + +@node patch Quoting Style +@subsection @command{patch} Quoting Style +@cindex quoting style + +When @command{patch} outputs a file name in a diagnostic message, it +can format the name in any of several ways. This can be useful to +output file names unambiguously, even if they contain punctuation or +special characters like newlines. The +@option{--quoting-style=@var{word}} option controls how names are +output. The @var{word} should be one of the following: + +@table @samp +@item literal +Output names as-is. +@item shell +Quote names for the shell if they contain shell metacharacters or would +cause ambiguous output. +@item shell-always +Quote names for the shell, even if they would normally not require quoting. +@item c +Quote names as for a C language string. +@item escape +Quote as with @samp{c} except omit the surrounding double-quote +characters. +@c The following are not yet implemented in patch 2.5.4. +@c @item clocale +@c Quote as with @samp{c} except use quotation marks appropriate for the +@c locale. +@c @item locale +@c @c Use @t instead of @samp to avoid duplicate quoting in some output styles. +@c Like @samp{clocale}, but quote @t{`like this'} instead of @t{"like +@c this"} in the default C locale. This looks nicer on many displays. +@end table + +@vindex QUOTING_STYLE +You can specify the default value of the @option{--quoting-style} +option with the environment variable @env{QUOTING_STYLE}. If that +environment variable is not set, the default value is @samp{shell}, +but this default may change in a future version of @command{patch}. + +@node patch and POSIX +@section @command{patch} and the @acronym{POSIX} Standard +@cindex @acronym{POSIX} + +@vindex POSIXLY_CORRECT +If you specify the @option{--posix} option, or set the +@env{POSIXLY_CORRECT} environment variable, @command{patch} conforms +more strictly to the @acronym{POSIX} standard, as follows: + +@itemize @bullet +@item +Take the first existing file from the list (old, new, index) +when intuiting file names from diff headers. @xref{Multiple Patches}. + +@item +Do not remove files that are removed by a diff. +@xref{Creating and Removing}. + +@item +Do not ask whether to get files from @acronym{RCS}, ClearCase, or +@acronym{SCCS}. @xref{Revision Control}. + +@item +Require that all options precede the files in the command line. + +@item +Do not backup files, even when there is a mismatch. @xref{Backups}. + +@end itemize + +@node patch and Tradition +@section @acronym{GNU} @command{patch} and Traditional @command{patch} +@cindex traditional @command{patch} + +The current version of @acronym{GNU} @command{patch} normally follows the +@acronym{POSIX} standard. @xref{patch and POSIX}, for the few exceptions +to this general rule. + +Unfortunately, @acronym{POSIX} redefined the behavior of @command{patch} in +several important ways. You should be aware of the following +differences if you must interoperate with traditional @command{patch}, +or with @acronym{GNU} @command{patch} version 2.1 and earlier. + +@itemize @bullet +@item +In traditional @command{patch}, the @option{-p} option's operand was +optional, and a bare @option{-p} was equivalent to @option{-p0}. The +@option{-p} option now requires an operand, and @option{-p@ 0} is now +equivalent to @option{-p0}. For maximum compatibility, use options +like @option{-p0} and @option{-p1}. + +Also, traditional @command{patch} simply counted slashes when +stripping path prefixes; @command{patch} now counts pathname +components. That is, a sequence of one or more adjacent slashes now +counts as a single slash. For maximum portability, avoid sending +patches containing @file{//} in file names. + +@item +In traditional @command{patch}, backups were enabled by default. This +behavior is now enabled with the @option{-b} or @option{--backup} +option. + +Conversely, in @acronym{POSIX} @command{patch}, backups are never made, +even when there is a mismatch. In @acronym{GNU} @command{patch}, this +behavior is enabled with the @option{--no-backup-if-mismatch} option, +or by conforming to @acronym{POSIX}. + +The @option{-b@ @var{suffix}} option of traditional @command{patch} is +equivalent to the @samp{-b -z@ @var{suffix}} options of @acronym{GNU} +@command{patch}. + +@item +Traditional @command{patch} used a complicated (and incompletely +documented) method to intuit the name of the file to be patched from +the patch header. This method did not conform to @acronym{POSIX}, and had +a few gotchas. Now @command{patch} uses a different, equally +complicated (but better documented) method that is optionally +@acronym{POSIX}-conforming; we hope it has fewer gotchas. The two methods +are compatible if the file names in the context diff header and the +@samp{Index:} line are all identical after prefix-stripping. Your +patch is normally compatible if each header's file names all contain +the same number of slashes. + +@item +When traditional @command{patch} asked the user a question, it sent +the question to standard error and looked for an answer from the first +file in the following list that was a terminal: standard error, +standard output, @file{/dev/tty}, and standard input. Now +@command{patch} sends questions to standard output and gets answers +from @file{/dev/tty}. Defaults for some answers have been changed so +that @command{patch} never goes into an infinite loop when using +default answers. + +@item +Traditional @command{patch} exited with a status value that counted +the number of bad hunks, or with status 1 if there was real trouble. +Now @command{patch} exits with status 1 if some hunks failed, or with +2 if there was real trouble. + +@item +Limit yourself to the following options when sending instructions +meant to be executed by anyone running @acronym{GNU} @command{patch}, +traditional @command{patch}, or a @command{patch} that conforms to +@acronym{POSIX}. Spaces are significant in the following list, and +operands are required. + +@example +@option{-c} +@option{-d @var{dir}} +@option{-D @var{define}} +@option{-e} +@option{-l} +@option{-n} +@option{-N} +@option{-o @var{outfile}} +@option{-p@var{num}} +@option{-R} +@option{-r @var{rejectfile}} +@end example + +@end itemize + +@node Making Patches +@chapter Tips for Making and Using Patches + +Use some common sense when making and using patches. For example, +when sending bug fixes to a program's maintainer, send several small +patches, one per independent subject, instead of one large, +harder-to-digest patch that covers all the subjects. + +Here are some other things you should keep in mind if you are going to +distribute patches for updating a software package. + +@menu +* Tips for Patch Producers:: Advice for making patches. +* Tips for Patch Consumers:: Advice for using patches. +* Avoiding Common Mistakes:: Avoiding common mistakes when using @command{patch}. +* Generating Smaller Patches:: How to generate smaller patches. +@end menu + +@node Tips for Patch Producers +@section Tips for Patch Producers +@cindex patch producer tips + +To create a patch that changes an older version of a package into a +newer version, first make a copy of the older and newer versions in +adjacent subdirectories. It is common to do that by unpacking +@command{tar} archives of the two versions. + +To generate the patch, use the command @samp{diff -Naur @var{old} +@var{new}} where @var{old} and @var{new} identify the old and new +directories. The names @var{old} and @var{new} should not contain any +slashes. The @option{-N} option lets the patch create and remove +files; @option{-a} lets the patch update non-text files; @option{-u} +generates useful time stamps and enough context; and @option{-r} lets +the patch update subdirectories. Here is an example command, using +Bourne shell syntax: + +@example +diff -Naur gcc-3.0.3 gcc-3.0.4 +@end example + +Tell your recipients how to apply the patches. This should include +which working directory to use, and which @command{patch} options to +use; the option @samp{-p1} is recommended. Test your procedure by +pretending to be a recipient and applying your patches to a copy of +the original files. + +@xref{Avoiding Common Mistakes}, for how to avoid common mistakes when +generating a patch. + +@node Tips for Patch Consumers +@section Tips for Patch Consumers +@cindex patch consumer tips + +A patch producer should tell recipients how to apply the patches, so +the first rule of thumb for a patch consumer is to follow the +instructions supplied with the patch. + +@acronym{GNU} @command{diff} can analyze files with arbitrarily long lines +and files that end in incomplete lines. However, older versions of +@command{patch} cannot patch such files. If you are having trouble +applying such patches, try upgrading to a recent version of @acronym{GNU} +@command{patch}. + +@node Avoiding Common Mistakes +@section Avoiding Common Mistakes +@cindex common mistakes with patches +@cindex patch, common mistakes + +When producing a patch for multiple files, apply @command{diff} to +directories whose names do not have slashes. This reduces confusion +when the patch consumer specifies the @option{-p@var{number}} option, +since this option can have surprising results when the old and new +file names have different numbers of slashes. For example, do not +send a patch with a header that looks like this: + +@example +diff -Naur v2.0.29/prog/README prog/README +--- v2.0.29/prog/README 2002-03-10 23:30:39.942229878 -0800 ++++ prog/README 2002-03-17 20:49:32.442260588 -0800 +@end example + +@noindent +because the two file names have different numbers of slashes, and +different versions of @command{patch} interpret the file names +differently. To avoid confusion, send output that looks like this +instead: + +@example +diff -Naur v2.0.29/prog/README v2.0.30/prog/README +--- v2.0.29/prog/README 2002-03-10 23:30:39.942229878 -0800 ++++ v2.0.30/prog/README 2002-03-17 20:49:32.442260588 -0800 +@end example + +Make sure you have specified the file names correctly, either in a +context diff header or with an @samp{Index:} line. Take care to not send out +reversed patches, since these make people wonder whether they have +already applied the patch. + +Avoid sending patches that compare backup file names like +@file{README.orig} or @file{README~}, since this might confuse +@command{patch} into patching a backup file instead of the real file. +Instead, send patches that compare the same base file names in +different directories, e.g.@: @file{old/README} and @file{new/README}. + +To save people from partially applying a patch before other patches that +should have gone before it, you can make the first patch in the patch +file update a file with a name like @file{patchlevel.h} or +@file{version.c}, which contains a patch level or version number. If +the input file contains the wrong version number, @command{patch} will +complain immediately. + +An even clearer way to prevent this problem is to put a @samp{Prereq:} +line before the patch. If the leading text in the patch file contains a +line that starts with @samp{Prereq:}, @command{patch} takes the next word +from that line (normally a version number) and checks whether the next +input file contains that word, preceded and followed by either +white space or a newline. If not, @command{patch} prompts you for +confirmation before proceeding. This makes it difficult to accidentally +apply patches in the wrong order. + +@node Generating Smaller Patches +@section Generating Smaller Patches +@cindex patches, shrinking + +The simplest way to generate a patch is to use @samp{diff -Naur} +(@pxref{Tips for Patch Producers}), but you might be able to reduce +the size of the patch by renaming or removing some files before making +the patch. If the older version of the package contains any files +that the newer version does not, or if any files have been renamed +between the two versions, make a list of @command{rm} and @command{mv} +commands for the user to execute in the old version directory before +applying the patch. Then run those commands yourself in the scratch +directory. + +If there are any files that you don't need to include in the patch +because they can easily be rebuilt from other files (for example, +@file{TAGS} and output from @command{yacc} and @command{makeinfo}), +exclude them from the patch by giving @command{diff} the @option{-x +@var{pattern}} option (@pxref{Comparing Directories}). If you want +your patch to modify a derived file because your recipients lack tools +to build it, make sure that the patch for the derived file follows any +patches for files that it depends on, so that the recipients' time +stamps will not confuse @command{make}. + +Now you can create the patch using @samp{diff -Naur}. Make sure to +specify the scratch directory first and the newer directory second. + +Add to the top of the patch a note telling the user any @command{rm} and +@command{mv} commands to run before applying the patch. Then you can +remove the scratch directory. + +You can also shrink the patch size by using fewer lines of context, +but bear in mind that @command{patch} typically needs at least two +lines for proper operation when patches do not exactly match the input +files. + +@node Invoking cmp +@chapter Invoking @command{cmp} +@cindex invoking @command{cmp} +@cindex @command{cmp} invocation + +The @command{cmp} command compares two files, and if they differ, +tells the first byte and line number where they differ or reports +that one file is a prefix of the other. Bytes and +lines are numbered starting with 1. The arguments of @command{cmp} +are as follows: + +@example +cmp @var{options}@dots{} @var{from-file} @r{[}@var{to-file} @r{[}@var{from-skip} @r{[}@var{to-skip}@r{]}@r{]}@r{]} +@end example + +The file name @file{-} is always the standard input. @command{cmp} +also uses the standard input if one file name is omitted. The +@var{from-skip} and @var{to-skip} operands specify how many bytes to +ignore at the start of each file; they are equivalent to the +@option{--ignore-initial=@var{from-skip}:@var{to-skip}} option. + +By default, @command{cmp} outputs nothing if the two files have the +same contents. If one file is a prefix of the other, @command{cmp} +prints to standard error a message of the following form: + +@example +cmp: EOF on @var{shorter-file} +@end example + +Otherwise, @command{cmp} prints to standard output a message of the +following form: + +@example +@var{from-file} @var{to-file} differ: char @var{byte-number}, line @var{line-number} +@end example + +The message formats can differ outside the @acronym{POSIX} locale. +Also, @acronym{POSIX} allows the @acronym{EOF} message to be followed +by a blank and some additional information. + +An exit status of 0 means no differences were found, 1 means some +differences were found, and 2 means trouble. + +@menu +* cmp Options:: Summary of options to @command{cmp}. +@end menu + +@node cmp Options +@section Options to @command{cmp} +@cindex @command{cmp} options +@cindex options for @command{cmp} + +Below is a summary of all of the options that @acronym{GNU} +@command{cmp} accepts. Most options have two equivalent names, one of +which is a single letter preceded by @samp{-}, and the other of which +is a long name preceded by @samp{--}. Multiple single letter options +(unless they take an argument) can be combined into a single command +line word: @option{-bl} is equivalent to @option{-b -l}. + +@table @option +@item -b +@itemx --print-bytes +Print the differing bytes. Display control bytes as a +@samp{^} followed by a letter of the alphabet and precede bytes +that have the high bit set with @samp{M-} (which stands for ``meta''). + +@item --help +Output a summary of usage and then exit. + +@item -i @var{skip} +@itemx --ignore-initial=@var{skip} +Ignore any differences in the first @var{skip} bytes of the input +files. Treat files with fewer than @var{skip} bytes as if they are +empty. If @var{skip} is of the form +@option{@var{from-skip}:@var{to-skip}}, skip the first @var{from-skip} +bytes of the first input file and the first @var{to-skip} bytes of the +second. + +@item -l +@itemx --verbose +Output the (decimal) byte numbers and (octal) values of all differing bytes, +instead of the default standard output. + +@item -n @var{count} +@itemx --bytes=@var{count} +Compare at most @var{count} input bytes. + +@item -s +@itemx --quiet +@itemx --silent +Do not print anything; only return an exit status indicating whether +the files differ. + +@item -v +@itemx --version +Output version information and then exit. +@end table + +In the above table, operands that are byte counts are normally +decimal, but may be preceded by @samp{0} for octal and @samp{0x} for +hexadecimal. + +A byte count can be followed by a suffix to specify a multiple of that +count; in this case an omitted integer is understood to be 1. A bare +size letter, or one followed by @samp{iB}, specifies a multiple using +powers of 1024. A size letter followed by @samp{B} specifies powers +of 1000 instead. For example, @option{-n 4M} and @option{-n 4MiB} are +equivalent to @option{-n 4194304}, whereas @option{-n 4MB} is +equivalent to @option{-n 4000000}. This notation is upward compatible +with the @uref{http://www.bipm.fr/enus/3_SI/si-prefixes.html, SI +prefixes} for decimal multiples and with the +@uref{http://physics.nist.gov/cuu/Units/binary.html, IEC 60027-2 +prefixes for binary multiples}. + +The following suffixes are defined. Large sizes like @code{1Y} may be +rejected by your computer due to limitations of its arithmetic. + +@table @samp +@item kB +@cindex kilobyte, definition of +kilobyte: @math{10^3 = 1000}. +@item k +@itemx K +@itemx KiB +@cindex kibibyte, definition of +kibibyte: @math{2^10 = 1024}. @samp{K} is special: the SI prefix is +@samp{k} and the IEC 60027-2 prefix is @samp{Ki}, but tradition and +@acronym{POSIX} use @samp{k} to mean @samp{KiB}. +@item MB +@cindex megabyte, definition of +megabyte: @math{10^6 = 1,000,000}. +@item M +@itemx MiB +@cindex mebibyte, definition of +mebibyte: @math{2^20 = 1,048,576}. +@item GB +@cindex gigabyte, definition of +gigabyte: @math{10^9 = 1,000,000,000}. +@item G +@itemx GiB +@cindex gibibyte, definition of +gibibyte: @math{2^30 = 1,073,741,824}. +@item TB +@cindex terabyte, definition of +terabyte: @math{10^12 = 1,000,000,000,000}. +@item T +@itemx TiB +@cindex tebibyte, definition of +tebibyte: @math{2^40 = 1,099,511,627,776}. +@item PB +@cindex petabyte, definition of +petabyte: @math{10^15 = 1,000,000,000,000,000}. +@item P +@itemx PiB +@cindex pebibyte, definition of +pebibyte: @math{2^50 = 1,125,899,906,842,624}. +@item EB +@cindex exabyte, definition of +exabyte: @math{10^18 = 1,000,000,000,000,000,000}. +@item E +@itemx EiB +@cindex exbibyte, definition of +exbibyte: @math{2^60 = 1,152,921,504,606,846,976}. +@item ZB +@cindex zettabyte, definition of +zettabyte: @math{10^21 = 1,000,000,000,000,000,000,000} +@item Z +@itemx ZiB +@math{2^70 = 1,180,591,620,717,411,303,424}. +(@samp{Zi} is a GNU extension to IEC 60027-2.) +@item YB +@cindex yottabyte, definition of +yottabyte: @math{10^24 = 1,000,000,000,000,000,000,000,000}. +@item Y +@itemx YiB +@math{2^80 = 1,208,925,819,614,629,174,706,176}. +(@samp{Yi} is a GNU extension to IEC 60027-2.) +@end table + +@node Invoking diff +@chapter Invoking @command{diff} +@cindex invoking @command{diff} +@cindex @command{diff} invocation + +The format for running the @command{diff} command is: + +@example +diff @var{options}@dots{} @var{files}@dots{} +@end example + +In the simplest case, two file names @var{from-file} and +@var{to-file} are given, and @command{diff} compares the contents of +@var{from-file} and @var{to-file}. A file name of @file{-} stands for +text read from the standard input. As a special case, @samp{diff - -} +compares a copy of standard input to itself. + +If one file is a directory and the other is not, @command{diff} compares +the file in the directory whose name is that of the non-directory. +The non-directory file must not be @file{-}. + +If two file names are given and both are directories, +@command{diff} compares corresponding files in both directories, in +alphabetical order; this comparison is not recursive unless the +@option{-r} or @option{--recursive} option is given. @command{diff} never +compares the actual contents of a directory as if it were a file. The +file that is fully specified may not be standard input, because standard +input is nameless and the notion of ``file with the same name'' does not +apply. + +If the @option{--from-file=@var{file}} option is given, the number of +file names is arbitrary, and @var{file} is compared to each named file. +Similarly, if the @option{--to-file=@var{file}} option is given, each +named file is compared to @var{file}. + +@command{diff} options begin with @samp{-}, so normally file names +may not begin with @samp{-}. However, @option{--} as an +argument by itself treats the remaining arguments as file names even if +they begin with @samp{-}. + +An exit status of 0 means no differences were found, 1 means some +differences were found, and 2 means trouble. Normally, differing +binary files count as trouble, but this can be altered by using the +@option{-a} or @option{--text} option, or the @option{-q} or +@option{--brief} option. + +@menu +* diff Options:: Summary of options to @command{diff}. +@end menu + +@node diff Options +@section Options to @command{diff} +@cindex @command{diff} options +@cindex options for @command{diff} + +Below is a summary of all of the options that @acronym{GNU} +@command{diff} accepts. Most options have two equivalent names, one +of which is a single letter preceded by @samp{-}, and the other of +which is a long name preceded by @samp{--}. Multiple single letter +options (unless they take an argument) can be combined into a single +command line word: @option{-ac} is equivalent to @option{-a -c}. Long +named options can be abbreviated to any unique prefix of their name. +Brackets ([ and ]) indicate that an option takes an optional argument. + +@table @option +@item -a +@itemx --text +Treat all files as text and compare them line-by-line, even if they +do not seem to be text. @xref{Binary}. + +@item -b +@itemx --ignore-space-change +Ignore changes in amount of white space. @xref{White Space}. + +@item -B +@itemx --ignore-blank-lines +Ignore changes that just insert or delete blank lines. @xref{Blank +Lines}. + +@item --binary +Read and write data in binary mode. @xref{Binary}. + +@item -c +Use the context output format, showing three lines of context. +@xref{Context Format}. + +@item -C @var{lines} +@itemx --context@r{[}=@var{lines}@r{]} +Use the context output format, showing @var{lines} (an integer) lines of +context, or three if @var{lines} is not given. @xref{Context Format}. +For proper operation, @command{patch} typically needs at least two lines of +context. + +On older systems, @command{diff} supports an obsolete option +@option{-@var{lines}} that has effect when combined with @option{-c} +or @option{-p}. @acronym{POSIX} 1003.1-2001 (@pxref{Standards +conformance}) does not allow this; use @option{-C @var{lines}} +instead. + +@item --changed-group-format=@var{format} +Use @var{format} to output a line group containing differing lines from +both files in if-then-else format. @xref{Line Group Formats}. + +@item -d +@itemx --minimal +Change the algorithm perhaps find a smaller set of changes. This makes +@command{diff} slower (sometimes much slower). @xref{diff Performance}. + +@item -D @var{name} +@itemx --ifdef=@var{name} +Make merged @samp{#ifdef} format output, conditional on the preprocessor +macro @var{name}. @xref{If-then-else}. + +@item -e +@itemx --ed +Make output that is a valid @command{ed} script. @xref{ed Scripts}. + +@item -E +@itemx --ignore-tab-expansion +Ignore changes due to tab expansion. +@xref{White Space}. + +@item -f +@itemx --forward-ed +Make output that looks vaguely like an @command{ed} script but has changes +in the order they appear in the file. @xref{Forward ed}. + +@item -F @var{regexp} +@itemx --show-function-line=@var{regexp} +In context and unified format, for each hunk of differences, show some +of the last preceding line that matches @var{regexp}. @xref{Specified +Headings}. + +@item --from-file=@var{file} +Compare @var{file} to each operand; @var{file} may be a directory. + +@item --help +Output a summary of usage and then exit. + +@item --horizon-lines=@var{lines} +Do not discard the last @var{lines} lines of the common prefix +and the first @var{lines} lines of the common suffix. +@xref{diff Performance}. + +@item -i +@itemx --ignore-case +Ignore changes in case; consider upper- and lower-case letters +equivalent. @xref{Case Folding}. + +@item -I @var{regexp} +@itemx --ignore-matching-lines=@var{regexp} +Ignore changes that just insert or delete lines that match @var{regexp}. +@xref{Specified Lines}. + +@item --ignore-file-name-case +Ignore case when comparing file names during recursive comparison. +@xref{Comparing Directories}. + +@item -l +@itemx --paginate +Pass the output through @command{pr} to paginate it. @xref{Pagination}. + +@item --label=@var{label} +Use @var{label} instead of the file name in the context format +(@pxref{Context Format}) and unified format (@pxref{Unified Format}) +headers. @xref{RCS}. + +@item --left-column +Print only the left column of two common lines in side by side format. +@xref{Side by Side Format}. + +@item --line-format=@var{format} +Use @var{format} to output all input lines in if-then-else format. +@xref{Line Formats}. + +@item -n +@itemx --rcs +Output @acronym{RCS}-format diffs; like @option{-f} except that each command +specifies the number of lines affected. @xref{RCS}. + +@item -N +@itemx --new-file +In directory comparison, if a file is found in only one directory, +treat it as present but empty in the other directory. @xref{Comparing +Directories}. + +@item --new-group-format=@var{format} +Use @var{format} to output a group of lines taken from just the second +file in if-then-else format. @xref{Line Group Formats}. + +@item --new-line-format=@var{format} +Use @var{format} to output a line taken from just the second file in +if-then-else format. @xref{Line Formats}. + +@item --old-group-format=@var{format} +Use @var{format} to output a group of lines taken from just the first +file in if-then-else format. @xref{Line Group Formats}. + +@item --old-line-format=@var{format} +Use @var{format} to output a line taken from just the first file in +if-then-else format. @xref{Line Formats}. + +@item -p +@itemx --show-c-function +Show which C function each change is in. @xref{C Function Headings}. + +@item -q +@itemx --brief +Report only whether the files differ, not the details of the +differences. @xref{Brief}. + +@item -r +@itemx --recursive +When comparing directories, recursively compare any subdirectories +found. @xref{Comparing Directories}. + +@item -s +@itemx --report-identical-files +Report when two files are the same. @xref{Comparing Directories}. + +@item -S @var{file} +@itemx --starting-file=@var{file} +When comparing directories, start with the file @var{file}. This is +used for resuming an aborted comparison. @xref{Comparing Directories}. + +@item --speed-large-files +Use heuristics to speed handling of large files that have numerous +scattered small changes. @xref{diff Performance}. + +@item --strip-trailing-cr +Strip any trailing carriage return at the end of an input line. +@xref{Binary}. + +@item --suppress-common-lines +Do not print common lines in side by side format. +@xref{Side by Side Format}. + +@item -t +@itemx --expand-tabs +Expand tabs to spaces in the output, to preserve the alignment of tabs +in the input files. @xref{Tabs}. + +@item -T +@itemx --initial-tab +Output a tab rather than a space before the text of a line in normal or +context format. This causes the alignment of tabs in the line to look +normal. @xref{Tabs}. + +@item --tabsize=@var{columns} +Assume that tab stops are set every @var{columns} (default 8) print +columns. @xref{Tabs}. + +@item --to-file=@var{file} +Compare each operand to @var{file}; @var{file} may be a directory. + +@item -u +Use the unified output format, showing three lines of context. +@xref{Unified Format}. + +@item --unchanged-group-format=@var{format} +Use @var{format} to output a group of common lines taken from both files +in if-then-else format. @xref{Line Group Formats}. + +@item --unchanged-line-format=@var{format} +Use @var{format} to output a line common to both files in if-then-else +format. @xref{Line Formats}. + +@item --unidirectional-new-file +When comparing directories, if a file appears only in the second +directory of the two, treat it as present but empty in the other. +@xref{Comparing Directories}. + +@item -U @var{lines} +@itemx --unified@r{[}=@var{lines}@r{]} +Use the unified output format, showing @var{lines} (an integer) lines of +context, or three if @var{lines} is not given. @xref{Unified Format}. +For proper operation, @command{patch} typically needs at least two lines of +context. + +On older systems, @command{diff} supports an obsolete option +@option{-@var{lines}} that has effect when combined with @option{-u}. +@acronym{POSIX} 1003.1-2001 (@pxref{Standards conformance}) does not allow +this; use @option{-U @var{lines}} instead. + +@item -v +@itemx --version +Output version information and then exit. + +@item -w +@itemx --ignore-all-space +Ignore white space when comparing lines. @xref{White Space}. + +@item -W @var{columns} +@itemx --width=@var{columns} +Output at most @var{columns} (default 130) print columns per line in +side by side format. @xref{Side by Side Format}. + +@item -x @var{pattern} +@itemx --exclude=@var{pattern} +When comparing directories, ignore files and subdirectories whose basenames +match @var{pattern}. @xref{Comparing Directories}. + +@item -X @var{file} +@itemx --exclude-from=@var{file} +When comparing directories, ignore files and subdirectories whose basenames +match any pattern contained in @var{file}. @xref{Comparing Directories}. + +@item -y +@itemx --side-by-side +Use the side by side output format. @xref{Side by Side Format}. +@end table + +@node Invoking diff3 +@chapter Invoking @command{diff3} +@cindex invoking @command{diff3} +@cindex @command{diff3} invocation + +The @command{diff3} command compares three files and outputs descriptions +of their differences. Its arguments are as follows: + +@example +diff3 @var{options}@dots{} @var{mine} @var{older} @var{yours} +@end example + +The files to compare are @var{mine}, @var{older}, and @var{yours}. +At most one of these three file names may be @file{-}, +which tells @command{diff3} to read the standard input for that file. + +An exit status of 0 means @command{diff3} was successful, 1 means some +conflicts were found, and 2 means trouble. + +@menu +* diff3 Options:: Summary of options to @command{diff3}. +@end menu + +@node diff3 Options +@section Options to @command{diff3} +@cindex @command{diff3} options +@cindex options for @command{diff3} + +Below is a summary of all of the options that @acronym{GNU} @command{diff3} +accepts. Multiple single letter options (unless they take an argument) +can be combined into a single command line argument. + +@table @option +@item -a +@itemx --text +Treat all files as text and compare them line-by-line, even if they +do not appear to be text. @xref{Binary}. + +@item -A +@itemx --show-all +Incorporate all unmerged changes from @var{older} to @var{yours} into +@var{mine}, surrounding conflicts with bracket lines. +@xref{Marking Conflicts}. + +@item --diff-program=@var{program} +Use the compatible comparison program @var{program} to compare files +instead of @command{diff}. + +@item -e +@itemx --ed +Generate an @command{ed} script that incorporates all the changes from +@var{older} to @var{yours} into @var{mine}. @xref{Which Changes}. + +@item -E +@itemx --show-overlap +Like @option{-e}, except bracket lines from overlapping changes' first +and third files. +@xref{Marking Conflicts}. +With @option{-E}, an overlapping change looks like this: + +@example +<<<<<<< @var{mine} +@r{lines from @var{mine}} +======= +@r{lines from @var{yours}} +>>>>>>> @var{yours} +@end example + +@item --help +Output a summary of usage and then exit. + +@item -i +Generate @samp{w} and @samp{q} commands at the end of the @command{ed} +script for System V compatibility. This option must be combined with +one of the @option{-AeExX3} options, and may not be combined with @option{-m}. +@xref{Saving the Changed File}. + +@item --label=@var{label} +Use the label @var{label} for the brackets output by the @option{-A}, +@option{-E} and @option{-X} options. This option may be given up to three +times, one for each input file. The default labels are the names of +the input files. Thus @samp{diff3 --label X --label Y --label Z -m A +B C} acts like +@samp{diff3 -m A B C}, except that the output looks like it came from +files named @samp{X}, @samp{Y} and @samp{Z} rather than from files +named @samp{A}, @samp{B} and @samp{C}. @xref{Marking Conflicts}. + +@item -m +@itemx --merge +Apply the edit script to the first file and send the result to standard +output. Unlike piping the output from @command{diff3} to @command{ed}, this +works even for binary files and incomplete lines. @option{-A} is assumed +if no edit script option is specified. @xref{Bypassing ed}. + +@item --strip-trailing-cr +Strip any trailing carriage return at the end of an input line. +@xref{Binary}. + +@item -T +@itemx --initial-tab +Output a tab rather than two spaces before the text of a line in normal format. +This causes the alignment of tabs in the line to look normal. @xref{Tabs}. + +@item -v +@itemx --version +Output version information and then exit. + +@item -x +@itemx --overlap-only +Like @option{-e}, except output only the overlapping changes. +@xref{Which Changes}. + +@item -X +Like @option{-E}, except output only the overlapping changes. +In other words, like @option{-x}, except bracket changes as in @option{-E}. +@xref{Marking Conflicts}. + +@item -3 +@itemx --easy-only +Like @option{-e}, except output only the nonoverlapping changes. +@xref{Which Changes}. +@end table + +@node Invoking patch +@chapter Invoking @command{patch} +@cindex invoking @command{patch} +@cindex @command{patch} invocation + +Normally @command{patch} is invoked like this: + +@example +patch <@var{patchfile} +@end example + +The full format for invoking @command{patch} is: + +@example +patch @var{options}@dots{} @r{[}@var{origfile} @r{[}@var{patchfile}@r{]}@r{]} +@end example + +You can also specify where to read the patch from with the @option{-i +@var{patchfile}} or @option{--input=@var{patchfile}} option. +If you do not specify @var{patchfile}, or if @var{patchfile} is +@file{-}, @command{patch} reads the patch (that is, the @command{diff} output) +from the standard input. + +If you do not specify an input file on the command line, @command{patch} +tries to intuit from the @dfn{leading text} (any text in the patch +that comes before the @command{diff} output) which file to edit. +@xref{Multiple Patches}. + +By default, @command{patch} replaces the original input file with the +patched version, possibly after renaming the original file into a +backup file (@pxref{Backup Names}, for a description of how +@command{patch} names backup files). You can also specify where to +put the output with the @option{-o @var{file}} or +@option{--output=@var{file}} option; however, do not use this option +if @var{file} is one of the input files. + +@menu +* patch Options:: Summary table of options to @command{patch}. +@end menu + +@node patch Options +@section Options to @command{patch} +@cindex @command{patch} options +@cindex options for @command{patch} + +Here is a summary of all of the options that @acronym{GNU} @command{patch} +accepts. @xref{patch and Tradition}, for which of these options are +safe to use in older versions of @command{patch}. + +Multiple single-letter options that do not take an argument can be +combined into a single command line argument with only one dash. + +@table @option +@item -b +@itemx --backup +Back up the original contents of each file, even if backups would +normally not be made. @xref{Backups}. + +@item -B @var{prefix} +@itemx --prefix=@var{prefix} +Prepend @var{prefix} to backup file names. @xref{Backup Names}. + +@item --backup-if-mismatch +Back up the original contents of each file if the patch does not +exactly match the file. This is the default behavior when not +conforming to @acronym{POSIX}. @xref{Backups}. + +@item --binary +Read and write all files in binary mode, except for standard output +and @file{/dev/tty}. This option has no effect on +@acronym{POSIX}-conforming systems like @acronym{GNU}/Linux. On systems where +this option makes a difference, the patch should be generated by +@samp{diff -a --binary}. @xref{Binary}. + +@item -c +@itemx --context +Interpret the patch file as a context diff. @xref{patch Input}. + +@item -d @var{directory} +@itemx --directory=@var{directory} +Make directory @var{directory} the current directory for interpreting +both file names in the patch file, and file names given as arguments to +other options. @xref{patch Directories}. + +@item -D @var{name} +@itemx --ifdef=@var{name} +Make merged if-then-else output using @var{name}. @xref{If-then-else}. + +@item --dry-run +Print the results of applying the patches without actually changing +any files. @xref{Dry Runs}. + +@item -e +@itemx --ed +Interpret the patch file as an @command{ed} script. @xref{patch Input}. + +@item -E +@itemx --remove-empty-files +Remove output files that are empty after the patches have been applied. +@xref{Creating and Removing}. + +@item -f +@itemx --force +Assume that the user knows exactly what he or she is doing, and do not +ask any questions. @xref{patch Messages}. + +@item -F @var{lines} +@itemx --fuzz=@var{lines} +Set the maximum fuzz factor to @var{lines}. @xref{Inexact}. + +@item -g @var{num} +@itemx --get=@var{num} +If @var{num} is positive, get input files from a revision control +system as necessary; if zero, do not get the files; if negative, ask +the user whether to get the files. @xref{Revision Control}. + +@item --help +Output a summary of usage and then exit. + +@item -i @var{patchfile} +@itemx --input=@var{patchfile} +Read the patch from @var{patchfile} rather than from standard input. +@xref{patch Options}. + +@item -l +@itemx --ignore-white-space +Let any sequence of blanks (spaces or tabs) in the patch file match +any sequence of blanks in the input file. @xref{Changed White Space}. + +@item -n +@itemx --normal +Interpret the patch file as a normal diff. @xref{patch Input}. + +@item -N +@itemx --forward +Ignore patches that @command{patch} thinks are reversed or already applied. +See also @option{-R}. @xref{Reversed Patches}. + +@item --no-backup-if-mismatch +Do not back up the original contents of files. This is the default +behavior when conforming to @acronym{POSIX}. @xref{Backups}. + +@item -o @var{file} +@itemx --output=@var{file} +Use @var{file} as the output file name. @xref{patch Options}. + +@item -p@var{number} +@itemx --strip=@var{number} +Set the file name strip count to @var{number}. @xref{patch Directories}. + +@item --posix +Conform to @acronym{POSIX}, as if the @env{POSIXLY_CORRECT} environment +variable had been set. @xref{patch and POSIX}. + +@item --quoting-style=@var{word} +Use style @var{word} to quote names in diagnostics, as if the +@env{QUOTING_STYLE} environment variable had been set to @var{word}. +@xref{patch Quoting Style}. + +@item -r @var{reject-file} +@itemx --reject-file=@var{reject-file} +Use @var{reject-file} as the reject file name. @xref{Reject Names}. + +@item -R +@itemx --reverse +Assume that this patch was created with the old and new files swapped. +@xref{Reversed Patches}. + +@item -s +@itemx --quiet +@itemx --silent +Work silently unless an error occurs. @xref{patch Messages}. + +@item -t +@itemx --batch +Do not ask any questions. @xref{patch Messages}. + +@item -T +@itemx --set-time +Set the modification and access times of patched files from time +stamps given in context diff headers, assuming that the context diff +headers use local time. @xref{Patching Time Stamps}. + +@item -u +@itemx --unified +Interpret the patch file as a unified diff. @xref{patch Input}. + +@item -v +@itemx --version +Output version information and then exit. + +@item -V @var{backup-style} +@itemx --version=control=@var{backup-style} +Select the naming convention for backup file names. @xref{Backup Names}. + +@item --verbose +Print more diagnostics than usual. @xref{patch Messages}. + +@item -x @var{number} +@itemx --debug=@var{number} +Set internal debugging flags. Of interest only to @command{patch} +patchers. + +@item -Y @var{prefix} +@itemx --basename-prefix=@var{prefix} +Prepend @var{prefix} to base names of backup files. @xref{Backup Names}. + +@item -z @var{suffix} +@itemx --suffix=@var{suffix} +Use @var{suffix} as the backup extension instead of @samp{.orig} or +@samp{~}. @xref{Backup Names}. + +@item -Z +@itemx --set-utc +Set the modification and access times of patched files from time +stamps given in context diff headers, assuming that the context diff +headers use @acronym{UTC}. @xref{Patching Time Stamps}. + +@end table + +@node Invoking sdiff +@chapter Invoking @command{sdiff} +@cindex invoking @command{sdiff} +@cindex @command{sdiff} invocation + +The @command{sdiff} command merges two files and interactively outputs the +results. Its arguments are as follows: + +@example +sdiff -o @var{outfile} @var{options}@dots{} @var{from-file} @var{to-file} +@end example + +This merges @var{from-file} with @var{to-file}, with output to @var{outfile}. +If @var{from-file} is a directory and @var{to-file} is not, @command{sdiff} +compares the file in @var{from-file} whose file name is that of @var{to-file}, +and vice versa. @var{from-file} and @var{to-file} may not both be +directories. + +@command{sdiff} options begin with @samp{-}, so normally @var{from-file} +and @var{to-file} may not begin with @samp{-}. However, @option{--} as an +argument by itself treats the remaining arguments as file names even if +they begin with @samp{-}. You may not use @file{-} as an input file. + +@command{sdiff} without @option{-o} (or @option{--output}) produces a +side-by-side difference. This usage is obsolete; use the @option{-y} +or @option{--side-by-side} option of @command{diff} instead. + +An exit status of 0 means no differences were found, 1 means some +differences were found, and 2 means trouble. + +@menu +* sdiff Options:: Summary of options to @command{diff}. +@end menu + +@node sdiff Options +@section Options to @command{sdiff} +@cindex @command{sdiff} options +@cindex options for @command{sdiff} + +Below is a summary of all of the options that @acronym{GNU} +@command{sdiff} accepts. Each option has two equivalent names, one of +which is a single letter preceded by @samp{-}, and the other of which +is a long name preceded by @samp{--}. Multiple single letter options +(unless they take an argument) can be combined into a single command +line argument. Long named options can be abbreviated to any unique +prefix of their name. + +@table @option +@item -a +@itemx --text +Treat all files as text and compare them line-by-line, even if they +do not appear to be text. @xref{Binary}. + +@item -b +@itemx --ignore-space-change +Ignore changes in amount of white space. @xref{White Space}. + +@item -B +@itemx --ignore-blank-lines +Ignore changes that just insert or delete blank lines. @xref{Blank +Lines}. + +@item -d +@itemx --minimal +Change the algorithm to perhaps find a smaller set of changes. This +makes @command{sdiff} slower (sometimes much slower). @xref{diff +Performance}. + +@item --diff-program=@var{program} +Use the compatible comparison program @var{program} to compare files +instead of @command{diff}. + +@item -E +@itemx --ignore-tab-expansion +Ignore changes due to tab expansion. +@xref{White Space}. + +@item --help +Output a summary of usage and then exit. + +@item -i +@itemx --ignore-case +Ignore changes in case; consider upper- and lower-case to be the same. +@xref{Case Folding}. + +@item -I @var{regexp} +@itemx --ignore-matching-lines=@var{regexp} +Ignore changes that just insert or delete lines that match @var{regexp}. +@xref{Specified Lines}. + +@item -l +@itemx --left-column +Print only the left column of two common lines. +@xref{Side by Side Format}. + +@item -o @var{file} +@itemx --output=@var{file} +Put merged output into @var{file}. This option is required for merging. + +@item -s +@itemx --suppress-common-lines +Do not print common lines. @xref{Side by Side Format}. + +@item --speed-large-files +Use heuristics to speed handling of large files that have numerous +scattered small changes. @xref{diff Performance}. + +@item --strip-trailing-cr +Strip any trailing carriage return at the end of an input line. +@xref{Binary}. + +@item -t +@itemx --expand-tabs +Expand tabs to spaces in the output, to preserve the alignment of tabs +in the input files. @xref{Tabs}. + +@item --tabsize=@var{columns} +Assume that tab stops are set every @var{columns} (default 8) print +columns. @xref{Tabs}. + +@item -v +@itemx --version +Output version information and then exit. + +@item -w @var{columns} +@itemx --width=@var{columns} +Output at most @var{columns} (default 130) print columns per line. +@xref{Side by Side Format}. Note that for historical reasons, this +option is @option{-W} in @command{diff}, @option{-w} in @command{sdiff}. + +@item -W +@itemx --ignore-all-space +Ignore white space when comparing lines. @xref{White Space}. +Note that for historical reasons, this option is @option{-w} in @command{diff}, +@option{-W} in @command{sdiff}. +@end table + +@node Standards conformance +@chapter Standards conformance +@cindex @acronym{POSIX} + +@vindex POSIXLY_CORRECT +In a few cases, the @acronym{GNU} utilities' default behavior is +incompatible with the @acronym{POSIX} standard. To suppress these +incompatibilities, define the @env{POSIXLY_CORRECT} environment +variable. Unless you are checking for @acronym{POSIX} conformance, you +probably do not need to define @env{POSIXLY_CORRECT}. + +Normally options and operands can appear in any order, and programs act +as if all the options appear before any operands. For example, +@samp{diff lao tzu -C 2} acts like @samp{diff -C 2 lao tzu}, since +@samp{2} is an option-argument of @option{-C}. However, if the +@env{POSIXLY_CORRECT} environment variable is set, options must appear +before operands, unless otherwise specified for a particular command. + +Newer versions of @acronym{POSIX} are occasionally incompatible with older +versions. For example, older versions of @acronym{POSIX} allowed the +command @samp{diff -c -10} to have the same meaning as @samp{diff -C +10}, but @acronym{POSIX} 1003.1-2001 @samp{diff} no longer allows +digit-string options like @option{-10}. + +@vindex _POSIX2_VERSION +The @acronym{GNU} utilities normally conform to the version of @acronym{POSIX} +that is standard for your system. To cause them to conform to a +different version of @acronym{POSIX}, define the @env{_POSIX2_VERSION} +environment variable to a value of the form @var{yyyymm} specifying +the year and month the standard was adopted. Two values are currently +supported for @env{_POSIX2_VERSION}: @samp{199209} stands for +@acronym{POSIX} 1003.2-1992, and @samp{200112} stands for @acronym{POSIX} +1003.1-2001. For example, if you are running older software that +assumes an older version of @acronym{POSIX} and uses @samp{diff -c -10}, +you can work around the compatibility problems by setting +@samp{_POSIX2_VERSION=199209} in your environment. + +@node Projects +@chapter Future Projects + +Here are some ideas for improving @acronym{GNU} @command{diff} and +@command{patch}. The @acronym{GNU} project has identified some +improvements as potential programming projects for volunteers. You +can also help by reporting any bugs that you find. + +If you are a programmer and would like to contribute something to the +@acronym{GNU} project, please consider volunteering for one of these +projects. If you are seriously contemplating work, please write to +@email{gvc@@gnu.org} to coordinate with other volunteers. + +@menu +* Shortcomings:: Suggested projects for improvements. +* Bugs:: Reporting bugs. +@end menu + +@node Shortcomings +@section Suggested Projects for Improving @acronym{GNU} @command{diff} and @command{patch} +@cindex projects for directories + +One should be able to use @acronym{GNU} @command{diff} to generate a +patch from any pair of directory trees, and given the patch and a copy +of one such tree, use @command{patch} to generate a faithful copy of +the other. Unfortunately, some changes to directory trees cannot be +expressed using current patch formats; also, @command{patch} does not +handle some of the existing formats. These shortcomings motivate the +following suggested projects. + +@menu +* Internationalization:: Handling multibyte and varying-width characters. +* Changing Structure:: Handling changes to the directory structure. +* Special Files:: Handling symbolic links, device special files, etc. +* Unusual File Names:: Handling file names that contain unusual characters. +* Time Stamp Order:: Outputting diffs in time stamp order. +* Ignoring Changes:: Ignoring certain changes while showing others. +* Speedups:: Improving performance. +@end menu + +@node Internationalization +@subsection Handling Multibyte and Varying-Width Characters +@cindex multibyte characters +@cindex varying-width characters + +@command{diff}, @command{diff3} and @command{sdiff} treat each line of +input as a string of unibyte characters. This can mishandle multibyte +characters in some cases. For example, when asked to ignore spaces, +@command{diff} does not properly ignore a multibyte space character. + +Also, @command{diff} currently assumes that each byte is one column +wide, and this assumption is incorrect in some locales, e.g., locales +that use UTF-8 encoding. This causes problems with the @option{-y} or +@option{--side-by-side} option of @command{diff}. + +These problems need to be fixed without unduly affecting the +performance of the utilities in unibyte environments. + +The IBM GNU/Linux Technology Center Internationalization Team has +proposed +@uref{http://oss.software.ibm.com/developer/opensource/linux/patches/i18n/diffutils-2.7.2-i18n-0.1.patch.gz,patches +to support internationalized @command{diff}}. +Unfortunately, these patches are incomplete and are to an older +version of @command{diff}, so more work needs to be done in this area. + +@node Changing Structure +@subsection Handling Changes to the Directory Structure +@cindex directory structure changes + +@command{diff} and @command{patch} do not handle some changes to directory +structure. For example, suppose one directory tree contains a directory +named @samp{D} with some subsidiary files, and another contains a file +with the same name @samp{D}. @samp{diff -r} does not output enough +information for @command{patch} to transform the directory subtree into +the file. + +There should be a way to specify that a file has been removed without +having to include its entire contents in the patch file. There should +also be a way to tell @command{patch} that a file was renamed, even if +there is no way for @command{diff} to generate such information. +There should be a way to tell @command{patch} that a file's time stamp +has changed, even if its contents have not changed. + +These problems can be fixed by extending the @command{diff} output format +to represent changes in directory structure, and extending @command{patch} +to understand these extensions. + +@node Special Files +@subsection Files that are Neither Directories Nor Regular Files +@cindex special files + +Some files are neither directories nor regular files: they are unusual +files like symbolic links, device special files, named pipes, and +sockets. Currently, @command{diff} treats symbolic links as if they +were the pointed-to files, except that a recursive @command{diff} +reports an error if it detects infinite loops of symbolic links (e.g., +symbolic links to @file{..}). @command{diff} treats other special +files like regular files if they are specified at the top level, but +simply reports their presence when comparing directories. This means +that @command{patch} cannot represent changes to such files. For +example, if you change which file a symbolic link points to, +@command{diff} outputs the difference between the two files, instead +of the change to the symbolic link. + +@c This might not be a good idea; is it wise for root to install devices +@c this way? +@command{diff} should optionally report changes to special files specially, +and @command{patch} should be extended to understand these extensions. + +@node Unusual File Names +@subsection File Names that Contain Unusual Characters +@cindex file names with unusual characters + +When a file name contains an unusual character like a newline or +white space, @samp{diff -r} generates a patch that @command{patch} cannot +parse. The problem is with format of @command{diff} output, not just with +@command{patch}, because with odd enough file names one can cause +@command{diff} to generate a patch that is syntactically correct but +patches the wrong files. The format of @command{diff} output should be +extended to handle all possible file names. + +@node Time Stamp Order +@subsection Outputting Diffs in Time Stamp Order + +Applying @command{patch} to a multiple-file diff can result in files +whose time stamps are out of order. @acronym{GNU} @command{patch} has +options to restore the time stamps of the updated files +(@pxref{Patching Time Stamps}), but sometimes it is useful to generate +a patch that works even if the recipient does not have @acronym{GNU} patch, +or does not use these options. One way to do this would be to +implement a @command{diff} option to output diffs in time stamp order. + +@node Ignoring Changes +@subsection Ignoring Certain Changes + +It would be nice to have a feature for specifying two strings, one in +@var{from-file} and one in @var{to-file}, which should be considered to +match. Thus, if the two strings are @samp{foo} and @samp{bar}, then if +two lines differ only in that @samp{foo} in file 1 corresponds to +@samp{bar} in file 2, the lines are treated as identical. + +It is not clear how general this feature can or should be, or +what syntax should be used for it. + +A partial substitute is to filter one or both files before comparing, +e.g.: + +@example +sed 's/foo/bar/g' file1 | diff - file2 +@end example + +However, this outputs the filtered text, not the original. + +@node Speedups +@subsection Improving Performance + +When comparing two large directory structures, one of which was +originally copied from the other with time stamps preserved (e.g., +with @samp{cp -pR}), it would greatly improve performance if an option +told @command{diff} to assume that two files with the same size and +time stamps have the same content. @xref{diff Performance}. + +@node Bugs +@section Reporting Bugs +@cindex bug reports +@cindex reporting bugs + +If you think you have found a bug in @acronym{GNU} @command{cmp}, +@command{diff}, @command{diff3}, or @command{sdiff}, please report it +by electronic mail to the +@uref{http://mail.gnu.org/mailman/listinfo/bug-gnu-utils,GNU utilities +bug report mailing list} @email{bug-gnu-utils@@gnu.org}. Please send +bug reports for @acronym{GNU} @command{patch} to +@email{bug-patch@@gnu.org}. Send as precise a description of the +problem as you can, including the output of the @option{--version} +option and sample input files that produce the bug, if applicable. If +you have a nontrivial fix for the bug, please send it as well. If you +have a patch, please send it too. It may simplify the maintainer's +job if the patch is relative to a recent test release, which you can +find in the directory @uref{ftp://alpha.gnu.org/gnu/diffutils/}. + +@node Copying This Manual +@appendix Copying This Manual + +@menu +* GNU Free Documentation License:: License for copying this manual. +@end menu + +@include fdl.texi + +@node Translations +@appendix Translations of This Manual + +Nishio Futoshi of the GNUjdoc project has prepared a Japanese +translation of this manual. Its most recent version can be found at +@uref{http://openlab.ring.gr.jp/gnujdoc/cvsweb/cvsweb.cgi/gnujdoc/}. + +@node Index +@appendix Index + +@printindex cp + +@bye diff --git a/contrib/diff/doc/fdl.texi b/contrib/diff/doc/fdl.texi new file mode 100644 index 0000000..11737cc --- /dev/null +++ b/contrib/diff/doc/fdl.texi @@ -0,0 +1,452 @@ + +@node GNU Free Documentation License +@appendixsec GNU Free Documentation License + +@cindex FDL, GNU Free Documentation License +@center Version 1.2, November 2002 + +@display +Copyright @copyright{} 2000,2001,2002 Free Software Foundation, Inc. +59 Temple Place, Suite 330, Boston, MA 02111-1307, USA + +Everyone is permitted to copy and distribute verbatim copies +of this license document, but changing it is not allowed. +@end display + +@enumerate 0 +@item +PREAMBLE + +The purpose of this License is to make a manual, textbook, or other +functional and useful document @dfn{free} in the sense of freedom: to +assure everyone the effective freedom to copy and redistribute it, +with or without modifying it, either commercially or noncommercially. +Secondarily, this License preserves for the author and publisher a way +to get credit for their work, while not being considered responsible +for modifications made by others. + +This License is a kind of ``copyleft'', which means that derivative +works of the document must themselves be free in the same sense. It +complements the GNU General Public License, which is a copyleft +license designed for free software. + +We have designed this License in order to use it for manuals for free +software, because free software needs free documentation: a free +program should come with manuals providing the same freedoms that the +software does. But this License is not limited to software manuals; +it can be used for any textual work, regardless of subject matter or +whether it is published as a printed book. We recommend this License +principally for works whose purpose is instruction or reference. + +@item +APPLICABILITY AND DEFINITIONS + +This License applies to any manual or other work, in any medium, that +contains a notice placed by the copyright holder saying it can be +distributed under the terms of this License. Such a notice grants a +world-wide, royalty-free license, unlimited in duration, to use that +work under the conditions stated herein. The ``Document'', below, +refers to any such manual or work. Any member of the public is a +licensee, and is addressed as ``you''. You accept the license if you +copy, modify or distribute the work in a way requiring permission +under copyright law. + +A ``Modified Version'' of the Document means any work containing the +Document or a portion of it, either copied verbatim, or with +modifications and/or translated into another language. + +A ``Secondary Section'' is a named appendix or a front-matter section +of the Document that deals exclusively with the relationship of the +publishers or authors of the Document to the Document's overall +subject (or to related matters) and contains nothing that could fall +directly within that overall subject. (Thus, if the Document is in +part a textbook of mathematics, a Secondary Section may not explain +any mathematics.) The relationship could be a matter of historical +connection with the subject or with related matters, or of legal, +commercial, philosophical, ethical or political position regarding +them. + +The ``Invariant Sections'' are certain Secondary Sections whose titles +are designated, as being those of Invariant Sections, in the notice +that says that the Document is released under this License. If a +section does not fit the above definition of Secondary then it is not +allowed to be designated as Invariant. The Document may contain zero +Invariant Sections. If the Document does not identify any Invariant +Sections then there are none. + +The ``Cover Texts'' are certain short passages of text that are listed, +as Front-Cover Texts or Back-Cover Texts, in the notice that says that +the Document is released under this License. A Front-Cover Text may +be at most 5 words, and a Back-Cover Text may be at most 25 words. + +A ``Transparent'' copy of the Document means a machine-readable copy, +represented in a format whose specification is available to the +general public, that is suitable for revising the document +straightforwardly with generic text editors or (for images composed of +pixels) generic paint programs or (for drawings) some widely available +drawing editor, and that is suitable for input to text formatters or +for automatic translation to a variety of formats suitable for input +to text formatters. A copy made in an otherwise Transparent file +format whose markup, or absence of markup, has been arranged to thwart +or discourage subsequent modification by readers is not Transparent. +An image format is not Transparent if used for any substantial amount +of text. A copy that is not ``Transparent'' is called ``Opaque''. + +Examples of suitable formats for Transparent copies include plain +@sc{ascii} without markup, Texinfo input format, La@TeX{} input +format, @acronym{SGML} or @acronym{XML} using a publicly available +@acronym{DTD}, and standard-conforming simple @acronym{HTML}, +PostScript or @acronym{PDF} designed for human modification. Examples +of transparent image formats include @acronym{PNG}, @acronym{XCF} and +@acronym{JPG}. Opaque formats include proprietary formats that can be +read and edited only by proprietary word processors, @acronym{SGML} or +@acronym{XML} for which the @acronym{DTD} and/or processing tools are +not generally available, and the machine-generated @acronym{HTML}, +PostScript or @acronym{PDF} produced by some word processors for +output purposes only. + +The ``Title Page'' means, for a printed book, the title page itself, +plus such following pages as are needed to hold, legibly, the material +this License requires to appear in the title page. For works in +formats which do not have any title page as such, ``Title Page'' means +the text near the most prominent appearance of the work's title, +preceding the beginning of the body of the text. + +A section ``Entitled XYZ'' means a named subunit of the Document whose +title either is precisely XYZ or contains XYZ in parentheses following +text that translates XYZ in another language. (Here XYZ stands for a +specific section name mentioned below, such as ``Acknowledgements'', +``Dedications'', ``Endorsements'', or ``History''.) To ``Preserve the Title'' +of such a section when you modify the Document means that it remains a +section ``Entitled XYZ'' according to this definition. + +The Document may include Warranty Disclaimers next to the notice which +states that this License applies to the Document. These Warranty +Disclaimers are considered to be included by reference in this +License, but only as regards disclaiming warranties: any other +implication that these Warranty Disclaimers may have is void and has +no effect on the meaning of this License. + +@item +VERBATIM COPYING + +You may copy and distribute the Document in any medium, either +commercially or noncommercially, provided that this License, the +copyright notices, and the license notice saying this License applies +to the Document are reproduced in all copies, and that you add no other +conditions whatsoever to those of this License. You may not use +technical measures to obstruct or control the reading or further +copying of the copies you make or distribute. However, you may accept +compensation in exchange for copies. If you distribute a large enough +number of copies you must also follow the conditions in section 3. + +You may also lend copies, under the same conditions stated above, and +you may publicly display copies. + +@item +COPYING IN QUANTITY + +If you publish printed copies (or copies in media that commonly have +printed covers) of the Document, numbering more than 100, and the +Document's license notice requires Cover Texts, you must enclose the +copies in covers that carry, clearly and legibly, all these Cover +Texts: Front-Cover Texts on the front cover, and Back-Cover Texts on +the back cover. Both covers must also clearly and legibly identify +you as the publisher of these copies. The front cover must present +the full title with all words of the title equally prominent and +visible. You may add other material on the covers in addition. +Copying with changes limited to the covers, as long as they preserve +the title of the Document and satisfy these conditions, can be treated +as verbatim copying in other respects. + +If the required texts for either cover are too voluminous to fit +legibly, you should put the first ones listed (as many as fit +reasonably) on the actual cover, and continue the rest onto adjacent +pages. + +If you publish or distribute Opaque copies of the Document numbering +more than 100, you must either include a machine-readable Transparent +copy along with each Opaque copy, or state in or with each Opaque copy +a computer-network location from which the general network-using +public has access to download using public-standard network protocols +a complete Transparent copy of the Document, free of added material. +If you use the latter option, you must take reasonably prudent steps, +when you begin distribution of Opaque copies in quantity, to ensure +that this Transparent copy will remain thus accessible at the stated +location until at least one year after the last time you distribute an +Opaque copy (directly or through your agents or retailers) of that +edition to the public. + +It is requested, but not required, that you contact the authors of the +Document well before redistributing any large number of copies, to give +them a chance to provide you with an updated version of the Document. + +@item +MODIFICATIONS + +You may copy and distribute a Modified Version of the Document under +the conditions of sections 2 and 3 above, provided that you release +the Modified Version under precisely this License, with the Modified +Version filling the role of the Document, thus licensing distribution +and modification of the Modified Version to whoever possesses a copy +of it. In addition, you must do these things in the Modified Version: + +@enumerate A +@item +Use in the Title Page (and on the covers, if any) a title distinct +from that of the Document, and from those of previous versions +(which should, if there were any, be listed in the History section +of the Document). You may use the same title as a previous version +if the original publisher of that version gives permission. + +@item +List on the Title Page, as authors, one or more persons or entities +responsible for authorship of the modifications in the Modified +Version, together with at least five of the principal authors of the +Document (all of its principal authors, if it has fewer than five), +unless they release you from this requirement. + +@item +State on the Title page the name of the publisher of the +Modified Version, as the publisher. + +@item +Preserve all the copyright notices of the Document. + +@item +Add an appropriate copyright notice for your modifications +adjacent to the other copyright notices. + +@item +Include, immediately after the copyright notices, a license notice +giving the public permission to use the Modified Version under the +terms of this License, in the form shown in the Addendum below. + +@item +Preserve in that license notice the full lists of Invariant Sections +and required Cover Texts given in the Document's license notice. + +@item +Include an unaltered copy of this License. + +@item +Preserve the section Entitled ``History'', Preserve its Title, and add +to it an item stating at least the title, year, new authors, and +publisher of the Modified Version as given on the Title Page. If +there is no section Entitled ``History'' in the Document, create one +stating the title, year, authors, and publisher of the Document as +given on its Title Page, then add an item describing the Modified +Version as stated in the previous sentence. + +@item +Preserve the network location, if any, given in the Document for +public access to a Transparent copy of the Document, and likewise +the network locations given in the Document for previous versions +it was based on. These may be placed in the ``History'' section. +You may omit a network location for a work that was published at +least four years before the Document itself, or if the original +publisher of the version it refers to gives permission. + +@item +For any section Entitled ``Acknowledgements'' or ``Dedications'', Preserve +the Title of the section, and preserve in the section all the +substance and tone of each of the contributor acknowledgements and/or +dedications given therein. + +@item +Preserve all the Invariant Sections of the Document, +unaltered in their text and in their titles. Section numbers +or the equivalent are not considered part of the section titles. + +@item +Delete any section Entitled ``Endorsements''. Such a section +may not be included in the Modified Version. + +@item +Do not retitle any existing section to be Entitled ``Endorsements'' or +to conflict in title with any Invariant Section. + +@item +Preserve any Warranty Disclaimers. +@end enumerate + +If the Modified Version includes new front-matter sections or +appendices that qualify as Secondary Sections and contain no material +copied from the Document, you may at your option designate some or all +of these sections as invariant. To do this, add their titles to the +list of Invariant Sections in the Modified Version's license notice. +These titles must be distinct from any other section titles. + +You may add a section Entitled ``Endorsements'', provided it contains +nothing but endorsements of your Modified Version by various +parties---for example, statements of peer review or that the text has +been approved by an organization as the authoritative definition of a +standard. + +You may add a passage of up to five words as a Front-Cover Text, and a +passage of up to 25 words as a Back-Cover Text, to the end of the list +of Cover Texts in the Modified Version. Only one passage of +Front-Cover Text and one of Back-Cover Text may be added by (or +through arrangements made by) any one entity. If the Document already +includes a cover text for the same cover, previously added by you or +by arrangement made by the same entity you are acting on behalf of, +you may not add another; but you may replace the old one, on explicit +permission from the previous publisher that added the old one. + +The author(s) and publisher(s) of the Document do not by this License +give permission to use their names for publicity for or to assert or +imply endorsement of any Modified Version. + +@item +COMBINING DOCUMENTS + +You may combine the Document with other documents released under this +License, under the terms defined in section 4 above for modified +versions, provided that you include in the combination all of the +Invariant Sections of all of the original documents, unmodified, and +list them all as Invariant Sections of your combined work in its +license notice, and that you preserve all their Warranty Disclaimers. + +The combined work need only contain one copy of this License, and +multiple identical Invariant Sections may be replaced with a single +copy. If there are multiple Invariant Sections with the same name but +different contents, make the title of each such section unique by +adding at the end of it, in parentheses, the name of the original +author or publisher of that section if known, or else a unique number. +Make the same adjustment to the section titles in the list of +Invariant Sections in the license notice of the combined work. + +In the combination, you must combine any sections Entitled ``History'' +in the various original documents, forming one section Entitled +``History''; likewise combine any sections Entitled ``Acknowledgements'', +and any sections Entitled ``Dedications''. You must delete all +sections Entitled ``Endorsements.'' + +@item +COLLECTIONS OF DOCUMENTS + +You may make a collection consisting of the Document and other documents +released under this License, and replace the individual copies of this +License in the various documents with a single copy that is included in +the collection, provided that you follow the rules of this License for +verbatim copying of each of the documents in all other respects. + +You may extract a single document from such a collection, and distribute +it individually under this License, provided you insert a copy of this +License into the extracted document, and follow this License in all +other respects regarding verbatim copying of that document. + +@item +AGGREGATION WITH INDEPENDENT WORKS + +A compilation of the Document or its derivatives with other separate +and independent documents or works, in or on a volume of a storage or +distribution medium, is called an ``aggregate'' if the copyright +resulting from the compilation is not used to limit the legal rights +of the compilation's users beyond what the individual works permit. +When the Document is included in an aggregate, this License does not +apply to the other works in the aggregate which are not themselves +derivative works of the Document. + +If the Cover Text requirement of section 3 is applicable to these +copies of the Document, then if the Document is less than one half of +the entire aggregate, the Document's Cover Texts may be placed on +covers that bracket the Document within the aggregate, or the +electronic equivalent of covers if the Document is in electronic form. +Otherwise they must appear on printed covers that bracket the whole +aggregate. + +@item +TRANSLATION + +Translation is considered a kind of modification, so you may +distribute translations of the Document under the terms of section 4. +Replacing Invariant Sections with translations requires special +permission from their copyright holders, but you may include +translations of some or all Invariant Sections in addition to the +original versions of these Invariant Sections. You may include a +translation of this License, and all the license notices in the +Document, and any Warranty Disclaimers, provided that you also include +the original English version of this License and the original versions +of those notices and disclaimers. In case of a disagreement between +the translation and the original version of this License or a notice +or disclaimer, the original version will prevail. + +If a section in the Document is Entitled ``Acknowledgements'', +``Dedications'', or ``History'', the requirement (section 4) to Preserve +its Title (section 1) will typically require changing the actual +title. + +@item +TERMINATION + +You may not copy, modify, sublicense, or distribute the Document except +as expressly provided for under this License. Any other attempt to +copy, modify, sublicense or distribute the Document is void, and will +automatically terminate your rights under this License. However, +parties who have received copies, or rights, from you under this +License will not have their licenses terminated so long as such +parties remain in full compliance. + +@item +FUTURE REVISIONS OF THIS LICENSE + +The Free Software Foundation may publish new, revised versions +of the GNU Free Documentation License from time to time. Such new +versions will be similar in spirit to the present version, but may +differ in detail to address new problems or concerns. See +@uref{http://www.gnu.org/copyleft/}. + +Each version of the License is given a distinguishing version number. +If the Document specifies that a particular numbered version of this +License ``or any later version'' applies to it, you have the option of +following the terms and conditions either of that specified version or +of any later version that has been published (not as a draft) by the +Free Software Foundation. If the Document does not specify a version +number of this License, you may choose any version ever published (not +as a draft) by the Free Software Foundation. +@end enumerate + +@page +@appendixsubsec ADDENDUM: How to use this License for your documents + +To use this License in a document you have written, include a copy of +the License in the document and put the following copyright and +license notices just after the title page: + +@smallexample +@group + Copyright (C) @var{year} @var{your name}. + Permission is granted to copy, distribute and/or modify this document + under the terms of the GNU Free Documentation License, Version 1.2 + or any later version published by the Free Software Foundation; + with no Invariant Sections, no Front-Cover Texts, and no Back-Cover + Texts. A copy of the license is included in the section entitled ``GNU + Free Documentation License''. +@end group +@end smallexample + +If you have Invariant Sections, Front-Cover Texts and Back-Cover Texts, +replace the ``with...Texts.'' line with this: + +@smallexample +@group + with the Invariant Sections being @var{list their titles}, with + the Front-Cover Texts being @var{list}, and with the Back-Cover Texts + being @var{list}. +@end group +@end smallexample + +If you have Invariant Sections without Cover Texts, or some other +combination of the three, merge those two alternatives to suit the +situation. + +If your document contains nontrivial examples of program code, we +recommend releasing these examples in parallel under your choice of +free software license, such as the GNU General Public License, +to permit their use in free software. + +@c Local Variables: +@c ispell-local-pdict: "ispell-dict" +@c End: + diff --git a/contrib/diff/doc/stamp-vti b/contrib/diff/doc/stamp-vti new file mode 100644 index 0000000..cb009ce --- /dev/null +++ b/contrib/diff/doc/stamp-vti @@ -0,0 +1,4 @@ +@set UPDATED 12 April 2004 +@set UPDATED-MONTH April 2004 +@set EDITION 2.8.7 +@set VERSION 2.8.7 diff --git a/contrib/diff/doc/version.texi b/contrib/diff/doc/version.texi new file mode 100644 index 0000000..cb009ce --- /dev/null +++ b/contrib/diff/doc/version.texi @@ -0,0 +1,4 @@ +@set UPDATED 12 April 2004 +@set UPDATED-MONTH April 2004 +@set EDITION 2.8.7 +@set VERSION 2.8.7 diff --git a/contrib/diff/exgettext b/contrib/diff/exgettext new file mode 100755 index 0000000..4fef5ed --- /dev/null +++ b/contrib/diff/exgettext @@ -0,0 +1,115 @@ +#! /bin/sh +# Wrapper around gettext for programs using the msgid convention. + +# Copyright (C) 1998, 2001, 2004 Free Software Foundation, Inc. + +# Written by Paul Eggert . + +# This program is free software; you can redistribute it and/or modify +# it under the terms of the GNU General Public License as published by +# the Free Software Foundation; either version 2, or (at your option) +# any later version. + +# This program is distributed in the hope that it will be useful, +# but WITHOUT ANY WARRANTY; without even the implied warranty of +# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +# GNU General Public License for more details. + +# You should have received a copy of the GNU General Public License +# along with GNU CC; see the file COPYING. If not, write to +# the Free Software Foundation, 59 Temple Place - Suite 330, +# Boston, MA 02111-1307, USA. + +# Always operate in the C locale. +LANG=C +LANGUAGE=C +LC_ALL=C +export LANG LANGUAGE LC_ALL + +# Set AWK if environment has not already set it. +AWK=${AWK-awk} + +# The argument to this wrapper is the xgettext command to be executed. +# Extract the xgettext program name from the rest of the command. +xgettext=${1?} +shift + +# Save work if we're just wrapping a no-op. +case $xgettext in +:) exit;; +esac + +# Find the files to be scanned, and the directory to scan them from. +directory=. +files= +for i +do + case $i in + --directory=*) + directory=`expr " $i" : ' --directory=\(.*\)'`;; + --files-from=*) + files_from=`expr " $i" : ' --files-from=\(.*\)'` + files=`$AWK '/^[^#]/ { print }' $files_from`;; + esac +done + +# Generate keyword options for xgettext, +# by scanning for declarations of functions +# whose parameter names end in "msgid". +generate_keyword_options=' + /^[A-Z_a-z].*\(.*msgid[,)]/ { + + paren_index = index($0, "(") + + name = substr($0, 1, paren_index - 1) + sub(/[^0-9A-Z_a-z]*$/, "", name) + sub(/[ ]+PARAMS/, "", name) + sub(/[ ]+VPARAMS/, "", name) + sub(/.*[^0-9A-Z_a-z]/, "", name) + + args = substr($0, paren_index) + sub(/msgid[,)].*/, "", args) + for (n = 1; sub(/^[^,]*,/, "", args); n++) { + continue; + } + + if (n == 1) { + keyword = name + } else { + keyword = name ":" n + } + + if (! keyword_seen[keyword]++) { + print "--keyword=" keyword + } + } +' +keyword_options=`( + cd $directory && + $AWK "$generate_keyword_options" $files < /dev/null +)` || exit + +# Run the xgettext command, with extra input containing the extra +# msgids that it wouldn't normally find. +generate_emsgids=' + /%e.*}/ { + line = $0 + while ((percent_index = index(line, "%e")) != 0) { + line = substr(line, percent_index + 2) + bracket_index = index(line, "}") + if (bracket_index == 0) { + continue + } + msgid = substr(line, 1, bracket_index - 1) + if (index(msgid, "%") != 0) { + continue + } + printf "#line %d \"%s\"\n", FNR, FILENAME + printf "_(\"%s\")\n", msgid + line = substr(line, bracket_index + 1) + } + } +' +(cd $directory && + $AWK "$generate_emsgids" $files < /dev/null +) | "$xgettext" $keyword_options ${1+"$@"} - diff --git a/contrib/diff/lib/Makefile.am b/contrib/diff/lib/Makefile.am new file mode 100644 index 0000000..dbf8b0f --- /dev/null +++ b/contrib/diff/lib/Makefile.am @@ -0,0 +1,35 @@ +# Automakefile for GNU Diffutils library. + +# Copyright (C) 2001, 2002, 2004 Free Software Foundation, Inc. + +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; either version 2, or (at your option) +## any later version. + +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. + +## You should have received a copy of the GNU General Public License +## along with this program; if not, write to the Free Software +## Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA +## 02111-1307, USA. + +noinst_LIBRARIES = libdiffutils.a + +noinst_HEADERS = cmpbuf.h prepargs.h quotesys.h + +libdiffutils_a_SOURCES = $(lib_SOURCES) +lib_SOURCES = cmpbuf.c prepargs.c quotesys.c + +libdiffutils_a_LIBADD = @ALLOCA@ @LIBOBJS@ +libdiffutils_a_DEPENDENCIES = $(libdiffutils_a_LIBADD) + +BUILT_SOURCES = +DISTCLEANFILES = +EXTRA_DIST = setmode.h waitpid.c +MOSTLYCLEANFILES = + +include gnulib.mk diff --git a/contrib/diff/lib/alloca.c b/contrib/diff/lib/alloca.c new file mode 100644 index 0000000..a580078 --- /dev/null +++ b/contrib/diff/lib/alloca.c @@ -0,0 +1,489 @@ +/* alloca.c -- allocate automatically reclaimed memory + (Mostly) portable public-domain implementation -- D A Gwyn + + This implementation of the PWB library alloca function, + which is used to allocate space off the run-time stack so + that it is automatically reclaimed upon procedure exit, + was inspired by discussions with J. Q. Johnson of Cornell. + J.Otto Tennant contributed the Cray support. + + There are some preprocessor constants that can + be defined when compiling for your specific system, for + improved efficiency; however, the defaults should be okay. + + The general concept of this implementation is to keep + track of all alloca-allocated blocks, and reclaim any + that are found to be deeper in the stack than the current + invocation. This heuristic does not reclaim storage as + soon as it becomes invalid, but it will do so eventually. + + As a special case, alloca(0) reclaims storage without + allocating any. It is a good idea to use alloca(0) in + your main control loop, etc. to force garbage collection. */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include +#include + +#ifdef emacs +# include "lisp.h" +# include "blockinput.h" +# ifdef EMACS_FREE +# undef free +# define free EMACS_FREE +# endif +#else +# define memory_full() abort () +#endif + +/* If compiling with GCC 2, this file's not needed. */ +#if !defined (__GNUC__) || __GNUC__ < 2 + +/* If someone has defined alloca as a macro, + there must be some other way alloca is supposed to work. */ +# ifndef alloca + +# ifdef emacs +# ifdef static +/* actually, only want this if static is defined as "" + -- this is for usg, in which emacs must undefine static + in order to make unexec workable + */ +# ifndef STACK_DIRECTION +you +lose +-- must know STACK_DIRECTION at compile-time +/* Using #error here is not wise since this file should work for + old and obscure compilers. */ +# endif /* STACK_DIRECTION undefined */ +# endif /* static */ +# endif /* emacs */ + +/* If your stack is a linked list of frames, you have to + provide an "address metric" ADDRESS_FUNCTION macro. */ + +# if defined (CRAY) && defined (CRAY_STACKSEG_END) +long i00afunc (); +# define ADDRESS_FUNCTION(arg) (char *) i00afunc (&(arg)) +# else +# define ADDRESS_FUNCTION(arg) &(arg) +# endif + +/* Define STACK_DIRECTION if you know the direction of stack + growth for your system; otherwise it will be automatically + deduced at run-time. + + STACK_DIRECTION > 0 => grows toward higher addresses + STACK_DIRECTION < 0 => grows toward lower addresses + STACK_DIRECTION = 0 => direction of growth unknown */ + +# ifndef STACK_DIRECTION +# define STACK_DIRECTION 0 /* Direction unknown. */ +# endif + +# if STACK_DIRECTION != 0 + +# define STACK_DIR STACK_DIRECTION /* Known at compile-time. */ + +# else /* STACK_DIRECTION == 0; need run-time code. */ + +static int stack_dir; /* 1 or -1 once known. */ +# define STACK_DIR stack_dir + +static void +find_stack_direction (void) +{ + static char *addr = NULL; /* Address of first `dummy', once known. */ + auto char dummy; /* To get stack address. */ + + if (addr == NULL) + { /* Initial entry. */ + addr = ADDRESS_FUNCTION (dummy); + + find_stack_direction (); /* Recurse once. */ + } + else + { + /* Second entry. */ + if (ADDRESS_FUNCTION (dummy) > addr) + stack_dir = 1; /* Stack grew upward. */ + else + stack_dir = -1; /* Stack grew downward. */ + } +} + +# endif /* STACK_DIRECTION == 0 */ + +/* An "alloca header" is used to: + (a) chain together all alloca'ed blocks; + (b) keep track of stack depth. + + It is very important that sizeof(header) agree with malloc + alignment chunk size. The following default should work okay. */ + +# ifndef ALIGN_SIZE +# define ALIGN_SIZE sizeof(double) +# endif + +typedef union hdr +{ + char align[ALIGN_SIZE]; /* To force sizeof(header). */ + struct + { + union hdr *next; /* For chaining headers. */ + char *deep; /* For stack depth measure. */ + } h; +} header; + +static header *last_alloca_header = NULL; /* -> last alloca header. */ + +/* Return a pointer to at least SIZE bytes of storage, + which will be automatically reclaimed upon exit from + the procedure that called alloca. Originally, this space + was supposed to be taken from the current stack frame of the + caller, but that method cannot be made to work for some + implementations of C, for example under Gould's UTX/32. */ + +void * +alloca (size_t size) +{ + auto char probe; /* Probes stack depth: */ + register char *depth = ADDRESS_FUNCTION (probe); + +# if STACK_DIRECTION == 0 + if (STACK_DIR == 0) /* Unknown growth direction. */ + find_stack_direction (); +# endif + + /* Reclaim garbage, defined as all alloca'd storage that + was allocated from deeper in the stack than currently. */ + + { + register header *hp; /* Traverses linked list. */ + +# ifdef emacs + BLOCK_INPUT; +# endif + + for (hp = last_alloca_header; hp != NULL;) + if ((STACK_DIR > 0 && hp->h.deep > depth) + || (STACK_DIR < 0 && hp->h.deep < depth)) + { + register header *np = hp->h.next; + + free (hp); /* Collect garbage. */ + + hp = np; /* -> next header. */ + } + else + break; /* Rest are not deeper. */ + + last_alloca_header = hp; /* -> last valid storage. */ + +# ifdef emacs + UNBLOCK_INPUT; +# endif + } + + if (size == 0) + return NULL; /* No allocation required. */ + + /* Allocate combined header + user data storage. */ + + { + /* Address of header. */ + register header *new; + + size_t combined_size = sizeof (header) + size; + if (combined_size < sizeof (header)) + memory_full (); + + new = malloc (combined_size); + + if (! new) + memory_full (); + + new->h.next = last_alloca_header; + new->h.deep = depth; + + last_alloca_header = new; + + /* User storage begins just after header. */ + + return (void *) (new + 1); + } +} + +# if defined (CRAY) && defined (CRAY_STACKSEG_END) + +# ifdef DEBUG_I00AFUNC +# include +# endif + +# ifndef CRAY_STACK +# define CRAY_STACK +# ifndef CRAY2 +/* Stack structures for CRAY-1, CRAY X-MP, and CRAY Y-MP */ +struct stack_control_header + { + long shgrow:32; /* Number of times stack has grown. */ + long shaseg:32; /* Size of increments to stack. */ + long shhwm:32; /* High water mark of stack. */ + long shsize:32; /* Current size of stack (all segments). */ + }; + +/* The stack segment linkage control information occurs at + the high-address end of a stack segment. (The stack + grows from low addresses to high addresses.) The initial + part of the stack segment linkage control information is + 0200 (octal) words. This provides for register storage + for the routine which overflows the stack. */ + +struct stack_segment_linkage + { + long ss[0200]; /* 0200 overflow words. */ + long sssize:32; /* Number of words in this segment. */ + long ssbase:32; /* Offset to stack base. */ + long:32; + long sspseg:32; /* Offset to linkage control of previous + segment of stack. */ + long:32; + long sstcpt:32; /* Pointer to task common address block. */ + long sscsnm; /* Private control structure number for + microtasking. */ + long ssusr1; /* Reserved for user. */ + long ssusr2; /* Reserved for user. */ + long sstpid; /* Process ID for pid based multi-tasking. */ + long ssgvup; /* Pointer to multitasking thread giveup. */ + long sscray[7]; /* Reserved for Cray Research. */ + long ssa0; + long ssa1; + long ssa2; + long ssa3; + long ssa4; + long ssa5; + long ssa6; + long ssa7; + long sss0; + long sss1; + long sss2; + long sss3; + long sss4; + long sss5; + long sss6; + long sss7; + }; + +# else /* CRAY2 */ +/* The following structure defines the vector of words + returned by the STKSTAT library routine. */ +struct stk_stat + { + long now; /* Current total stack size. */ + long maxc; /* Amount of contiguous space which would + be required to satisfy the maximum + stack demand to date. */ + long high_water; /* Stack high-water mark. */ + long overflows; /* Number of stack overflow ($STKOFEN) calls. */ + long hits; /* Number of internal buffer hits. */ + long extends; /* Number of block extensions. */ + long stko_mallocs; /* Block allocations by $STKOFEN. */ + long underflows; /* Number of stack underflow calls ($STKRETN). */ + long stko_free; /* Number of deallocations by $STKRETN. */ + long stkm_free; /* Number of deallocations by $STKMRET. */ + long segments; /* Current number of stack segments. */ + long maxs; /* Maximum number of stack segments so far. */ + long pad_size; /* Stack pad size. */ + long current_address; /* Current stack segment address. */ + long current_size; /* Current stack segment size. This + number is actually corrupted by STKSTAT to + include the fifteen word trailer area. */ + long initial_address; /* Address of initial segment. */ + long initial_size; /* Size of initial segment. */ + }; + +/* The following structure describes the data structure which trails + any stack segment. I think that the description in 'asdef' is + out of date. I only describe the parts that I am sure about. */ + +struct stk_trailer + { + long this_address; /* Address of this block. */ + long this_size; /* Size of this block (does not include + this trailer). */ + long unknown2; + long unknown3; + long link; /* Address of trailer block of previous + segment. */ + long unknown5; + long unknown6; + long unknown7; + long unknown8; + long unknown9; + long unknown10; + long unknown11; + long unknown12; + long unknown13; + long unknown14; + }; + +# endif /* CRAY2 */ +# endif /* not CRAY_STACK */ + +# ifdef CRAY2 +/* Determine a "stack measure" for an arbitrary ADDRESS. + I doubt that "lint" will like this much. */ + +static long +i00afunc (long *address) +{ + struct stk_stat status; + struct stk_trailer *trailer; + long *block, size; + long result = 0; + + /* We want to iterate through all of the segments. The first + step is to get the stack status structure. We could do this + more quickly and more directly, perhaps, by referencing the + $LM00 common block, but I know that this works. */ + + STKSTAT (&status); + + /* Set up the iteration. */ + + trailer = (struct stk_trailer *) (status.current_address + + status.current_size + - 15); + + /* There must be at least one stack segment. Therefore it is + a fatal error if "trailer" is null. */ + + if (trailer == 0) + abort (); + + /* Discard segments that do not contain our argument address. */ + + while (trailer != 0) + { + block = (long *) trailer->this_address; + size = trailer->this_size; + if (block == 0 || size == 0) + abort (); + trailer = (struct stk_trailer *) trailer->link; + if ((block <= address) && (address < (block + size))) + break; + } + + /* Set the result to the offset in this segment and add the sizes + of all predecessor segments. */ + + result = address - block; + + if (trailer == 0) + { + return result; + } + + do + { + if (trailer->this_size <= 0) + abort (); + result += trailer->this_size; + trailer = (struct stk_trailer *) trailer->link; + } + while (trailer != 0); + + /* We are done. Note that if you present a bogus address (one + not in any segment), you will get a different number back, formed + from subtracting the address of the first block. This is probably + not what you want. */ + + return (result); +} + +# else /* not CRAY2 */ +/* Stack address function for a CRAY-1, CRAY X-MP, or CRAY Y-MP. + Determine the number of the cell within the stack, + given the address of the cell. The purpose of this + routine is to linearize, in some sense, stack addresses + for alloca. */ + +static long +i00afunc (long address) +{ + long stkl = 0; + + long size, pseg, this_segment, stack; + long result = 0; + + struct stack_segment_linkage *ssptr; + + /* Register B67 contains the address of the end of the + current stack segment. If you (as a subprogram) store + your registers on the stack and find that you are past + the contents of B67, you have overflowed the segment. + + B67 also points to the stack segment linkage control + area, which is what we are really interested in. */ + + stkl = CRAY_STACKSEG_END (); + ssptr = (struct stack_segment_linkage *) stkl; + + /* If one subtracts 'size' from the end of the segment, + one has the address of the first word of the segment. + + If this is not the first segment, 'pseg' will be + nonzero. */ + + pseg = ssptr->sspseg; + size = ssptr->sssize; + + this_segment = stkl - size; + + /* It is possible that calling this routine itself caused + a stack overflow. Discard stack segments which do not + contain the target address. */ + + while (!(this_segment <= address && address <= stkl)) + { +# ifdef DEBUG_I00AFUNC + fprintf (stderr, "%011o %011o %011o\n", this_segment, address, stkl); +# endif + if (pseg == 0) + break; + stkl = stkl - pseg; + ssptr = (struct stack_segment_linkage *) stkl; + size = ssptr->sssize; + pseg = ssptr->sspseg; + this_segment = stkl - size; + } + + result = address - this_segment; + + /* If you subtract pseg from the current end of the stack, + you get the address of the previous stack segment's end. + This seems a little convoluted to me, but I'll bet you save + a cycle somewhere. */ + + while (pseg != 0) + { +# ifdef DEBUG_I00AFUNC + fprintf (stderr, "%011o %011o\n", pseg, size); +# endif + stkl = stkl - pseg; + ssptr = (struct stack_segment_linkage *) stkl; + size = ssptr->sssize; + pseg = ssptr->sspseg; + result += size; + } + return (result); +} + +# endif /* not CRAY2 */ +# endif /* CRAY */ + +# endif /* no alloca */ +#endif /* not GCC version 2 */ diff --git a/contrib/diff/lib/alloca_.h b/contrib/diff/lib/alloca_.h new file mode 100644 index 0000000..8ed34fd --- /dev/null +++ b/contrib/diff/lib/alloca_.h @@ -0,0 +1,68 @@ +/* Memory allocation on the stack. + Copyright (C) 1995, 1999, 2001-2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published + by the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU + General Public License for more details. + + You should have received a copy of the GNU General Public + License along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, + USA. */ + +/* When this file is included, it may be preceded only by preprocessor + declarations. Thanks to AIX. Therefore we include it right after + "config.h", not later. */ + +#ifndef _ALLOCA_H +# define _ALLOCA_H + +/* alloca(N) returns a pointer (void* or char*) to N bytes of memory + allocated on the stack, and which will last until the function returns. + Use of alloca should be avoided: + - inside arguments of function calls - undefined behaviour, + - in inline functions - the allocation may actually last until the + calling function returns, + - for huge N (say, N >= 65536) - you never know how large (or small) + the stack is, and when the stack cannot fulfill the memory allocation + request, the program just crashes. + */ + +# ifdef __GNUC__ +# ifndef alloca +# define alloca __builtin_alloca +# endif +# else +# ifdef _MSC_VER +# include +# define alloca _alloca +# else +# if HAVE_ALLOCA_H +# include +# else +# ifdef _AIX + # pragma alloca +# else +# ifdef __hpux /* This section must match that of bison generated files. */ +# ifdef __cplusplus +extern "C" void *alloca (unsigned int); +# else /* not __cplusplus */ +extern void *alloca (); +# endif /* not __cplusplus */ +# else /* not __hpux */ +# ifndef alloca +extern char *alloca (); +# endif +# endif /* __hpux */ +# endif +# endif +# endif +# endif + +#endif /* _ALLOCA_H */ diff --git a/contrib/diff/lib/basename.c b/contrib/diff/lib/basename.c new file mode 100644 index 0000000..5ff2989 --- /dev/null +++ b/contrib/diff/lib/basename.c @@ -0,0 +1,79 @@ +/* basename.c -- return the last element in a path + + Copyright (C) 1990, 1998, 1999, 2000, 2001, 2003 Free Software + Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "dirname.h" +#include + +/* In general, we can't use the builtin `basename' function if available, + since it has different meanings in different environments. + In some environments the builtin `basename' modifies its argument. + + Return the address of the last file name component of NAME. If + NAME has no file name components because it is all slashes, return + NAME if it is empty, the address of its last slash otherwise. */ + +char * +base_name (char const *name) +{ + char const *base = name + FILESYSTEM_PREFIX_LEN (name); + char const *p; + + for (p = base; *p; p++) + { + if (ISSLASH (*p)) + { + /* Treat multiple adjacent slashes like a single slash. */ + do p++; + while (ISSLASH (*p)); + + /* If the file name ends in slash, use the trailing slash as + the basename if no non-slashes have been found. */ + if (! *p) + { + if (ISSLASH (*base)) + base = p - 1; + break; + } + + /* *P is a non-slash preceded by a slash. */ + base = p; + } + } + + return (char *) base; +} + +/* Return the length of of the basename NAME. Typically NAME is the + value returned by base_name. Act like strlen (NAME), except omit + redundant trailing slashes. */ + +size_t +base_len (char const *name) +{ + size_t len; + + for (len = strlen (name); 1 < len && ISSLASH (name[len - 1]); len--) + continue; + + return len; +} diff --git a/contrib/diff/lib/c-stack.c b/contrib/diff/lib/c-stack.c new file mode 100644 index 0000000..a018e51 --- /dev/null +++ b/contrib/diff/lib/c-stack.c @@ -0,0 +1,311 @@ +/* Stack overflow handling. + + Copyright (C) 2002, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +/* NOTES: + + A program that uses alloca, dynamic arrays, or large local + variables may extend the stack by more than a page at a time. If + so, when the stack overflows the operating system may not detect + the overflow until the program uses the array, and this module may + incorrectly report a program error instead of a stack overflow. + + To avoid this problem, allocate only small objects on the stack; a + program should be OK if it limits single allocations to a page or + less. Allocate larger arrays in static storage, or on the heap + (e.g., with malloc). Yes, this is a pain, but we don't know of any + better solution that is portable. + + No attempt has been made to deal with multithreaded applications. */ + +#if HAVE_CONFIG_H +# include +#endif + +#ifndef __attribute__ +# if __GNUC__ < 3 || __STRICT_ANSI__ +# define __attribute__(x) +# endif +#endif + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +#include +#ifndef ENOTSUP +# define ENOTSUP EINVAL +#endif +#ifndef EOVERFLOW +# define EOVERFLOW EINVAL +#endif + +#include +#if ! HAVE_STACK_T && ! defined stack_t +typedef struct sigaltstack stack_t; +#endif + +#include +#include + +#if HAVE_SYS_RESOURCE_H +/* Include sys/time.h here, because... + SunOS-4.1.x fails to include . + This gives "incomplete type" errors for ru_utime and tu_stime. */ +# if HAVE_SYS_TIME_H +# include +# endif +# include +#endif + +#if HAVE_UCONTEXT_H +# include +#endif + +#if HAVE_UNISTD_H +# include +#endif +#ifndef STDERR_FILENO +# define STDERR_FILENO 2 +#endif + +#if DEBUG +# include +#endif + +#include "c-stack.h" +#include "exitfail.h" + +#if (HAVE_STRUCT_SIGACTION_SA_SIGACTION && defined SA_NODEFER \ + && defined SA_ONSTACK && defined SA_RESETHAND && defined SA_SIGINFO) +# define SIGACTION_WORKS 1 +#else +# define SIGACTION_WORKS 0 +#endif + +extern char *program_name; + +/* The user-specified action to take when a SEGV-related program error + or stack overflow occurs. */ +static void (* volatile segv_action) (int); + +/* Translated messages for program errors and stack overflow. Do not + translate them in the signal handler, since gettext is not + async-signal-safe. */ +static char const * volatile program_error_message; +static char const * volatile stack_overflow_message; + +/* Output an error message, then exit with status EXIT_FAILURE if it + appears to have been a stack overflow, or with a core dump + otherwise. This function is async-signal-safe. */ + +static void die (int) __attribute__ ((noreturn)); +static void +die (int signo) +{ + char const *message; + segv_action (signo); + message = signo ? program_error_message : stack_overflow_message; + write (STDERR_FILENO, program_name, strlen (program_name)); + write (STDERR_FILENO, ": ", 2); + write (STDERR_FILENO, message, strlen (message)); + write (STDERR_FILENO, "\n", 1); + if (! signo) + _exit (exit_failure); + kill (getpid (), signo); + abort (); +} + +#if HAVE_SIGALTSTACK && HAVE_DECL_SIGALTSTACK + +/* Direction of the C runtime stack. This function is + async-signal-safe. */ + +# if STACK_DIRECTION +# define find_stack_direction(ptr) STACK_DIRECTION +# else +static int +find_stack_direction (char const *addr) +{ + char dummy; + return ! addr ? find_stack_direction (&dummy) : addr < &dummy ? 1 : -1; +} +# endif + +/* Storage for the alternate signal stack. */ +static union +{ + char buffer[SIGSTKSZ]; + + /* These other members are for proper alignment. There's no + standard way to guarantee stack alignment, but this seems enough + in practice. */ + long double ld; + long l; + void *p; +} alternate_signal_stack; + +# if SIGACTION_WORKS + +/* Handle a segmentation violation and exit. This function is + async-signal-safe. */ + +static void segv_handler (int, siginfo_t *, void *) __attribute__((noreturn)); +static void +segv_handler (int signo, siginfo_t *info, + void *context __attribute__ ((unused))) +{ + /* Clear SIGNO if it seems to have been a stack overflow. */ + if (0 < info->si_code) + { +# if ! HAVE_XSI_STACK_OVERFLOW_HEURISTIC + /* We can't easily determine whether it is a stack overflow; so + assume that the rest of our program is perfect (!) and that + this segmentation violation is a stack overflow. */ + signo = 0; +# else + /* If the faulting address is within the stack, or within one + page of the stack end, assume that it is a stack + overflow. */ + ucontext_t const *user_context = context; + char const *stack_base = user_context->uc_stack.ss_sp; + size_t stack_size = user_context->uc_stack.ss_size; + char const *faulting_address = info->si_addr; + size_t s = faulting_address - stack_base; + size_t page_size = sysconf (_SC_PAGESIZE); + if (find_stack_direction (0) < 0) + s += page_size; + if (s < stack_size + page_size) + signo = 0; + +# if DEBUG + { + char buf[1024]; + sprintf (buf, + "segv_handler fault=%p base=%p size=%lx page=%lx signo=%d\n", + faulting_address, stack_base, (unsigned long) stack_size, + (unsigned long) page_size, signo); + write (STDERR_FILENO, buf, strlen (buf)); + } +# endif +# endif + } + + die (signo); +} +# endif + +static void +null_action (int signo __attribute__ ((unused))) +{ +} + +/* Set up ACTION so that it is invoked on C stack overflow. Return -1 + (setting errno) if this cannot be done. + + When ACTION is called, it is passed an argument equal to SIGSEGV + for a segmentation violation that does not appear related to stack + overflow, and is passed zero otherwise. On many platforms it is + hard to tell; when in doubt, zero is passed. + + A null ACTION acts like an action that does nothing. + + ACTION must be async-signal-safe. ACTION together with its callees + must not require more than SIGSTKSZ bytes of stack space. */ + +int +c_stack_action (void (*action) (int)) +{ + int r; + stack_t st; + st.ss_flags = 0; + st.ss_sp = alternate_signal_stack.buffer; + st.ss_size = sizeof alternate_signal_stack.buffer; + r = sigaltstack (&st, 0); + if (r != 0) + return r; + + segv_action = action ? action : null_action; + program_error_message = _("program error"); + stack_overflow_message = _("stack overflow"); + + { +# if SIGACTION_WORKS + struct sigaction act; + sigemptyset (&act.sa_mask); + + /* POSIX 1003.1-2001 says SA_RESETHAND implies SA_NODEFER, but + this is not true on Solaris 8 at least. It doesn't hurt to use + SA_NODEFER here, so leave it in. */ + act.sa_flags = SA_NODEFER | SA_ONSTACK | SA_RESETHAND | SA_SIGINFO; + + act.sa_sigaction = segv_handler; + + return sigaction (SIGSEGV, &act, 0); +# else + return signal (SIGSEGV, die) == SIG_ERR ? -1 : 0; +# endif + } +} + +#else /* ! (HAVE_SIGALTSTACK && HAVE_DECL_SIGALTSTACK) */ + +int +c_stack_action (void (*action) (int) __attribute__ ((unused))) +{ + errno = ENOTSUP; + return -1; +} + +#endif + + + +#if DEBUG + +int volatile exit_failure; + +static long +recurse (char *p) +{ + char array[500]; + array[0] = 1; + return *p + recurse (array); +} + +char *program_name; + +int +main (int argc __attribute__ ((unused)), char **argv) +{ + program_name = argv[0]; + fprintf (stderr, + "The last output line should contain \"stack overflow\".\n"); + if (c_stack_action (0) == 0) + return recurse ("\1"); + perror ("c_stack_action"); + return 1; +} + +#endif /* DEBUG */ + +/* +Local Variables: +compile-command: "gcc -DDEBUG -DHAVE_CONFIG_H -I.. -g -O -Wall -W c-stack.c" +End: +*/ diff --git a/contrib/diff/lib/c-stack.h b/contrib/diff/lib/c-stack.h new file mode 100644 index 0000000..968fd52 --- /dev/null +++ b/contrib/diff/lib/c-stack.h @@ -0,0 +1,19 @@ +/* Stack overflow handling. + + Copyright (C) 2002, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +int c_stack_action (void (*) (int)); diff --git a/contrib/diff/lib/cmpbuf.c b/contrib/diff/lib/cmpbuf.c new file mode 100644 index 0000000..858aa39 --- /dev/null +++ b/contrib/diff/lib/cmpbuf.c @@ -0,0 +1,147 @@ +/* Buffer primitives for comparison operations. + + Copyright (C) 1993, 1995, 1998, 2001, 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include +#include + +#include +#ifndef SA_RESTART +# ifdef SA_INTERRUPT /* e.g. SunOS 4.1.x */ +# define SA_RESTART SA_INTERRUPT +# else +# define SA_RESTART 0 +# endif +#endif + +#if HAVE_UNISTD_H +# include +#endif + +#if HAVE_INTTYPES_H +# include +#endif + +#include +#include "cmpbuf.h" + +/* Determine whether an integer type is signed, and its bounds. + This code assumes two's (or one's!) complement with no holes. */ + +/* The extra casts work around common compiler bugs, + e.g. Cray C 5.0.3.0 when t == time_t. */ +#ifndef TYPE_SIGNED +# define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) +#endif +#ifndef TYPE_MINIMUM +# define TYPE_MINIMUM(t) ((t) (TYPE_SIGNED (t) \ + ? ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1) \ + : (t) 0)) +#endif +#ifndef TYPE_MAXIMUM +# define TYPE_MAXIMUM(t) ((t) (~ (t) 0 - TYPE_MINIMUM (t))) +#endif + +#ifndef PTRDIFF_MAX +# define PTRDIFF_MAX TYPE_MAXIMUM (ptrdiff_t) +#endif +#ifndef SIZE_MAX +# define SIZE_MAX TYPE_MAXIMUM (size_t) +#endif +#ifndef SSIZE_MAX +# define SSIZE_MAX TYPE_MAXIMUM (ssize_t) +#endif + +#undef MIN +#define MIN(a, b) ((a) <= (b) ? (a) : (b)) + +/* Read NBYTES bytes from descriptor FD into BUF. + NBYTES must not be SIZE_MAX. + Return the number of characters successfully read. + On error, return SIZE_MAX, setting errno. + The number returned is always NBYTES unless end-of-file or error. */ + +size_t +block_read (int fd, char *buf, size_t nbytes) +{ + char *bp = buf; + char const *buflim = buf + nbytes; + size_t readlim = SSIZE_MAX; + + do + { + size_t bytes_to_read = MIN (buflim - bp, readlim); + ssize_t nread = read (fd, bp, bytes_to_read); + if (nread <= 0) + { + if (nread == 0) + break; + + /* Accommodate Tru64 5.1, which can't read more than INT_MAX + bytes at a time. They call that a 64-bit OS? */ + if (errno == EINVAL && INT_MAX < bytes_to_read) + { + readlim = INT_MAX; + continue; + } + + /* This is needed for programs that have signal handlers on + older hosts without SA_RESTART. It also accommodates + ancient AIX hosts that set errno to EINTR after uncaught + SIGCONT. See + (1993-04-22). */ + if (! SA_RESTART && errno == EINTR) + continue; + + return SIZE_MAX; + } + bp += nread; + } + while (bp < buflim); + + return bp - buf; +} + +/* Least common multiple of two buffer sizes A and B. However, if + either A or B is zero, or if the multiple is greater than LCM_MAX, + return a reasonable buffer size. */ + +size_t +buffer_lcm (size_t a, size_t b, size_t lcm_max) +{ + size_t lcm, m, n, q, r; + + /* Yield reasonable values if buffer sizes are zero. */ + if (!a) + return b ? b : 8 * 1024; + if (!b) + return a; + + /* n = gcd (a, b) */ + for (m = a, n = b; (r = m % n) != 0; m = n, n = r) + continue; + + /* Yield a if there is an overflow. */ + q = a / n; + lcm = q * b; + return lcm <= lcm_max && lcm / b == q ? lcm : a; +} diff --git a/contrib/diff/lib/cmpbuf.h b/contrib/diff/lib/cmpbuf.h new file mode 100644 index 0000000..9b29cd0 --- /dev/null +++ b/contrib/diff/lib/cmpbuf.h @@ -0,0 +1,21 @@ +/* Buffer primitives for comparison operations. + + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +size_t block_read (int, char *, size_t); +size_t buffer_lcm (size_t, size_t, size_t); diff --git a/contrib/diff/lib/dirname.c b/contrib/diff/lib/dirname.c new file mode 100644 index 0000000..815192a --- /dev/null +++ b/contrib/diff/lib/dirname.c @@ -0,0 +1,119 @@ +/* dirname.c -- return all but the last element in a path + Copyright (C) 1990, 1998, 2000, 2001, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "dirname.h" + +#include +#include "xalloc.h" + +/* Return the length of `dirname (PATH)', or zero if PATH is + in the working directory. Works properly even if + there are trailing slashes (by effectively ignoring them). */ +size_t +dir_len (char const *path) +{ + size_t prefix_length = FILESYSTEM_PREFIX_LEN (path); + size_t length; + + /* Strip the basename and any redundant slashes before it. */ + for (length = base_name (path) - path; prefix_length < length; length--) + if (! ISSLASH (path[length - 1])) + return length; + + /* But don't strip the only slash from "/". */ + return prefix_length + ISSLASH (path[prefix_length]); +} + +/* Return the leading directories part of PATH, + allocated with xmalloc. + Works properly even if there are trailing slashes + (by effectively ignoring them). */ + +char * +dir_name (char const *path) +{ + size_t length = dir_len (path); + int append_dot = (length == FILESYSTEM_PREFIX_LEN (path)); + char *newpath = xmalloc (length + append_dot + 1); + memcpy (newpath, path, length); + if (append_dot) + newpath[length++] = '.'; + newpath[length] = 0; + return newpath; +} + +#ifdef TEST_DIRNAME +/* + +Run the test like this (expect no output): + gcc -DHAVE_CONFIG_H -DTEST_DIRNAME -I.. -O -Wall \ + basename.c dirname.c xmalloc.c error.c + sed -n '/^BEGIN-DATA$/,/^END-DATA$/p' dirname.c|grep -v DATA|./a.out + +If it's been built on a DOS or Windows platforms, run another test like +this (again, expect no output): + sed -n '/^BEGIN-DOS-DATA$/,/^END-DOS-DATA$/p' dirname.c|grep -v DATA|./a.out + +BEGIN-DATA +foo//// . +bar/foo//// bar +foo/ . +/ / +. . +a . +END-DATA + +BEGIN-DOS-DATA +c:///// c:/ +c:/ c:/ +c:/. c:/ +c:foo c:. +c:foo/bar c:foo +END-DOS-DATA + +*/ + +# define MAX_BUFF_LEN 1024 +# include + +char *program_name; + +int +main (int argc, char *argv[]) +{ + char buff[MAX_BUFF_LEN + 1]; + + program_name = argv[0]; + + buff[MAX_BUFF_LEN] = 0; + while (fgets (buff, MAX_BUFF_LEN, stdin) && buff[0]) + { + char path[MAX_BUFF_LEN]; + char expected_result[MAX_BUFF_LEN]; + char const *result; + sscanf (buff, "%s %s", path, expected_result); + result = dir_name (path); + if (strcmp (result, expected_result)) + printf ("%s: got %s, expected %s\n", path, result, expected_result); + } + return 0; +} +#endif diff --git a/contrib/diff/lib/dirname.h b/contrib/diff/lib/dirname.h new file mode 100644 index 0000000..62da37d --- /dev/null +++ b/contrib/diff/lib/dirname.h @@ -0,0 +1,43 @@ +/* Take file names apart into directory and base names. + + Copyright (C) 1998, 2001, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef DIRNAME_H_ +# define DIRNAME_H_ 1 + +# include + +# ifndef DIRECTORY_SEPARATOR +# define DIRECTORY_SEPARATOR '/' +# endif + +# ifndef ISSLASH +# define ISSLASH(C) ((C) == DIRECTORY_SEPARATOR) +# endif + +# ifndef FILESYSTEM_PREFIX_LEN +# define FILESYSTEM_PREFIX_LEN(Filename) 0 +# endif + +char *base_name (char const *path); +char *dir_name (char const *path); +size_t base_len (char const *path); +size_t dir_len (char const *path); + +int strip_trailing_slashes (char *path); + +#endif /* not DIRNAME_H_ */ diff --git a/contrib/diff/lib/error.c b/contrib/diff/lib/error.c new file mode 100644 index 0000000..1149235 --- /dev/null +++ b/contrib/diff/lib/error.c @@ -0,0 +1,310 @@ +/* Error handler for noninteractive utilities + Copyright (C) 1990-1998, 2000-2002, 2003 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by David MacKenzie . */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include "error.h" + +#include +#include +#include +#include + +#ifdef _LIBC +# include +#else +# include "gettext.h" +#endif + +#ifdef _LIBC +# include +# define mbsrtowcs __mbsrtowcs +#endif + +#if !_LIBC +# include "unlocked-io.h" +#endif + +#ifndef _ +# define _(String) String +#endif + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +void (*error_print_progname) (void); + +/* This variable is incremented each time `error' is called. */ +unsigned int error_message_count; + +#ifdef _LIBC +/* In the GNU C library, there is a predefined variable for this. */ + +# define program_name program_invocation_name +# include +# include + +/* In GNU libc we want do not want to use the common name `error' directly. + Instead make it a weak alias. */ +extern void __error (int status, int errnum, const char *message, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); +extern void __error_at_line (int status, int errnum, const char *file_name, + unsigned int line_number, const char *message, + ...) + __attribute__ ((__format__ (__printf__, 5, 6)));; +# define error __error +# define error_at_line __error_at_line + +# include +# define fflush(s) INTUSE(_IO_fflush) (s) +# undef putc +# define putc(c, fp) INTUSE(_IO_putc) (c, fp) + +# include + +#else /* not _LIBC */ + +# if !HAVE_DECL_STRERROR_R && STRERROR_R_CHAR_P +# ifndef HAVE_DECL_STRERROR_R +"this configure-time declaration test was not run" +# endif +char *strerror_r (); +# endif + +# ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +# endif + +/* The calling program should define program_name and set it to the + name of the executing program. */ +extern char *program_name; + +# if HAVE_STRERROR_R || defined strerror_r +# define __strerror_r strerror_r +# endif +#endif /* not _LIBC */ + +static void +print_errno_message (int errnum) +{ + char const *s; + +#if defined HAVE_STRERROR_R || _LIBC + char errbuf[1024]; +# if STRERROR_R_CHAR_P || _LIBC + s = __strerror_r (errnum, errbuf, sizeof errbuf); +# else + if (__strerror_r (errnum, errbuf, sizeof errbuf) == 0) + s = errbuf; + else + s = 0; +# endif +#else + s = strerror (errnum); +#endif + +#if !_LIBC + if (! s) + s = _("Unknown system error"); +#endif + +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + { + __fwprintf (stderr, L": %s", s); + return; + } +#endif + + fprintf (stderr, ": %s", s); +} + +static void +error_tail (int status, int errnum, const char *message, va_list args) +{ +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + { +# define ALLOCA_LIMIT 2000 + size_t len = strlen (message) + 1; + const wchar_t *wmessage = L"out of memory"; + wchar_t *wbuf = (len < ALLOCA_LIMIT + ? alloca (len * sizeof *wbuf) + : len <= SIZE_MAX / sizeof *wbuf + ? malloc (len * sizeof *wbuf) + : NULL); + + if (wbuf) + { + size_t res; + mbstate_t st; + const char *tmp = message; + memset (&st, '\0', sizeof (st)); + res = mbsrtowcs (wbuf, &tmp, len, &st); + wmessage = res == (size_t) -1 ? L"???" : wbuf; + } + + __vfwprintf (stderr, wmessage, args); + if (! (len < ALLOCA_LIMIT)) + free (wbuf); + } + else +#endif + vfprintf (stderr, message, args); + va_end (args); + + ++error_message_count; + if (errnum) + print_errno_message (errnum); +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + putwc (L'\n', stderr); + else +#endif + putc ('\n', stderr); + fflush (stderr); + if (status) + exit (status); +} + + +/* Print the program name and error message MESSAGE, which is a printf-style + format string with optional args. + If ERRNUM is nonzero, print its corresponding system error message. + Exit with status STATUS if it is nonzero. */ +void +error (int status, int errnum, const char *message, ...) +{ + va_list args; + +#if defined _LIBC && defined __libc_ptf_call + /* We do not want this call to be cut short by a thread + cancellation. Therefore disable cancellation for now. */ + int state = PTHREAD_CANCEL_ENABLE; + __libc_ptf_call (pthread_setcancelstate, (PTHREAD_CANCEL_DISABLE, &state), + 0); +#endif + + fflush (stdout); +#ifdef _LIBC + _IO_flockfile (stderr); +#endif + if (error_print_progname) + (*error_print_progname) (); + else + { +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s: ", program_name); + else +#endif + fprintf (stderr, "%s: ", program_name); + } + + va_start (args, message); + error_tail (status, errnum, message, args); + +#ifdef _LIBC + _IO_funlockfile (stderr); +# ifdef __libc_ptf_call + __libc_ptf_call (pthread_setcancelstate, (state, NULL), 0); +# endif +#endif +} + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +int error_one_per_line; + +void +error_at_line (int status, int errnum, const char *file_name, + unsigned int line_number, const char *message, ...) +{ + va_list args; + + if (error_one_per_line) + { + static const char *old_file_name; + static unsigned int old_line_number; + + if (old_line_number == line_number + && (file_name == old_file_name + || strcmp (old_file_name, file_name) == 0)) + /* Simply return and print nothing. */ + return; + + old_file_name = file_name; + old_line_number = line_number; + } + +#if defined _LIBC && defined __libc_ptf_call + /* We do not want this call to be cut short by a thread + cancellation. Therefore disable cancellation for now. */ + int state = PTHREAD_CANCEL_ENABLE; + __libc_ptf_call (pthread_setcancelstate, (PTHREAD_CANCEL_DISABLE, &state), + 0); +#endif + + fflush (stdout); +#ifdef _LIBC + _IO_flockfile (stderr); +#endif + if (error_print_progname) + (*error_print_progname) (); + else + { +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s: ", program_name); + else +#endif + fprintf (stderr, "%s:", program_name); + } + + if (file_name != NULL) + { +#if _LIBC + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s:%d: ", file_name, line_number); + else +#endif + fprintf (stderr, "%s:%d: ", file_name, line_number); + } + + va_start (args, message); + error_tail (status, errnum, message, args); + +#ifdef _LIBC + _IO_funlockfile (stderr); +# ifdef __libc_ptf_call + __libc_ptf_call (pthread_setcancelstate, (state, NULL), 0); +# endif +#endif +} + +#ifdef _LIBC +/* Make the weak alias. */ +# undef error +# undef error_at_line +weak_alias (__error, error) +weak_alias (__error_at_line, error_at_line) +#endif diff --git a/contrib/diff/lib/error.h b/contrib/diff/lib/error.h new file mode 100644 index 0000000..8ed6359 --- /dev/null +++ b/contrib/diff/lib/error.h @@ -0,0 +1,66 @@ +/* Declaration for error-reporting function + Copyright (C) 1995, 1996, 1997, 2003 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _ERROR_H +#define _ERROR_H 1 + +#ifndef __attribute__ +/* This feature is available in gcc versions 2.5 and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 5) +# define __attribute__(Spec) /* empty */ +# endif +/* The __-protected variants of `format' and `printf' attributes + are accepted by gcc versions 2.6.4 (effectively 2.7) and later. */ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 7) +# define __format__ format +# define __printf__ printf +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* Print a message with `fprintf (stderr, FORMAT, ...)'; + if ERRNUM is nonzero, follow it with ": " and strerror (ERRNUM). + If STATUS is nonzero, terminate the program with `exit (STATUS)'. */ + +extern void error (int __status, int __errnum, const char *__format, ...) + __attribute__ ((__format__ (__printf__, 3, 4))); + +extern void error_at_line (int __status, int __errnum, const char *__fname, + unsigned int __lineno, const char *__format, ...) + __attribute__ ((__format__ (__printf__, 5, 6))); + +/* If NULL, error will flush stdout, then print on stderr the program + name, a colon and a space. Otherwise, error will call this + function without parameters instead. */ +extern void (*error_print_progname) (void); + +/* This variable is incremented each time `error' is called. */ +extern unsigned int error_message_count; + +/* Sometimes we want to have at most one error per line. This + variable controls whether this mode is selected or not. */ +extern int error_one_per_line; + +#ifdef __cplusplus +} +#endif + +#endif /* error.h */ diff --git a/contrib/diff/lib/exclude.c b/contrib/diff/lib/exclude.c new file mode 100644 index 0000000..55f1a39 --- /dev/null +++ b/contrib/diff/lib/exclude.c @@ -0,0 +1,263 @@ +/* exclude.c -- exclude file names + + Copyright (C) 1992, 1993, 1994, 1997, 1999, 2000, 2001, 2002, 2003 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert */ + +#if HAVE_CONFIG_H +# include +#endif + +#include + +#include +#include +#ifndef errno +extern int errno; +#endif +#include +#include +#include +#include + +#include "exclude.h" +#include "fnmatch.h" +#include "unlocked-io.h" +#include "xalloc.h" + +#if STDC_HEADERS || (! defined isascii && ! HAVE_ISASCII) +# define IN_CTYPE_DOMAIN(c) true +#else +# define IN_CTYPE_DOMAIN(c) isascii (c) +#endif + +static inline bool +is_space (unsigned char c) +{ + return IN_CTYPE_DOMAIN (c) && isspace (c); +} + +/* Verify a requirement at compile-time (unlike assert, which is runtime). */ +#define verify(name, assertion) struct name { char a[(assertion) ? 1 : -1]; } + +/* Non-GNU systems lack these options, so we don't need to check them. */ +#ifndef FNM_CASEFOLD +# define FNM_CASEFOLD 0 +#endif +#ifndef FNM_LEADING_DIR +# define FNM_LEADING_DIR 0 +#endif + +verify (EXCLUDE_macros_do_not_collide_with_FNM_macros, + (((EXCLUDE_ANCHORED | EXCLUDE_INCLUDE | EXCLUDE_WILDCARDS) + & (FNM_PATHNAME | FNM_NOESCAPE | FNM_PERIOD | FNM_LEADING_DIR + | FNM_CASEFOLD)) + == 0)); + +/* An exclude pattern-options pair. The options are fnmatch options + ORed with EXCLUDE_* options. */ + +struct patopts + { + char const *pattern; + int options; + }; + +/* An exclude list, of pattern-options pairs. */ + +struct exclude + { + struct patopts *exclude; + size_t exclude_alloc; + size_t exclude_count; + }; + +/* Return a newly allocated and empty exclude list. */ + +struct exclude * +new_exclude (void) +{ + return xzalloc (sizeof *new_exclude ()); +} + +/* Free the storage associated with an exclude list. */ + +void +free_exclude (struct exclude *ex) +{ + free (ex->exclude); + free (ex); +} + +/* Return zero if PATTERN matches F, obeying OPTIONS, except that + (unlike fnmatch) wildcards are disabled in PATTERN. */ + +static int +fnmatch_no_wildcards (char const *pattern, char const *f, int options) +{ + if (! (options & FNM_LEADING_DIR)) + return ((options & FNM_CASEFOLD) + ? strcasecmp (pattern, f) + : strcmp (pattern, f)); + else + { + size_t patlen = strlen (pattern); + int r = ((options & FNM_CASEFOLD) + ? strncasecmp (pattern, f, patlen) + : strncmp (pattern, f, patlen)); + if (! r) + { + r = f[patlen]; + if (r == '/') + r = 0; + } + return r; + } +} + +/* Return true if EX excludes F. */ + +bool +excluded_filename (struct exclude const *ex, char const *f) +{ + size_t exclude_count = ex->exclude_count; + + /* If no options are given, the default is to include. */ + if (exclude_count == 0) + return false; + else + { + struct patopts const *exclude = ex->exclude; + size_t i; + + /* Otherwise, the default is the opposite of the first option. */ + bool excluded = !! (exclude[0].options & EXCLUDE_INCLUDE); + + /* Scan through the options, seeing whether they change F from + excluded to included or vice versa. */ + for (i = 0; i < exclude_count; i++) + { + char const *pattern = exclude[i].pattern; + int options = exclude[i].options; + if (excluded == !! (options & EXCLUDE_INCLUDE)) + { + int (*matcher) (char const *, char const *, int) = + (options & EXCLUDE_WILDCARDS + ? fnmatch + : fnmatch_no_wildcards); + bool matched = ((*matcher) (pattern, f, options) == 0); + char const *p; + + if (! (options & EXCLUDE_ANCHORED)) + for (p = f; *p && ! matched; p++) + if (*p == '/' && p[1] != '/') + matched = ((*matcher) (pattern, p + 1, options) == 0); + + excluded ^= matched; + } + } + + return excluded; + } +} + +/* Append to EX the exclusion PATTERN with OPTIONS. */ + +void +add_exclude (struct exclude *ex, char const *pattern, int options) +{ + struct patopts *patopts; + + if (ex->exclude_count == ex->exclude_alloc) + ex->exclude = x2nrealloc (ex->exclude, &ex->exclude_alloc, + sizeof *ex->exclude); + + patopts = &ex->exclude[ex->exclude_count++]; + patopts->pattern = pattern; + patopts->options = options; +} + +/* Use ADD_FUNC to append to EX the patterns in FILENAME, each with + OPTIONS. LINE_END terminates each pattern in the file. If + LINE_END is a space character, ignore trailing spaces and empty + lines in FILE. Return -1 on failure, 0 on success. */ + +int +add_exclude_file (void (*add_func) (struct exclude *, char const *, int), + struct exclude *ex, char const *filename, int options, + char line_end) +{ + bool use_stdin = filename[0] == '-' && !filename[1]; + FILE *in; + char *buf = NULL; + char *p; + char const *pattern; + char const *lim; + size_t buf_alloc = 0; + size_t buf_count = 0; + int c; + int e = 0; + + if (use_stdin) + in = stdin; + else if (! (in = fopen (filename, "r"))) + return -1; + + while ((c = getc (in)) != EOF) + { + if (buf_count == buf_alloc) + buf = x2realloc (buf, &buf_alloc); + buf[buf_count++] = c; + } + + if (ferror (in)) + e = errno; + + if (!use_stdin && fclose (in) != 0) + e = errno; + + buf = xrealloc (buf, buf_count + 1); + buf[buf_count] = line_end; + lim = buf + buf_count + ! (buf_count == 0 || buf[buf_count - 1] == line_end); + pattern = buf; + + for (p = buf; p < lim; p++) + if (*p == line_end) + { + char *pattern_end = p; + + if (is_space (line_end)) + { + for (; ; pattern_end--) + if (pattern_end == pattern) + goto next_pattern; + else if (! is_space (pattern_end[-1])) + break; + } + + *pattern_end = '\0'; + (*add_func) (ex, pattern, options); + + next_pattern: + pattern = p + 1; + } + + errno = e; + return e ? -1 : 0; +} diff --git a/contrib/diff/lib/exclude.h b/contrib/diff/lib/exclude.h new file mode 100644 index 0000000..a7e2971 --- /dev/null +++ b/contrib/diff/lib/exclude.h @@ -0,0 +1,43 @@ +/* exclude.h -- declarations for excluding file names + + Copyright (C) 1992, 1993, 1994, 1997, 1999, 2001, 2002, 2003 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert */ + +/* Exclude options, which can be ORed with fnmatch options. */ + +/* Patterns must match the start of file names, instead of matching + anywhere after a '/'. */ +#define EXCLUDE_ANCHORED (1 << 30) + +/* Include instead of exclude. */ +#define EXCLUDE_INCLUDE (1 << 29) + +/* '?', '*', '[', and '\\' are special in patterns. Without this + option, these characters are ordinary and fnmatch is not used. */ +#define EXCLUDE_WILDCARDS (1 << 28) + +struct exclude; + +struct exclude *new_exclude (void); +void free_exclude (struct exclude *); +void add_exclude (struct exclude *, char const *, int); +int add_exclude_file (void (*) (struct exclude *, char const *, int), + struct exclude *, char const *, int, char); +bool excluded_filename (struct exclude const *, char const *); diff --git a/contrib/diff/lib/exit.h b/contrib/diff/lib/exit.h new file mode 100644 index 0000000..4e8d465 --- /dev/null +++ b/contrib/diff/lib/exit.h @@ -0,0 +1,32 @@ +/* exit() function. + Copyright (C) 1995, 2001 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _EXIT_H +#define _EXIT_H + +/* Get exit() declaration. */ +#include + +/* Some systems do not define EXIT_*, even with STDC_HEADERS. */ +#ifndef EXIT_SUCCESS +# define EXIT_SUCCESS 0 +#endif +#ifndef EXIT_FAILURE +# define EXIT_FAILURE 1 +#endif + +#endif /* _EXIT_H */ diff --git a/contrib/diff/lib/exitfail.c b/contrib/diff/lib/exitfail.c new file mode 100644 index 0000000..2ae5f69 --- /dev/null +++ b/contrib/diff/lib/exitfail.c @@ -0,0 +1,27 @@ +/* Failure exit status + + Copyright (C) 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "exitfail.h" +#include "exit.h" + +int volatile exit_failure = EXIT_FAILURE; diff --git a/contrib/diff/lib/exitfail.h b/contrib/diff/lib/exitfail.h new file mode 100644 index 0000000..cf5ab71 --- /dev/null +++ b/contrib/diff/lib/exitfail.h @@ -0,0 +1,20 @@ +/* Failure exit status + + Copyright (C) 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +extern int volatile exit_failure; diff --git a/contrib/diff/lib/file-type.c b/contrib/diff/lib/file-type.c new file mode 100644 index 0000000..3c58c7f --- /dev/null +++ b/contrib/diff/lib/file-type.c @@ -0,0 +1,75 @@ +/* Return a string describing the type of a file. + + Copyright (C) 1993, 1994, 2001, 2002, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include +#include +#include "file-type.h" + +#include +#define _(text) gettext (text) + +char const * +file_type (struct stat const *st) +{ + /* See POSIX 1003.1-2001 XCU Table 4-8 lines 17093-17107 for some of + these formats. + + To keep diagnostics grammatical in English, the returned string + must start with a consonant. */ + + if (S_ISREG (st->st_mode)) + return st->st_size == 0 ? _("regular empty file") : _("regular file"); + + if (S_ISDIR (st->st_mode)) + return _("directory"); + + if (S_ISBLK (st->st_mode)) + return _("block special file"); + + if (S_ISCHR (st->st_mode)) + return _("character special file"); + + if (S_ISFIFO (st->st_mode)) + return _("fifo"); + + if (S_ISLNK (st->st_mode)) + return _("symbolic link"); + + if (S_ISSOCK (st->st_mode)) + return _("socket"); + + if (S_TYPEISMQ (st)) + return _("message queue"); + + if (S_TYPEISSEM (st)) + return _("semaphore"); + + if (S_TYPEISSHM (st)) + return _("shared memory object"); + + if (S_TYPEISTMO (st)) + return _("typed memory object"); + + return _("weird file"); +} diff --git a/contrib/diff/lib/file-type.h b/contrib/diff/lib/file-type.h new file mode 100644 index 0000000..502f091 --- /dev/null +++ b/contrib/diff/lib/file-type.h @@ -0,0 +1,166 @@ +/* Return a string describing the type of a file. + + Copyright (C) 1993, 1994, 2001, 2002, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert and Jim Meyering. */ + +#ifndef FILE_TYPE_H +# define FILE_TYPE_H 1 + +# if ! defined S_ISREG && ! defined S_IFREG +you must include before including this file +# endif + +char const *file_type (struct stat const *); + +# ifndef S_IFMT +# define S_IFMT 0170000 +# endif + +# if STAT_MACROS_BROKEN +# undef S_ISBLK +# undef S_ISCHR +# undef S_ISDIR +# undef S_ISDOOR +# undef S_ISFIFO +# undef S_ISLNK +# undef S_ISNAM +# undef S_ISMPB +# undef S_ISMPC +# undef S_ISNWK +# undef S_ISREG +# undef S_ISSOCK +# endif + + +# ifndef S_ISBLK +# ifdef S_IFBLK +# define S_ISBLK(m) (((m) & S_IFMT) == S_IFBLK) +# else +# define S_ISBLK(m) 0 +# endif +# endif + +# ifndef S_ISCHR +# ifdef S_IFCHR +# define S_ISCHR(m) (((m) & S_IFMT) == S_IFCHR) +# else +# define S_ISCHR(m) 0 +# endif +# endif + +# ifndef S_ISDIR +# ifdef S_IFDIR +# define S_ISDIR(m) (((m) & S_IFMT) == S_IFDIR) +# else +# define S_ISDIR(m) 0 +# endif +# endif + +# ifndef S_ISDOOR /* Solaris 2.5 and up */ +# ifdef S_IFDOOR +# define S_ISDOOR(m) (((m) & S_IFMT) == S_IFDOOR) +# else +# define S_ISDOOR(m) 0 +# endif +# endif + +# ifndef S_ISFIFO +# ifdef S_IFIFO +# define S_ISFIFO(m) (((m) & S_IFMT) == S_IFIFO) +# else +# define S_ISFIFO(m) 0 +# endif +# endif + +# ifndef S_ISLNK +# ifdef S_IFLNK +# define S_ISLNK(m) (((m) & S_IFMT) == S_IFLNK) +# else +# define S_ISLNK(m) 0 +# endif +# endif + +# ifndef S_ISMPB /* V7 */ +# ifdef S_IFMPB +# define S_ISMPB(m) (((m) & S_IFMT) == S_IFMPB) +# define S_ISMPC(m) (((m) & S_IFMT) == S_IFMPC) +# else +# define S_ISMPB(m) 0 +# define S_ISMPC(m) 0 +# endif +# endif + +# ifndef S_ISNAM /* Xenix */ +# ifdef S_IFNAM +# define S_ISNAM(m) (((m) & S_IFMT) == S_IFNAM) +# else +# define S_ISNAM(m) 0 +# endif +# endif + +# ifndef S_ISNWK /* HP/UX */ +# ifdef S_IFNWK +# define S_ISNWK(m) (((m) & S_IFMT) == S_IFNWK) +# else +# define S_ISNWK(m) 0 +# endif +# endif + +# ifndef S_ISREG +# ifdef S_IFREG +# define S_ISREG(m) (((m) & S_IFMT) == S_IFREG) +# else +# define S_ISREG(m) 0 +# endif +# endif + +# ifndef S_ISSOCK +# ifdef S_IFSOCK +# define S_ISSOCK(m) (((m) & S_IFMT) == S_IFSOCK) +# else +# define S_ISSOCK(m) 0 +# endif +# endif + + +# ifndef S_TYPEISMQ +# define S_TYPEISMQ(p) 0 +# endif + +# ifndef S_TYPEISTMO +# define S_TYPEISTMO(p) 0 +# endif + + +# ifndef S_TYPEISSEM +# ifdef S_INSEM +# define S_TYPEISSEM(p) (S_ISNAM ((p)->st_mode) && (p)->st_rdev == S_INSEM) +# else +# define S_TYPEISSEM(p) 0 +# endif +# endif + +# ifndef S_TYPEISSHM +# ifdef S_INSHD +# define S_TYPEISSHM(p) (S_ISNAM ((p)->st_mode) && (p)->st_rdev == S_INSHD) +# else +# define S_TYPEISSHM(p) 0 +# endif +# endif + +#endif /* FILE_TYPE_H */ diff --git a/contrib/diff/lib/fnmatch.c b/contrib/diff/lib/fnmatch.c new file mode 100644 index 0000000..d0cd8cd --- /dev/null +++ b/contrib/diff/lib/fnmatch.c @@ -0,0 +1,403 @@ +/* Copyright (C) 1991,1992,1993,1996,1997,1998,1999,2000,2001,2002,2003,2004 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +/* Enable GNU extensions in fnmatch.h. */ +#ifndef _GNU_SOURCE +# define _GNU_SOURCE 1 +#endif + +#ifdef __GNUC__ +# define alloca __builtin_alloca +# define HAVE_ALLOCA 1 +#else +# if defined HAVE_ALLOCA_H || defined _LIBC +# include +# else +# ifdef _AIX + # pragma alloca +# else +# ifndef alloca +char *alloca (); +# endif +# endif +# endif +#endif + +#if ! defined __builtin_expect && __GNUC__ < 3 +# define __builtin_expect(expr, expected) (expr) +#endif + +#include + +#include +#include +#include +#include +#include +#include + +#define WIDE_CHAR_SUPPORT (HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_BTOWC) + +/* For platform which support the ISO C amendement 1 functionality we + support user defined character classes. */ +#if defined _LIBC || WIDE_CHAR_SUPPORT +/* Solaris 2.5 has a bug: must be included before . */ +# include +# include +#endif + +/* We need some of the locale data (the collation sequence information) + but there is no interface to get this information in general. Therefore + we support a correct implementation only in glibc. */ +#ifdef _LIBC +# include "../locale/localeinfo.h" +# include "../locale/elem-hash.h" +# include "../locale/coll-lookup.h" +# include + +# define CONCAT(a,b) __CONCAT(a,b) +# define mbsrtowcs __mbsrtowcs +# define fnmatch __fnmatch +extern int fnmatch (const char *pattern, const char *string, int flags); +#endif + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +/* We often have to test for FNM_FILE_NAME and FNM_PERIOD being both set. */ +#define NO_LEADING_PERIOD(flags) \ + ((flags & (FNM_FILE_NAME | FNM_PERIOD)) == (FNM_FILE_NAME | FNM_PERIOD)) + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself, and have not detected a bug + in the library. This code is part of the GNU C + Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object files, + it is simpler to just do this in the source for each such file. */ + +#if defined _LIBC || !defined __GNU_LIBRARY__ || !HAVE_FNMATCH_GNU + + +# if defined STDC_HEADERS || !defined isascii +# define ISASCII(c) 1 +# else +# define ISASCII(c) isascii(c) +# endif + +# ifdef isblank +# define ISBLANK(c) (ISASCII (c) && isblank (c)) +# else +# define ISBLANK(c) ((c) == ' ' || (c) == '\t') +# endif +# ifdef isgraph +# define ISGRAPH(c) (ISASCII (c) && isgraph (c)) +# else +# define ISGRAPH(c) (ISASCII (c) && isprint (c) && !isspace (c)) +# endif + +# define ISPRINT(c) (ISASCII (c) && isprint (c)) +# define ISDIGIT(c) (ISASCII (c) && isdigit (c)) +# define ISALNUM(c) (ISASCII (c) && isalnum (c)) +# define ISALPHA(c) (ISASCII (c) && isalpha (c)) +# define ISCNTRL(c) (ISASCII (c) && iscntrl (c)) +# define ISLOWER(c) (ISASCII (c) && islower (c)) +# define ISPUNCT(c) (ISASCII (c) && ispunct (c)) +# define ISSPACE(c) (ISASCII (c) && isspace (c)) +# define ISUPPER(c) (ISASCII (c) && isupper (c)) +# define ISXDIGIT(c) (ISASCII (c) && isxdigit (c)) + +# define STREQ(s1, s2) ((strcmp (s1, s2) == 0)) + +# if defined _LIBC || WIDE_CHAR_SUPPORT +/* The GNU C library provides support for user-defined character classes + and the functions from ISO C amendement 1. */ +# ifdef CHARCLASS_NAME_MAX +# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX +# else +/* This shouldn't happen but some implementation might still have this + problem. Use a reasonable default value. */ +# define CHAR_CLASS_MAX_LENGTH 256 +# endif + +# ifdef _LIBC +# define IS_CHAR_CLASS(string) __wctype (string) +# else +# define IS_CHAR_CLASS(string) wctype (string) +# endif + +# ifdef _LIBC +# define ISWCTYPE(WC, WT) __iswctype (WC, WT) +# else +# define ISWCTYPE(WC, WT) iswctype (WC, WT) +# endif + +# if (HAVE_MBSTATE_T && HAVE_MBSRTOWCS) || _LIBC +/* In this case we are implementing the multibyte character handling. */ +# define HANDLE_MULTIBYTE 1 +# endif + +# else +# define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */ + +# define IS_CHAR_CLASS(string) \ + (STREQ (string, "alpha") || STREQ (string, "upper") \ + || STREQ (string, "lower") || STREQ (string, "digit") \ + || STREQ (string, "alnum") || STREQ (string, "xdigit") \ + || STREQ (string, "space") || STREQ (string, "print") \ + || STREQ (string, "punct") || STREQ (string, "graph") \ + || STREQ (string, "cntrl") || STREQ (string, "blank")) +# endif + +/* Avoid depending on library functions or files + whose names are inconsistent. */ + +# ifndef errno +extern int errno; +# endif + +/* Global variable. */ +static int posixly_correct; + +# ifndef internal_function +/* Inside GNU libc we mark some function in a special way. In other + environments simply ignore the marking. */ +# define internal_function +# endif + +/* Note that this evaluates C many times. */ +# ifdef _LIBC +# define FOLD(c) ((flags & FNM_CASEFOLD) ? tolower (c) : (c)) +# else +# define FOLD(c) ((flags & FNM_CASEFOLD) && ISUPPER (c) ? tolower (c) : (c)) +# endif +# define CHAR char +# define UCHAR unsigned char +# define INT int +# define FCT internal_fnmatch +# define EXT ext_match +# define END end_pattern +# define L(CS) CS +# ifdef _LIBC +# define BTOWC(C) __btowc (C) +# else +# define BTOWC(C) btowc (C) +# endif +# define STRLEN(S) strlen (S) +# define STRCAT(D, S) strcat (D, S) +# ifdef _LIBC +# define MEMPCPY(D, S, N) __mempcpy (D, S, N) +# else +# if HAVE_MEMPCPY +# define MEMPCPY(D, S, N) mempcpy (D, S, N) +# else +# define MEMPCPY(D, S, N) ((void *) ((char *) memcpy (D, S, N) + (N))) +# endif +# endif +# define MEMCHR(S, C, N) memchr (S, C, N) +# define STRCOLL(S1, S2) strcoll (S1, S2) +# include "fnmatch_loop.c" + + +# if HANDLE_MULTIBYTE +# define FOLD(c) ((flags & FNM_CASEFOLD) ? towlower (c) : (c)) +# define CHAR wchar_t +# define UCHAR wint_t +# define INT wint_t +# define FCT internal_fnwmatch +# define EXT ext_wmatch +# define END end_wpattern +# define L(CS) L##CS +# define BTOWC(C) (C) +# ifdef _LIBC +# define STRLEN(S) __wcslen (S) +# define STRCAT(D, S) __wcscat (D, S) +# define MEMPCPY(D, S, N) __wmempcpy (D, S, N) +# else +# define STRLEN(S) wcslen (S) +# define STRCAT(D, S) wcscat (D, S) +# if HAVE_WMEMPCPY +# define MEMPCPY(D, S, N) wmempcpy (D, S, N) +# else +# define MEMPCPY(D, S, N) (wmemcpy (D, S, N) + (N)) +# endif +# endif +# define MEMCHR(S, C, N) wmemchr (S, C, N) +# define STRCOLL(S1, S2) wcscoll (S1, S2) +# define WIDE_CHAR_VERSION 1 + +# undef IS_CHAR_CLASS +/* We have to convert the wide character string in a multibyte string. But + we know that the character class names consist of alphanumeric characters + from the portable character set, and since the wide character encoding + for a member of the portable character set is the same code point as + its single-byte encoding, we can use a simplified method to convert the + string to a multibyte character string. */ +static wctype_t +is_char_class (const wchar_t *wcs) +{ + char s[CHAR_CLASS_MAX_LENGTH + 1]; + char *cp = s; + + do + { + /* Test for a printable character from the portable character set. */ +# ifdef _LIBC + if (*wcs < 0x20 || *wcs > 0x7e + || *wcs == 0x24 || *wcs == 0x40 || *wcs == 0x60) + return (wctype_t) 0; +# else + switch (*wcs) + { + case L' ': case L'!': case L'"': case L'#': case L'%': + case L'&': case L'\'': case L'(': case L')': case L'*': + case L'+': case L',': case L'-': case L'.': case L'/': + case L'0': case L'1': case L'2': case L'3': case L'4': + case L'5': case L'6': case L'7': case L'8': case L'9': + case L':': case L';': case L'<': case L'=': case L'>': + case L'?': + case L'A': case L'B': case L'C': case L'D': case L'E': + case L'F': case L'G': case L'H': case L'I': case L'J': + case L'K': case L'L': case L'M': case L'N': case L'O': + case L'P': case L'Q': case L'R': case L'S': case L'T': + case L'U': case L'V': case L'W': case L'X': case L'Y': + case L'Z': + case L'[': case L'\\': case L']': case L'^': case L'_': + case L'a': case L'b': case L'c': case L'd': case L'e': + case L'f': case L'g': case L'h': case L'i': case L'j': + case L'k': case L'l': case L'm': case L'n': case L'o': + case L'p': case L'q': case L'r': case L's': case L't': + case L'u': case L'v': case L'w': case L'x': case L'y': + case L'z': case L'{': case L'|': case L'}': case L'~': + break; + default: + return (wctype_t) 0; + } +# endif + + /* Avoid overrunning the buffer. */ + if (cp == s + CHAR_CLASS_MAX_LENGTH) + return (wctype_t) 0; + + *cp++ = (char) *wcs++; + } + while (*wcs != L'\0'); + + *cp = '\0'; + +# ifdef _LIBC + return __wctype (s); +# else + return wctype (s); +# endif +} +# define IS_CHAR_CLASS(string) is_char_class (string) + +# include "fnmatch_loop.c" +# endif + + +int +fnmatch (const char *pattern, const char *string, int flags) +{ +# if HANDLE_MULTIBYTE +# define ALLOCA_LIMIT 2000 + if (__builtin_expect (MB_CUR_MAX, 1) != 1) + { + mbstate_t ps; + size_t patsize; + size_t strsize; + size_t totsize; + wchar_t *wpattern; + wchar_t *wstring; + int res; + + /* Calculate the size needed to convert the strings to + wide characters. */ + memset (&ps, '\0', sizeof (ps)); + patsize = mbsrtowcs (NULL, &pattern, 0, &ps) + 1; + if (__builtin_expect (patsize == 0, 0)) + /* Something wrong. + XXX Do we have to set `errno' to something which mbsrtows hasn't + already done? */ + return -1; + assert (mbsinit (&ps)); + strsize = mbsrtowcs (NULL, &string, 0, &ps) + 1; + if (__builtin_expect (strsize == 0, 0)) + /* Something wrong. + XXX Do we have to set `errno' to something which mbsrtows hasn't + already done? */ + return -1; + assert (mbsinit (&ps)); + totsize = patsize + strsize; + if (__builtin_expect (! (patsize <= totsize + && totsize <= SIZE_MAX / sizeof (wchar_t)), + 0)) + { + errno = ENOMEM; + return -1; + } + + /* Allocate room for the wide characters. */ + if (__builtin_expect (totsize < ALLOCA_LIMIT, 1)) + wpattern = (wchar_t *) alloca (totsize * sizeof (wchar_t)); + else + { + wpattern = malloc (totsize * sizeof (wchar_t)); + if (__builtin_expect (! wpattern, 0)) + { + errno = ENOMEM; + return -1; + } + } + wstring = wpattern + patsize; + + /* Convert the strings into wide characters. */ + mbsrtowcs (wpattern, &pattern, patsize, &ps); + assert (mbsinit (&ps)); + mbsrtowcs (wstring, &string, strsize, &ps); + + res = internal_fnwmatch (wpattern, wstring, wstring + strsize - 1, + flags & FNM_PERIOD, flags); + + if (__builtin_expect (! (totsize < ALLOCA_LIMIT), 0)) + free (wpattern); + return res; + } +# endif /* HANDLE_MULTIBYTE */ + + return internal_fnmatch (pattern, string, string + strlen (string), + flags & FNM_PERIOD, flags); +} + +# ifdef _LIBC +# undef fnmatch +versioned_symbol (libc, __fnmatch, fnmatch, GLIBC_2_2_3); +# if SHLIB_COMPAT(libc, GLIBC_2_0, GLIBC_2_2_3) +strong_alias (__fnmatch, __fnmatch_old) +compat_symbol (libc, __fnmatch_old, fnmatch, GLIBC_2_0); +# endif +libc_hidden_ver (__fnmatch, fnmatch) +# endif + +#endif /* _LIBC or not __GNU_LIBRARY__. */ diff --git a/contrib/diff/lib/fnmatch_.h b/contrib/diff/lib/fnmatch_.h new file mode 100644 index 0000000..87e661f --- /dev/null +++ b/contrib/diff/lib/fnmatch_.h @@ -0,0 +1,63 @@ +/* Copyright (C) 1991, 1992, 1993, 1996, 1997, 1998, 1999, 2001, 2002, 2003 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _FNMATCH_H +# define _FNMATCH_H 1 + +# ifdef __cplusplus +extern "C" { +# endif + +/* We #undef these before defining them because some losing systems + (HP-UX A.08.07 for example) define these in . */ +# undef FNM_PATHNAME +# undef FNM_NOESCAPE +# undef FNM_PERIOD + +/* Bits set in the FLAGS argument to `fnmatch'. */ +# define FNM_PATHNAME (1 << 0) /* No wildcard can ever match `/'. */ +# define FNM_NOESCAPE (1 << 1) /* Backslashes don't quote special chars. */ +# define FNM_PERIOD (1 << 2) /* Leading `.' is matched only explicitly. */ + +# if !defined _POSIX_C_SOURCE || _POSIX_C_SOURCE < 2 || defined _GNU_SOURCE +# define FNM_FILE_NAME FNM_PATHNAME /* Preferred GNU name. */ +# define FNM_LEADING_DIR (1 << 3) /* Ignore `/...' after a match. */ +# define FNM_CASEFOLD (1 << 4) /* Compare without regard to case. */ +# define FNM_EXTMATCH (1 << 5) /* Use ksh-like extended matching. */ +# endif + +/* Value returned by `fnmatch' if STRING does not match PATTERN. */ +# define FNM_NOMATCH 1 + +/* This value is returned if the implementation does not support + `fnmatch'. Since this is not the case here it will never be + returned but the conformance test suites still require the symbol + to be defined. */ +# ifdef _XOPEN_SOURCE +# define FNM_NOSYS (-1) +# endif + +/* Match NAME against the filename pattern PATTERN, + returning zero if it matches, FNM_NOMATCH if not. */ +extern int fnmatch (const char *__pattern, const char *__name, + int __flags); + +# ifdef __cplusplus +} +# endif + +#endif /* fnmatch.h */ diff --git a/contrib/diff/lib/fnmatch_loop.c b/contrib/diff/lib/fnmatch_loop.c new file mode 100644 index 0000000..7798b59 --- /dev/null +++ b/contrib/diff/lib/fnmatch_loop.c @@ -0,0 +1,1192 @@ +/* Copyright (C) 1991,1992,1993,1996,1997,1998,1999,2000,2001,2002,2003,2004 + Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Match STRING against the filename pattern PATTERN, returning zero if + it matches, nonzero if not. */ +static int EXT (INT opt, const CHAR *pattern, const CHAR *string, + const CHAR *string_end, int no_leading_period, int flags) + internal_function; +static const CHAR *END (const CHAR *patternp) internal_function; + +static int +internal_function +FCT (const CHAR *pattern, const CHAR *string, const CHAR *string_end, + int no_leading_period, int flags) +{ + register const CHAR *p = pattern, *n = string; + register UCHAR c; +#ifdef _LIBC +# if WIDE_CHAR_VERSION + const char *collseq = (const char *) + _NL_CURRENT(LC_COLLATE, _NL_COLLATE_COLLSEQWC); +# else + const UCHAR *collseq = (const UCHAR *) + _NL_CURRENT(LC_COLLATE, _NL_COLLATE_COLLSEQMB); +# endif +#endif + + while ((c = *p++) != L('\0')) + { + int new_no_leading_period = 0; + c = FOLD (c); + + switch (c) + { + case L('?'): + if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') + { + int res; + + res = EXT (c, p, n, string_end, no_leading_period, + flags); + if (res != -1) + return res; + } + + if (n == string_end) + return FNM_NOMATCH; + else if (*n == L('/') && (flags & FNM_FILE_NAME)) + return FNM_NOMATCH; + else if (*n == L('.') && no_leading_period) + return FNM_NOMATCH; + break; + + case L('\\'): + if (!(flags & FNM_NOESCAPE)) + { + c = *p++; + if (c == L('\0')) + /* Trailing \ loses. */ + return FNM_NOMATCH; + c = FOLD (c); + } + if (n == string_end || FOLD ((UCHAR) *n) != c) + return FNM_NOMATCH; + break; + + case L('*'): + if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') + { + int res; + + res = EXT (c, p, n, string_end, no_leading_period, + flags); + if (res != -1) + return res; + } + + if (n != string_end && *n == L('.') && no_leading_period) + return FNM_NOMATCH; + + for (c = *p++; c == L('?') || c == L('*'); c = *p++) + { + if (*p == L('(') && (flags & FNM_EXTMATCH) != 0) + { + const CHAR *endp = END (p); + if (endp != p) + { + /* This is a pattern. Skip over it. */ + p = endp; + continue; + } + } + + if (c == L('?')) + { + /* A ? needs to match one character. */ + if (n == string_end) + /* There isn't another character; no match. */ + return FNM_NOMATCH; + else if (*n == L('/') + && __builtin_expect (flags & FNM_FILE_NAME, 0)) + /* A slash does not match a wildcard under + FNM_FILE_NAME. */ + return FNM_NOMATCH; + else + /* One character of the string is consumed in matching + this ? wildcard, so *??? won't match if there are + less than three characters. */ + ++n; + } + } + + if (c == L('\0')) + /* The wildcard(s) is/are the last element of the pattern. + If the name is a file name and contains another slash + this means it cannot match, unless the FNM_LEADING_DIR + flag is set. */ + { + int result = (flags & FNM_FILE_NAME) == 0 ? 0 : FNM_NOMATCH; + + if (flags & FNM_FILE_NAME) + { + if (flags & FNM_LEADING_DIR) + result = 0; + else + { + if (MEMCHR (n, L('/'), string_end - n) == NULL) + result = 0; + } + } + + return result; + } + else + { + const CHAR *endp; + + endp = MEMCHR (n, (flags & FNM_FILE_NAME) ? L('/') : L('\0'), + string_end - n); + if (endp == NULL) + endp = string_end; + + if (c == L('[') + || (__builtin_expect (flags & FNM_EXTMATCH, 0) != 0 + && (c == L('@') || c == L('+') || c == L('!')) + && *p == L('('))) + { + int flags2 = ((flags & FNM_FILE_NAME) + ? flags : (flags & ~FNM_PERIOD)); + int no_leading_period2 = no_leading_period; + + for (--p; n < endp; ++n, no_leading_period2 = 0) + if (FCT (p, n, string_end, no_leading_period2, flags2) + == 0) + return 0; + } + else if (c == L('/') && (flags & FNM_FILE_NAME)) + { + while (n < string_end && *n != L('/')) + ++n; + if (n < string_end && *n == L('/') + && (FCT (p, n + 1, string_end, flags & FNM_PERIOD, flags) + == 0)) + return 0; + } + else + { + int flags2 = ((flags & FNM_FILE_NAME) + ? flags : (flags & ~FNM_PERIOD)); + int no_leading_period2 = no_leading_period; + + if (c == L('\\') && !(flags & FNM_NOESCAPE)) + c = *p; + c = FOLD (c); + for (--p; n < endp; ++n, no_leading_period2 = 0) + if (FOLD ((UCHAR) *n) == c + && (FCT (p, n, string_end, no_leading_period2, flags2) + == 0)) + return 0; + } + } + + /* If we come here no match is possible with the wildcard. */ + return FNM_NOMATCH; + + case L('['): + { + /* Nonzero if the sense of the character class is inverted. */ + register int not; + CHAR cold; + UCHAR fn; + + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + if (n == string_end) + return FNM_NOMATCH; + + if (*n == L('.') && no_leading_period) + return FNM_NOMATCH; + + if (*n == L('/') && (flags & FNM_FILE_NAME)) + /* `/' cannot be matched. */ + return FNM_NOMATCH; + + not = (*p == L('!') || (posixly_correct < 0 && *p == L('^'))); + if (not) + ++p; + + fn = FOLD ((UCHAR) *n); + + c = *p++; + for (;;) + { + if (!(flags & FNM_NOESCAPE) && c == L('\\')) + { + if (*p == L('\0')) + return FNM_NOMATCH; + c = FOLD ((UCHAR) *p); + ++p; + + if (c == fn) + goto matched; + } + else if (c == L('[') && *p == L(':')) + { + /* Leave room for the null. */ + CHAR str[CHAR_CLASS_MAX_LENGTH + 1]; + size_t c1 = 0; +#if defined _LIBC || WIDE_CHAR_SUPPORT + wctype_t wt; +#endif + const CHAR *startp = p; + + for (;;) + { + if (c1 == CHAR_CLASS_MAX_LENGTH) + /* The name is too long and therefore the pattern + is ill-formed. */ + return FNM_NOMATCH; + + c = *++p; + if (c == L(':') && p[1] == L(']')) + { + p += 2; + break; + } + if (c < L('a') || c >= L('z')) + { + /* This cannot possibly be a character class name. + Match it as a normal range. */ + p = startp; + c = L('['); + goto normal_bracket; + } + str[c1++] = c; + } + str[c1] = L('\0'); + +#if defined _LIBC || WIDE_CHAR_SUPPORT + wt = IS_CHAR_CLASS (str); + if (wt == 0) + /* Invalid character class name. */ + return FNM_NOMATCH; + +# if defined _LIBC && ! WIDE_CHAR_VERSION + /* The following code is glibc specific but does + there a good job in speeding up the code since + we can avoid the btowc() call. */ + if (_ISCTYPE ((UCHAR) *n, wt)) + goto matched; +# else + if (ISWCTYPE (BTOWC ((UCHAR) *n), wt)) + goto matched; +# endif +#else + if ((STREQ (str, L("alnum")) && ISALNUM ((UCHAR) *n)) + || (STREQ (str, L("alpha")) && ISALPHA ((UCHAR) *n)) + || (STREQ (str, L("blank")) && ISBLANK ((UCHAR) *n)) + || (STREQ (str, L("cntrl")) && ISCNTRL ((UCHAR) *n)) + || (STREQ (str, L("digit")) && ISDIGIT ((UCHAR) *n)) + || (STREQ (str, L("graph")) && ISGRAPH ((UCHAR) *n)) + || (STREQ (str, L("lower")) && ISLOWER ((UCHAR) *n)) + || (STREQ (str, L("print")) && ISPRINT ((UCHAR) *n)) + || (STREQ (str, L("punct")) && ISPUNCT ((UCHAR) *n)) + || (STREQ (str, L("space")) && ISSPACE ((UCHAR) *n)) + || (STREQ (str, L("upper")) && ISUPPER ((UCHAR) *n)) + || (STREQ (str, L("xdigit")) && ISXDIGIT ((UCHAR) *n))) + goto matched; +#endif + c = *p++; + } +#ifdef _LIBC + else if (c == L('[') && *p == L('=')) + { + UCHAR str[1]; + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + const CHAR *startp = p; + + c = *++p; + if (c == L('\0')) + { + p = startp; + c = L('['); + goto normal_bracket; + } + str[0] = c; + + c = *++p; + if (c != L('=') || p[1] != L(']')) + { + p = startp; + c = L('['); + goto normal_bracket; + } + p += 2; + + if (nrules == 0) + { + if ((UCHAR) *n == str[0]) + goto matched; + } + else + { + const int32_t *table; +# if WIDE_CHAR_VERSION + const int32_t *weights; + const int32_t *extra; +# else + const unsigned char *weights; + const unsigned char *extra; +# endif + const int32_t *indirect; + int32_t idx; + const UCHAR *cp = (const UCHAR *) str; + + /* This #include defines a local function! */ +# if WIDE_CHAR_VERSION +# include +# else +# include +# endif + +# if WIDE_CHAR_VERSION + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEWC); + weights = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTWC); + extra = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAWC); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTWC); +# else + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); +# endif + + idx = findidx (&cp); + if (idx != 0) + { + /* We found a table entry. Now see whether the + character we are currently at has the same + equivalance class value. */ + int len = weights[idx]; + int32_t idx2; + const UCHAR *np = (const UCHAR *) n; + + idx2 = findidx (&np); + if (idx2 != 0 && len == weights[idx2]) + { + int cnt = 0; + + while (cnt < len + && (weights[idx + 1 + cnt] + == weights[idx2 + 1 + cnt])) + ++cnt; + + if (cnt == len) + goto matched; + } + } + } + + c = *p++; + } +#endif + else if (c == L('\0')) + /* [ (unterminated) loses. */ + return FNM_NOMATCH; + else + { + int is_range = 0; + +#ifdef _LIBC + int is_seqval = 0; + + if (c == L('[') && *p == L('.')) + { + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + const CHAR *startp = p; + size_t c1 = 0; + + while (1) + { + c = *++p; + if (c == L('.') && p[1] == L(']')) + { + p += 2; + break; + } + if (c == '\0') + return FNM_NOMATCH; + ++c1; + } + + /* We have to handling the symbols differently in + ranges since then the collation sequence is + important. */ + is_range = *p == L('-') && p[1] != L('\0'); + + if (nrules == 0) + { + /* There are no names defined in the collation + data. Therefore we only accept the trivial + names consisting of the character itself. */ + if (c1 != 1) + return FNM_NOMATCH; + + if (!is_range && *n == startp[1]) + goto matched; + + cold = startp[1]; + c = *p++; + } + else + { + int32_t table_size; + const int32_t *symb_table; +# ifdef WIDE_CHAR_VERSION + char str[c1]; + unsigned int strcnt; +# else +# define str (startp + 1) +# endif + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + +# ifdef WIDE_CHAR_VERSION + /* We have to convert the name to a single-byte + string. This is possible since the names + consist of ASCII characters and the internal + representation is UCS4. */ + for (strcnt = 0; strcnt < c1; ++strcnt) + str[strcnt] = startp[1 + strcnt]; +# endif + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing table. */ + hash = elem_hash (str, c1); + + idx = 0; + elem = hash % table_size; + second = hash % (table_size - 2); + while (symb_table[2 * elem] != 0) + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && c1 == extra[symb_table[2 * elem + 1]] + && memcmp (str, + &extra[symb_table[2 * elem + 1] + + 1], c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + + if (symb_table[2 * elem] != 0) + { + /* Compare the byte sequence but only if + this is not part of a range. */ +# ifdef WIDE_CHAR_VERSION + int32_t *wextra; + + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + + wextra = (int32_t *) &extra[idx + 4]; +# endif + + if (! is_range) + { +# ifdef WIDE_CHAR_VERSION + for (c1 = 0; + (int32_t) c1 < wextra[idx]; + ++c1) + if (n[c1] != wextra[1 + c1]) + break; + + if ((int32_t) c1 == wextra[idx]) + goto matched; +# else + for (c1 = 0; c1 < extra[idx]; ++c1) + if (n[c1] != extra[1 + c1]) + break; + + if (c1 == extra[idx]) + goto matched; +# endif + } + + /* Get the collation sequence value. */ + is_seqval = 1; +# ifdef WIDE_CHAR_VERSION + cold = wextra[1 + wextra[idx]]; +# else + /* Adjust for the alignment. */ + idx += 1 + extra[idx]; + idx = (idx + 3) & ~4; + cold = *((int32_t *) &extra[idx]); +# endif + + c = *p++; + } + else if (c1 == 1) + { + /* No valid character. Match it as a + single byte. */ + if (!is_range && *n == str[0]) + goto matched; + + cold = str[0]; + c = *p++; + } + else + return FNM_NOMATCH; + } + } + else +# undef str +#endif + { + c = FOLD (c); + normal_bracket: + + /* We have to handling the symbols differently in + ranges since then the collation sequence is + important. */ + is_range = (*p == L('-') && p[1] != L('\0') + && p[1] != L(']')); + + if (!is_range && c == fn) + goto matched; + + cold = c; + c = *p++; + } + + if (c == L('-') && *p != L(']')) + { +#if _LIBC + /* We have to find the collation sequence + value for C. Collation sequence is nothing + we can regularly access. The sequence + value is defined by the order in which the + definitions of the collation values for the + various characters appear in the source + file. A strange concept, nowhere + documented. */ + uint32_t fcollseq; + uint32_t lcollseq; + UCHAR cend = *p++; + +# ifdef WIDE_CHAR_VERSION + /* Search in the `names' array for the characters. */ + fcollseq = __collseq_table_lookup (collseq, fn); + if (fcollseq == ~((uint32_t) 0)) + /* XXX We don't know anything about the character + we are supposed to match. This means we are + failing. */ + goto range_not_matched; + + if (is_seqval) + lcollseq = cold; + else + lcollseq = __collseq_table_lookup (collseq, cold); +# else + fcollseq = collseq[fn]; + lcollseq = is_seqval ? cold : collseq[(UCHAR) cold]; +# endif + + is_seqval = 0; + if (cend == L('[') && *p == L('.')) + { + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_NRULES); + const CHAR *startp = p; + size_t c1 = 0; + + while (1) + { + c = *++p; + if (c == L('.') && p[1] == L(']')) + { + p += 2; + break; + } + if (c == '\0') + return FNM_NOMATCH; + ++c1; + } + + if (nrules == 0) + { + /* There are no names defined in the + collation data. Therefore we only + accept the trivial names consisting + of the character itself. */ + if (c1 != 1) + return FNM_NOMATCH; + + cend = startp[1]; + } + else + { + int32_t table_size; + const int32_t *symb_table; +# ifdef WIDE_CHAR_VERSION + char str[c1]; + unsigned int strcnt; +# else +# define str (startp + 1) +# endif + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + +# ifdef WIDE_CHAR_VERSION + /* We have to convert the name to a single-byte + string. This is possible since the names + consist of ASCII characters and the internal + representation is UCS4. */ + for (strcnt = 0; strcnt < c1; ++strcnt) + str[strcnt] = startp[1 + strcnt]; +# endif + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing + table. */ + hash = elem_hash (str, c1); + + idx = 0; + elem = hash % table_size; + second = hash % (table_size - 2); + while (symb_table[2 * elem] != 0) + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && (c1 + == extra[symb_table[2 * elem + 1]]) + && memcmp (str, + &extra[symb_table[2 * elem + 1] + + 1], c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + + if (symb_table[2 * elem] != 0) + { + /* Compare the byte sequence but only if + this is not part of a range. */ +# ifdef WIDE_CHAR_VERSION + int32_t *wextra; + + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~4; + + wextra = (int32_t *) &extra[idx + 4]; +# endif + /* Get the collation sequence value. */ + is_seqval = 1; +# ifdef WIDE_CHAR_VERSION + cend = wextra[1 + wextra[idx]]; +# else + /* Adjust for the alignment. */ + idx += 1 + extra[idx]; + idx = (idx + 3) & ~4; + cend = *((int32_t *) &extra[idx]); +# endif + } + else if (symb_table[2 * elem] != 0 && c1 == 1) + { + cend = str[0]; + c = *p++; + } + else + return FNM_NOMATCH; + } +# undef str + } + else + { + if (!(flags & FNM_NOESCAPE) && cend == L('\\')) + cend = *p++; + if (cend == L('\0')) + return FNM_NOMATCH; + cend = FOLD (cend); + } + + /* XXX It is not entirely clear to me how to handle + characters which are not mentioned in the + collation specification. */ + if ( +# ifdef WIDE_CHAR_VERSION + lcollseq == 0xffffffff || +# endif + lcollseq <= fcollseq) + { + /* We have to look at the upper bound. */ + uint32_t hcollseq; + + if (is_seqval) + hcollseq = cend; + else + { +# ifdef WIDE_CHAR_VERSION + hcollseq = + __collseq_table_lookup (collseq, cend); + if (hcollseq == ~((uint32_t) 0)) + { + /* Hum, no information about the upper + bound. The matching succeeds if the + lower bound is matched exactly. */ + if (lcollseq != fcollseq) + goto range_not_matched; + + goto matched; + } +# else + hcollseq = collseq[cend]; +# endif + } + + if (lcollseq <= hcollseq && fcollseq <= hcollseq) + goto matched; + } +# ifdef WIDE_CHAR_VERSION + range_not_matched: +# endif +#else + /* We use a boring value comparison of the character + values. This is better than comparing using + `strcoll' since the latter would have surprising + and sometimes fatal consequences. */ + UCHAR cend = *p++; + + if (!(flags & FNM_NOESCAPE) && cend == L('\\')) + cend = *p++; + if (cend == L('\0')) + return FNM_NOMATCH; + + /* It is a range. */ + if (cold <= fn && fn <= cend) + goto matched; +#endif + + c = *p++; + } + } + + if (c == L(']')) + break; + } + + if (!not) + return FNM_NOMATCH; + break; + + matched: + /* Skip the rest of the [...] that already matched. */ + do + { + ignore_next: + c = *p++; + + if (c == L('\0')) + /* [... (unterminated) loses. */ + return FNM_NOMATCH; + + if (!(flags & FNM_NOESCAPE) && c == L('\\')) + { + if (*p == L('\0')) + return FNM_NOMATCH; + /* XXX 1003.2d11 is unclear if this is right. */ + ++p; + } + else if (c == L('[') && *p == L(':')) + { + int c1 = 0; + const CHAR *startp = p; + + while (1) + { + c = *++p; + if (++c1 == CHAR_CLASS_MAX_LENGTH) + return FNM_NOMATCH; + + if (*p == L(':') && p[1] == L(']')) + break; + + if (c < L('a') || c >= L('z')) + { + p = startp; + goto ignore_next; + } + } + p += 2; + c = *p++; + } + else if (c == L('[') && *p == L('=')) + { + c = *++p; + if (c == L('\0')) + return FNM_NOMATCH; + c = *++p; + if (c != L('=') || p[1] != L(']')) + return FNM_NOMATCH; + p += 2; + c = *p++; + } + else if (c == L('[') && *p == L('.')) + { + ++p; + while (1) + { + c = *++p; + if (c == '\0') + return FNM_NOMATCH; + + if (*p == L('.') && p[1] == L(']')) + break; + } + p += 2; + c = *p++; + } + } + while (c != L(']')); + if (not) + return FNM_NOMATCH; + } + break; + + case L('+'): + case L('@'): + case L('!'): + if (__builtin_expect (flags & FNM_EXTMATCH, 0) && *p == '(') + { + int res; + + res = EXT (c, p, n, string_end, no_leading_period, flags); + if (res != -1) + return res; + } + goto normal_match; + + case L('/'): + if (NO_LEADING_PERIOD (flags)) + { + if (n == string_end || c != (UCHAR) *n) + return FNM_NOMATCH; + + new_no_leading_period = 1; + break; + } + /* FALLTHROUGH */ + default: + normal_match: + if (n == string_end || c != FOLD ((UCHAR) *n)) + return FNM_NOMATCH; + } + + no_leading_period = new_no_leading_period; + ++n; + } + + if (n == string_end) + return 0; + + if ((flags & FNM_LEADING_DIR) && n != string_end && *n == L('/')) + /* The FNM_LEADING_DIR flag says that "foo*" matches "foobar/frobozz". */ + return 0; + + return FNM_NOMATCH; +} + + +static const CHAR * +internal_function +END (const CHAR *pattern) +{ + const CHAR *p = pattern; + + while (1) + if (*++p == L('\0')) + /* This is an invalid pattern. */ + return pattern; + else if (*p == L('[')) + { + /* Handle brackets special. */ + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + /* Skip the not sign. We have to recognize it because of a possibly + following ']'. */ + if (*++p == L('!') || (posixly_correct < 0 && *p == L('^'))) + ++p; + /* A leading ']' is recognized as such. */ + if (*p == L(']')) + ++p; + /* Skip over all characters of the list. */ + while (*p != L(']')) + if (*p++ == L('\0')) + /* This is no valid pattern. */ + return pattern; + } + else if ((*p == L('?') || *p == L('*') || *p == L('+') || *p == L('@') + || *p == L('!')) && p[1] == L('(')) + p = END (p + 1); + else if (*p == L(')')) + break; + + return p + 1; +} + + +static int +internal_function +EXT (INT opt, const CHAR *pattern, const CHAR *string, const CHAR *string_end, + int no_leading_period, int flags) +{ + const CHAR *startp; + int level; + struct patternlist + { + struct patternlist *next; + CHAR str[1]; + } *list = NULL; + struct patternlist **lastp = &list; + size_t pattern_len = STRLEN (pattern); + const CHAR *p; + const CHAR *rs; + enum { ALLOCA_LIMIT = 8000 }; + + /* Parse the pattern. Store the individual parts in the list. */ + level = 0; + for (startp = p = pattern + 1; level >= 0; ++p) + if (*p == L('\0')) + /* This is an invalid pattern. */ + return -1; + else if (*p == L('[')) + { + /* Handle brackets special. */ + if (posixly_correct == 0) + posixly_correct = getenv ("POSIXLY_CORRECT") != NULL ? 1 : -1; + + /* Skip the not sign. We have to recognize it because of a possibly + following ']'. */ + if (*++p == L('!') || (posixly_correct < 0 && *p == L('^'))) + ++p; + /* A leading ']' is recognized as such. */ + if (*p == L(']')) + ++p; + /* Skip over all characters of the list. */ + while (*p != L(']')) + if (*p++ == L('\0')) + /* This is no valid pattern. */ + return -1; + } + else if ((*p == L('?') || *p == L('*') || *p == L('+') || *p == L('@') + || *p == L('!')) && p[1] == L('(')) + /* Remember the nesting level. */ + ++level; + else if (*p == L(')')) + { + if (level-- == 0) + { + /* This means we found the end of the pattern. */ +#define NEW_PATTERN \ + struct patternlist *newp; \ + size_t plen; \ + size_t plensize; \ + size_t newpsize; \ + \ + plen = (opt == L('?') || opt == L('@') \ + ? pattern_len \ + : p - startp + 1); \ + plensize = plen * sizeof (CHAR); \ + newpsize = offsetof (struct patternlist, str) + plensize; \ + if ((size_t) -1 / sizeof (CHAR) < plen \ + || newpsize < offsetof (struct patternlist, str) \ + || ALLOCA_LIMIT <= newpsize) \ + return -1; \ + newp = (struct patternlist *) alloca (newpsize); \ + *((CHAR *) MEMPCPY (newp->str, startp, p - startp)) = L('\0'); \ + newp->next = NULL; \ + *lastp = newp; \ + lastp = &newp->next + NEW_PATTERN; + } + } + else if (*p == L('|')) + { + if (level == 0) + { + NEW_PATTERN; + startp = p + 1; + } + } + assert (list != NULL); + assert (p[-1] == L(')')); +#undef NEW_PATTERN + + switch (opt) + { + case L('*'): + if (FCT (p, string, string_end, no_leading_period, flags) == 0) + return 0; + /* FALLTHROUGH */ + + case L('+'): + do + { + for (rs = string; rs <= string_end; ++rs) + /* First match the prefix with the current pattern with the + current pattern. */ + if (FCT (list->str, string, rs, no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) == 0 + /* This was successful. Now match the rest with the rest + of the pattern. */ + && (FCT (p, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags) ? 1 : 0, + flags & FNM_FILE_NAME + ? flags : flags & ~FNM_PERIOD) == 0 + /* This didn't work. Try the whole pattern. */ + || (rs != string + && FCT (pattern - 1, rs, string_end, + rs == string + ? no_leading_period + : (rs[-1] == '/' && NO_LEADING_PERIOD (flags) + ? 1 : 0), + flags & FNM_FILE_NAME + ? flags : flags & ~FNM_PERIOD) == 0))) + /* It worked. Signal success. */ + return 0; + } + while ((list = list->next) != NULL); + + /* None of the patterns lead to a match. */ + return FNM_NOMATCH; + + case L('?'): + if (FCT (p, string, string_end, no_leading_period, flags) == 0) + return 0; + /* FALLTHROUGH */ + + case L('@'): + do + /* I cannot believe it but `strcat' is actually acceptable + here. Match the entire string with the prefix from the + pattern list and the rest of the pattern following the + pattern list. */ + if (FCT (STRCAT (list->str, p), string, string_end, + no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) == 0) + /* It worked. Signal success. */ + return 0; + while ((list = list->next) != NULL); + + /* None of the patterns lead to a match. */ + return FNM_NOMATCH; + + case L('!'): + for (rs = string; rs <= string_end; ++rs) + { + struct patternlist *runp; + + for (runp = list; runp != NULL; runp = runp->next) + if (FCT (runp->str, string, rs, no_leading_period, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) == 0) + break; + + /* If none of the patterns matched see whether the rest does. */ + if (runp == NULL + && (FCT (p, rs, string_end, + rs == string + ? no_leading_period + : rs[-1] == '/' && NO_LEADING_PERIOD (flags) ? 1 : 0, + flags & FNM_FILE_NAME ? flags : flags & ~FNM_PERIOD) + == 0)) + /* This is successful. */ + return 0; + } + + /* None of the patterns together with the rest of the pattern + lead to a match. */ + return FNM_NOMATCH; + + default: + assert (! "Invalid extended matching operator"); + break; + } + + return -1; +} + + +#undef FOLD +#undef CHAR +#undef UCHAR +#undef INT +#undef FCT +#undef EXT +#undef END +#undef MEMPCPY +#undef MEMCHR +#undef STRCOLL +#undef STRLEN +#undef STRCAT +#undef L +#undef BTOWC diff --git a/contrib/diff/lib/getopt.c b/contrib/diff/lib/getopt.c new file mode 100644 index 0000000..8f017f8 --- /dev/null +++ b/contrib/diff/lib/getopt.c @@ -0,0 +1,1254 @@ +/* Getopt for GNU. + NOTE: getopt is now part of the C library, so if you don't know what + "Keep this file name-space clean" means, talk to drepper@gnu.org + before changing it! + Copyright (C) 1987,88,89,90,91,92,93,94,95,96,98,99,2000,2001,2002,2003,2004 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* This tells Alpha OSF/1 not to define a getopt prototype in . + Ditto for AIX 3.2 and . */ +#ifndef _NO_PROTO +# define _NO_PROTO +#endif + +#ifdef HAVE_CONFIG_H +# include +#endif + +#include + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself. This code is part of the GNU C + Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object files, + it is simpler to just do this in the source for each such file. */ + +#define GETOPT_INTERFACE_VERSION 2 +#if !defined _LIBC && defined __GLIBC__ && __GLIBC__ >= 2 +# include +# if _GNU_GETOPT_INTERFACE_VERSION == GETOPT_INTERFACE_VERSION +# define ELIDE_CODE +# endif +#endif + +#ifndef ELIDE_CODE + + +/* This needs to come after some library #include + to get __GNU_LIBRARY__ defined. */ +#ifdef __GNU_LIBRARY__ +/* Don't include stdlib.h for non-GNU C libraries because some of them + contain conflicting prototypes for getopt. */ +# include +# include +#endif /* GNU C library. */ + +#include + +#ifdef VMS +# include +#endif + +#ifdef _LIBC +# include +#else +# include "gettext.h" +# define _(msgid) gettext (msgid) +#endif + +#if defined _LIBC && defined USE_IN_LIBIO +# include +#endif + +#ifndef attribute_hidden +# define attribute_hidden +#endif + +/* This version of `getopt' appears to the caller like standard Unix `getopt' + but it behaves differently for the user, since it allows the user + to intersperse the options with the other arguments. + + As `getopt' works, it permutes the elements of ARGV so that, + when it is done, all the options precede everything else. Thus + all application programs are extended to handle flexible argument order. + + Setting the environment variable POSIXLY_CORRECT disables permutation. + Then the behavior is completely standard. + + GNU application programs can use a third alternative mode in which + they can distinguish the relative order of options and other arguments. */ + +#include "getopt.h" +#include "getopt_int.h" + +/* For communication from `getopt' to the caller. + When `getopt' finds an option that takes an argument, + the argument value is returned here. + Also, when `ordering' is RETURN_IN_ORDER, + each non-option ARGV-element is returned here. */ + +char *optarg; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +/* 1003.2 says this must be 1 before any call. */ +int optind = 1; + +/* Callers store zero here to inhibit the error message + for unrecognized options. */ + +int opterr = 1; + +/* Set to an option character which was unrecognized. + This must be initialized on some systems to avoid linking in the + system's own getopt implementation. */ + +int optopt = '?'; + +/* Keep a global copy of all internal members of getopt_data. */ + +static struct _getopt_data getopt_data; + + +#ifndef __GNU_LIBRARY__ + +/* Avoid depending on library functions or files + whose names are inconsistent. */ + +#ifndef getenv +extern char *getenv (); +#endif + +#endif /* not __GNU_LIBRARY__ */ + +#ifdef _LIBC +/* Stored original parameters. + XXX This is no good solution. We should rather copy the args so + that we can compare them later. But we must not use malloc(3). */ +extern int __libc_argc; +extern char **__libc_argv; + +/* Bash 2.0 gives us an environment variable containing flags + indicating ARGV elements that should not be considered arguments. */ + +# ifdef USE_NONOPTION_FLAGS +/* Defined in getopt_init.c */ +extern char *__getopt_nonoption_flags; +# endif + +# ifdef USE_NONOPTION_FLAGS +# define SWAP_FLAGS(ch1, ch2) \ + if (d->__nonoption_flags_len > 0) \ + { \ + char __tmp = __getopt_nonoption_flags[ch1]; \ + __getopt_nonoption_flags[ch1] = __getopt_nonoption_flags[ch2]; \ + __getopt_nonoption_flags[ch2] = __tmp; \ + } +# else +# define SWAP_FLAGS(ch1, ch2) +# endif +#else /* !_LIBC */ +# define SWAP_FLAGS(ch1, ch2) +#endif /* _LIBC */ + +/* Exchange two adjacent subsequences of ARGV. + One subsequence is elements [first_nonopt,last_nonopt) + which contains all the non-options that have been skipped so far. + The other is elements [last_nonopt,optind), which contains all + the options processed since those non-options were skipped. + + `first_nonopt' and `last_nonopt' are relocated so that they describe + the new indices of the non-options in ARGV after they are moved. */ + +static void +exchange (char **argv, struct _getopt_data *d) +{ + int bottom = d->__first_nonopt; + int middle = d->__last_nonopt; + int top = d->optind; + char *tem; + + /* Exchange the shorter segment with the far end of the longer segment. + That puts the shorter segment into the right place. + It leaves the longer segment in the right place overall, + but it consists of two parts that need to be swapped next. */ + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + /* First make sure the handling of the `__getopt_nonoption_flags' + string can work normally. Our top argument must be in the range + of the string. */ + if (d->__nonoption_flags_len > 0 && top >= d->__nonoption_flags_max_len) + { + /* We must extend the array. The user plays games with us and + presents new arguments. */ + char *new_str = malloc (top + 1); + if (new_str == NULL) + d->__nonoption_flags_len = d->__nonoption_flags_max_len = 0; + else + { + memset (__mempcpy (new_str, __getopt_nonoption_flags, + d->__nonoption_flags_max_len), + '\0', top + 1 - d->__nonoption_flags_max_len); + d->__nonoption_flags_max_len = top + 1; + __getopt_nonoption_flags = new_str; + } + } +#endif + + while (top > middle && middle > bottom) + { + if (top - middle > middle - bottom) + { + /* Bottom segment is the short one. */ + int len = middle - bottom; + register int i; + + /* Swap it with the top part of the top segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[top - (middle - bottom) + i]; + argv[top - (middle - bottom) + i] = tem; + SWAP_FLAGS (bottom + i, top - (middle - bottom) + i); + } + /* Exclude the moved bottom segment from further swapping. */ + top -= len; + } + else + { + /* Top segment is the short one. */ + int len = top - middle; + register int i; + + /* Swap it with the bottom part of the bottom segment. */ + for (i = 0; i < len; i++) + { + tem = argv[bottom + i]; + argv[bottom + i] = argv[middle + i]; + argv[middle + i] = tem; + SWAP_FLAGS (bottom + i, middle + i); + } + /* Exclude the moved top segment from further swapping. */ + bottom += len; + } + } + + /* Update records for the slots the non-options now occupy. */ + + d->__first_nonopt += (d->optind - d->__last_nonopt); + d->__last_nonopt = d->optind; +} + +/* Initialize the internal data when the first call is made. */ + +static const char * +_getopt_initialize (int argc, char *const *argv, const char *optstring, + struct _getopt_data *d) +{ + /* Start processing options with ARGV-element 1 (since ARGV-element 0 + is the program name); the sequence of previously skipped + non-option ARGV-elements is empty. */ + + d->__first_nonopt = d->__last_nonopt = d->optind; + + d->__nextchar = NULL; + + d->__posixly_correct = !!getenv ("POSIXLY_CORRECT"); + + /* Determine how to handle the ordering of options and nonoptions. */ + + if (optstring[0] == '-') + { + d->__ordering = RETURN_IN_ORDER; + ++optstring; + } + else if (optstring[0] == '+') + { + d->__ordering = REQUIRE_ORDER; + ++optstring; + } + else if (d->__posixly_correct) + d->__ordering = REQUIRE_ORDER; + else + d->__ordering = PERMUTE; + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + if (!d->__posixly_correct + && argc == __libc_argc && argv == __libc_argv) + { + if (d->__nonoption_flags_max_len == 0) + { + if (__getopt_nonoption_flags == NULL + || __getopt_nonoption_flags[0] == '\0') + d->__nonoption_flags_max_len = -1; + else + { + const char *orig_str = __getopt_nonoption_flags; + int len = d->__nonoption_flags_max_len = strlen (orig_str); + if (d->__nonoption_flags_max_len < argc) + d->__nonoption_flags_max_len = argc; + __getopt_nonoption_flags = + (char *) malloc (d->__nonoption_flags_max_len); + if (__getopt_nonoption_flags == NULL) + d->__nonoption_flags_max_len = -1; + else + memset (__mempcpy (__getopt_nonoption_flags, orig_str, len), + '\0', d->__nonoption_flags_max_len - len); + } + } + d->__nonoption_flags_len = d->__nonoption_flags_max_len; + } + else + d->__nonoption_flags_len = 0; +#endif + + return optstring; +} + +/* Scan elements of ARGV (whose length is ARGC) for option characters + given in OPTSTRING. + + If an element of ARGV starts with '-', and is not exactly "-" or "--", + then it is an option element. The characters of this element + (aside from the initial '-') are option characters. If `getopt' + is called repeatedly, it returns successively each of the option characters + from each of the option elements. + + If `getopt' finds another option character, it returns that character, + updating `optind' and `nextchar' so that the next call to `getopt' can + resume the scan with the following option character or ARGV-element. + + If there are no more option characters, `getopt' returns -1. + Then `optind' is the index in ARGV of the first ARGV-element + that is not an option. (The ARGV-elements have been permuted + so that those that are not options now come last.) + + OPTSTRING is a string containing the legitimate option characters. + If an option character is seen that is not listed in OPTSTRING, + return '?' after printing an error message. If you set `opterr' to + zero, the error message is suppressed but we still return '?'. + + If a char in OPTSTRING is followed by a colon, that means it wants an arg, + so the following text in the same ARGV-element, or the text of the following + ARGV-element, is returned in `optarg'. Two colons mean an option that + wants an optional arg; if there is text in the current ARGV-element, + it is returned in `optarg', otherwise `optarg' is set to zero. + + If OPTSTRING starts with `-' or `+', it requests different methods of + handling the non-option ARGV-elements. + See the comments about RETURN_IN_ORDER and REQUIRE_ORDER, above. + + Long-named options begin with `--' instead of `-'. + Their names may be abbreviated as long as the abbreviation is unique + or is an exact match for some defined option. If they have an + argument, it follows the option name in the same ARGV-element, separated + from the option name by a `=', or else the in next ARGV-element. + When `getopt' finds a long-named option, it returns 0 if that option's + `flag' field is nonzero, the value of the option's `val' field + if the `flag' field is zero. + + The elements of ARGV aren't really const, because we permute them. + But we pretend they're const in the prototype to be compatible + with other systems. + + LONGOPTS is a vector of `struct option' terminated by an + element containing a name which is zero. + + LONGIND returns the index in LONGOPT of the long-named option found. + It is only valid when a long-named option has been found by the most + recent call. + + If LONG_ONLY is nonzero, '-' as well as '--' can introduce + long-named options. */ + +int +_getopt_internal_r (int argc, char *const *argv, const char *optstring, + const struct option *longopts, int *longind, + int long_only, struct _getopt_data *d) +{ + int print_errors = d->opterr; + if (optstring[0] == ':') + print_errors = 0; + + if (argc < 1) + return -1; + + d->optarg = NULL; + + if (d->optind == 0 || !d->__initialized) + { + if (d->optind == 0) + d->optind = 1; /* Don't scan ARGV[0], the program name. */ + optstring = _getopt_initialize (argc, argv, optstring, d); + d->__initialized = 1; + } + + /* Test whether ARGV[optind] points to a non-option argument. + Either it does not have option syntax, or there is an environment flag + from the shell indicating it is not an option. The later information + is only used when the used in the GNU libc. */ +#if defined _LIBC && defined USE_NONOPTION_FLAGS +# define NONOPTION_P (argv[d->optind][0] != '-' || argv[d->optind][1] == '\0' \ + || (d->optind < d->__nonoption_flags_len \ + && __getopt_nonoption_flags[d->optind] == '1')) +#else +# define NONOPTION_P (argv[d->optind][0] != '-' || argv[d->optind][1] == '\0') +#endif + + if (d->__nextchar == NULL || *d->__nextchar == '\0') + { + /* Advance to the next ARGV-element. */ + + /* Give FIRST_NONOPT & LAST_NONOPT rational values if OPTIND has been + moved back by the user (who may also have changed the arguments). */ + if (d->__last_nonopt > d->optind) + d->__last_nonopt = d->optind; + if (d->__first_nonopt > d->optind) + d->__first_nonopt = d->optind; + + if (d->__ordering == PERMUTE) + { + /* If we have just processed some options following some non-options, + exchange them so that the options come first. */ + + if (d->__first_nonopt != d->__last_nonopt + && d->__last_nonopt != d->optind) + exchange ((char **) argv, d); + else if (d->__last_nonopt != d->optind) + d->__first_nonopt = d->optind; + + /* Skip any additional non-options + and extend the range of non-options previously skipped. */ + + while (d->optind < argc && NONOPTION_P) + d->optind++; + d->__last_nonopt = d->optind; + } + + /* The special ARGV-element `--' means premature end of options. + Skip it like a null option, + then exchange with previous non-options as if it were an option, + then skip everything else like a non-option. */ + + if (d->optind != argc && !strcmp (argv[d->optind], "--")) + { + d->optind++; + + if (d->__first_nonopt != d->__last_nonopt + && d->__last_nonopt != d->optind) + exchange ((char **) argv, d); + else if (d->__first_nonopt == d->__last_nonopt) + d->__first_nonopt = d->optind; + d->__last_nonopt = argc; + + d->optind = argc; + } + + /* If we have done all the ARGV-elements, stop the scan + and back over any non-options that we skipped and permuted. */ + + if (d->optind == argc) + { + /* Set the next-arg-index to point at the non-options + that we previously skipped, so the caller will digest them. */ + if (d->__first_nonopt != d->__last_nonopt) + d->optind = d->__first_nonopt; + return -1; + } + + /* If we have come to a non-option and did not permute it, + either stop the scan or describe it to the caller and pass it by. */ + + if (NONOPTION_P) + { + if (d->__ordering == REQUIRE_ORDER) + return -1; + d->optarg = argv[d->optind++]; + return 1; + } + + /* We have found another option-ARGV-element. + Skip the initial punctuation. */ + + d->__nextchar = (argv[d->optind] + 1 + + (longopts != NULL && argv[d->optind][1] == '-')); + } + + /* Decode the current option-ARGV-element. */ + + /* Check whether the ARGV-element is a long option. + + If long_only and the ARGV-element has the form "-f", where f is + a valid short option, don't consider it an abbreviated form of + a long option that starts with f. Otherwise there would be no + way to give the -f short option. + + On the other hand, if there's a long option "fubar" and + the ARGV-element is "-fu", do consider that an abbreviation of + the long option, just like "--fu", and not "-f" with arg "u". + + This distinction seems to be the most useful approach. */ + + if (longopts != NULL + && (argv[d->optind][1] == '-' + || (long_only && (argv[d->optind][2] + || !strchr (optstring, argv[d->optind][1]))))) + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound = -1; + int option_index; + + for (nameend = d->__nextchar; *nameend && *nameend != '='; nameend++) + /* Do nothing. */ ; + + /* Test all long options for either exact match + or abbreviated matches. */ + for (p = longopts, option_index = 0; p->name; p++, option_index++) + if (!strncmp (p->name, d->__nextchar, nameend - d->__nextchar)) + { + if ((unsigned int) (nameend - d->__nextchar) + == (unsigned int) strlen (p->name)) + { + /* Exact match found. */ + pfound = p; + indfound = option_index; + exact = 1; + break; + } + else if (pfound == NULL) + { + /* First nonexact match found. */ + pfound = p; + indfound = option_index; + } + else if (long_only + || pfound->has_arg != p->has_arg + || pfound->flag != p->flag + || pfound->val != p->val) + /* Second or later nonexact match found. */ + ambig = 1; + } + + if (ambig && !exact) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("%s: option `%s' is ambiguous\n"), + argv[0], argv[d->optind]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option `%s' is ambiguous\n"), + argv[0], argv[d->optind]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optind++; + d->optopt = 0; + return '?'; + } + + if (pfound != NULL) + { + option_index = indfound; + d->optind++; + if (*nameend) + { + /* Don't test has_arg with >, because some C compilers don't + allow it to be used on enums. */ + if (pfound->has_arg) + d->optarg = nameend + 1; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (argv[d->optind - 1][1] == '-') + { + /* --option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("\ +%s: option `--%s' doesn't allow an argument\n"), + argv[0], pfound->name); +#else + fprintf (stderr, _("\ +%s: option `--%s' doesn't allow an argument\n"), + argv[0], pfound->name); +#endif + } + else + { + /* +option or -option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("\ +%s: option `%c%s' doesn't allow an argument\n"), + argv[0], argv[d->optind - 1][0], + pfound->name); +#else + fprintf (stderr, _("\ +%s: option `%c%s' doesn't allow an argument\n"), + argv[0], argv[d->optind - 1][0], + pfound->name); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + + d->__nextchar += strlen (d->__nextchar); + + d->optopt = pfound->val; + return '?'; + } + } + else if (pfound->has_arg == 1) + { + if (d->optind < argc) + d->optarg = argv[d->optind++]; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optopt = pfound->val; + return optstring[0] == ':' ? ':' : '?'; + } + } + d->__nextchar += strlen (d->__nextchar); + if (longind != NULL) + *longind = option_index; + if (pfound->flag) + { + *(pfound->flag) = pfound->val; + return 0; + } + return pfound->val; + } + + /* Can't find it as a long option. If this is not getopt_long_only, + or the option starts with '--' or is not a valid short + option, then it's an error. + Otherwise interpret it as a short option. */ + if (!long_only || argv[d->optind][1] == '-' + || strchr (optstring, *d->__nextchar) == NULL) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (argv[d->optind][1] == '-') + { + /* --option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: unrecognized option `--%s'\n"), + argv[0], d->__nextchar); +#else + fprintf (stderr, _("%s: unrecognized option `--%s'\n"), + argv[0], d->__nextchar); +#endif + } + else + { + /* +option or -option */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: unrecognized option `%c%s'\n"), + argv[0], argv[d->optind][0], d->__nextchar); +#else + fprintf (stderr, _("%s: unrecognized option `%c%s'\n"), + argv[0], argv[d->optind][0], d->__nextchar); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + d->__nextchar = (char *) ""; + d->optind++; + d->optopt = 0; + return '?'; + } + } + + /* Look at and handle the next short option-character. */ + + { + char c = *d->__nextchar++; + char *temp = strchr (optstring, c); + + /* Increment `optind' when we start to process its last character. */ + if (*d->__nextchar == '\0') + ++d->optind; + + if (temp == NULL || c == ':') + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + int n; +#endif + + if (d->__posixly_correct) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: illegal option -- %c\n"), + argv[0], c); +#else + fprintf (stderr, _("%s: illegal option -- %c\n"), argv[0], c); +#endif + } + else + { +#if defined _LIBC && defined USE_IN_LIBIO + n = __asprintf (&buf, _("%s: invalid option -- %c\n"), + argv[0], c); +#else + fprintf (stderr, _("%s: invalid option -- %c\n"), argv[0], c); +#endif + } + +#if defined _LIBC && defined USE_IN_LIBIO + if (n >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#endif + } + d->optopt = c; + return '?'; + } + /* Convenience. Treat POSIX -W foo same as long option --foo */ + if (temp[0] == 'W' && temp[1] == ';') + { + char *nameend; + const struct option *p; + const struct option *pfound = NULL; + int exact = 0; + int ambig = 0; + int indfound = 0; + int option_index; + + /* This is an option that requires an argument. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + /* If we end this ARGV-element by taking the rest as an arg, + we must advance to the next element now. */ + d->optind++; + } + else if (d->optind == argc) + { + if (print_errors) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, + _("%s: option requires an argument -- %c\n"), + argv[0], c) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option requires an argument -- %c\n"), + argv[0], c); +#endif + } + d->optopt = c; + if (optstring[0] == ':') + c = ':'; + else + c = '?'; + return c; + } + else + /* We already incremented `d->optind' once; + increment it again when taking next ARGV-elt as argument. */ + d->optarg = argv[d->optind++]; + + /* optarg is now the argument, see if it's in the + table of longopts. */ + + for (d->__nextchar = nameend = d->optarg; *nameend && *nameend != '='; + nameend++) + /* Do nothing. */ ; + + /* Test all long options for either exact match + or abbreviated matches. */ + for (p = longopts, option_index = 0; p->name; p++, option_index++) + if (!strncmp (p->name, d->__nextchar, nameend - d->__nextchar)) + { + if ((unsigned int) (nameend - d->__nextchar) == strlen (p->name)) + { + /* Exact match found. */ + pfound = p; + indfound = option_index; + exact = 1; + break; + } + else if (pfound == NULL) + { + /* First nonexact match found. */ + pfound = p; + indfound = option_index; + } + else + /* Second or later nonexact match found. */ + ambig = 1; + } + if (ambig && !exact) + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("%s: option `-W %s' is ambiguous\n"), + argv[0], argv[d->optind]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("%s: option `-W %s' is ambiguous\n"), + argv[0], argv[d->optind]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + d->optind++; + return '?'; + } + if (pfound != NULL) + { + option_index = indfound; + if (*nameend) + { + /* Don't test has_arg with >, because some C compilers don't + allow it to be used on enums. */ + if (pfound->has_arg) + d->optarg = nameend + 1; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `-W %s' doesn't allow an argument\n"), + argv[0], pfound->name) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, _("\ +%s: option `-W %s' doesn't allow an argument\n"), + argv[0], pfound->name); +#endif + } + + d->__nextchar += strlen (d->__nextchar); + return '?'; + } + } + else if (pfound->has_arg == 1) + { + if (d->optind < argc) + d->optarg = argv[d->optind++]; + else + { + if (print_errors) + { +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 + |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option `%s' requires an argument\n"), + argv[0], argv[d->optind - 1]); +#endif + } + d->__nextchar += strlen (d->__nextchar); + return optstring[0] == ':' ? ':' : '?'; + } + } + d->__nextchar += strlen (d->__nextchar); + if (longind != NULL) + *longind = option_index; + if (pfound->flag) + { + *(pfound->flag) = pfound->val; + return 0; + } + return pfound->val; + } + d->__nextchar = NULL; + return 'W'; /* Let the application handle it. */ + } + if (temp[1] == ':') + { + if (temp[2] == ':') + { + /* This is an option that accepts an argument optionally. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + d->optind++; + } + else + d->optarg = NULL; + d->__nextchar = NULL; + } + else + { + /* This is an option that requires an argument. */ + if (*d->__nextchar != '\0') + { + d->optarg = d->__nextchar; + /* If we end this ARGV-element by taking the rest as an arg, + we must advance to the next element now. */ + d->optind++; + } + else if (d->optind == argc) + { + if (print_errors) + { + /* 1003.2 specifies the format of this message. */ +#if defined _LIBC && defined USE_IN_LIBIO + char *buf; + + if (__asprintf (&buf, _("\ +%s: option requires an argument -- %c\n"), + argv[0], c) >= 0) + { + _IO_flockfile (stderr); + + int old_flags2 = ((_IO_FILE *) stderr)->_flags2; + ((_IO_FILE *) stderr)->_flags2 |= _IO_FLAGS2_NOTCANCEL; + + if (_IO_fwide (stderr, 0) > 0) + __fwprintf (stderr, L"%s", buf); + else + fputs (buf, stderr); + + ((_IO_FILE *) stderr)->_flags2 = old_flags2; + _IO_funlockfile (stderr); + + free (buf); + } +#else + fprintf (stderr, + _("%s: option requires an argument -- %c\n"), + argv[0], c); +#endif + } + d->optopt = c; + if (optstring[0] == ':') + c = ':'; + else + c = '?'; + } + else + /* We already incremented `optind' once; + increment it again when taking next ARGV-elt as argument. */ + d->optarg = argv[d->optind++]; + d->__nextchar = NULL; + } + } + return c; + } +} + +int +_getopt_internal (int argc, char *const *argv, const char *optstring, + const struct option *longopts, int *longind, int long_only) +{ + int result; + + getopt_data.optind = optind; + getopt_data.opterr = opterr; + + result = _getopt_internal_r (argc, argv, optstring, longopts, + longind, long_only, &getopt_data); + + optind = getopt_data.optind; + optarg = getopt_data.optarg; + optopt = getopt_data.optopt; + + return result; +} + +int +getopt (int argc, char *const *argv, const char *optstring) +{ + return _getopt_internal (argc, argv, optstring, + (const struct option *) 0, + (int *) 0, + 0); +} + +#endif /* Not ELIDE_CODE. */ + +#ifdef TEST + +/* Compile with -DTEST to make an executable for use in testing + the above definition of `getopt'. */ + +int +main (int argc, char **argv) +{ + int c; + int digit_optind = 0; + + while (1) + { + int this_option_optind = optind ? optind : 1; + + c = getopt (argc, argv, "abc:d:0123456789"); + if (c == -1) + break; + + switch (c) + { + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (digit_optind != 0 && digit_optind != this_option_optind) + printf ("digits occur in two different argv-elements.\n"); + digit_optind = this_option_optind; + printf ("option %c\n", c); + break; + + case 'a': + printf ("option a\n"); + break; + + case 'b': + printf ("option b\n"); + break; + + case 'c': + printf ("option c with value `%s'\n", optarg); + break; + + case '?': + break; + + default: + printf ("?? getopt returned character code 0%o ??\n", c); + } + } + + if (optind < argc) + { + printf ("non-option ARGV-elements: "); + while (optind < argc) + printf ("%s ", argv[optind++]); + printf ("\n"); + } + + exit (0); +} + +#endif /* TEST */ diff --git a/contrib/diff/lib/getopt.h b/contrib/diff/lib/getopt.h new file mode 100644 index 0000000..c61768c --- /dev/null +++ b/contrib/diff/lib/getopt.h @@ -0,0 +1,176 @@ +/* Declarations for getopt. + Copyright (C) 1989-1994,1996-1999,2001,2003,2004 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _GETOPT_H + +#ifndef __need_getopt +# define _GETOPT_H 1 +#endif + +/* If __GNU_LIBRARY__ is not already defined, either we are being used + standalone, or this is the first header included in the source file. + If we are being used with glibc, we need to include , but + that does not exist if we are standalone. So: if __GNU_LIBRARY__ is + not defined, include , which will pull in for us + if it's from glibc. (Why ctype.h? It's guaranteed to exist and it + doesn't flood the namespace with stuff the way some other headers do.) */ +#if !defined __GNU_LIBRARY__ +# include +#endif + +#ifndef __THROW +# ifndef __GNUC_PREREQ +# define __GNUC_PREREQ(maj, min) (0) +# endif +# if defined __cplusplus && __GNUC_PREREQ (2,8) +# define __THROW throw () +# else +# define __THROW +# endif +#endif + +#ifdef __cplusplus +extern "C" { +#endif + +/* For communication from `getopt' to the caller. + When `getopt' finds an option that takes an argument, + the argument value is returned here. + Also, when `ordering' is RETURN_IN_ORDER, + each non-option ARGV-element is returned here. */ + +extern char *optarg; + +/* Index in ARGV of the next element to be scanned. + This is used for communication to and from the caller + and for communication between successive calls to `getopt'. + + On entry to `getopt', zero means this is the first call; initialize. + + When `getopt' returns -1, this is the index of the first of the + non-option elements that the caller should itself scan. + + Otherwise, `optind' communicates from one call to the next + how much of ARGV has been scanned so far. */ + +extern int optind; + +/* Callers store zero here to inhibit the error message `getopt' prints + for unrecognized options. */ + +extern int opterr; + +/* Set to an option character which was unrecognized. */ + +extern int optopt; + +#ifndef __need_getopt +/* Describe the long-named options requested by the application. + The LONG_OPTIONS argument to getopt_long or getopt_long_only is a vector + of `struct option' terminated by an element containing a name which is + zero. + + The field `has_arg' is: + no_argument (or 0) if the option does not take an argument, + required_argument (or 1) if the option requires an argument, + optional_argument (or 2) if the option takes an optional argument. + + If the field `flag' is not NULL, it points to a variable that is set + to the value given in the field `val' when the option is found, but + left unchanged if the option is not found. + + To have a long-named option do something other than set an `int' to + a compiled-in constant, such as set a value from `optarg', set the + option's `flag' field to zero and its `val' field to a nonzero + value (the equivalent single-letter option character, if there is + one). For long options that have a zero `flag' field, `getopt' + returns the contents of the `val' field. */ + +struct option +{ + const char *name; + /* has_arg can't be an enum because some compilers complain about + type mismatches in all the code that assumes it is an int. */ + int has_arg; + int *flag; + int val; +}; + +/* Names for the values of the `has_arg' field of `struct option'. */ + +# define no_argument 0 +# define required_argument 1 +# define optional_argument 2 +#endif /* need getopt */ + + +/* Get definitions and prototypes for functions to process the + arguments in ARGV (ARGC of them, minus the program name) for + options given in OPTS. + + Return the option character from OPTS just read. Return -1 when + there are no more options. For unrecognized options, or options + missing arguments, `optopt' is set to the option letter, and '?' is + returned. + + The OPTS string is a list of characters which are recognized option + letters, optionally followed by colons, specifying that that letter + takes an argument, to be placed in `optarg'. + + If a letter in OPTS is followed by two colons, its argument is + optional. This behavior is specific to the GNU `getopt'. + + The argument `--' causes premature termination of argument + scanning, explicitly telling `getopt' that there are no more + options. + + If OPTS begins with `--', then non-option arguments are treated as + arguments to the option '\0'. This behavior is specific to the GNU + `getopt'. */ + +#ifdef __GNU_LIBRARY__ +/* Many other libraries have conflicting prototypes for getopt, with + differences in the consts, in stdlib.h. To avoid compilation + errors, only prototype getopt for the GNU C library. */ +extern int getopt (int ___argc, char *const *___argv, const char *__shortopts) + __THROW; +#else /* not __GNU_LIBRARY__ */ +extern int getopt (); +#endif /* __GNU_LIBRARY__ */ + +#ifndef __need_getopt +extern int getopt_long (int ___argc, char *const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind) + __THROW; +extern int getopt_long_only (int ___argc, char *const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind) + __THROW; + +#endif + +#ifdef __cplusplus +} +#endif + +/* Make sure we later can get all the definitions and declarations. */ +#undef __need_getopt + +#endif /* getopt.h */ diff --git a/contrib/diff/lib/getopt1.c b/contrib/diff/lib/getopt1.c new file mode 100644 index 0000000..7671eba --- /dev/null +++ b/contrib/diff/lib/getopt1.c @@ -0,0 +1,191 @@ +/* getopt_long and getopt_long_only entry points for GNU getopt. + Copyright (C) 1987,88,89,90,91,92,93,94,96,97,98,2004 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifdef HAVE_CONFIG_H +#include +#endif + +#ifdef _LIBC +# include +#else +# include "getopt.h" +#endif +#include "getopt_int.h" + +#include + +/* Comment out all this code if we are using the GNU C Library, and are not + actually compiling the library itself. This code is part of the GNU C + Library, but also included in many other GNU distributions. Compiling + and linking in this code is a waste when using the GNU C library + (especially if it is a shared library). Rather than having every GNU + program understand `configure --with-gnu-libc' and omit the object files, + it is simpler to just do this in the source for each such file. */ + +#define GETOPT_INTERFACE_VERSION 2 +#if !defined _LIBC && defined __GLIBC__ && __GLIBC__ >= 2 +#include +#if _GNU_GETOPT_INTERFACE_VERSION == GETOPT_INTERFACE_VERSION +#define ELIDE_CODE +#endif +#endif + +#ifndef ELIDE_CODE + + +/* This needs to come after some library #include + to get __GNU_LIBRARY__ defined. */ +#ifdef __GNU_LIBRARY__ +#include +#endif + +#ifndef NULL +#define NULL 0 +#endif + +int +getopt_long (int argc, char *const *argv, const char *options, + const struct option *long_options, int *opt_index) +{ + return _getopt_internal (argc, argv, options, long_options, opt_index, 0); +} + +int +_getopt_long_r (int argc, char *const *argv, const char *options, + const struct option *long_options, int *opt_index, + struct _getopt_data *d) +{ + return _getopt_internal_r (argc, argv, options, long_options, opt_index, + 0, d); +} + +/* Like getopt_long, but '-' as well as '--' can indicate a long option. + If an option that starts with '-' (not '--') doesn't match a long option, + but does match a short option, it is parsed as a short option + instead. */ + +int +getopt_long_only (int argc, char *const *argv, const char *options, + const struct option *long_options, int *opt_index) +{ + return _getopt_internal (argc, argv, options, long_options, opt_index, 1); +} + +int +_getopt_long_only_r (int argc, char *const *argv, const char *options, + const struct option *long_options, int *opt_index, + struct _getopt_data *d) +{ + return _getopt_internal_r (argc, argv, options, long_options, opt_index, + 1, d); +} + +#endif /* Not ELIDE_CODE. */ + +#ifdef TEST + +#include + +int +main (int argc, char **argv) +{ + int c; + int digit_optind = 0; + + while (1) + { + int this_option_optind = optind ? optind : 1; + int option_index = 0; + static struct option long_options[] = + { + {"add", 1, 0, 0}, + {"append", 0, 0, 0}, + {"delete", 1, 0, 0}, + {"verbose", 0, 0, 0}, + {"create", 0, 0, 0}, + {"file", 1, 0, 0}, + {0, 0, 0, 0} + }; + + c = getopt_long (argc, argv, "abc:d:0123456789", + long_options, &option_index); + if (c == -1) + break; + + switch (c) + { + case 0: + printf ("option %s", long_options[option_index].name); + if (optarg) + printf (" with arg %s", optarg); + printf ("\n"); + break; + + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (digit_optind != 0 && digit_optind != this_option_optind) + printf ("digits occur in two different argv-elements.\n"); + digit_optind = this_option_optind; + printf ("option %c\n", c); + break; + + case 'a': + printf ("option a\n"); + break; + + case 'b': + printf ("option b\n"); + break; + + case 'c': + printf ("option c with value `%s'\n", optarg); + break; + + case 'd': + printf ("option d with value `%s'\n", optarg); + break; + + case '?': + break; + + default: + printf ("?? getopt returned character code 0%o ??\n", c); + } + } + + if (optind < argc) + { + printf ("non-option ARGV-elements: "); + while (optind < argc) + printf ("%s ", argv[optind++]); + printf ("\n"); + } + + exit (0); +} + +#endif /* TEST */ diff --git a/contrib/diff/lib/getopt_int.h b/contrib/diff/lib/getopt_int.h new file mode 100644 index 0000000..0c5edde --- /dev/null +++ b/contrib/diff/lib/getopt_int.h @@ -0,0 +1,129 @@ +/* Internal declarations for getopt. + Copyright (C) 1989-1994,1996-1999,2001,2003,2004 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _GETOPT_INT_H +#define _GETOPT_INT_H 1 + +extern int _getopt_internal (int ___argc, char *const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + int __long_only); + + +/* Reentrant versions which can handle parsing multiple argument + vectors at the same time. */ + +/* Data type for reentrant functions. */ +struct _getopt_data +{ + /* These have exactly the same meaning as the corresponding global + variables, except that they are used for the reentrant + versions of getopt. */ + int optind; + int opterr; + int optopt; + char *optarg; + + /* Internal members. */ + + /* True if the internal members have been initialized. */ + int __initialized; + + /* The next char to be scanned in the option-element + in which the last option character we returned was found. + This allows us to pick up the scan where we left off. + + If this is zero, or a null string, it means resume the scan + by advancing to the next ARGV-element. */ + char *__nextchar; + + /* Describe how to deal with options that follow non-option ARGV-elements. + + If the caller did not specify anything, + the default is REQUIRE_ORDER if the environment variable + POSIXLY_CORRECT is defined, PERMUTE otherwise. + + REQUIRE_ORDER means don't recognize them as options; + stop option processing when the first non-option is seen. + This is what Unix does. + This mode of operation is selected by either setting the environment + variable POSIXLY_CORRECT, or using `+' as the first character + of the list of option characters. + + PERMUTE is the default. We permute the contents of ARGV as we + scan, so that eventually all the non-options are at the end. + This allows options to be given in any order, even with programs + that were not written to expect this. + + RETURN_IN_ORDER is an option available to programs that were + written to expect options and other ARGV-elements in any order + and that care about the ordering of the two. We describe each + non-option ARGV-element as if it were the argument of an option + with character code 1. Using `-' as the first character of the + list of option characters selects this mode of operation. + + The special argument `--' forces an end of option-scanning regardless + of the value of `ordering'. In the case of RETURN_IN_ORDER, only + `--' can cause `getopt' to return -1 with `optind' != ARGC. */ + + enum + { + REQUIRE_ORDER, PERMUTE, RETURN_IN_ORDER + } __ordering; + + /* If the POSIXLY_CORRECT environment variable is set. */ + int __posixly_correct; + + + /* Handle permutation of arguments. */ + + /* Describe the part of ARGV that contains non-options that have + been skipped. `first_nonopt' is the index in ARGV of the first + of them; `last_nonopt' is the index after the last of them. */ + + int __first_nonopt; + int __last_nonopt; + +#if defined _LIBC && defined USE_NONOPTION_FLAGS + int __nonoption_flags_max_len; + int __nonoption_flags_len; +# endif +}; + +/* The initializer is necessary to set OPTIND and OPTERR to their + default values and to clear the initialization flag. */ +#define _GETOPT_DATA_INITIALIZER { 1, 1 } + +extern int _getopt_internal_r (int ___argc, char *const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + int __long_only, struct _getopt_data *__data); + +extern int _getopt_long_r (int ___argc, char *const *___argv, + const char *__shortopts, + const struct option *__longopts, int *__longind, + struct _getopt_data *__data); + +extern int _getopt_long_only_r (int ___argc, char *const *___argv, + const char *__shortopts, + const struct option *__longopts, + int *__longind, + struct _getopt_data *__data); + +#endif /* getopt_int.h */ diff --git a/contrib/diff/lib/gettext.h b/contrib/diff/lib/gettext.h new file mode 100644 index 0000000..835732e --- /dev/null +++ b/contrib/diff/lib/gettext.h @@ -0,0 +1,68 @@ +/* Convenience header for conditional use of GNU . + Copyright (C) 1995-1998, 2000-2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _LIBGETTEXT_H +#define _LIBGETTEXT_H 1 + +/* NLS can be disabled through the configure --disable-nls option. */ +#if ENABLE_NLS + +/* Get declarations of GNU message catalog functions. */ +# include + +#else + +/* Solaris /usr/include/locale.h includes /usr/include/libintl.h, which + chokes if dcgettext is defined as a macro. So include it now, to make + later inclusions of a NOP. We don't include + as well because people using "gettext.h" will not include , + and also including would fail on SunOS 4, whereas + is OK. */ +#if defined(__sun) +# include +#endif + +/* Disabled NLS. + The casts to 'const char *' serve the purpose of producing warnings + for invalid uses of the value returned from these functions. + On pre-ANSI systems without 'const', the config.h file is supposed to + contain "#define const". */ +# define gettext(Msgid) ((const char *) (Msgid)) +# define dgettext(Domainname, Msgid) ((const char *) (Msgid)) +# define dcgettext(Domainname, Msgid, Category) ((const char *) (Msgid)) +# define ngettext(Msgid1, Msgid2, N) \ + ((N) == 1 ? (const char *) (Msgid1) : (const char *) (Msgid2)) +# define dngettext(Domainname, Msgid1, Msgid2, N) \ + ((N) == 1 ? (const char *) (Msgid1) : (const char *) (Msgid2)) +# define dcngettext(Domainname, Msgid1, Msgid2, N, Category) \ + ((N) == 1 ? (const char *) (Msgid1) : (const char *) (Msgid2)) +# define textdomain(Domainname) ((const char *) (Domainname)) +# define bindtextdomain(Domainname, Dirname) ((const char *) (Dirname)) +# define bind_textdomain_codeset(Domainname, Codeset) ((const char *) (Codeset)) + +#endif + +/* A pseudo function call that serves as a marker for the automated + extraction of messages, but does not call gettext(). The run-time + translation is done at a different place in the code. + The argument, String, should be a literal string. Concatenated strings + and other string expressions won't work. + The macro's expansion is not parenthesized, so that it is suitable as + initializer for static 'char[]' or 'const char[]' variables. */ +#define gettext_noop(String) String + +#endif /* _LIBGETTEXT_H */ diff --git a/contrib/diff/lib/gettimeofday.c b/contrib/diff/lib/gettimeofday.c new file mode 100644 index 0000000..1ebe4d0 --- /dev/null +++ b/contrib/diff/lib/gettimeofday.c @@ -0,0 +1,121 @@ +/* Work around the bug in some systems whereby gettimeofday clobbers the + static buffer that localtime uses for it's return value. The gettimeofday + function from Mac OS X 10.0.4, i.e. Darwin 1.3.7 has this problem. + The tzset replacement is necessary for at least Solaris 2.5, 2.5.1, and 2.6. + Copyright (C) 2001, 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* written by Jim Meyering */ + +#include + +/* Disable the definitions of these functions (from config.h) + so we can use the library versions here. */ +#undef gettimeofday +#undef gmtime +#undef localtime +#undef tzset + +#include + +#if TIME_WITH_SYS_TIME +# include +# include +#else +# if HAVE_SYS_TIME_H +# include +# else +# include +# endif +#endif + +#include + +static struct tm *localtime_buffer_addr; + +/* This is a wrapper for localtime. It is used only on systems for which + gettimeofday clobbers the static buffer used for localtime's result. + + On the first call, record the address of the static buffer that + localtime uses for its result. */ + +struct tm * +rpl_localtime (const time_t *timep) +{ + struct tm *tm = localtime (timep); + + if (! localtime_buffer_addr) + localtime_buffer_addr = tm; + + return tm; +} + +/* Same as above, since gmtime and localtime use the same buffer. */ +struct tm * +rpl_gmtime (const time_t *timep) +{ + struct tm *tm = gmtime (timep); + + if (! localtime_buffer_addr) + localtime_buffer_addr = tm; + + return tm; +} + +/* This is a wrapper for gettimeofday. It is used only on systems for which + gettimeofday clobbers the static buffer used for localtime's result. + + Save and restore the contents of the buffer used for localtime's result + around the call to gettimeofday. */ + +int +rpl_gettimeofday (struct timeval *tv, struct timezone *tz) +{ + struct tm save; + int result; + + if (! localtime_buffer_addr) + { + time_t t = 0; + localtime_buffer_addr = localtime (&t); + } + + save = *localtime_buffer_addr; + result = gettimeofday (tv, tz); + *localtime_buffer_addr = save; + + return result; +} + +/* This is a wrapper for tzset. It is used only on systems for which + tzset may clobber the static buffer used for localtime's result. + Save and restore the contents of the buffer used for localtime's + result around the call to tzset. */ +void +rpl_tzset (void) +{ + struct tm save; + + if (! localtime_buffer_addr) + { + time_t t = 0; + localtime_buffer_addr = localtime (&t); + } + + save = *localtime_buffer_addr; + tzset (); + *localtime_buffer_addr = save; +} diff --git a/contrib/diff/lib/gnulib.mk b/contrib/diff/lib/gnulib.mk new file mode 100644 index 0000000..6e7f05a --- /dev/null +++ b/contrib/diff/lib/gnulib.mk @@ -0,0 +1,90 @@ +# This file is generated automatically by "bootstrap". +BUILT_SOURCES += $(ALLOCA_H) +EXTRA_DIST += alloca_.h + +# We need the following in order to create an when the system +# doesn't have one that works with the given compiler. +all-local $(lib_OBJECTS): $(ALLOCA_H) +alloca.h: alloca_.h + cp $(srcdir)/alloca_.h $@-t + mv $@-t $@ +MOSTLYCLEANFILES += alloca.h alloca.h-t + +lib_SOURCES += c-stack.h c-stack.c + +lib_SOURCES += dirname.h dirname.c basename.c stripslash.c + + +lib_SOURCES += exclude.h exclude.c + +lib_SOURCES += exit.h + +lib_SOURCES += exitfail.h exitfail.c + + +lib_SOURCES += file-type.h file-type.c + +BUILT_SOURCES += $(FNMATCH_H) +EXTRA_DIST += fnmatch_.h fnmatch_loop.c + +# We need the following in order to create an when the system +# doesn't have one that supports the required API. +all-local $(lib_OBJECTS): $(FNMATCH_H) +fnmatch.h: fnmatch_.h + cp $(srcdir)/fnmatch_.h $@-t + mv $@-t $@ +MOSTLYCLEANFILES += fnmatch.h fnmatch.h-t + + +lib_SOURCES += getopt.h getopt.c getopt1.c getopt_int.h + +lib_SOURCES += gettext.h + + +lib_SOURCES += hard-locale.h hard-locale.c + +EXTRA_DIST += inttostr.c +lib_SOURCES += imaxtostr.c inttostr.h offtostr.c umaxtostr.c + + + +lib_SOURCES += posixver.h posixver.c + + +lib_SOURCES += regex.h + + +BUILT_SOURCES += $(STDBOOL_H) +EXTRA_DIST += stdbool_.h + +# We need the following in order to create an when the system +# doesn't have one that works. +all-local $(lib_OBJECTS): $(STDBOOL_H) +stdbool.h: stdbool_.h + sed -e 's/@''HAVE__BOOL''@/$(HAVE__BOOL)/g' < $(srcdir)/stdbool_.h > $@-t + mv $@-t $@ +MOSTLYCLEANFILES += stdbool.h stdbool.h-t + +lib_SOURCES += strcase.h + +lib_SOURCES += strftime.c + + + + + + + +lib_SOURCES += time_r.h + + +lib_SOURCES += unlocked-io.h + +lib_SOURCES += version-etc.h version-etc.c + +lib_SOURCES += xalloc.h xmalloc.c xstrdup.c + +lib_SOURCES += xstrtol.h xstrtol.c xstrtoul.c + +lib_SOURCES += xstrtoumax.c + diff --git a/contrib/diff/lib/hard-locale.c b/contrib/diff/lib/hard-locale.c new file mode 100644 index 0000000..8c0ee8c --- /dev/null +++ b/contrib/diff/lib/hard-locale.c @@ -0,0 +1,74 @@ +/* hard-locale.c -- Determine whether a locale is hard. + + Copyright (C) 1997, 1998, 1999, 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "hard-locale.h" + +#if HAVE_LOCALE_H +# include +#endif + +#include +#include + +/* Return nonzero if the current CATEGORY locale is hard, i.e. if you + can't get away with assuming traditional C or POSIX behavior. */ +int +hard_locale (int category) +{ +#if ! HAVE_SETLOCALE + return 0; +#else + + int hard = 1; + char const *p = setlocale (category, 0); + + if (p) + { +# if defined __GLIBC__ && 2 <= __GLIBC__ + if (strcmp (p, "C") == 0 || strcmp (p, "POSIX") == 0) + hard = 0; +# else + char *locale = malloc (strlen (p) + 1); + if (locale) + { + strcpy (locale, p); + + /* Temporarily set the locale to the "C" and "POSIX" locales + to find their names, so that we can determine whether one + or the other is the caller's locale. */ + if (((p = setlocale (category, "C")) + && strcmp (p, locale) == 0) + || ((p = setlocale (category, "POSIX")) + && strcmp (p, locale) == 0)) + hard = 0; + + /* Restore the caller's locale. */ + setlocale (category, locale); + free (locale); + } +# endif + } + + return hard; + +#endif +} diff --git a/contrib/diff/lib/hard-locale.h b/contrib/diff/lib/hard-locale.h new file mode 100644 index 0000000..ddc15d0 --- /dev/null +++ b/contrib/diff/lib/hard-locale.h @@ -0,0 +1,24 @@ +/* Determine whether a locale is hard. + + Copyright (C) 1999, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef HARD_LOCALE_H_ +# define HARD_LOCALE_H_ 1 + +int hard_locale (int); + +#endif /* HARD_LOCALE_H_ */ diff --git a/contrib/diff/lib/imaxtostr.c b/contrib/diff/lib/imaxtostr.c new file mode 100644 index 0000000..5e87ad5 --- /dev/null +++ b/contrib/diff/lib/imaxtostr.c @@ -0,0 +1,3 @@ +#define inttostr imaxtostr +#define inttype intmax_t +#include "inttostr.c" diff --git a/contrib/diff/lib/inttostr.c b/contrib/diff/lib/inttostr.c new file mode 100644 index 0000000..78a48af --- /dev/null +++ b/contrib/diff/lib/inttostr.c @@ -0,0 +1,49 @@ +/* inttostr.c -- convert integers to printable strings + + Copyright (C) 2001 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert */ + +#include "inttostr.h" + +/* Convert I to a printable string in BUF, which must be at least + INT_BUFSIZE_BOUND (INTTYPE) bytes long. Return the address of the + printable string, which need not start at BUF. */ + +char * +inttostr (inttype i, char *buf) +{ + char *p = buf + INT_STRLEN_BOUND (inttype); + *p = 0; + + if (i < 0) + { + do + *--p = '0' - i % 10; + while ((i /= 10) != 0); + + *--p = '-'; + } + else + { + do + *--p = '0' + i % 10; + while ((i /= 10) != 0); + } + + return p; +} diff --git a/contrib/diff/lib/inttostr.h b/contrib/diff/lib/inttostr.h new file mode 100644 index 0000000..6f2416b --- /dev/null +++ b/contrib/diff/lib/inttostr.h @@ -0,0 +1,47 @@ +/* inttostr.h -- convert integers to printable strings + + Copyright (C) 2001, 2002, 2003, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert */ + +#if HAVE_CONFIG_H +# include +#endif + +#if HAVE_INTTYPES_H +# include +#endif +#if HAVE_STDINT_H +# include +#endif + +#include + +#if HAVE_SYS_TYPES_H +# include +#endif + +/* Upper bound on the string length of an integer converted to string. + 302 / 1000 is ceil (log10 (2.0)). Subtract 1 for the sign bit; + add 1 for integer division truncation; add 1 more for a minus sign. */ +#define INT_STRLEN_BOUND(t) ((sizeof (t) * CHAR_BIT - 1) * 302 / 1000 + 2) + +#define INT_BUFSIZE_BOUND(t) (INT_STRLEN_BOUND (t) + 1) + +char *offtostr (off_t, char *); +char *imaxtostr (intmax_t, char *); +char *umaxtostr (uintmax_t, char *); diff --git a/contrib/diff/lib/malloc.c b/contrib/diff/lib/malloc.c new file mode 100644 index 0000000..a43d169 --- /dev/null +++ b/contrib/diff/lib/malloc.c @@ -0,0 +1,36 @@ +/* Work around bug on some systems where malloc (0) fails. + Copyright (C) 1997, 1998 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* written by Jim Meyering */ + +#if HAVE_CONFIG_H +# include +#endif +#undef malloc + +#include + +/* Allocate an N-byte block of memory from the heap. + If N is zero, allocate a 1-byte block. */ + +void * +rpl_malloc (size_t n) +{ + if (n == 0) + n = 1; + return malloc (n); +} diff --git a/contrib/diff/lib/mkstemp.c b/contrib/diff/lib/mkstemp.c new file mode 100644 index 0000000..f6312b6 --- /dev/null +++ b/contrib/diff/lib/mkstemp.c @@ -0,0 +1,42 @@ +/* Copyright (C) 1998, 1999, 2001 Free Software Foundation, Inc. + This file is derived from the one in the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include + +/* Disable the definition of mkstemp to rpl_mkstemp (from config.h) in this + file. Otherwise, we'd get conflicting prototypes for rpl_mkstemp on + most systems. */ +#undef mkstemp + +#include +#include + +#ifndef __GT_FILE +# define __GT_FILE 0 +#endif + +int __gen_tempname (); + +/* Generate a unique temporary file name from TEMPLATE. + The last six characters of TEMPLATE must be "XXXXXX"; + they are replaced with a string that makes the filename unique. + Then open the file and return a fd. */ +int +rpl_mkstemp (char *template) +{ + return __gen_tempname (template, __GT_FILE); +} diff --git a/contrib/diff/lib/offtostr.c b/contrib/diff/lib/offtostr.c new file mode 100644 index 0000000..45196e2 --- /dev/null +++ b/contrib/diff/lib/offtostr.c @@ -0,0 +1,3 @@ +#define inttostr offtostr +#define inttype off_t +#include "inttostr.c" diff --git a/contrib/diff/lib/posixver.c b/contrib/diff/lib/posixver.c new file mode 100644 index 0000000..754d7ac --- /dev/null +++ b/contrib/diff/lib/posixver.c @@ -0,0 +1,59 @@ +/* Which POSIX version to conform to, for utilities. + + Copyright (C) 2002, 2003, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "posixver.h" + +#include +#include + +#if HAVE_UNISTD_H +# include +#endif +#ifndef _POSIX2_VERSION +# define _POSIX2_VERSION 0 +#endif + +#ifndef DEFAULT_POSIX2_VERSION +# define DEFAULT_POSIX2_VERSION _POSIX2_VERSION +#endif + +/* The POSIX version that utilities should conform to. The default is + specified by the system. */ + +int +posix2_version (void) +{ + long int v = DEFAULT_POSIX2_VERSION; + char const *s = getenv ("_POSIX2_VERSION"); + + if (s && *s) + { + char *e; + long int i = strtol (s, &e, 10); + if (! *e) + v = i; + } + + return v < INT_MIN ? INT_MIN : v < INT_MAX ? v : INT_MAX; +} diff --git a/contrib/diff/lib/posixver.h b/contrib/diff/lib/posixver.h new file mode 100644 index 0000000..b64f6a2 --- /dev/null +++ b/contrib/diff/lib/posixver.h @@ -0,0 +1 @@ +int posix2_version (void); diff --git a/contrib/diff/lib/prepargs.c b/contrib/diff/lib/prepargs.c new file mode 100644 index 0000000..f5735ac --- /dev/null +++ b/contrib/diff/lib/prepargs.c @@ -0,0 +1,93 @@ +/* Parse arguments from a string and prepend them to an argv. + + Copyright (C) 1999, 2000, 2001, 2002 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software + Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA + 02111-1307, USA. */ + +/* Written by Paul Eggert . */ + +#ifdef HAVE_CONFIG_H +# include +#endif +#include "prepargs.h" +#include +#include +#include + +#include + +/* IN_CTYPE_DOMAIN (C) is nonzero if the unsigned char C can safely be given + as an argument to macros like "isspace". */ +#ifdef STDC_HEADERS +# define IN_CTYPE_DOMAIN(c) 1 +#else +# define IN_CTYPE_DOMAIN(c) ((c) <= 0177) +#endif + +#define ISSPACE(c) (IN_CTYPE_DOMAIN (c) && isspace (c)) + +/* Find the white-space-separated options specified by OPTIONS, and + using BUF to store copies of these options, set ARGV[0], ARGV[1], + etc. to the option copies. Return the number N of options found. + Do not set ARGV[N]. If ARGV is zero, do not store ARGV[0] etc. + Backslash can be used to escape whitespace (and backslashes). */ +static int +prepend_args (char const *options, char *buf, char **argv) +{ + char const *o = options; + char *b = buf; + int n = 0; + + for (;;) + { + while (ISSPACE ((unsigned char) *o)) + o++; + if (!*o) + return n; + if (argv) + argv[n] = b; + n++; + + do + if ((*b++ = *o++) == '\\' && *o) + b[-1] = *o++; + while (*o && ! ISSPACE ((unsigned char) *o)); + + *b++ = '\0'; + } +} + +/* Prepend the whitespace-separated options in OPTIONS to the argument + vector of a main program with argument count *PARGC and argument + vector *PARGV. */ +void +prepend_default_options (char const *options, int *pargc, char ***pargv) +{ + if (options) + { + char *buf = xmalloc (strlen (options) + 1); + int prepended = prepend_args (options, buf, (char **) 0); + int argc = *pargc; + char * const *argv = *pargv; + char **pp = (char **) xmalloc ((prepended + argc + 1) * sizeof *pp); + *pargc = prepended + argc; + *pargv = pp; + *pp++ = *argv++; + pp += prepend_args (options, buf, pp); + while ((*pp++ = *argv++)) + continue; + } +} diff --git a/contrib/diff/lib/prepargs.h b/contrib/diff/lib/prepargs.h new file mode 100644 index 0000000..ce93ea8 --- /dev/null +++ b/contrib/diff/lib/prepargs.h @@ -0,0 +1,3 @@ +/* Parse arguments from a string and prepend them to an argv. */ + +void prepend_default_options (char const *, int *, char ***); diff --git a/contrib/diff/lib/quotesys.c b/contrib/diff/lib/quotesys.c new file mode 100644 index 0000000..528f382 --- /dev/null +++ b/contrib/diff/lib/quotesys.c @@ -0,0 +1,125 @@ +/* Shell command argument quoting. + Copyright (C) 1994, 1995, 1997 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert */ + +#if HAVE_CONFIG_H +# include +#endif + +#include +#include + +/* Place into QUOTED a quoted version of ARG suitable for `system'. + Return the length of the resulting string (which is not null-terminated). + If QUOTED is null, return the length without any side effects. */ + +size_t +quote_system_arg (quoted, arg) + char *quoted; + char const *arg; +{ + char const *a; + size_t len = 0; + + /* Scan ARG, copying it to QUOTED if QUOTED is not null, + looking for shell metacharacters. */ + + for (a = arg; ; a++) + { + char c = *a; + switch (c) + { + case 0: + /* ARG has no shell metacharacters. */ + return len; + + case '=': + if (*arg == '-') + break; + /* Fall through. */ + case '\t': case '\n': case ' ': + case '!': case '"': case '#': case '$': case '%': case '&': case '\'': + case '(': case ')': case '*': case ';': + case '<': case '>': case '?': case '[': case '\\': + case '^': case '`': case '|': case '~': + { + /* ARG has a shell metacharacter. + Start over, quoting it this time. */ + + len = 0; + c = *arg++; + + /* If ARG is an option, quote just its argument. + This is not necessary, but it looks nicer. */ + if (c == '-' && arg < a) + { + c = *arg++; + + if (quoted) + { + quoted[len] = '-'; + quoted[len + 1] = c; + } + len += 2; + + if (c == '-') + while (arg < a) + { + c = *arg++; + if (quoted) + quoted[len] = c; + len++; + if (c == '=') + break; + } + c = *arg++; + } + + if (quoted) + quoted[len] = '\''; + len++; + + for (; c; c = *arg++) + { + if (c == '\'') + { + if (quoted) + { + quoted[len] = '\''; + quoted[len + 1] = '\\'; + quoted[len + 2] = '\''; + } + len += 3; + } + if (quoted) + quoted[len] = c; + len++; + } + + if (quoted) + quoted[len] = '\''; + return len + 1; + } + } + + if (quoted) + quoted[len] = c; + len++; + } +} diff --git a/contrib/diff/lib/quotesys.h b/contrib/diff/lib/quotesys.h new file mode 100644 index 0000000..0d0d825 --- /dev/null +++ b/contrib/diff/lib/quotesys.h @@ -0,0 +1,9 @@ +/* quotesys.h -- declarations for quoting system arguments */ + +#if defined __STDC__ || __GNUC__ +# define __QUOTESYS_P(args) args +#else +# define __QUOTESYS_P(args) () +#endif + +size_t quote_system_arg __QUOTESYS_P ((char *, char const *)); diff --git a/contrib/diff/lib/realloc.c b/contrib/diff/lib/realloc.c new file mode 100644 index 0000000..ccbf991 --- /dev/null +++ b/contrib/diff/lib/realloc.c @@ -0,0 +1,39 @@ +/* Work around bug on some systems where realloc (NULL, 0) fails. + Copyright (C) 1997, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* written by Jim Meyering */ + +#if HAVE_CONFIG_H +# include +#endif +#undef realloc + +#include + +/* Change the size of an allocated block of memory P to N bytes, + with error checking. If N is zero, change it to 1. If P is NULL, + use malloc. */ + +void * +rpl_realloc (void *p, size_t n) +{ + if (n == 0) + n = 1; + if (p == 0) + return malloc (n); + return realloc (p, n); +} diff --git a/contrib/diff/lib/regex.c b/contrib/diff/lib/regex.c new file mode 100644 index 0000000..ac70b90 --- /dev/null +++ b/contrib/diff/lib/regex.c @@ -0,0 +1,8298 @@ +/* Extended regular expression matching and search library, + version 0.12. + (Implements POSIX draft P1003.2/D11.2, except for some of the + internationalization features.) + + Copyright (C) 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, + 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* AIX requires this to be the first thing in the file. */ +#if defined _AIX && !defined REGEX_MALLOC + #pragma alloca +#endif + +#undef _GNU_SOURCE +#define _GNU_SOURCE + +#ifdef HAVE_CONFIG_H +# include +#endif + +#ifndef INSIDE_RECURSION + +# include + +# define WIDE_CHAR_SUPPORT (HAVE_WCTYPE_H && HAVE_WCHAR_H && HAVE_BTOWC) + +/* For platform which support the ISO C amendement 1 functionality we + support user defined character classes. */ +# if defined _LIBC || WIDE_CHAR_SUPPORT +/* Solaris 2.5 has a bug: must be included before . */ +# include +# include +# endif + +# ifdef _LIBC +/* We have to keep the namespace clean. */ +# define regfree(preg) __regfree (preg) +# define regexec(pr, st, nm, pm, ef) __regexec (pr, st, nm, pm, ef) +# define regcomp(preg, pattern, cflags) __regcomp (preg, pattern, cflags) +# define regerror(errcode, preg, errbuf, errbuf_size) \ + __regerror(errcode, preg, errbuf, errbuf_size) +# define re_set_registers(bu, re, nu, st, en) \ + __re_set_registers (bu, re, nu, st, en) +# define re_match_2(bufp, string1, size1, string2, size2, pos, regs, stop) \ + __re_match_2 (bufp, string1, size1, string2, size2, pos, regs, stop) +# define re_match(bufp, string, size, pos, regs) \ + __re_match (bufp, string, size, pos, regs) +# define re_search(bufp, string, size, startpos, range, regs) \ + __re_search (bufp, string, size, startpos, range, regs) +# define re_compile_pattern(pattern, length, bufp) \ + __re_compile_pattern (pattern, length, bufp) +# define re_set_syntax(syntax) __re_set_syntax (syntax) +# define re_search_2(bufp, st1, s1, st2, s2, startpos, range, regs, stop) \ + __re_search_2 (bufp, st1, s1, st2, s2, startpos, range, regs, stop) +# define re_compile_fastmap(bufp) __re_compile_fastmap (bufp) + +# define btowc __btowc +# define iswctype __iswctype +# define mbrtowc __mbrtowc +# define wcslen __wcslen +# define wcscoll __wcscoll +# define wcrtomb __wcrtomb + +/* We are also using some library internals. */ +# include +# include +# include +# include +# endif + +# ifdef _LIBC +# include +# undef gettext +# define gettext(msgid) __dcgettext ("libc", msgid, LC_MESSAGES) + /* This define is so xgettext can find the internationalizable strings. */ +# define gettext_noop(msgid) msgid +# else +/* This is for other GNU distributions with internationalized messages. */ +# include "gettext.h" +# endif + +/* Support for bounded pointers. */ +# if !defined _LIBC && !defined __BOUNDED_POINTERS__ +# define __bounded /* nothing */ +# define __unbounded /* nothing */ +# define __ptrvalue /* nothing */ +# endif + +/* The `emacs' switch turns on certain matching commands + that make sense only in Emacs. */ +# ifdef emacs + +# include "lisp.h" +# include "buffer.h" +# include "syntax.h" + +# else /* not emacs */ + +/* If we are not linking with Emacs proper, + we can't use the relocating allocator + even if config.h says that we can. */ +# undef REL_ALLOC + +# include + +/* When used in Emacs's lib-src, we need to get bzero and bcopy somehow. + If nothing else has been done, use the method below. */ +# ifdef INHIBIT_STRING_HEADER +# if !(defined HAVE_BZERO && defined HAVE_BCOPY) +# if !defined bzero && !defined bcopy +# undef INHIBIT_STRING_HEADER +# endif +# endif +# endif + +/* This is the normal way of making sure we have a bcopy and a bzero. + This is used in most programs--a few other programs avoid this + by defining INHIBIT_STRING_HEADER. */ +# ifndef INHIBIT_STRING_HEADER +# include +# ifndef bzero +# ifndef _LIBC +# define bzero(s, n) (memset (s, '\0', n), (s)) +# else +# define bzero(s, n) __bzero (s, n) +# endif +# endif +# endif + +/* Define the syntax stuff for \<, \>, etc. */ + +/* This must be nonzero for the wordchar and notwordchar pattern + commands in re_match_2. */ +# ifndef Sword +# define Sword 1 +# endif + +# ifdef SWITCH_ENUM_BUG +# define SWITCH_ENUM_CAST(x) ((int)(x)) +# else +# define SWITCH_ENUM_CAST(x) (x) +# endif + +# endif /* not emacs */ + +# include + +# ifndef MB_LEN_MAX +# define MB_LEN_MAX 1 +# endif + +/* Get the interface, including the syntax bits. */ +# include + +/* isalpha etc. are used for the character classes. */ +# include + +/* Jim Meyering writes: + + "... Some ctype macros are valid only for character codes that + isascii says are ASCII (SGI's IRIX-4.0.5 is one such system --when + using /bin/cc or gcc but without giving an ansi option). So, all + ctype uses should be through macros like ISPRINT... If + STDC_HEADERS is defined, then autoconf has verified that the ctype + macros don't need to be guarded with references to isascii. ... + Defining isascii to 1 should let any compiler worth its salt + eliminate the && through constant folding." + Solaris defines some of these symbols so we must undefine them first. */ + +# if defined STDC_HEADERS || (!defined isascii && !defined HAVE_ISASCII) +# define IN_CTYPE_DOMAIN(c) 1 +# else +# define IN_CTYPE_DOMAIN(c) isascii(c) +# endif + +# ifdef isblank +# define ISBLANK(c) (IN_CTYPE_DOMAIN (c) && isblank (c)) +# else +# define ISBLANK(c) ((c) == ' ' || (c) == '\t') +# endif +# ifdef isgraph +# define ISGRAPH(c) (IN_CTYPE_DOMAIN (c) && isgraph (c)) +# else +# define ISGRAPH(c) (IN_CTYPE_DOMAIN (c) && isprint (c) && !isspace (c)) +# endif + +# undef ISPRINT +# define ISPRINT(c) (IN_CTYPE_DOMAIN (c) && isprint (c)) +# define ISDIGIT(c) (IN_CTYPE_DOMAIN (c) && isdigit (c)) +# define ISALNUM(c) (IN_CTYPE_DOMAIN (c) && isalnum (c)) +# define ISALPHA(c) (IN_CTYPE_DOMAIN (c) && isalpha (c)) +# define ISCNTRL(c) (IN_CTYPE_DOMAIN (c) && iscntrl (c)) +# define ISLOWER(c) (IN_CTYPE_DOMAIN (c) && islower (c)) +# define ISPUNCT(c) (IN_CTYPE_DOMAIN (c) && ispunct (c)) +# define ISSPACE(c) (IN_CTYPE_DOMAIN (c) && isspace (c)) +# define ISUPPER(c) (IN_CTYPE_DOMAIN (c) && isupper (c)) +# define ISXDIGIT(c) (IN_CTYPE_DOMAIN (c) && isxdigit (c)) + +# ifdef _tolower +# define TOLOWER(c) _tolower(c) +# else +# define TOLOWER(c) tolower(c) +# endif + +# ifndef emacs +/* How many characters in the character set. */ +# define CHAR_SET_SIZE 256 + +# ifdef SYNTAX_TABLE + +extern char *re_syntax_table; + +# else /* not SYNTAX_TABLE */ + +static char re_syntax_table[CHAR_SET_SIZE]; + +static void +init_syntax_once (void) +{ + register int c; + static int done = 0; + + if (done) + return; + bzero (re_syntax_table, sizeof re_syntax_table); + + for (c = 0; c < CHAR_SET_SIZE; ++c) + if (ISALNUM (c)) + re_syntax_table[c] = Sword; + + re_syntax_table['_'] = Sword; + + done = 1; +} + +# endif /* not SYNTAX_TABLE */ + +# define SYNTAX(c) re_syntax_table[(unsigned char) (c)] + +# endif /* emacs */ + +/* Should we use malloc or alloca? If REGEX_MALLOC is not defined, we + use `alloca' instead of `malloc'. This is because using malloc in + re_search* or re_match* could cause memory leaks when C-g is used in + Emacs; also, malloc is slower and causes storage fragmentation. On + the other hand, malloc is more portable, and easier to debug. + + Because we sometimes use alloca, some routines have to be macros, + not functions -- `alloca'-allocated space disappears at the end of the + function it is called in. */ + +# ifdef REGEX_MALLOC + +# define REGEX_ALLOCATE malloc +# define REGEX_REALLOCATE(source, osize, nsize) realloc (source, nsize) +# define REGEX_FREE free + +# else /* not REGEX_MALLOC */ + +/* Emacs already defines alloca, sometimes. */ +# ifndef alloca + +/* Make alloca work the best possible way. */ +# ifdef __GNUC__ +# define alloca __builtin_alloca +# else /* not __GNUC__ */ +# if HAVE_ALLOCA_H +# include +# endif /* HAVE_ALLOCA_H */ +# endif /* not __GNUC__ */ + +# endif /* not alloca */ + +# define REGEX_ALLOCATE alloca + +/* Assumes a `char *destination' variable. */ +# define REGEX_REALLOCATE(source, osize, nsize) \ + (destination = (char *) alloca (nsize), \ + memcpy (destination, source, osize)) + +/* No need to do anything to free, after alloca. */ +# define REGEX_FREE(arg) ((void)0) /* Do nothing! But inhibit gcc warning. */ + +# endif /* not REGEX_MALLOC */ + +/* Define how to allocate the failure stack. */ + +# if defined REL_ALLOC && defined REGEX_MALLOC + +# define REGEX_ALLOCATE_STACK(size) \ + r_alloc (&failure_stack_ptr, (size)) +# define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + r_re_alloc (&failure_stack_ptr, (nsize)) +# define REGEX_FREE_STACK(ptr) \ + r_alloc_free (&failure_stack_ptr) + +# else /* not using relocating allocator */ + +# ifdef REGEX_MALLOC + +# define REGEX_ALLOCATE_STACK malloc +# define REGEX_REALLOCATE_STACK(source, osize, nsize) realloc (source, nsize) +# define REGEX_FREE_STACK free + +# else /* not REGEX_MALLOC */ + +# define REGEX_ALLOCATE_STACK alloca + +# define REGEX_REALLOCATE_STACK(source, osize, nsize) \ + REGEX_REALLOCATE (source, osize, nsize) +/* No need to explicitly free anything. */ +# define REGEX_FREE_STACK(arg) + +# endif /* not REGEX_MALLOC */ +# endif /* not using relocating allocator */ + + +/* True if `size1' is non-NULL and PTR is pointing anywhere inside + `string1' or just past its end. This works if PTR is NULL, which is + a good thing. */ +# define FIRST_STRING_P(ptr) \ + (size1 && string1 <= (ptr) && (ptr) <= string1 + size1) + +/* (Re)Allocate N items of type T using malloc, or fail. */ +# define TALLOC(n, t) ((t *) malloc ((n) * sizeof (t))) +# define RETALLOC(addr, n, t) ((addr) = (t *) realloc (addr, (n) * sizeof (t))) +# define RETALLOC_IF(addr, n, t) \ + if (addr) RETALLOC((addr), (n), t); else (addr) = TALLOC ((n), t) +# define REGEX_TALLOC(n, t) ((t *) REGEX_ALLOCATE ((n) * sizeof (t))) + +# define BYTEWIDTH 8 /* In bits. */ + +# define STREQ(s1, s2) ((strcmp (s1, s2) == 0)) + +# undef MAX +# undef MIN +# define MAX(a, b) ((a) > (b) ? (a) : (b)) +# define MIN(a, b) ((a) < (b) ? (a) : (b)) + +typedef char boolean; +# define false 0 +# define true 1 + +static reg_errcode_t byte_regex_compile (const char *pattern, size_t size, + reg_syntax_t syntax, + struct re_pattern_buffer *bufp); + +static int byte_re_match_2_internal (struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int pos, + struct re_registers *regs, + int stop); +static int byte_re_search_2 (struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int startpos, int range, + struct re_registers *regs, int stop); +static int byte_re_compile_fastmap (struct re_pattern_buffer *bufp); + +#ifdef MBS_SUPPORT +static reg_errcode_t wcs_regex_compile (const char *pattern, size_t size, + reg_syntax_t syntax, + struct re_pattern_buffer *bufp); + + +static int wcs_re_match_2_internal (struct re_pattern_buffer *bufp, + const char *cstring1, int csize1, + const char *cstring2, int csize2, + int pos, + struct re_registers *regs, + int stop, + wchar_t *string1, int size1, + wchar_t *string2, int size2, + int *mbs_offset1, int *mbs_offset2); +static int wcs_re_search_2 (struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int startpos, int range, + struct re_registers *regs, int stop); +static int wcs_re_compile_fastmap (struct re_pattern_buffer *bufp); +#endif + +/* These are the command codes that appear in compiled regular + expressions. Some opcodes are followed by argument bytes. A + command code can specify any interpretation whatsoever for its + arguments. Zero bytes may appear in the compiled regular expression. */ + +typedef enum +{ + no_op = 0, + + /* Succeed right away--no more backtracking. */ + succeed, + + /* Followed by one byte giving n, then by n literal bytes. */ + exactn, + +# ifdef MBS_SUPPORT + /* Same as exactn, but contains binary data. */ + exactn_bin, +# endif + + /* Matches any (more or less) character. */ + anychar, + + /* Matches any one char belonging to specified set. First + following byte is number of bitmap bytes. Then come bytes + for a bitmap saying which chars are in. Bits in each byte + are ordered low-bit-first. A character is in the set if its + bit is 1. A character too large to have a bit in the map is + automatically not in the set. */ + /* ifdef MBS_SUPPORT, following element is length of character + classes, length of collating symbols, length of equivalence + classes, length of character ranges, and length of characters. + Next, character class element, collating symbols elements, + equivalence class elements, range elements, and character + elements follow. + See regex_compile function. */ + charset, + + /* Same parameters as charset, but match any character that is + not one of those specified. */ + charset_not, + + /* Start remembering the text that is matched, for storing in a + register. Followed by one byte with the register number, in + the range 0 to one less than the pattern buffer's re_nsub + field. Then followed by one byte with the number of groups + inner to this one. (This last has to be part of the + start_memory only because we need it in the on_failure_jump + of re_match_2.) */ + start_memory, + + /* Stop remembering the text that is matched and store it in a + memory register. Followed by one byte with the register + number, in the range 0 to one less than `re_nsub' in the + pattern buffer, and one byte with the number of inner groups, + just like `start_memory'. (We need the number of inner + groups here because we don't have any easy way of finding the + corresponding start_memory when we're at a stop_memory.) */ + stop_memory, + + /* Match a duplicate of something remembered. Followed by one + byte containing the register number. */ + duplicate, + + /* Fail unless at beginning of line. */ + begline, + + /* Fail unless at end of line. */ + endline, + + /* Succeeds if at beginning of buffer (if emacs) or at beginning + of string to be matched (if not). */ + begbuf, + + /* Analogously, for end of buffer/string. */ + endbuf, + + /* Followed by two byte relative address to which to jump. */ + jump, + + /* Same as jump, but marks the end of an alternative. */ + jump_past_alt, + + /* Followed by two-byte relative address of place to resume at + in case of failure. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + on_failure_jump, + + /* Like on_failure_jump, but pushes a placeholder instead of the + current string position when executed. */ + on_failure_keep_string_jump, + + /* Throw away latest failure point and then jump to following + two-byte relative address. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + pop_failure_jump, + + /* Change to pop_failure_jump if know won't have to backtrack to + match; otherwise change to jump. This is used to jump + back to the beginning of a repeat. If what follows this jump + clearly won't match what the repeat does, such that we can be + sure that there is no use backtracking out of repetitions + already matched, then we change it to a pop_failure_jump. + Followed by two-byte address. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + maybe_pop_jump, + + /* Jump to following two-byte address, and push a dummy failure + point. This failure point will be thrown away if an attempt + is made to use it for a failure. A `+' construct makes this + before the first repeat. Also used as an intermediary kind + of jump when compiling an alternative. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + dummy_failure_jump, + + /* Push a dummy failure point and continue. Used at the end of + alternatives. */ + push_dummy_failure, + + /* Followed by two-byte relative address and two-byte number n. + After matching N times, jump to the address upon failure. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + succeed_n, + + /* Followed by two-byte relative address, and two-byte number n. + Jump to the address N times, then fail. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + jump_n, + + /* Set the following two-byte relative address to the + subsequent two-byte number. The address *includes* the two + bytes of number. */ + /* ifdef MBS_SUPPORT, the size of address is 1. */ + set_number_at, + + wordchar, /* Matches any word-constituent character. */ + notwordchar, /* Matches any char that is not a word-constituent. */ + + wordbeg, /* Succeeds if at word beginning. */ + wordend, /* Succeeds if at word end. */ + + wordbound, /* Succeeds if at a word boundary. */ + notwordbound /* Succeeds if not at a word boundary. */ + +# ifdef emacs + ,before_dot, /* Succeeds if before point. */ + at_dot, /* Succeeds if at point. */ + after_dot, /* Succeeds if after point. */ + + /* Matches any character whose syntax is specified. Followed by + a byte which contains a syntax code, e.g., Sword. */ + syntaxspec, + + /* Matches any character whose syntax is not that specified. */ + notsyntaxspec +# endif /* emacs */ +} re_opcode_t; +#endif /* not INSIDE_RECURSION */ + + +#ifdef BYTE +# define CHAR_T char +# define UCHAR_T unsigned char +# define COMPILED_BUFFER_VAR bufp->buffer +# define OFFSET_ADDRESS_SIZE 2 +# define PREFIX(name) byte_##name +# define ARG_PREFIX(name) name +# define PUT_CHAR(c) putchar (c) +#else +# ifdef WCHAR +# define CHAR_T wchar_t +# define UCHAR_T wchar_t +# define COMPILED_BUFFER_VAR wc_buffer +# define OFFSET_ADDRESS_SIZE 1 /* the size which STORE_NUMBER macro use */ +# define CHAR_CLASS_SIZE ((__alignof__(wctype_t)+sizeof(wctype_t))/sizeof(CHAR_T)+1) +# define PREFIX(name) wcs_##name +# define ARG_PREFIX(name) c##name +/* Should we use wide stream?? */ +# define PUT_CHAR(c) printf ("%C", c); +# define TRUE 1 +# define FALSE 0 +# else +# ifdef MBS_SUPPORT +# define WCHAR +# define INSIDE_RECURSION +# include "regex.c" +# undef INSIDE_RECURSION +# endif +# define BYTE +# define INSIDE_RECURSION +# include "regex.c" +# undef INSIDE_RECURSION +# endif +#endif +#include "unlocked-io.h" + +#ifdef INSIDE_RECURSION +/* Common operations on the compiled pattern. */ + +/* Store NUMBER in two contiguous bytes starting at DESTINATION. */ +/* ifdef MBS_SUPPORT, we store NUMBER in 1 element. */ + +# ifdef WCHAR +# define STORE_NUMBER(destination, number) \ + do { \ + *(destination) = (UCHAR_T)(number); \ + } while (0) +# else /* BYTE */ +# define STORE_NUMBER(destination, number) \ + do { \ + (destination)[0] = (number) & 0377; \ + (destination)[1] = (number) >> 8; \ + } while (0) +# endif /* WCHAR */ + +/* Same as STORE_NUMBER, except increment DESTINATION to + the byte after where the number is stored. Therefore, DESTINATION + must be an lvalue. */ +/* ifdef MBS_SUPPORT, we store NUMBER in 1 element. */ + +# define STORE_NUMBER_AND_INCR(destination, number) \ + do { \ + STORE_NUMBER (destination, number); \ + (destination) += OFFSET_ADDRESS_SIZE; \ + } while (0) + +/* Put into DESTINATION a number stored in two contiguous bytes starting + at SOURCE. */ +/* ifdef MBS_SUPPORT, we store NUMBER in 1 element. */ + +# ifdef WCHAR +# define EXTRACT_NUMBER(destination, source) \ + do { \ + (destination) = *(source); \ + } while (0) +# else /* BYTE */ +# define EXTRACT_NUMBER(destination, source) \ + do { \ + (destination) = *(source) & 0377; \ + (destination) += (signed char) (*((source) + 1)) << 8; \ + } while (0) +# endif + +# ifdef DEBUG +static void +PREFIX(extract_number) (int *dest, UCHAR_T *source) +{ +# ifdef WCHAR + *dest = *source; +# else /* BYTE */ + signed char temp = source[1]; + *dest = *source & 0377; + *dest += temp << 8; +# endif +} + +# ifndef EXTRACT_MACROS /* To debug the macros. */ +# undef EXTRACT_NUMBER +# define EXTRACT_NUMBER(dest, src) PREFIX(extract_number) (&dest, src) +# endif /* not EXTRACT_MACROS */ + +# endif /* DEBUG */ + +/* Same as EXTRACT_NUMBER, except increment SOURCE to after the number. + SOURCE must be an lvalue. */ + +# define EXTRACT_NUMBER_AND_INCR(destination, source) \ + do { \ + EXTRACT_NUMBER (destination, source); \ + (source) += OFFSET_ADDRESS_SIZE; \ + } while (0) + +# ifdef DEBUG +static void +PREFIX(extract_number_and_incr) (int *destination, UCHAR_T **source) +{ + PREFIX(extract_number) (destination, *source); + *source += OFFSET_ADDRESS_SIZE; +} + +# ifndef EXTRACT_MACROS +# undef EXTRACT_NUMBER_AND_INCR +# define EXTRACT_NUMBER_AND_INCR(dest, src) \ + PREFIX(extract_number_and_incr) (&dest, &src) +# endif /* not EXTRACT_MACROS */ + +# endif /* DEBUG */ + + + +/* If DEBUG is defined, Regex prints many voluminous messages about what + it is doing (if the variable `debug' is nonzero). If linked with the + main program in `iregex.c', you can enter patterns and strings + interactively. And if linked with the main program in `main.c' and + the other test files, you can run the already-written tests. */ + +# ifdef DEBUG + +# ifndef DEFINED_ONCE + +/* We use standard I/O for debugging. */ +# include + +/* It is useful to test things that ``must'' be true when debugging. */ +# include + +static int debug; + +# define DEBUG_STATEMENT(e) e +# define DEBUG_PRINT1(x) if (debug) printf (x) +# define DEBUG_PRINT2(x1, x2) if (debug) printf (x1, x2) +# define DEBUG_PRINT3(x1, x2, x3) if (debug) printf (x1, x2, x3) +# define DEBUG_PRINT4(x1, x2, x3, x4) if (debug) printf (x1, x2, x3, x4) +# endif /* not DEFINED_ONCE */ + +# define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) \ + if (debug) PREFIX(print_partial_compiled_pattern) (s, e) +# define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) \ + if (debug) PREFIX(print_double_string) (w, s1, sz1, s2, sz2) + + +/* Print the fastmap in human-readable form. */ + +# ifndef DEFINED_ONCE +void +print_fastmap (char *fastmap) +{ + unsigned was_a_range = 0; + unsigned i = 0; + + while (i < (1 << BYTEWIDTH)) + { + if (fastmap[i++]) + { + was_a_range = 0; + putchar (i - 1); + while (i < (1 << BYTEWIDTH) && fastmap[i]) + { + was_a_range = 1; + i++; + } + if (was_a_range) + { + printf ("-"); + putchar (i - 1); + } + } + } + putchar ('\n'); +} +# endif /* not DEFINED_ONCE */ + + +/* Print a compiled pattern string in human-readable form, starting at + the START pointer into it and ending just before the pointer END. */ + +void +PREFIX(print_partial_compiled_pattern) (UCHAR_T *start, UCHAR_T *end) +{ + int mcnt, mcnt2; + UCHAR_T *p1; + UCHAR_T *p = start; + UCHAR_T *pend = end; + + if (start == NULL) + { + printf ("(null)\n"); + return; + } + + /* Loop over pattern commands. */ + while (p < pend) + { +# ifdef _LIBC + printf ("%td:\t", p - start); +# else + printf ("%ld:\t", (long int) (p - start)); +# endif + + switch ((re_opcode_t) *p++) + { + case no_op: + printf ("/no_op"); + break; + + case exactn: + mcnt = *p++; + printf ("/exactn/%d", mcnt); + do + { + putchar ('/'); + PUT_CHAR (*p++); + } + while (--mcnt); + break; + +# ifdef MBS_SUPPORT + case exactn_bin: + mcnt = *p++; + printf ("/exactn_bin/%d", mcnt); + do + { + printf("/%lx", (long int) *p++); + } + while (--mcnt); + break; +# endif /* MBS_SUPPORT */ + + case start_memory: + mcnt = *p++; + printf ("/start_memory/%d/%ld", mcnt, (long int) *p++); + break; + + case stop_memory: + mcnt = *p++; + printf ("/stop_memory/%d/%ld", mcnt, (long int) *p++); + break; + + case duplicate: + printf ("/duplicate/%ld", (long int) *p++); + break; + + case anychar: + printf ("/anychar"); + break; + + case charset: + case charset_not: + { +# ifdef WCHAR + int i, length; + wchar_t *workp = p; + printf ("/charset [%s", + (re_opcode_t) *(workp - 1) == charset_not ? "^" : ""); + p += 5; + length = *workp++; /* the length of char_classes */ + for (i=0 ; ibuffer; + + PREFIX(print_partial_compiled_pattern) (buffer, buffer + + bufp->used / sizeof(UCHAR_T)); + printf ("%ld bytes used/%ld bytes allocated.\n", + bufp->used, bufp->allocated); + + if (bufp->fastmap_accurate && bufp->fastmap) + { + printf ("fastmap: "); + print_fastmap (bufp->fastmap); + } + +# ifdef _LIBC + printf ("re_nsub: %Zd\t", bufp->re_nsub); +# else + printf ("re_nsub: %ld\t", (long int) bufp->re_nsub); +# endif + printf ("regs_alloc: %d\t", bufp->regs_allocated); + printf ("can_be_null: %d\t", bufp->can_be_null); + printf ("newline_anchor: %d\n", bufp->newline_anchor); + printf ("no_sub: %d\t", bufp->no_sub); + printf ("not_bol: %d\t", bufp->not_bol); + printf ("not_eol: %d\t", bufp->not_eol); + printf ("syntax: %lx\n", bufp->syntax); + /* Perhaps we should print the translate table? */ +} + + +void +PREFIX(print_double_string) (const CHAR_T *where, + const CHAR_T *string1, + const CHAR_T *string2, + int size1, + int size2) +{ + int this_char; + + if (where == NULL) + printf ("(null)"); + else + { + int cnt; + + if (FIRST_STRING_P (where)) + { + for (this_char = where - string1; this_char < size1; this_char++) + PUT_CHAR (string1[this_char]); + + where = string2; + } + + cnt = 0; + for (this_char = where - string2; this_char < size2; this_char++) + { + PUT_CHAR (string2[this_char]); + if (++cnt > 100) + { + fputs ("...", stdout); + break; + } + } + } +} + +# ifndef DEFINED_ONCE +void +printchar (c) + int c; +{ + putc (c, stderr); +} +# endif + +# else /* not DEBUG */ + +# ifndef DEFINED_ONCE +# undef assert +# define assert(e) + +# define DEBUG_STATEMENT(e) +# define DEBUG_PRINT1(x) +# define DEBUG_PRINT2(x1, x2) +# define DEBUG_PRINT3(x1, x2, x3) +# define DEBUG_PRINT4(x1, x2, x3, x4) +# endif /* not DEFINED_ONCE */ +# define DEBUG_PRINT_COMPILED_PATTERN(p, s, e) +# define DEBUG_PRINT_DOUBLE_STRING(w, s1, sz1, s2, sz2) + +# endif /* not DEBUG */ + + + +# ifdef WCHAR +/* This convert a multibyte string to a wide character string. + And write their correspondances to offset_buffer(see below) + and write whether each wchar_t is binary data to is_binary. + This assume invalid multibyte sequences as binary data. + We assume offset_buffer and is_binary is already allocated + enough space. */ + +static size_t +convert_mbs_to_wcs (CHAR_T *dest, + const unsigned char* src, + + /* The length of multibyte string. */ + size_t len, + + /* Correspondences between src(char string) and + dest(wchar_t string) for optimization. E.g.: + src = "xxxyzz" + dest = {'X', 'Y', 'Z'} + (each "xxx", "y" and "zz" represent one + multibyte character corresponding to 'X', + 'Y' and 'Z'.) + offset_buffer = {0, 0+3("xxx"), 0+3+1("y"), + 0+3+1+2("zz")} + = {0, 3, 4, 6} */ + int *offset_buffer, + + char *is_binary) +{ + wchar_t *pdest = dest; + const unsigned char *psrc = src; + size_t wc_count = 0; + + mbstate_t mbs; + int i, consumed; + size_t mb_remain = len; + size_t mb_count = 0; + + /* Initialize the conversion state. */ + memset (&mbs, 0, sizeof (mbstate_t)); + + offset_buffer[0] = 0; + for( ; mb_remain > 0 ; ++wc_count, ++pdest, mb_remain -= consumed, + psrc += consumed) + { + consumed = mbrtowc (pdest, psrc, mb_remain, &mbs); + + if (consumed <= 0) + /* failed to convert. maybe src contains binary data. + So we consume 1 byte manualy. */ + { + *pdest = *psrc; + consumed = 1; + is_binary[wc_count] = TRUE; + } + else + is_binary[wc_count] = FALSE; + /* In sjis encoding, we use yen sign as escape character in + place of reverse solidus. So we convert 0x5c(yen sign in + sjis) to not 0xa5(yen sign in UCS2) but 0x5c(reverse + solidus in UCS2). */ + if (consumed == 1 && (int) *psrc == 0x5c && (int) *pdest == 0xa5) + *pdest = (wchar_t) *psrc; + + offset_buffer[wc_count + 1] = mb_count += consumed; + } + + /* Fill remain of the buffer with sentinel. */ + for (i = wc_count + 1 ; i <= len ; i++) + offset_buffer[i] = mb_count + 1; + + return wc_count; +} + +# endif /* WCHAR */ + +#else /* not INSIDE_RECURSION */ + +/* Set by `re_set_syntax' to the current regexp syntax to recognize. Can + also be assigned to arbitrarily: each pattern buffer stores its own + syntax, so it can be changed between regex compilations. */ +/* This has no initializer because initialized variables in Emacs + become read-only after dumping. */ +reg_syntax_t re_syntax_options; + + +/* Specify the precise syntax of regexps for compilation. This provides + for compatibility for various utilities which historically have + different, incompatible syntaxes. + + The argument SYNTAX is a bit mask comprised of the various bits + defined in regex.h. We return the old syntax. */ + +reg_syntax_t +re_set_syntax (reg_syntax_t syntax) +{ + reg_syntax_t ret = re_syntax_options; + + re_syntax_options = syntax; +# ifdef DEBUG + if (syntax & RE_DEBUG) + debug = 1; + else if (debug) /* was on but now is not */ + debug = 0; +# endif /* DEBUG */ + return ret; +} +# ifdef _LIBC +weak_alias (__re_set_syntax, re_set_syntax) +# endif + +/* This table gives an error message for each of the error codes listed + in regex.h. Obviously the order here has to be same as there. + POSIX doesn't require that we do anything for REG_NOERROR, + but why not be nice? */ + +static const char re_error_msgid[] = + { +# define REG_NOERROR_IDX 0 + gettext_noop ("Success") /* REG_NOERROR */ + "\0" +# define REG_NOMATCH_IDX (REG_NOERROR_IDX + sizeof "Success") + gettext_noop ("No match") /* REG_NOMATCH */ + "\0" +# define REG_BADPAT_IDX (REG_NOMATCH_IDX + sizeof "No match") + gettext_noop ("Invalid regular expression") /* REG_BADPAT */ + "\0" +# define REG_ECOLLATE_IDX (REG_BADPAT_IDX + sizeof "Invalid regular expression") + gettext_noop ("Invalid collation character") /* REG_ECOLLATE */ + "\0" +# define REG_ECTYPE_IDX (REG_ECOLLATE_IDX + sizeof "Invalid collation character") + gettext_noop ("Invalid character class name") /* REG_ECTYPE */ + "\0" +# define REG_EESCAPE_IDX (REG_ECTYPE_IDX + sizeof "Invalid character class name") + gettext_noop ("Trailing backslash") /* REG_EESCAPE */ + "\0" +# define REG_ESUBREG_IDX (REG_EESCAPE_IDX + sizeof "Trailing backslash") + gettext_noop ("Invalid back reference") /* REG_ESUBREG */ + "\0" +# define REG_EBRACK_IDX (REG_ESUBREG_IDX + sizeof "Invalid back reference") + gettext_noop ("Unmatched [ or [^") /* REG_EBRACK */ + "\0" +# define REG_EPAREN_IDX (REG_EBRACK_IDX + sizeof "Unmatched [ or [^") + gettext_noop ("Unmatched ( or \\(") /* REG_EPAREN */ + "\0" +# define REG_EBRACE_IDX (REG_EPAREN_IDX + sizeof "Unmatched ( or \\(") + gettext_noop ("Unmatched \\{") /* REG_EBRACE */ + "\0" +# define REG_BADBR_IDX (REG_EBRACE_IDX + sizeof "Unmatched \\{") + gettext_noop ("Invalid content of \\{\\}") /* REG_BADBR */ + "\0" +# define REG_ERANGE_IDX (REG_BADBR_IDX + sizeof "Invalid content of \\{\\}") + gettext_noop ("Invalid range end") /* REG_ERANGE */ + "\0" +# define REG_ESPACE_IDX (REG_ERANGE_IDX + sizeof "Invalid range end") + gettext_noop ("Memory exhausted") /* REG_ESPACE */ + "\0" +# define REG_BADRPT_IDX (REG_ESPACE_IDX + sizeof "Memory exhausted") + gettext_noop ("Invalid preceding regular expression") /* REG_BADRPT */ + "\0" +# define REG_EEND_IDX (REG_BADRPT_IDX + sizeof "Invalid preceding regular expression") + gettext_noop ("Premature end of regular expression") /* REG_EEND */ + "\0" +# define REG_ESIZE_IDX (REG_EEND_IDX + sizeof "Premature end of regular expression") + gettext_noop ("Regular expression too big") /* REG_ESIZE */ + "\0" +# define REG_ERPAREN_IDX (REG_ESIZE_IDX + sizeof "Regular expression too big") + gettext_noop ("Unmatched ) or \\)") /* REG_ERPAREN */ + }; + +static const size_t re_error_msgid_idx[] = + { + REG_NOERROR_IDX, + REG_NOMATCH_IDX, + REG_BADPAT_IDX, + REG_ECOLLATE_IDX, + REG_ECTYPE_IDX, + REG_EESCAPE_IDX, + REG_ESUBREG_IDX, + REG_EBRACK_IDX, + REG_EPAREN_IDX, + REG_EBRACE_IDX, + REG_BADBR_IDX, + REG_ERANGE_IDX, + REG_ESPACE_IDX, + REG_BADRPT_IDX, + REG_EEND_IDX, + REG_ESIZE_IDX, + REG_ERPAREN_IDX + }; + +#endif /* INSIDE_RECURSION */ + +#ifndef DEFINED_ONCE +/* Avoiding alloca during matching, to placate r_alloc. */ + +/* Define MATCH_MAY_ALLOCATE unless we need to make sure that the + searching and matching functions should not call alloca. On some + systems, alloca is implemented in terms of malloc, and if we're + using the relocating allocator routines, then malloc could cause a + relocation, which might (if the strings being searched are in the + ralloc heap) shift the data out from underneath the regexp + routines. + + Here's another reason to avoid allocation: Emacs + processes input from X in a signal handler; processing X input may + call malloc; if input arrives while a matching routine is calling + malloc, then we're scrod. But Emacs can't just block input while + calling matching routines; then we don't notice interrupts when + they come in. So, Emacs blocks input around all regexp calls + except the matching calls, which it leaves unprotected, in the + faith that they will not malloc. */ + +/* Normally, this is fine. */ +# define MATCH_MAY_ALLOCATE + +/* When using GNU C, we are not REALLY using the C alloca, no matter + what config.h may say. So don't take precautions for it. */ +# ifdef __GNUC__ +# undef C_ALLOCA +# endif + +/* The match routines may not allocate if (1) they would do it with malloc + and (2) it's not safe for them to use malloc. + Note that if REL_ALLOC is defined, matching would not use malloc for the + failure stack, but we would still use it for the register vectors; + so REL_ALLOC should not affect this. */ +# if (defined C_ALLOCA || defined REGEX_MALLOC) && defined emacs +# undef MATCH_MAY_ALLOCATE +# endif +#endif /* not DEFINED_ONCE */ + +#ifdef INSIDE_RECURSION +/* Failure stack declarations and macros; both re_compile_fastmap and + re_match_2 use a failure stack. These have to be macros because of + REGEX_ALLOCATE_STACK. */ + + +/* Number of failure points for which to initially allocate space + when matching. If this number is exceeded, we allocate more + space, so it is not a hard limit. */ +# ifndef INIT_FAILURE_ALLOC +# define INIT_FAILURE_ALLOC 5 +# endif + +/* Roughly the maximum number of failure points on the stack. Would be + exactly that if always used MAX_FAILURE_ITEMS items each time we failed. + This is a variable only so users of regex can assign to it; we never + change it ourselves. */ + +# ifdef INT_IS_16BIT + +# ifndef DEFINED_ONCE +# if defined MATCH_MAY_ALLOCATE +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +long int re_max_failures = 4000; +# else +long int re_max_failures = 2000; +# endif +# endif + +union PREFIX(fail_stack_elt) +{ + UCHAR_T *pointer; + long int integer; +}; + +typedef union PREFIX(fail_stack_elt) PREFIX(fail_stack_elt_t); + +typedef struct +{ + PREFIX(fail_stack_elt_t) *stack; + unsigned long int size; + unsigned long int avail; /* Offset of next open position. */ +} PREFIX(fail_stack_type); + +# else /* not INT_IS_16BIT */ + +# ifndef DEFINED_ONCE +# if defined MATCH_MAY_ALLOCATE +/* 4400 was enough to cause a crash on Alpha OSF/1, + whose default stack limit is 2mb. */ +int re_max_failures = 4000; +# else +int re_max_failures = 2000; +# endif +# endif + +union PREFIX(fail_stack_elt) +{ + UCHAR_T *pointer; + int integer; +}; + +typedef union PREFIX(fail_stack_elt) PREFIX(fail_stack_elt_t); + +typedef struct +{ + PREFIX(fail_stack_elt_t) *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} PREFIX(fail_stack_type); + +# endif /* INT_IS_16BIT */ + +# ifndef DEFINED_ONCE +# define FAIL_STACK_EMPTY() (fail_stack.avail == 0) +# define FAIL_STACK_PTR_EMPTY() (fail_stack_ptr->avail == 0) +# define FAIL_STACK_FULL() (fail_stack.avail == fail_stack.size) +# endif + + +/* Define macros to initialize and free the failure stack. + Do `return -2' if the alloc fails. */ + +# ifdef MATCH_MAY_ALLOCATE +# define INIT_FAIL_STACK() \ + do { \ + fail_stack.stack = (PREFIX(fail_stack_elt_t) *) \ + REGEX_ALLOCATE_STACK (INIT_FAILURE_ALLOC * sizeof (PREFIX(fail_stack_elt_t))); \ + \ + if (fail_stack.stack == NULL) \ + return -2; \ + \ + fail_stack.size = INIT_FAILURE_ALLOC; \ + fail_stack.avail = 0; \ + } while (0) + +# define RESET_FAIL_STACK() REGEX_FREE_STACK (fail_stack.stack) +# else +# define INIT_FAIL_STACK() \ + do { \ + fail_stack.avail = 0; \ + } while (0) + +# define RESET_FAIL_STACK() +# endif + + +/* Double the size of FAIL_STACK, up to approximately `re_max_failures' items. + + Return 1 if succeeds, and 0 if either ran out of memory + allocating space for it or it was already too large. + + REGEX_REALLOCATE_STACK requires `destination' be declared. */ + +# define DOUBLE_FAIL_STACK(fail_stack) \ + ((fail_stack).size > (unsigned) (re_max_failures * MAX_FAILURE_ITEMS) \ + ? 0 \ + : ((fail_stack).stack = (PREFIX(fail_stack_elt_t) *) \ + REGEX_REALLOCATE_STACK ((fail_stack).stack, \ + (fail_stack).size * sizeof (PREFIX(fail_stack_elt_t)), \ + ((fail_stack).size << 1) * sizeof (PREFIX(fail_stack_elt_t))),\ + \ + (fail_stack).stack == NULL \ + ? 0 \ + : ((fail_stack).size <<= 1, \ + 1))) + + +/* Push pointer POINTER on FAIL_STACK. + Return 1 if was able to do so and 0 if ran out of memory allocating + space to do so. */ +# define PUSH_PATTERN_OP(POINTER, FAIL_STACK) \ + ((FAIL_STACK_FULL () \ + && !DOUBLE_FAIL_STACK (FAIL_STACK)) \ + ? 0 \ + : ((FAIL_STACK).stack[(FAIL_STACK).avail++].pointer = POINTER, \ + 1)) + +/* Push a pointer value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +# define PUSH_FAILURE_POINTER(item) \ + fail_stack.stack[fail_stack.avail++].pointer = (UCHAR_T *) (item) + +/* This pushes an integer-valued item onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +# define PUSH_FAILURE_INT(item) \ + fail_stack.stack[fail_stack.avail++].integer = (item) + +/* Push a fail_stack_elt_t value onto the failure stack. + Assumes the variable `fail_stack'. Probably should only + be called from within `PUSH_FAILURE_POINT'. */ +# define PUSH_FAILURE_ELT(item) \ + fail_stack.stack[fail_stack.avail++] = (item) + +/* These three POP... operations complement the three PUSH... operations. + All assume that `fail_stack' is nonempty. */ +# define POP_FAILURE_POINTER() fail_stack.stack[--fail_stack.avail].pointer +# define POP_FAILURE_INT() fail_stack.stack[--fail_stack.avail].integer +# define POP_FAILURE_ELT() fail_stack.stack[--fail_stack.avail] + +/* Used to omit pushing failure point id's when we're not debugging. */ +# ifdef DEBUG +# define DEBUG_PUSH PUSH_FAILURE_INT +# define DEBUG_POP(item_addr) *(item_addr) = POP_FAILURE_INT () +# else +# define DEBUG_PUSH(item) +# define DEBUG_POP(item_addr) +# endif + + +/* Push the information about the state we will need + if we ever fail back to it. + + Requires variables fail_stack, regstart, regend, reg_info, and + num_regs_pushed be declared. DOUBLE_FAIL_STACK requires `destination' + be declared. + + Does `return FAILURE_CODE' if runs out of memory. */ + +# define PUSH_FAILURE_POINT(pattern_place, string_place, failure_code) \ + do { \ + char *destination; \ + /* Must be int, so when we don't save any registers, the arithmetic \ + of 0 + -1 isn't done as unsigned. */ \ + /* Can't be int, since there is not a shred of a guarantee that int \ + is wide enough to hold a value of something to which pointer can \ + be assigned */ \ + active_reg_t this_reg; \ + \ + DEBUG_STATEMENT (failure_id++); \ + DEBUG_STATEMENT (nfailure_points_pushed++); \ + DEBUG_PRINT2 ("\nPUSH_FAILURE_POINT #%u:\n", failure_id); \ + DEBUG_PRINT2 (" Before push, next avail: %d\n", (fail_stack).avail);\ + DEBUG_PRINT2 (" size: %d\n", (fail_stack).size);\ + \ + DEBUG_PRINT2 (" slots needed: %ld\n", NUM_FAILURE_ITEMS); \ + DEBUG_PRINT2 (" available: %d\n", REMAINING_AVAIL_SLOTS); \ + \ + /* Ensure we have enough space allocated for what we will push. */ \ + while (REMAINING_AVAIL_SLOTS < NUM_FAILURE_ITEMS) \ + { \ + if (!DOUBLE_FAIL_STACK (fail_stack)) \ + return failure_code; \ + \ + DEBUG_PRINT2 ("\n Doubled stack; size now: %d\n", \ + (fail_stack).size); \ + DEBUG_PRINT2 (" slots available: %d\n", REMAINING_AVAIL_SLOTS);\ + } \ + \ + /* Push the info, starting with the registers. */ \ + DEBUG_PRINT1 ("\n"); \ + \ + if (1) \ + for (this_reg = lowest_active_reg; this_reg <= highest_active_reg; \ + this_reg++) \ + { \ + DEBUG_PRINT2 (" Pushing reg: %lu\n", this_reg); \ + DEBUG_STATEMENT (num_regs_pushed++); \ + \ + DEBUG_PRINT2 (" start: %p\n", regstart[this_reg]); \ + PUSH_FAILURE_POINTER (regstart[this_reg]); \ + \ + DEBUG_PRINT2 (" end: %p\n", regend[this_reg]); \ + PUSH_FAILURE_POINTER (regend[this_reg]); \ + \ + DEBUG_PRINT2 (" info: %p\n ", \ + reg_info[this_reg].word.pointer); \ + DEBUG_PRINT2 (" match_null=%d", \ + REG_MATCH_NULL_STRING_P (reg_info[this_reg])); \ + DEBUG_PRINT2 (" active=%d", IS_ACTIVE (reg_info[this_reg])); \ + DEBUG_PRINT2 (" matched_something=%d", \ + MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT2 (" ever_matched=%d", \ + EVER_MATCHED_SOMETHING (reg_info[this_reg])); \ + DEBUG_PRINT1 ("\n"); \ + PUSH_FAILURE_ELT (reg_info[this_reg].word); \ + } \ + \ + DEBUG_PRINT2 (" Pushing low active reg: %ld\n", lowest_active_reg);\ + PUSH_FAILURE_INT (lowest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing high active reg: %ld\n", highest_active_reg);\ + PUSH_FAILURE_INT (highest_active_reg); \ + \ + DEBUG_PRINT2 (" Pushing pattern %p:\n", pattern_place); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pattern_place, pend); \ + PUSH_FAILURE_POINTER (pattern_place); \ + \ + DEBUG_PRINT2 (" Pushing string %p: `", string_place); \ + DEBUG_PRINT_DOUBLE_STRING (string_place, string1, size1, string2, \ + size2); \ + DEBUG_PRINT1 ("'\n"); \ + PUSH_FAILURE_POINTER (string_place); \ + \ + DEBUG_PRINT2 (" Pushing failure id: %u\n", failure_id); \ + DEBUG_PUSH (failure_id); \ + } while (0) + +# ifndef DEFINED_ONCE +/* This is the number of items that are pushed and popped on the stack + for each register. */ +# define NUM_REG_ITEMS 3 + +/* Individual items aside from the registers. */ +# ifdef DEBUG +# define NUM_NONREG_ITEMS 5 /* Includes failure point id. */ +# else +# define NUM_NONREG_ITEMS 4 +# endif + +/* We push at most this many items on the stack. */ +/* We used to use (num_regs - 1), which is the number of registers + this regexp will save; but that was changed to 5 + to avoid stack overflow for a regexp with lots of parens. */ +# define MAX_FAILURE_ITEMS (5 * NUM_REG_ITEMS + NUM_NONREG_ITEMS) + +/* We actually push this many items. */ +# define NUM_FAILURE_ITEMS \ + (((0 \ + ? 0 : highest_active_reg - lowest_active_reg + 1) \ + * NUM_REG_ITEMS) \ + + NUM_NONREG_ITEMS) + +/* How many items can still be added to the stack without overflowing it. */ +# define REMAINING_AVAIL_SLOTS ((fail_stack).size - (fail_stack).avail) +# endif /* not DEFINED_ONCE */ + + +/* Pops what PUSH_FAIL_STACK pushes. + + We restore into the parameters, all of which should be lvalues: + STR -- the saved data position. + PAT -- the saved pattern position. + LOW_REG, HIGH_REG -- the highest and lowest active registers. + REGSTART, REGEND -- arrays of string positions. + REG_INFO -- array of information about each subexpression. + + Also assumes the variables `fail_stack' and (if debugging), `bufp', + `pend', `string1', `size1', `string2', and `size2'. */ +# define POP_FAILURE_POINT(str, pat, low_reg, high_reg, regstart, regend, reg_info)\ +{ \ + DEBUG_STATEMENT (unsigned failure_id;) \ + active_reg_t this_reg; \ + const UCHAR_T *string_temp; \ + \ + assert (!FAIL_STACK_EMPTY ()); \ + \ + /* Remove failure points and point to how many regs pushed. */ \ + DEBUG_PRINT1 ("POP_FAILURE_POINT:\n"); \ + DEBUG_PRINT2 (" Before pop, next avail: %d\n", fail_stack.avail); \ + DEBUG_PRINT2 (" size: %d\n", fail_stack.size); \ + \ + assert (fail_stack.avail >= NUM_NONREG_ITEMS); \ + \ + DEBUG_POP (&failure_id); \ + DEBUG_PRINT2 (" Popping failure id: %u\n", failure_id); \ + \ + /* If the saved string location is NULL, it came from an \ + on_failure_keep_string_jump opcode, and we want to throw away the \ + saved NULL, thus retaining our current position in the string. */ \ + string_temp = POP_FAILURE_POINTER (); \ + if (string_temp != NULL) \ + str = (const CHAR_T *) string_temp; \ + \ + DEBUG_PRINT2 (" Popping string %p: `", str); \ + DEBUG_PRINT_DOUBLE_STRING (str, string1, size1, string2, size2); \ + DEBUG_PRINT1 ("'\n"); \ + \ + pat = (UCHAR_T *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" Popping pattern %p:\n", pat); \ + DEBUG_PRINT_COMPILED_PATTERN (bufp, pat, pend); \ + \ + /* Restore register info. */ \ + high_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping high active reg: %ld\n", high_reg); \ + \ + low_reg = (active_reg_t) POP_FAILURE_INT (); \ + DEBUG_PRINT2 (" Popping low active reg: %ld\n", low_reg); \ + \ + if (1) \ + for (this_reg = high_reg; this_reg >= low_reg; this_reg--) \ + { \ + DEBUG_PRINT2 (" Popping reg: %ld\n", this_reg); \ + \ + reg_info[this_reg].word = POP_FAILURE_ELT (); \ + DEBUG_PRINT2 (" info: %p\n", \ + reg_info[this_reg].word.pointer); \ + \ + regend[this_reg] = (const CHAR_T *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" end: %p\n", regend[this_reg]); \ + \ + regstart[this_reg] = (const CHAR_T *) POP_FAILURE_POINTER (); \ + DEBUG_PRINT2 (" start: %p\n", regstart[this_reg]); \ + } \ + else \ + { \ + for (this_reg = highest_active_reg; this_reg > high_reg; this_reg--) \ + { \ + reg_info[this_reg].word.integer = 0; \ + regend[this_reg] = 0; \ + regstart[this_reg] = 0; \ + } \ + highest_active_reg = high_reg; \ + } \ + \ + set_regs_matched_done = 0; \ + DEBUG_STATEMENT (nfailure_points_popped++); \ +} /* POP_FAILURE_POINT */ + +/* Structure for per-register (a.k.a. per-group) information. + Other register information, such as the + starting and ending positions (which are addresses), and the list of + inner groups (which is a bits list) are maintained in separate + variables. + + We are making a (strictly speaking) nonportable assumption here: that + the compiler will pack our bit fields into something that fits into + the type of `word', i.e., is something that fits into one item on the + failure stack. */ + + +/* Declarations and macros for re_match_2. */ + +typedef union +{ + PREFIX(fail_stack_elt_t) word; + struct + { + /* This field is one if this group can match the empty string, + zero if not. If not yet determined, `MATCH_NULL_UNSET_VALUE'. */ +# define MATCH_NULL_UNSET_VALUE 3 + unsigned match_null_string_p : 2; + unsigned is_active : 1; + unsigned matched_something : 1; + unsigned ever_matched_something : 1; + } bits; +} PREFIX(register_info_type); + +# ifndef DEFINED_ONCE +# define REG_MATCH_NULL_STRING_P(R) ((R).bits.match_null_string_p) +# define IS_ACTIVE(R) ((R).bits.is_active) +# define MATCHED_SOMETHING(R) ((R).bits.matched_something) +# define EVER_MATCHED_SOMETHING(R) ((R).bits.ever_matched_something) + + +/* Call this when have matched a real character; it sets `matched' flags + for the subexpressions which we are currently inside. Also records + that those subexprs have matched. */ +# define SET_REGS_MATCHED() \ + do \ + { \ + if (!set_regs_matched_done) \ + { \ + active_reg_t r; \ + set_regs_matched_done = 1; \ + for (r = lowest_active_reg; r <= highest_active_reg; r++) \ + { \ + MATCHED_SOMETHING (reg_info[r]) \ + = EVER_MATCHED_SOMETHING (reg_info[r]) \ + = 1; \ + } \ + } \ + } \ + while (0) +# endif /* not DEFINED_ONCE */ + +/* Registers are set to a sentinel when they haven't yet matched. */ +static CHAR_T PREFIX(reg_unset_dummy); +# define REG_UNSET_VALUE (&PREFIX(reg_unset_dummy)) +# define REG_UNSET(e) ((e) == REG_UNSET_VALUE) + +/* Subroutine declarations and macros for regex_compile. */ +static void PREFIX(store_op1) (re_opcode_t op, UCHAR_T *loc, int arg); +static void PREFIX(store_op2) (re_opcode_t op, UCHAR_T *loc, + int arg1, int arg2); +static void PREFIX(insert_op1) (re_opcode_t op, UCHAR_T *loc, + int arg, UCHAR_T *end); +static void PREFIX(insert_op2) (re_opcode_t op, UCHAR_T *loc, + int arg1, int arg2, UCHAR_T *end); +static boolean PREFIX(at_begline_loc_p) (const CHAR_T *pattern, + const CHAR_T *p, + reg_syntax_t syntax); +static boolean PREFIX(at_endline_loc_p) (const CHAR_T *p, + const CHAR_T *pend, + reg_syntax_t syntax); +# ifdef WCHAR +static reg_errcode_t wcs_compile_range (CHAR_T range_start, + const CHAR_T **p_ptr, + const CHAR_T *pend, + char *translate, + reg_syntax_t syntax, + UCHAR_T *b, + CHAR_T *char_set); +static void insert_space (int num, CHAR_T *loc, CHAR_T *end); +# else /* BYTE */ +static reg_errcode_t byte_compile_range (unsigned int range_start, + const char **p_ptr, + const char *pend, + char *translate, + reg_syntax_t syntax, + unsigned char *b); +# endif /* WCHAR */ + +/* Fetch the next character in the uncompiled pattern---translating it + if necessary. Also cast from a signed character in the constant + string passed to us by the user to an unsigned char that we can use + as an array index (in, e.g., `translate'). */ +/* ifdef MBS_SUPPORT, we translate only if character <= 0xff, + because it is impossible to allocate 4GB array for some encodings + which have 4 byte character_set like UCS4. */ +# ifndef PATFETCH +# ifdef WCHAR +# define PATFETCH(c) \ + do {if (p == pend) return REG_EEND; \ + c = (UCHAR_T) *p++; \ + if (translate && (c <= 0xff)) c = (UCHAR_T) translate[c]; \ + } while (0) +# else /* BYTE */ +# define PATFETCH(c) \ + do {if (p == pend) return REG_EEND; \ + c = (unsigned char) *p++; \ + if (translate) c = (unsigned char) translate[c]; \ + } while (0) +# endif /* WCHAR */ +# endif + +/* Fetch the next character in the uncompiled pattern, with no + translation. */ +# define PATFETCH_RAW(c) \ + do {if (p == pend) return REG_EEND; \ + c = (UCHAR_T) *p++; \ + } while (0) + +/* Go backwards one character in the pattern. */ +# define PATUNFETCH p-- + + +/* If `translate' is non-null, return translate[D], else just D. We + cast the subscript to translate because some data is declared as + `char *', to avoid warnings when a string constant is passed. But + when we use a character as a subscript we must make it unsigned. */ +/* ifdef MBS_SUPPORT, we translate only if character <= 0xff, + because it is impossible to allocate 4GB array for some encodings + which have 4 byte character_set like UCS4. */ + +# ifndef TRANSLATE +# ifdef WCHAR +# define TRANSLATE(d) \ + ((translate && ((UCHAR_T) (d)) <= 0xff) \ + ? (char) translate[(unsigned char) (d)] : (d)) +# else /* BYTE */ +# define TRANSLATE(d) \ + (translate ? (char) translate[(unsigned char) (d)] : (d)) +# endif /* WCHAR */ +# endif + + +/* Macros for outputting the compiled pattern into `buffer'. */ + +/* If the buffer isn't allocated when it comes in, use this. */ +# define INIT_BUF_SIZE (32 * sizeof(UCHAR_T)) + +/* Make sure we have at least N more bytes of space in buffer. */ +# ifdef WCHAR +# define GET_BUFFER_SPACE(n) \ + while (((unsigned long)b - (unsigned long)COMPILED_BUFFER_VAR \ + + (n)*sizeof(CHAR_T)) > bufp->allocated) \ + EXTEND_BUFFER () +# else /* BYTE */ +# define GET_BUFFER_SPACE(n) \ + while ((unsigned long) (b - bufp->buffer + (n)) > bufp->allocated) \ + EXTEND_BUFFER () +# endif /* WCHAR */ + +/* Make sure we have one more byte of buffer space and then add C to it. */ +# define BUF_PUSH(c) \ + do { \ + GET_BUFFER_SPACE (1); \ + *b++ = (UCHAR_T) (c); \ + } while (0) + + +/* Ensure we have two more bytes of buffer space and then append C1 and C2. */ +# define BUF_PUSH_2(c1, c2) \ + do { \ + GET_BUFFER_SPACE (2); \ + *b++ = (UCHAR_T) (c1); \ + *b++ = (UCHAR_T) (c2); \ + } while (0) + + +/* As with BUF_PUSH_2, except for three bytes. */ +# define BUF_PUSH_3(c1, c2, c3) \ + do { \ + GET_BUFFER_SPACE (3); \ + *b++ = (UCHAR_T) (c1); \ + *b++ = (UCHAR_T) (c2); \ + *b++ = (UCHAR_T) (c3); \ + } while (0) + +/* Store a jump with opcode OP at LOC to location TO. We store a + relative address offset by the three bytes the jump itself occupies. */ +# define STORE_JUMP(op, loc, to) \ + PREFIX(store_op1) (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE))) + +/* Likewise, for a two-argument jump. */ +# define STORE_JUMP2(op, loc, to, arg) \ + PREFIX(store_op2) (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE)), arg) + +/* Like `STORE_JUMP', but for inserting. Assume `b' is the buffer end. */ +# define INSERT_JUMP(op, loc, to) \ + PREFIX(insert_op1) (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE)), b) + +/* Like `STORE_JUMP2', but for inserting. Assume `b' is the buffer end. */ +# define INSERT_JUMP2(op, loc, to, arg) \ + PREFIX(insert_op2) (op, loc, (int) ((to) - (loc) - (1 + OFFSET_ADDRESS_SIZE)),\ + arg, b) + +/* This is not an arbitrary limit: the arguments which represent offsets + into the pattern are two bytes long. So if 2^16 bytes turns out to + be too small, many things would have to change. */ +/* Any other compiler which, like MSC, has allocation limit below 2^16 + bytes will have to use approach similar to what was done below for + MSC and drop MAX_BUF_SIZE a bit. Otherwise you may end up + reallocating to 0 bytes. Such thing is not going to work too well. + You have been warned!! */ +# ifndef DEFINED_ONCE +# if defined _MSC_VER && !defined WIN32 +/* Microsoft C 16-bit versions limit malloc to approx 65512 bytes. + The REALLOC define eliminates a flurry of conversion warnings, + but is not required. */ +# define MAX_BUF_SIZE 65500L +# define REALLOC(p,s) realloc ((p), (size_t) (s)) +# else +# define MAX_BUF_SIZE (1L << 16) +# define REALLOC(p,s) realloc ((p), (s)) +# endif + +/* Extend the buffer by twice its current size via realloc and + reset the pointers that pointed into the old block to point to the + correct places in the new one. If extending the buffer results in it + being larger than MAX_BUF_SIZE, then flag memory exhausted. */ +# if __BOUNDED_POINTERS__ +# define SET_HIGH_BOUND(P) (__ptrhigh (P) = __ptrlow (P) + bufp->allocated) +# define MOVE_BUFFER_POINTER(P) \ + (__ptrlow (P) += incr, SET_HIGH_BOUND (P), __ptrvalue (P) += incr) +# define ELSE_EXTEND_BUFFER_HIGH_BOUND \ + else \ + { \ + SET_HIGH_BOUND (b); \ + SET_HIGH_BOUND (begalt); \ + if (fixup_alt_jump) \ + SET_HIGH_BOUND (fixup_alt_jump); \ + if (laststart) \ + SET_HIGH_BOUND (laststart); \ + if (pending_exact) \ + SET_HIGH_BOUND (pending_exact); \ + } +# else +# define MOVE_BUFFER_POINTER(P) (P) += incr +# define ELSE_EXTEND_BUFFER_HIGH_BOUND +# endif +# endif /* not DEFINED_ONCE */ + +# ifdef WCHAR +# define EXTEND_BUFFER() \ + do { \ + UCHAR_T *old_buffer = COMPILED_BUFFER_VAR; \ + int wchar_count; \ + if (bufp->allocated + sizeof(UCHAR_T) > MAX_BUF_SIZE) \ + return REG_ESIZE; \ + bufp->allocated <<= 1; \ + if (bufp->allocated > MAX_BUF_SIZE) \ + bufp->allocated = MAX_BUF_SIZE; \ + /* How many characters the new buffer can have? */ \ + wchar_count = bufp->allocated / sizeof(UCHAR_T); \ + if (wchar_count == 0) wchar_count = 1; \ + /* Truncate the buffer to CHAR_T align. */ \ + bufp->allocated = wchar_count * sizeof(UCHAR_T); \ + RETALLOC (COMPILED_BUFFER_VAR, wchar_count, UCHAR_T); \ + bufp->buffer = (char*)COMPILED_BUFFER_VAR; \ + if (COMPILED_BUFFER_VAR == NULL) \ + return REG_ESPACE; \ + /* If the buffer moved, move all the pointers into it. */ \ + if (old_buffer != COMPILED_BUFFER_VAR) \ + { \ + int incr = COMPILED_BUFFER_VAR - old_buffer; \ + MOVE_BUFFER_POINTER (b); \ + MOVE_BUFFER_POINTER (begalt); \ + if (fixup_alt_jump) \ + MOVE_BUFFER_POINTER (fixup_alt_jump); \ + if (laststart) \ + MOVE_BUFFER_POINTER (laststart); \ + if (pending_exact) \ + MOVE_BUFFER_POINTER (pending_exact); \ + } \ + ELSE_EXTEND_BUFFER_HIGH_BOUND \ + } while (0) +# else /* BYTE */ +# define EXTEND_BUFFER() \ + do { \ + UCHAR_T *old_buffer = COMPILED_BUFFER_VAR; \ + if (bufp->allocated == MAX_BUF_SIZE) \ + return REG_ESIZE; \ + bufp->allocated <<= 1; \ + if (bufp->allocated > MAX_BUF_SIZE) \ + bufp->allocated = MAX_BUF_SIZE; \ + bufp->buffer = REALLOC (COMPILED_BUFFER_VAR, bufp->allocated); \ + if (COMPILED_BUFFER_VAR == NULL) \ + return REG_ESPACE; \ + /* If the buffer moved, move all the pointers into it. */ \ + if (old_buffer != COMPILED_BUFFER_VAR) \ + { \ + int incr = COMPILED_BUFFER_VAR - old_buffer; \ + MOVE_BUFFER_POINTER (b); \ + MOVE_BUFFER_POINTER (begalt); \ + if (fixup_alt_jump) \ + MOVE_BUFFER_POINTER (fixup_alt_jump); \ + if (laststart) \ + MOVE_BUFFER_POINTER (laststart); \ + if (pending_exact) \ + MOVE_BUFFER_POINTER (pending_exact); \ + } \ + ELSE_EXTEND_BUFFER_HIGH_BOUND \ + } while (0) +# endif /* WCHAR */ + +# ifndef DEFINED_ONCE +/* Since we have one byte reserved for the register number argument to + {start,stop}_memory, the maximum number of groups we can report + things about is what fits in that byte. */ +# define MAX_REGNUM 255 + +/* But patterns can have more than `MAX_REGNUM' registers. We just + ignore the excess. */ +typedef unsigned regnum_t; + + +/* Macros for the compile stack. */ + +/* Since offsets can go either forwards or backwards, this type needs to + be able to hold values from -(MAX_BUF_SIZE - 1) to MAX_BUF_SIZE - 1. */ +/* int may be not enough when sizeof(int) == 2. */ +typedef long pattern_offset_t; + +typedef struct +{ + pattern_offset_t begalt_offset; + pattern_offset_t fixup_alt_jump; + pattern_offset_t inner_group_offset; + pattern_offset_t laststart_offset; + regnum_t regnum; +} compile_stack_elt_t; + + +typedef struct +{ + compile_stack_elt_t *stack; + unsigned size; + unsigned avail; /* Offset of next open position. */ +} compile_stack_type; + + +# define INIT_COMPILE_STACK_SIZE 32 + +# define COMPILE_STACK_EMPTY (compile_stack.avail == 0) +# define COMPILE_STACK_FULL (compile_stack.avail == compile_stack.size) + +/* The next available element. */ +# define COMPILE_STACK_TOP (compile_stack.stack[compile_stack.avail]) + +# endif /* not DEFINED_ONCE */ + +/* Set the bit for character C in a list. */ +# ifndef DEFINED_ONCE +# define SET_LIST_BIT(c) \ + (b[((unsigned char) (c)) / BYTEWIDTH] \ + |= 1 << (((unsigned char) c) % BYTEWIDTH)) +# endif /* DEFINED_ONCE */ + +/* Get the next unsigned number in the uncompiled pattern. */ +# define GET_UNSIGNED_NUMBER(num) \ + { \ + while (p != pend) \ + { \ + PATFETCH (c); \ + if (c < '0' || c > '9') \ + break; \ + if (num <= RE_DUP_MAX) \ + { \ + if (num < 0) \ + num = 0; \ + num = num * 10 + c - '0'; \ + } \ + } \ + } + +# ifndef DEFINED_ONCE +# if defined _LIBC || WIDE_CHAR_SUPPORT +/* The GNU C library provides support for user-defined character classes + and the functions from ISO C amendement 1. */ +# ifdef CHARCLASS_NAME_MAX +# define CHAR_CLASS_MAX_LENGTH CHARCLASS_NAME_MAX +# else +/* This shouldn't happen but some implementation might still have this + problem. Use a reasonable default value. */ +# define CHAR_CLASS_MAX_LENGTH 256 +# endif + +# ifdef _LIBC +# define IS_CHAR_CLASS(string) __wctype (string) +# else +# define IS_CHAR_CLASS(string) wctype (string) +# endif +# else +# define CHAR_CLASS_MAX_LENGTH 6 /* Namely, `xdigit'. */ + +# define IS_CHAR_CLASS(string) \ + (STREQ (string, "alpha") || STREQ (string, "upper") \ + || STREQ (string, "lower") || STREQ (string, "digit") \ + || STREQ (string, "alnum") || STREQ (string, "xdigit") \ + || STREQ (string, "space") || STREQ (string, "print") \ + || STREQ (string, "punct") || STREQ (string, "graph") \ + || STREQ (string, "cntrl") || STREQ (string, "blank")) +# endif +# endif /* DEFINED_ONCE */ + +# ifndef MATCH_MAY_ALLOCATE + +/* If we cannot allocate large objects within re_match_2_internal, + we make the fail stack and register vectors global. + The fail stack, we grow to the maximum size when a regexp + is compiled. + The register vectors, we adjust in size each time we + compile a regexp, according to the number of registers it needs. */ + +static PREFIX(fail_stack_type) fail_stack; + +/* Size with which the following vectors are currently allocated. + That is so we can make them bigger as needed, + but never make them smaller. */ +# ifdef DEFINED_ONCE +static int regs_allocated_size; + +static const char ** regstart, ** regend; +static const char ** old_regstart, ** old_regend; +static const char **best_regstart, **best_regend; +static const char **reg_dummy; +# endif /* DEFINED_ONCE */ + +static PREFIX(register_info_type) *PREFIX(reg_info); +static PREFIX(register_info_type) *PREFIX(reg_info_dummy); + +/* Make the register vectors big enough for NUM_REGS registers, + but don't make them smaller. */ + +static void +PREFIX(regex_grow_registers) (int num_regs) +{ + if (num_regs > regs_allocated_size) + { + RETALLOC_IF (regstart, num_regs, const char *); + RETALLOC_IF (regend, num_regs, const char *); + RETALLOC_IF (old_regstart, num_regs, const char *); + RETALLOC_IF (old_regend, num_regs, const char *); + RETALLOC_IF (best_regstart, num_regs, const char *); + RETALLOC_IF (best_regend, num_regs, const char *); + RETALLOC_IF (PREFIX(reg_info), num_regs, PREFIX(register_info_type)); + RETALLOC_IF (reg_dummy, num_regs, const char *); + RETALLOC_IF (PREFIX(reg_info_dummy), num_regs, PREFIX(register_info_type)); + + regs_allocated_size = num_regs; + } +} + +# endif /* not MATCH_MAY_ALLOCATE */ + +# ifndef DEFINED_ONCE +static boolean group_in_compile_stack (compile_stack_type + compile_stack, + regnum_t regnum); +# endif /* not DEFINED_ONCE */ + +/* `regex_compile' compiles PATTERN (of length SIZE) according to SYNTAX. + Returns one of error codes defined in `regex.h', or zero for success. + + Assumes the `allocated' (and perhaps `buffer') and `translate' + fields are set in BUFP on entry. + + If it succeeds, results are put in BUFP (if it returns an error, the + contents of BUFP are undefined): + `buffer' is the compiled pattern; + `syntax' is set to SYNTAX; + `used' is set to the length of the compiled pattern; + `fastmap_accurate' is zero; + `re_nsub' is the number of subexpressions in PATTERN; + `not_bol' and `not_eol' are zero; + + The `fastmap' and `newline_anchor' fields are neither + examined nor set. */ + +/* Return, freeing storage we allocated. */ +# ifdef WCHAR +# define FREE_STACK_RETURN(value) \ + return (free(pattern), free(mbs_offset), free(is_binary), free (compile_stack.stack), value) +# else +# define FREE_STACK_RETURN(value) \ + return (free (compile_stack.stack), value) +# endif /* WCHAR */ + +static reg_errcode_t +PREFIX(regex_compile) (const char *ARG_PREFIX(pattern), + size_t ARG_PREFIX(size), + reg_syntax_t syntax, + struct re_pattern_buffer *bufp) +{ + /* We fetch characters from PATTERN here. Even though PATTERN is + `char *' (i.e., signed), we declare these variables as unsigned, so + they can be reliably used as array indices. */ + register UCHAR_T c, c1; + +#ifdef WCHAR + /* A temporary space to keep wchar_t pattern and compiled pattern. */ + CHAR_T *pattern, *COMPILED_BUFFER_VAR; + size_t size; + /* offset buffer for optimization. See convert_mbs_to_wc. */ + int *mbs_offset = NULL; + /* It hold whether each wchar_t is binary data or not. */ + char *is_binary = NULL; + /* A flag whether exactn is handling binary data or not. */ + char is_exactn_bin = FALSE; +#endif /* WCHAR */ + + /* A random temporary spot in PATTERN. */ + const CHAR_T *p1; + + /* Points to the end of the buffer, where we should append. */ + register UCHAR_T *b; + + /* Keeps track of unclosed groups. */ + compile_stack_type compile_stack; + + /* Points to the current (ending) position in the pattern. */ +#ifdef WCHAR + const CHAR_T *p; + const CHAR_T *pend; +#else /* BYTE */ + const CHAR_T *p = pattern; + const CHAR_T *pend = pattern + size; +#endif /* WCHAR */ + + /* How to translate the characters in the pattern. */ + RE_TRANSLATE_TYPE translate = bufp->translate; + + /* Address of the count-byte of the most recently inserted `exactn' + command. This makes it possible to tell if a new exact-match + character can be added to that command or if the character requires + a new `exactn' command. */ + UCHAR_T *pending_exact = 0; + + /* Address of start of the most recently finished expression. + This tells, e.g., postfix * where to find the start of its + operand. Reset at the beginning of groups and alternatives. */ + UCHAR_T *laststart = 0; + + /* Address of beginning of regexp, or inside of last group. */ + UCHAR_T *begalt; + + /* Address of the place where a forward jump should go to the end of + the containing expression. Each alternative of an `or' -- except the + last -- ends with a forward jump of this sort. */ + UCHAR_T *fixup_alt_jump = 0; + + /* Counts open-groups as they are encountered. Remembered for the + matching close-group on the compile stack, so the same register + number is put in the stop_memory as the start_memory. */ + regnum_t regnum = 0; + +#ifdef WCHAR + /* Initialize the wchar_t PATTERN and offset_buffer. */ + p = pend = pattern = TALLOC(csize + 1, CHAR_T); + mbs_offset = TALLOC(csize + 1, int); + is_binary = TALLOC(csize + 1, char); + if (pattern == NULL || mbs_offset == NULL || is_binary == NULL) + { + free(pattern); + free(mbs_offset); + free(is_binary); + return REG_ESPACE; + } + pattern[csize] = L'\0'; /* sentinel */ + size = convert_mbs_to_wcs(pattern, cpattern, csize, mbs_offset, is_binary); + pend = p + size; + if (size < 0) + { + free(pattern); + free(mbs_offset); + free(is_binary); + return REG_BADPAT; + } +#endif + +#ifdef DEBUG + DEBUG_PRINT1 ("\nCompiling pattern: "); + if (debug) + { + unsigned debug_count; + + for (debug_count = 0; debug_count < size; debug_count++) + PUT_CHAR (pattern[debug_count]); + putchar ('\n'); + } +#endif /* DEBUG */ + + /* Initialize the compile stack. */ + compile_stack.stack = TALLOC (INIT_COMPILE_STACK_SIZE, compile_stack_elt_t); + if (compile_stack.stack == NULL) + { +#ifdef WCHAR + free(pattern); + free(mbs_offset); + free(is_binary); +#endif + return REG_ESPACE; + } + + compile_stack.size = INIT_COMPILE_STACK_SIZE; + compile_stack.avail = 0; + + /* Initialize the pattern buffer. */ + bufp->syntax = syntax; + bufp->fastmap_accurate = 0; + bufp->not_bol = bufp->not_eol = 0; + + /* Set `used' to zero, so that if we return an error, the pattern + printer (for debugging) will think there's no pattern. We reset it + at the end. */ + bufp->used = 0; + + /* Always count groups, whether or not bufp->no_sub is set. */ + bufp->re_nsub = 0; + +#if !defined emacs && !defined SYNTAX_TABLE + /* Initialize the syntax table. */ + init_syntax_once (); +#endif + + if (bufp->allocated == 0) + { + if (bufp->buffer) + { /* If zero allocated, but buffer is non-null, try to realloc + enough space. This loses if buffer's address is bogus, but + that is the user's responsibility. */ +#ifdef WCHAR + /* Free bufp->buffer and allocate an array for wchar_t pattern + buffer. */ + free(bufp->buffer); + COMPILED_BUFFER_VAR = TALLOC (INIT_BUF_SIZE/sizeof(UCHAR_T), + UCHAR_T); +#else + RETALLOC (COMPILED_BUFFER_VAR, INIT_BUF_SIZE, UCHAR_T); +#endif /* WCHAR */ + } + else + { /* Caller did not allocate a buffer. Do it for them. */ + COMPILED_BUFFER_VAR = TALLOC (INIT_BUF_SIZE / sizeof(UCHAR_T), + UCHAR_T); + } + + if (!COMPILED_BUFFER_VAR) FREE_STACK_RETURN (REG_ESPACE); +#ifdef WCHAR + bufp->buffer = (char*)COMPILED_BUFFER_VAR; +#endif /* WCHAR */ + bufp->allocated = INIT_BUF_SIZE; + } +#ifdef WCHAR + else + COMPILED_BUFFER_VAR = (UCHAR_T*) bufp->buffer; +#endif + + begalt = b = COMPILED_BUFFER_VAR; + + /* Loop through the uncompiled pattern until we're at the end. */ + while (p != pend) + { + PATFETCH (c); + + switch (c) + { + case '^': + { + if ( /* If at start of pattern, it's an operator. */ + p == pattern + 1 + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's come before. */ + || PREFIX(at_begline_loc_p) (pattern, p, syntax)) + BUF_PUSH (begline); + else + goto normal_char; + } + break; + + + case '$': + { + if ( /* If at end of pattern, it's an operator. */ + p == pend + /* If context independent, it's an operator. */ + || syntax & RE_CONTEXT_INDEP_ANCHORS + /* Otherwise, depends on what's next. */ + || PREFIX(at_endline_loc_p) (p, pend, syntax)) + BUF_PUSH (endline); + else + goto normal_char; + } + break; + + + case '+': + case '?': + if ((syntax & RE_BK_PLUS_QM) + || (syntax & RE_LIMITED_OPS)) + goto normal_char; + handle_plus: + case '*': + /* If there is no previous pattern... */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS) + FREE_STACK_RETURN (REG_BADRPT); + else if (!(syntax & RE_CONTEXT_INDEP_OPS)) + goto normal_char; + } + + { + /* Are we optimizing this jump? */ + boolean keep_string_p = false; + + /* 1 means zero (many) matches is allowed. */ + char zero_times_ok = 0, many_times_ok = 0; + + /* If there is a sequence of repetition chars, collapse it + down to just one (the right one). We can't combine + interval operators with these because of, e.g., `a{2}*', + which should only match an even number of `a's. */ + + for (;;) + { + zero_times_ok |= c != '+'; + many_times_ok |= c != '?'; + + if (p == pend) + break; + + PATFETCH (c); + + if (c == '*' + || (!(syntax & RE_BK_PLUS_QM) && (c == '+' || c == '?'))) + ; + + else if (syntax & RE_BK_PLUS_QM && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + if (!(c1 == '+' || c1 == '?')) + { + PATUNFETCH; + PATUNFETCH; + break; + } + + c = c1; + } + else + { + PATUNFETCH; + break; + } + + /* If we get here, we found another repeat character. */ + } + + /* Star, etc. applied to an empty pattern is equivalent + to an empty pattern. */ + if (!laststart) + break; + + /* Now we know whether or not zero matches is allowed + and also whether or not two or more matches is allowed. */ + if (many_times_ok) + { /* More than one repetition is allowed, so put in at the + end a backward relative jump from `b' to before the next + jump we're going to put in below (which jumps from + laststart to after this jump). + + But if we are at the `*' in the exact sequence `.*\n', + insert an unconditional jump backwards to the ., + instead of the beginning of the loop. This way we only + push a failure point once, instead of every time + through the loop. */ + assert (p - 1 > pattern); + + /* Allocate the space for the jump. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + + /* We know we are not at the first character of the pattern, + because laststart was nonzero. And we've already + incremented `p', by the way, to be the character after + the `*'. Do we have to do something analogous here + for null bytes, because of RE_DOT_NOT_NULL? */ + if (TRANSLATE (*(p - 2)) == TRANSLATE ('.') + && zero_times_ok + && p < pend && TRANSLATE (*p) == TRANSLATE ('\n') + && !(syntax & RE_DOT_NEWLINE)) + { /* We have .*\n. */ + STORE_JUMP (jump, b, laststart); + keep_string_p = true; + } + else + /* Anything else. */ + STORE_JUMP (maybe_pop_jump, b, laststart - + (1 + OFFSET_ADDRESS_SIZE)); + + /* We've added more stuff to the buffer. */ + b += 1 + OFFSET_ADDRESS_SIZE; + } + + /* On failure, jump from laststart to b + 3, which will be the + end of the buffer after this jump is inserted. */ + /* ifdef WCHAR, 'b + 1 + OFFSET_ADDRESS_SIZE' instead of + 'b + 3'. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (keep_string_p ? on_failure_keep_string_jump + : on_failure_jump, + laststart, b + 1 + OFFSET_ADDRESS_SIZE); + pending_exact = 0; + b += 1 + OFFSET_ADDRESS_SIZE; + + if (!zero_times_ok) + { + /* At least one repetition is required, so insert a + `dummy_failure_jump' before the initial + `on_failure_jump' instruction of the loop. This + effects a skip over that instruction the first time + we hit that loop. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (dummy_failure_jump, laststart, laststart + + 2 + 2 * OFFSET_ADDRESS_SIZE); + b += 1 + OFFSET_ADDRESS_SIZE; + } + } + break; + + + case '.': + laststart = b; + BUF_PUSH (anychar); + break; + + + case '[': + { + boolean had_char_class = false; +#ifdef WCHAR + CHAR_T range_start = 0xffffffff; +#else + unsigned int range_start = 0xffffffff; +#endif + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + +#ifdef WCHAR + /* We assume a charset(_not) structure as a wchar_t array. + charset[0] = (re_opcode_t) charset(_not) + charset[1] = l (= length of char_classes) + charset[2] = m (= length of collating_symbols) + charset[3] = n (= length of equivalence_classes) + charset[4] = o (= length of char_ranges) + charset[5] = p (= length of chars) + + charset[6] = char_class (wctype_t) + charset[6+CHAR_CLASS_SIZE] = char_class (wctype_t) + ... + charset[l+5] = char_class (wctype_t) + + charset[l+6] = collating_symbol (wchar_t) + ... + charset[l+m+5] = collating_symbol (wchar_t) + ifdef _LIBC we use the index if + _NL_COLLATE_SYMB_EXTRAMB instead of + wchar_t string. + + charset[l+m+6] = equivalence_classes (wchar_t) + ... + charset[l+m+n+5] = equivalence_classes (wchar_t) + ifdef _LIBC we use the index in + _NL_COLLATE_WEIGHT instead of + wchar_t string. + + charset[l+m+n+6] = range_start + charset[l+m+n+7] = range_end + ... + charset[l+m+n+2o+4] = range_start + charset[l+m+n+2o+5] = range_end + ifdef _LIBC we use the value looked up + in _NL_COLLATE_COLLSEQ instead of + wchar_t character. + + charset[l+m+n+2o+6] = char + ... + charset[l+m+n+2o+p+5] = char + + */ + + /* We need at least 6 spaces: the opcode, the length of + char_classes, the length of collating_symbols, the length of + equivalence_classes, the length of char_ranges, the length of + chars. */ + GET_BUFFER_SPACE (6); + + /* Save b as laststart. And We use laststart as the pointer + to the first element of the charset here. + In other words, laststart[i] indicates charset[i]. */ + laststart = b; + + /* We test `*p == '^' twice, instead of using an if + statement, so we only need one BUF_PUSH. */ + BUF_PUSH (*p == '^' ? charset_not : charset); + if (*p == '^') + p++; + + /* Push the length of char_classes, the length of + collating_symbols, the length of equivalence_classes, the + length of char_ranges and the length of chars. */ + BUF_PUSH_3 (0, 0, 0); + BUF_PUSH_2 (0, 0); + + /* Remember the first position in the bracket expression. */ + p1 = p; + + /* charset_not matches newline according to a syntax bit. */ + if ((re_opcode_t) b[-6] == charset_not + && (syntax & RE_HAT_LISTS_NOT_NEWLINE)) + { + BUF_PUSH('\n'); + laststart[5]++; /* Update the length of characters */ + } + + /* Read in characters and ranges, setting map bits. */ + for (;;) + { + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + PATFETCH (c); + + /* \ might escape characters inside [...] and [^...]. */ + if ((syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + BUF_PUSH(c1); + laststart[5]++; /* Update the length of chars */ + range_start = c1; + continue; + } + + /* Could be the end of the bracket expression. If it's + not (i.e., when the bracket expression is `[]' so + far), the ']' character bit gets set way below. */ + if (c == ']' && p != p1 + 1) + break; + + /* Look ahead to see if it's a range when the last thing + was a character class. */ + if (had_char_class && c == '-' && *p != ']') + FREE_STACK_RETURN (REG_ERANGE); + + /* Look ahead to see if it's a range when the last thing + was a character: if this is a hyphen not at the + beginning or the end of a list, then it's the range + operator. */ + if (c == '-' + && !(p - 2 >= pattern && p[-2] == '[') + && !(p - 3 >= pattern && p[-3] == '[' && p[-2] == '^') + && *p != ']') + { + reg_errcode_t ret; + /* Allocate the space for range_start and range_end. */ + GET_BUFFER_SPACE (2); + /* Update the pointer to indicate end of buffer. */ + b += 2; + ret = wcs_compile_range (range_start, &p, pend, translate, + syntax, b, laststart); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + else if (p[0] == '-' && p[1] != ']') + { /* This handles ranges made up of characters only. */ + reg_errcode_t ret; + + /* Move past the `-'. */ + PATFETCH (c1); + /* Allocate the space for range_start and range_end. */ + GET_BUFFER_SPACE (2); + /* Update the pointer to indicate end of buffer. */ + b += 2; + ret = wcs_compile_range (c, &p, pend, translate, syntax, b, + laststart); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + + /* See if we're at the beginning of a possible character + class. */ + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == ':') + { /* Leave room for the null. */ + char str[CHAR_CLASS_MAX_LENGTH + 1]; + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[:'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == ':' && *p == ']') || p == pend) + break; + if (c1 < CHAR_CLASS_MAX_LENGTH) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + /* If isn't a word bracketed by `[:' and `:]': + undo the ending character, the letters, and leave + the leading `:' and `[' (but store them as character). */ + if (c == ':' && *p == ']') + { + wctype_t wt; + uintptr_t alignedp; + + /* Query the character class as wctype_t. */ + wt = IS_CHAR_CLASS (str); + if (wt == 0) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + /* Allocate the space for character class. */ + GET_BUFFER_SPACE(CHAR_CLASS_SIZE); + /* Update the pointer to indicate end of buffer. */ + b += CHAR_CLASS_SIZE; + /* Move data which follow character classes + not to violate the data. */ + insert_space(CHAR_CLASS_SIZE, + laststart + 6 + laststart[1], + b - 1); + alignedp = ((uintptr_t)(laststart + 6 + laststart[1]) + + __alignof__(wctype_t) - 1) + & ~(uintptr_t)(__alignof__(wctype_t) - 1); + /* Store the character class. */ + *((wctype_t*)alignedp) = wt; + /* Update length of char_classes */ + laststart[1] += CHAR_CLASS_SIZE; + + had_char_class = true; + } + else + { + c1++; + while (c1--) + PATUNFETCH; + BUF_PUSH ('['); + BUF_PUSH (':'); + laststart[5] += 2; /* Update the length of characters */ + range_start = ':'; + had_char_class = false; + } + } + else if (syntax & RE_CHAR_CLASSES && c == '[' && (*p == '=' + || *p == '.')) + { + CHAR_T str[128]; /* Should be large enough. */ + CHAR_T delim = *p; /* '=' or '.' */ +# ifdef _LIBC + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[=' or '[[.'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == delim && *p == ']') || p == pend) + break; + if (c1 < sizeof (str) - 1) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + if (c == delim && *p == ']' && str[0] != '\0') + { + unsigned int i, offset; + /* If we have no collation data we use the default + collation in which each character is in a class + by itself. It also means that ASCII is the + character set and therefore we cannot have character + with more than one byte in the multibyte + representation. */ + + /* If not defined _LIBC, we push the name and + `\0' for the sake of matching performance. */ + int datasize = c1 + 1; + +# ifdef _LIBC + int32_t idx = 0; + if (nrules == 0) +# endif + { + if (c1 != 1) + FREE_STACK_RETURN (REG_ECOLLATE); + } +# ifdef _LIBC + else + { + const int32_t *table; + const int32_t *weights; + const int32_t *extra; + const int32_t *indirect; + wint_t *cp; + + /* This #include defines a local function! */ +# include + + if(delim == '=') + { + /* We push the index for equivalence class. */ + cp = (wint_t*)str; + + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_TABLEWC); + weights = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_WEIGHTWC); + extra = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_EXTRAWC); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_INDIRECTWC); + + idx = findidx ((const wint_t**)&cp); + if (idx == 0 || cp < (wint_t*) str + c1) + /* This is no valid character. */ + FREE_STACK_RETURN (REG_ECOLLATE); + + str[0] = (wchar_t)idx; + } + else /* delim == '.' */ + { + /* We push collation sequence value + for collating symbol. */ + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + char char_str[c1]; + + /* We have to convert the name to a single-byte + string. This is possible since the names + consist of ASCII characters and the internal + representation is UCS4. */ + for (i = 0; i < c1; ++i) + char_str[i] = str[i]; + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing table. */ + hash = elem_hash (char_str, c1); + + idx = 0; + elem = hash % table_size; + second = hash % (table_size - 2); + while (symb_table[2 * elem] != 0) + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && c1 == extra[symb_table[2 * elem + 1]] + && memcmp (char_str, + &extra[symb_table[2 * elem + 1] + + 1], c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + + if (symb_table[2 * elem] != 0) + { + /* Compute the index of the byte sequence + in the table. */ + idx += 1 + extra[idx]; + /* Adjust for the alignment. */ + idx = (idx + 3) & ~3; + + str[0] = (wchar_t) idx + 4; + } + else if (symb_table[2 * elem] == 0 && c1 == 1) + { + /* No valid character. Match it as a + single byte character. */ + had_char_class = false; + BUF_PUSH(str[0]); + /* Update the length of characters */ + laststart[5]++; + range_start = str[0]; + + /* Throw away the ] at the end of the + collating symbol. */ + PATFETCH (c); + /* exit from the switch block. */ + continue; + } + else + FREE_STACK_RETURN (REG_ECOLLATE); + } + datasize = 1; + } +# endif + /* Throw away the ] at the end of the equivalence + class (or collating symbol). */ + PATFETCH (c); + + /* Allocate the space for the equivalence class + (or collating symbol) (and '\0' if needed). */ + GET_BUFFER_SPACE(datasize); + /* Update the pointer to indicate end of buffer. */ + b += datasize; + + if (delim == '=') + { /* equivalence class */ + /* Calculate the offset of char_ranges, + which is next to equivalence_classes. */ + offset = laststart[1] + laststart[2] + + laststart[3] +6; + /* Insert space. */ + insert_space(datasize, laststart + offset, b - 1); + + /* Write the equivalence_class and \0. */ + for (i = 0 ; i < datasize ; i++) + laststart[offset + i] = str[i]; + + /* Update the length of equivalence_classes. */ + laststart[3] += datasize; + had_char_class = true; + } + else /* delim == '.' */ + { /* collating symbol */ + /* Calculate the offset of the equivalence_classes, + which is next to collating_symbols. */ + offset = laststart[1] + laststart[2] + 6; + /* Insert space and write the collationg_symbol + and \0. */ + insert_space(datasize, laststart + offset, b-1); + for (i = 0 ; i < datasize ; i++) + laststart[offset + i] = str[i]; + + /* In re_match_2_internal if range_start < -1, we + assume -range_start is the offset of the + collating symbol which is specified as + the character of the range start. So we assign + -(laststart[1] + laststart[2] + 6) to + range_start. */ + range_start = -(laststart[1] + laststart[2] + 6); + /* Update the length of collating_symbol. */ + laststart[2] += datasize; + had_char_class = false; + } + } + else + { + c1++; + while (c1--) + PATUNFETCH; + BUF_PUSH ('['); + BUF_PUSH (delim); + laststart[5] += 2; /* Update the length of characters */ + range_start = delim; + had_char_class = false; + } + } + else + { + had_char_class = false; + BUF_PUSH(c); + laststart[5]++; /* Update the length of characters */ + range_start = c; + } + } + +#else /* BYTE */ + /* Ensure that we have enough space to push a charset: the + opcode, the length count, and the bitset; 34 bytes in all. */ + GET_BUFFER_SPACE (34); + + laststart = b; + + /* We test `*p == '^' twice, instead of using an if + statement, so we only need one BUF_PUSH. */ + BUF_PUSH (*p == '^' ? charset_not : charset); + if (*p == '^') + p++; + + /* Remember the first position in the bracket expression. */ + p1 = p; + + /* Push the number of bytes in the bitmap. */ + BUF_PUSH ((1 << BYTEWIDTH) / BYTEWIDTH); + + /* Clear the whole map. */ + bzero (b, (1 << BYTEWIDTH) / BYTEWIDTH); + + /* charset_not matches newline according to a syntax bit. */ + if ((re_opcode_t) b[-2] == charset_not + && (syntax & RE_HAT_LISTS_NOT_NEWLINE)) + SET_LIST_BIT ('\n'); + + /* Read in characters and ranges, setting map bits. */ + for (;;) + { + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + PATFETCH (c); + + /* \ might escape characters inside [...] and [^...]. */ + if ((syntax & RE_BACKSLASH_ESCAPE_IN_LISTS) && c == '\\') + { + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + PATFETCH (c1); + SET_LIST_BIT (c1); + range_start = c1; + continue; + } + + /* Could be the end of the bracket expression. If it's + not (i.e., when the bracket expression is `[]' so + far), the ']' character bit gets set way below. */ + if (c == ']' && p != p1 + 1) + break; + + /* Look ahead to see if it's a range when the last thing + was a character class. */ + if (had_char_class && c == '-' && *p != ']') + FREE_STACK_RETURN (REG_ERANGE); + + /* Look ahead to see if it's a range when the last thing + was a character: if this is a hyphen not at the + beginning or the end of a list, then it's the range + operator. */ + if (c == '-' + && !(p - 2 >= pattern && p[-2] == '[') + && !(p - 3 >= pattern && p[-3] == '[' && p[-2] == '^') + && *p != ']') + { + reg_errcode_t ret + = byte_compile_range (range_start, &p, pend, translate, + syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + + else if (p[0] == '-' && p[1] != ']') + { /* This handles ranges made up of characters only. */ + reg_errcode_t ret; + + /* Move past the `-'. */ + PATFETCH (c1); + + ret = byte_compile_range (c, &p, pend, translate, syntax, b); + if (ret != REG_NOERROR) FREE_STACK_RETURN (ret); + range_start = 0xffffffff; + } + + /* See if we're at the beginning of a possible character + class. */ + + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == ':') + { /* Leave room for the null. */ + char str[CHAR_CLASS_MAX_LENGTH + 1]; + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[:'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == ':' && *p == ']') || p == pend) + break; + if (c1 < CHAR_CLASS_MAX_LENGTH) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + /* If isn't a word bracketed by `[:' and `:]': + undo the ending character, the letters, and leave + the leading `:' and `[' (but set bits for them). */ + if (c == ':' && *p == ']') + { +# if defined _LIBC || WIDE_CHAR_SUPPORT + boolean is_lower = STREQ (str, "lower"); + boolean is_upper = STREQ (str, "upper"); + wctype_t wt; + int ch; + + wt = IS_CHAR_CLASS (str); + if (wt == 0) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (ch = 0; ch < 1 << BYTEWIDTH; ++ch) + { + if (iswctype (btowc (ch), wt)) + SET_LIST_BIT (ch); + + if (translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + + had_char_class = true; +# else + int ch; + boolean is_alnum = STREQ (str, "alnum"); + boolean is_alpha = STREQ (str, "alpha"); + boolean is_blank = STREQ (str, "blank"); + boolean is_cntrl = STREQ (str, "cntrl"); + boolean is_digit = STREQ (str, "digit"); + boolean is_graph = STREQ (str, "graph"); + boolean is_lower = STREQ (str, "lower"); + boolean is_print = STREQ (str, "print"); + boolean is_punct = STREQ (str, "punct"); + boolean is_space = STREQ (str, "space"); + boolean is_upper = STREQ (str, "upper"); + boolean is_xdigit = STREQ (str, "xdigit"); + + if (!IS_CHAR_CLASS (str)) + FREE_STACK_RETURN (REG_ECTYPE); + + /* Throw away the ] at the end of the character + class. */ + PATFETCH (c); + + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (ch = 0; ch < 1 << BYTEWIDTH; ch++) + { + /* This was split into 3 if's to + avoid an arbitrary limit in some compiler. */ + if ( (is_alnum && ISALNUM (ch)) + || (is_alpha && ISALPHA (ch)) + || (is_blank && ISBLANK (ch)) + || (is_cntrl && ISCNTRL (ch))) + SET_LIST_BIT (ch); + if ( (is_digit && ISDIGIT (ch)) + || (is_graph && ISGRAPH (ch)) + || (is_lower && ISLOWER (ch)) + || (is_print && ISPRINT (ch))) + SET_LIST_BIT (ch); + if ( (is_punct && ISPUNCT (ch)) + || (is_space && ISSPACE (ch)) + || (is_upper && ISUPPER (ch)) + || (is_xdigit && ISXDIGIT (ch))) + SET_LIST_BIT (ch); + if ( translate && (is_upper || is_lower) + && (ISUPPER (ch) || ISLOWER (ch))) + SET_LIST_BIT (ch); + } + had_char_class = true; +# endif /* libc || wctype.h */ + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT (':'); + range_start = ':'; + had_char_class = false; + } + } + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == '=') + { + unsigned char str[MB_LEN_MAX + 1]; +# ifdef _LIBC + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[='. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == '=' && *p == ']') || p == pend) + break; + if (c1 < MB_LEN_MAX) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + if (c == '=' && *p == ']' && str[0] != '\0') + { + /* If we have no collation data we use the default + collation in which each character is in a class + by itself. It also means that ASCII is the + character set and therefore we cannot have character + with more than one byte in the multibyte + representation. */ +# ifdef _LIBC + if (nrules == 0) +# endif + { + if (c1 != 1) + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Set the bit for the character. */ + SET_LIST_BIT (str[0]); + } +# ifdef _LIBC + else + { + /* Try to match the byte sequence in `str' against + those known to the collate implementation. + First find out whether the bytes in `str' are + actually from exactly one character. */ + const int32_t *table; + const unsigned char *weights; + const unsigned char *extra; + const int32_t *indirect; + int32_t idx; + const unsigned char *cp = str; + int ch; + + /* This #include defines a local function! */ +# include + + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB); + weights = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTMB); + + idx = findidx (&cp); + if (idx == 0 || cp < str + c1) + /* This is no valid character. */ + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Now we have to go throught the whole table + and find all characters which have the same + first level weight. + + XXX Note that this is not entirely correct. + we would have to match multibyte sequences + but this is not possible with the current + implementation. */ + for (ch = 1; ch < 256; ++ch) + /* XXX This test would have to be changed if we + would allow matching multibyte sequences. */ + if (table[ch] > 0) + { + int32_t idx2 = table[ch]; + size_t len = weights[idx2]; + + /* Test whether the lenghts match. */ + if (weights[idx] == len) + { + /* They do. New compare the bytes of + the weight. */ + size_t cnt = 0; + + while (cnt < len + && (weights[idx + 1 + cnt] + == weights[idx2 + 1 + cnt])) + ++cnt; + + if (cnt == len) + /* They match. Mark the character as + acceptable. */ + SET_LIST_BIT (ch); + } + } + } +# endif + had_char_class = true; + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT ('='); + range_start = '='; + had_char_class = false; + } + } + else if (syntax & RE_CHAR_CLASSES && c == '[' && *p == '.') + { + unsigned char str[128]; /* Should be large enough. */ +# ifdef _LIBC + uint32_t nrules = + _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif + + PATFETCH (c); + c1 = 0; + + /* If pattern is `[[.'. */ + if (p == pend) FREE_STACK_RETURN (REG_EBRACK); + + for (;;) + { + PATFETCH (c); + if ((c == '.' && *p == ']') || p == pend) + break; + if (c1 < sizeof (str)) + str[c1++] = c; + else + /* This is in any case an invalid class name. */ + str[0] = '\0'; + } + str[c1] = '\0'; + + if (c == '.' && *p == ']' && str[0] != '\0') + { + /* If we have no collation data we use the default + collation in which each character is the name + for its own class which contains only the one + character. It also means that ASCII is the + character set and therefore we cannot have character + with more than one byte in the multibyte + representation. */ +# ifdef _LIBC + if (nrules == 0) +# endif + { + if (c1 != 1) + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Set the bit for the character. */ + SET_LIST_BIT (str[0]); + range_start = ((const unsigned char *) str)[0]; + } +# ifdef _LIBC + else + { + /* Try to match the byte sequence in `str' against + those known to the collate implementation. + First find out whether the bytes in `str' are + actually from exactly one character. */ + int32_t table_size; + const int32_t *symb_table; + const unsigned char *extra; + int32_t idx; + int32_t elem; + int32_t second; + int32_t hash; + + table_size = + _NL_CURRENT_WORD (LC_COLLATE, + _NL_COLLATE_SYMB_HASH_SIZEMB); + symb_table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_TABLEMB); + extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_SYMB_EXTRAMB); + + /* Locate the character in the hashing table. */ + hash = elem_hash (str, c1); + + idx = 0; + elem = hash % table_size; + second = hash % (table_size - 2); + while (symb_table[2 * elem] != 0) + { + /* First compare the hashing value. */ + if (symb_table[2 * elem] == hash + && c1 == extra[symb_table[2 * elem + 1]] + && memcmp (str, + &extra[symb_table[2 * elem + 1] + + 1], + c1) == 0) + { + /* Yep, this is the entry. */ + idx = symb_table[2 * elem + 1]; + idx += 1 + extra[idx]; + break; + } + + /* Next entry. */ + elem += second; + } + + if (symb_table[2 * elem] == 0) + /* This is no valid character. */ + FREE_STACK_RETURN (REG_ECOLLATE); + + /* Throw away the ] at the end of the equivalence + class. */ + PATFETCH (c); + + /* Now add the multibyte character(s) we found + to the accept list. + + XXX Note that this is not entirely correct. + we would have to match multibyte sequences + but this is not possible with the current + implementation. Also, we have to match + collating symbols, which expand to more than + one file, as a whole and not allow the + individual bytes. */ + c1 = extra[idx++]; + if (c1 == 1) + range_start = extra[idx]; + while (c1-- > 0) + { + SET_LIST_BIT (extra[idx]); + ++idx; + } + } +# endif + had_char_class = false; + } + else + { + c1++; + while (c1--) + PATUNFETCH; + SET_LIST_BIT ('['); + SET_LIST_BIT ('.'); + range_start = '.'; + had_char_class = false; + } + } + else + { + had_char_class = false; + SET_LIST_BIT (c); + range_start = c; + } + } + + /* Discard any (non)matching list bytes that are all 0 at the + end of the map. Decrease the map-length byte too. */ + while ((int) b[-1] > 0 && b[b[-1] - 1] == 0) + b[-1]--; + b += b[-1]; +#endif /* WCHAR */ + } + break; + + + case '(': + if (syntax & RE_NO_BK_PARENS) + goto handle_open; + else + goto normal_char; + + + case ')': + if (syntax & RE_NO_BK_PARENS) + goto handle_close; + else + goto normal_char; + + + case '\n': + if (syntax & RE_NEWLINE_ALT) + goto handle_alt; + else + goto normal_char; + + + case '|': + if (syntax & RE_NO_BK_VBAR) + goto handle_alt; + else + goto normal_char; + + + case '{': + if (syntax & RE_INTERVALS && syntax & RE_NO_BK_BRACES) + goto handle_interval; + else + goto normal_char; + + + case '\\': + if (p == pend) FREE_STACK_RETURN (REG_EESCAPE); + + /* Do not translate the character after the \, so that we can + distinguish, e.g., \B from \b, even if we normally would + translate, e.g., B to b. */ + PATFETCH_RAW (c); + + switch (c) + { + case '(': + if (syntax & RE_NO_BK_PARENS) + goto normal_backslash; + + handle_open: + bufp->re_nsub++; + regnum++; + + if (COMPILE_STACK_FULL) + { + RETALLOC (compile_stack.stack, compile_stack.size << 1, + compile_stack_elt_t); + if (compile_stack.stack == NULL) return REG_ESPACE; + + compile_stack.size <<= 1; + } + + /* These are the values to restore when we hit end of this + group. They are all relative offsets, so that if the + whole pattern moves because of realloc, they will still + be valid. */ + COMPILE_STACK_TOP.begalt_offset = begalt - COMPILED_BUFFER_VAR; + COMPILE_STACK_TOP.fixup_alt_jump + = fixup_alt_jump ? fixup_alt_jump - COMPILED_BUFFER_VAR + 1 : 0; + COMPILE_STACK_TOP.laststart_offset = b - COMPILED_BUFFER_VAR; + COMPILE_STACK_TOP.regnum = regnum; + + /* We will eventually replace the 0 with the number of + groups inner to this one. But do not push a + start_memory for groups beyond the last one we can + represent in the compiled pattern. */ + if (regnum <= MAX_REGNUM) + { + COMPILE_STACK_TOP.inner_group_offset = b + - COMPILED_BUFFER_VAR + 2; + BUF_PUSH_3 (start_memory, regnum, 0); + } + + compile_stack.avail++; + + fixup_alt_jump = 0; + laststart = 0; + begalt = b; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + break; + + + case ')': + if (syntax & RE_NO_BK_PARENS) goto normal_backslash; + + if (COMPILE_STACK_EMPTY) + { + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_backslash; + else + FREE_STACK_RETURN (REG_ERPAREN); + } + + handle_close: + if (fixup_alt_jump) + { /* Push a dummy failure point at the end of the + alternative for a possible future + `pop_failure_jump' to pop. See comments at + `push_dummy_failure' in `re_match_2'. */ + BUF_PUSH (push_dummy_failure); + + /* We allocated space for this jump when we assigned + to `fixup_alt_jump', in the `handle_alt' case below. */ + STORE_JUMP (jump_past_alt, fixup_alt_jump, b - 1); + } + + /* See similar code for backslashed left paren above. */ + if (COMPILE_STACK_EMPTY) + { + if (syntax & RE_UNMATCHED_RIGHT_PAREN_ORD) + goto normal_char; + else + FREE_STACK_RETURN (REG_ERPAREN); + } + + /* Since we just checked for an empty stack above, this + ``can't happen''. */ + assert (compile_stack.avail != 0); + { + /* We don't just want to restore into `regnum', because + later groups should continue to be numbered higher, + as in `(ab)c(de)' -- the second group is #2. */ + regnum_t this_group_regnum; + + compile_stack.avail--; + begalt = COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.begalt_offset; + fixup_alt_jump + = COMPILE_STACK_TOP.fixup_alt_jump + ? COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.fixup_alt_jump - 1 + : 0; + laststart = COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.laststart_offset; + this_group_regnum = COMPILE_STACK_TOP.regnum; + /* If we've reached MAX_REGNUM groups, then this open + won't actually generate any code, so we'll have to + clear pending_exact explicitly. */ + pending_exact = 0; + + /* We're at the end of the group, so now we know how many + groups were inside this one. */ + if (this_group_regnum <= MAX_REGNUM) + { + UCHAR_T *inner_group_loc + = COMPILED_BUFFER_VAR + COMPILE_STACK_TOP.inner_group_offset; + + *inner_group_loc = regnum - this_group_regnum; + BUF_PUSH_3 (stop_memory, this_group_regnum, + regnum - this_group_regnum); + } + } + break; + + + case '|': /* `\|'. */ + if (syntax & RE_LIMITED_OPS || syntax & RE_NO_BK_VBAR) + goto normal_backslash; + handle_alt: + if (syntax & RE_LIMITED_OPS) + goto normal_char; + + /* Insert before the previous alternative a jump which + jumps to this alternative if the former fails. */ + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (on_failure_jump, begalt, + b + 2 + 2 * OFFSET_ADDRESS_SIZE); + pending_exact = 0; + b += 1 + OFFSET_ADDRESS_SIZE; + + /* The alternative before this one has a jump after it + which gets executed if it gets matched. Adjust that + jump so it will jump to this alternative's analogous + jump (put in below, which in turn will jump to the next + (if any) alternative's such jump, etc.). The last such + jump jumps to the correct final destination. A picture: + _____ _____ + | | | | + | v | v + a | b | c + + If we are at `b', then fixup_alt_jump right now points to a + three-byte space after `a'. We'll put in the jump, set + fixup_alt_jump to right after `b', and leave behind three + bytes which we'll fill in when we get to after `c'. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + /* Mark and leave space for a jump after this alternative, + to be filled in later either by next alternative or + when know we're at the end of a series of alternatives. */ + fixup_alt_jump = b; + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + b += 1 + OFFSET_ADDRESS_SIZE; + + laststart = 0; + begalt = b; + break; + + + case '{': + /* If \{ is a literal. */ + if (!(syntax & RE_INTERVALS) + /* If we're at `\{' and it's not the open-interval + operator. */ + || (syntax & RE_NO_BK_BRACES)) + goto normal_backslash; + + handle_interval: + { + /* If got here, then the syntax allows intervals. */ + + /* At least (most) this many matches must be made. */ + int lower_bound = -1, upper_bound = -1; + + /* Place in the uncompiled pattern (i.e., just after + the '{') to go back to if the interval is invalid. */ + const CHAR_T *beg_interval = p; + + if (p == pend) + goto invalid_interval; + + GET_UNSIGNED_NUMBER (lower_bound); + + if (c == ',') + { + GET_UNSIGNED_NUMBER (upper_bound); + if (upper_bound < 0) + upper_bound = RE_DUP_MAX; + } + else + /* Interval such as `{1}' => match exactly once. */ + upper_bound = lower_bound; + + if (! (0 <= lower_bound && lower_bound <= upper_bound)) + goto invalid_interval; + + if (!(syntax & RE_NO_BK_BRACES)) + { + if (c != '\\' || p == pend) + goto invalid_interval; + PATFETCH (c); + } + + if (c != '}') + goto invalid_interval; + + /* If it's invalid to have no preceding re. */ + if (!laststart) + { + if (syntax & RE_CONTEXT_INVALID_OPS + && !(syntax & RE_INVALID_INTERVAL_ORD)) + FREE_STACK_RETURN (REG_BADRPT); + else if (syntax & RE_CONTEXT_INDEP_OPS) + laststart = b; + else + goto unfetch_interval; + } + + /* We just parsed a valid interval. */ + + if (RE_DUP_MAX < upper_bound) + FREE_STACK_RETURN (REG_BADBR); + + /* If the upper bound is zero, don't want to succeed at + all; jump from `laststart' to `b + 3', which will be + the end of the buffer after we insert the jump. */ + /* ifdef WCHAR, 'b + 1 + OFFSET_ADDRESS_SIZE' + instead of 'b + 3'. */ + if (upper_bound == 0) + { + GET_BUFFER_SPACE (1 + OFFSET_ADDRESS_SIZE); + INSERT_JUMP (jump, laststart, b + 1 + + OFFSET_ADDRESS_SIZE); + b += 1 + OFFSET_ADDRESS_SIZE; + } + + /* Otherwise, we have a nontrivial interval. When + we're all done, the pattern will look like: + set_number_at + set_number_at + succeed_n + + jump_n + (The upper bound and `jump_n' are omitted if + `upper_bound' is 1, though.) */ + else + { /* If the upper bound is > 1, we need to insert + more at the end of the loop. */ + unsigned nbytes = 2 + 4 * OFFSET_ADDRESS_SIZE + + (upper_bound > 1) * (2 + 4 * OFFSET_ADDRESS_SIZE); + + GET_BUFFER_SPACE (nbytes); + + /* Initialize lower bound of the `succeed_n', even + though it will be set during matching by its + attendant `set_number_at' (inserted next), + because `re_compile_fastmap' needs to know. + Jump to the `jump_n' we might insert below. */ + INSERT_JUMP2 (succeed_n, laststart, + b + 1 + 2 * OFFSET_ADDRESS_SIZE + + (upper_bound > 1) * (1 + 2 * OFFSET_ADDRESS_SIZE) + , lower_bound); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + + /* Code to initialize the lower bound. Insert + before the `succeed_n'. The `5' is the last two + bytes of this `set_number_at', plus 3 bytes of + the following `succeed_n'. */ + /* ifdef WCHAR, The '1+2*OFFSET_ADDRESS_SIZE' + is the 'set_number_at', plus '1+OFFSET_ADDRESS_SIZE' + of the following `succeed_n'. */ + PREFIX(insert_op2) (set_number_at, laststart, 1 + + 2 * OFFSET_ADDRESS_SIZE, lower_bound, b); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + + if (upper_bound > 1) + { /* More than one repetition is allowed, so + append a backward jump to the `succeed_n' + that starts this interval. + + When we've reached this during matching, + we'll have matched the interval once, so + jump back only `upper_bound - 1' times. */ + STORE_JUMP2 (jump_n, b, laststart + + 2 * OFFSET_ADDRESS_SIZE + 1, + upper_bound - 1); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + + /* The location we want to set is the second + parameter of the `jump_n'; that is `b-2' as + an absolute address. `laststart' will be + the `set_number_at' we're about to insert; + `laststart+3' the number to set, the source + for the relative address. But we are + inserting into the middle of the pattern -- + so everything is getting moved up by 5. + Conclusion: (b - 2) - (laststart + 3) + 5, + i.e., b - laststart. + + We insert this at the beginning of the loop + so that if we fail during matching, we'll + reinitialize the bounds. */ + PREFIX(insert_op2) (set_number_at, laststart, + b - laststart, + upper_bound - 1, b); + b += 1 + 2 * OFFSET_ADDRESS_SIZE; + } + } + pending_exact = 0; + break; + + invalid_interval: + if (!(syntax & RE_INVALID_INTERVAL_ORD)) + FREE_STACK_RETURN (p == pend ? REG_EBRACE : REG_BADBR); + unfetch_interval: + /* Match the characters as literals. */ + p = beg_interval; + c = '{'; + if (syntax & RE_NO_BK_BRACES) + goto normal_char; + else + goto normal_backslash; + } + +#ifdef emacs + /* There is no way to specify the before_dot and after_dot + operators. rms says this is ok. --karl */ + case '=': + BUF_PUSH (at_dot); + break; + + case 's': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (syntaxspec, syntax_spec_code[c]); + break; + + case 'S': + laststart = b; + PATFETCH (c); + BUF_PUSH_2 (notsyntaxspec, syntax_spec_code[c]); + break; +#endif /* emacs */ + + + case 'w': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (wordchar); + break; + + + case 'W': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + laststart = b; + BUF_PUSH (notwordchar); + break; + + + case '<': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbeg); + break; + + case '>': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordend); + break; + + case 'b': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (wordbound); + break; + + case 'B': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (notwordbound); + break; + + case '`': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (begbuf); + break; + + case '\'': + if (syntax & RE_NO_GNU_OPS) + goto normal_char; + BUF_PUSH (endbuf); + break; + + case '1': case '2': case '3': case '4': case '5': + case '6': case '7': case '8': case '9': + if (syntax & RE_NO_BK_REFS) + goto normal_char; + + c1 = c - '0'; + + if (c1 > regnum) + FREE_STACK_RETURN (REG_ESUBREG); + + /* Can't back reference to a subexpression if inside of it. */ + if (group_in_compile_stack (compile_stack, (regnum_t) c1)) + goto normal_char; + + laststart = b; + BUF_PUSH_2 (duplicate, c1); + break; + + + case '+': + case '?': + if (syntax & RE_BK_PLUS_QM) + goto handle_plus; + else + goto normal_backslash; + + default: + normal_backslash: + /* You might think it would be useful for \ to mean + not to translate; but if we don't translate it + it will never match anything. */ + c = TRANSLATE (c); + goto normal_char; + } + break; + + + default: + /* Expects the character in `c'. */ + normal_char: + /* If no exactn currently being built. */ + if (!pending_exact +#ifdef WCHAR + /* If last exactn handle binary(or character) and + new exactn handle character(or binary). */ + || is_exactn_bin != is_binary[p - 1 - pattern] +#endif /* WCHAR */ + + /* If last exactn not at current position. */ + || pending_exact + *pending_exact + 1 != b + + /* We have only one byte following the exactn for the count. */ + || *pending_exact == (1 << BYTEWIDTH) - 1 + + /* If followed by a repetition operator. */ + || *p == '*' || *p == '^' + || ((syntax & RE_BK_PLUS_QM) + ? *p == '\\' && (p[1] == '+' || p[1] == '?') + : (*p == '+' || *p == '?')) + || ((syntax & RE_INTERVALS) + && ((syntax & RE_NO_BK_BRACES) + ? *p == '{' + : (p[0] == '\\' && p[1] == '{')))) + { + /* Start building a new exactn. */ + + laststart = b; + +#ifdef WCHAR + /* Is this exactn binary data or character? */ + is_exactn_bin = is_binary[p - 1 - pattern]; + if (is_exactn_bin) + BUF_PUSH_2 (exactn_bin, 0); + else + BUF_PUSH_2 (exactn, 0); +#else + BUF_PUSH_2 (exactn, 0); +#endif /* WCHAR */ + pending_exact = b - 1; + } + + BUF_PUSH (c); + (*pending_exact)++; + break; + } /* switch (c) */ + } /* while p != pend */ + + + /* Through the pattern now. */ + + if (fixup_alt_jump) + STORE_JUMP (jump_past_alt, fixup_alt_jump, b); + + if (!COMPILE_STACK_EMPTY) + FREE_STACK_RETURN (REG_EPAREN); + + /* If we don't want backtracking, force success + the first time we reach the end of the compiled pattern. */ + if (syntax & RE_NO_POSIX_BACKTRACKING) + BUF_PUSH (succeed); + +#ifdef WCHAR + free (pattern); + free (mbs_offset); + free (is_binary); +#endif + free (compile_stack.stack); + + /* We have succeeded; set the length of the buffer. */ +#ifdef WCHAR + bufp->used = (uintptr_t) b - (uintptr_t) COMPILED_BUFFER_VAR; +#else + bufp->used = b - bufp->buffer; +#endif + +#ifdef DEBUG + if (debug) + { + DEBUG_PRINT1 ("\nCompiled pattern: \n"); + PREFIX(print_compiled_pattern) (bufp); + } +#endif /* DEBUG */ + +#ifndef MATCH_MAY_ALLOCATE + /* Initialize the failure stack to the largest possible stack. This + isn't necessary unless we're trying to avoid calling alloca in + the search and match routines. */ + { + int num_regs = bufp->re_nsub + 1; + + /* Since DOUBLE_FAIL_STACK refuses to double only if the current size + is strictly greater than re_max_failures, the largest possible stack + is 2 * re_max_failures failure points. */ + if (fail_stack.size < (2 * re_max_failures * MAX_FAILURE_ITEMS)) + { + fail_stack.size = (2 * re_max_failures * MAX_FAILURE_ITEMS); + +# ifdef emacs + if (! fail_stack.stack) + fail_stack.stack + = (PREFIX(fail_stack_elt_t) *) xmalloc (fail_stack.size + * sizeof (PREFIX(fail_stack_elt_t))); + else + fail_stack.stack + = (PREFIX(fail_stack_elt_t) *) xrealloc (fail_stack.stack, + (fail_stack.size + * sizeof (PREFIX(fail_stack_elt_t)))); +# else /* not emacs */ + if (! fail_stack.stack) + fail_stack.stack + = malloc (fail_stack.size * sizeof (PREFIX(fail_stack_elt_t))); + else + fail_stack.stack + = realloc (fail_stack.stack, + fail_stack.size * sizeof (PREFIX(fail_stack_elt_t))); +# endif /* not emacs */ + } + + PREFIX(regex_grow_registers) (num_regs); + } +#endif /* not MATCH_MAY_ALLOCATE */ + + return REG_NOERROR; +} /* regex_compile */ + +/* Subroutines for `regex_compile'. */ + +/* Store OP at LOC followed by two-byte integer parameter ARG. */ +/* ifdef WCHAR, integer parameter is 1 wchar_t. */ + +static void +PREFIX(store_op1) (re_opcode_t op, UCHAR_T *loc, int arg) +{ + *loc = (UCHAR_T) op; + STORE_NUMBER (loc + 1, arg); +} + + +/* Like `store_op1', but for two two-byte parameters ARG1 and ARG2. */ +/* ifdef WCHAR, integer parameter is 1 wchar_t. */ + +static void +PREFIX(store_op2) (re_opcode_t op, UCHAR_T *loc, int arg1, int arg2) +{ + *loc = (UCHAR_T) op; + STORE_NUMBER (loc + 1, arg1); + STORE_NUMBER (loc + 1 + OFFSET_ADDRESS_SIZE, arg2); +} + + +/* Copy the bytes from LOC to END to open up three bytes of space at LOC + for OP followed by two-byte integer parameter ARG. */ +/* ifdef WCHAR, integer parameter is 1 wchar_t. */ + +static void +PREFIX(insert_op1) (re_opcode_t op, UCHAR_T *loc, int arg, UCHAR_T *end) +{ + register UCHAR_T *pfrom = end; + register UCHAR_T *pto = end + 1 + OFFSET_ADDRESS_SIZE; + + while (pfrom != loc) + *--pto = *--pfrom; + + PREFIX(store_op1) (op, loc, arg); +} + + +/* Like `insert_op1', but for two two-byte parameters ARG1 and ARG2. */ +/* ifdef WCHAR, integer parameter is 1 wchar_t. */ + +static void +PREFIX(insert_op2) (re_opcode_t op, UCHAR_T *loc, int arg1, int arg2, + UCHAR_T *end) +{ + register UCHAR_T *pfrom = end; + register UCHAR_T *pto = end + 1 + 2 * OFFSET_ADDRESS_SIZE; + + while (pfrom != loc) + *--pto = *--pfrom; + + PREFIX(store_op2) (op, loc, arg1, arg2); +} + + +/* P points to just after a ^ in PATTERN. Return true if that ^ comes + after an alternative or a begin-subexpression. We assume there is at + least one character before the ^. */ + +static boolean +PREFIX(at_begline_loc_p) (const CHAR_T *pattern, const CHAR_T *p, + reg_syntax_t syntax) +{ + const CHAR_T *prev = p - 2; + boolean prev_prev_backslash = prev > pattern && prev[-1] == '\\'; + + return + /* After a subexpression? */ + (*prev == '(' && (syntax & RE_NO_BK_PARENS || prev_prev_backslash)) + /* After an alternative? */ + || (*prev == '|' && (syntax & RE_NO_BK_VBAR || prev_prev_backslash)); +} + + +/* The dual of at_begline_loc_p. This one is for $. We assume there is + at least one character after the $, i.e., `P < PEND'. */ + +static boolean +PREFIX(at_endline_loc_p) (const CHAR_T *p, const CHAR_T *pend, + reg_syntax_t syntax) +{ + const CHAR_T *next = p; + boolean next_backslash = *next == '\\'; + const CHAR_T *next_next = p + 1 < pend ? p + 1 : 0; + + return + /* Before a subexpression? */ + (syntax & RE_NO_BK_PARENS ? *next == ')' + : next_backslash && next_next && *next_next == ')') + /* Before an alternative? */ + || (syntax & RE_NO_BK_VBAR ? *next == '|' + : next_backslash && next_next && *next_next == '|'); +} + +#else /* not INSIDE_RECURSION */ + +/* Returns true if REGNUM is in one of COMPILE_STACK's elements and + false if it's not. */ + +static boolean +group_in_compile_stack (compile_stack_type compile_stack, + regnum_t regnum) +{ + int this_element; + + for (this_element = compile_stack.avail - 1; + this_element >= 0; + this_element--) + if (compile_stack.stack[this_element].regnum == regnum) + return true; + + return false; +} +#endif /* not INSIDE_RECURSION */ + +#ifdef INSIDE_RECURSION + +#ifdef WCHAR +/* This insert space, which size is "num", into the pattern at "loc". + "end" must point the end of the allocated buffer. */ +static void +insert_space (int num, CHAR_T *loc, CHAR_T *end) +{ + register CHAR_T *pto = end; + register CHAR_T *pfrom = end - num; + + while (pfrom >= loc) + *pto-- = *pfrom--; +} +#endif /* WCHAR */ + +#ifdef WCHAR +static reg_errcode_t +wcs_compile_range (CHAR_T range_start_char, + const CHAR_T **p_ptr, const CHAR_T *pend, + RE_TRANSLATE_TYPE translate, reg_syntax_t syntax, + CHAR_T *b, CHAR_T *char_set) +{ + const CHAR_T *p = *p_ptr; + CHAR_T range_start, range_end; + reg_errcode_t ret; +# ifdef _LIBC + uint32_t nrules; + uint32_t start_val, end_val; +# endif + if (p == pend) + return REG_ERANGE; + +# ifdef _LIBC + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); + if (nrules != 0) + { + const char *collseq = (const char *) _NL_CURRENT(LC_COLLATE, + _NL_COLLATE_COLLSEQWC); + const unsigned char *extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + + if (range_start_char < -1) + { + /* range_start is a collating symbol. */ + int32_t *wextra; + /* Retreive the index and get collation sequence value. */ + wextra = (int32_t*)(extra + char_set[-range_start_char]); + start_val = wextra[1 + *wextra]; + } + else + start_val = collseq_table_lookup(collseq, TRANSLATE(range_start_char)); + + end_val = collseq_table_lookup (collseq, TRANSLATE (p[0])); + + /* Report an error if the range is empty and the syntax prohibits + this. */ + ret = ((syntax & RE_NO_EMPTY_RANGES) + && (start_val > end_val))? REG_ERANGE : REG_NOERROR; + + /* Insert space to the end of the char_ranges. */ + insert_space(2, b - char_set[5] - 2, b - 1); + *(b - char_set[5] - 2) = (wchar_t)start_val; + *(b - char_set[5] - 1) = (wchar_t)end_val; + char_set[4]++; /* ranges_index */ + } + else +# endif + { + range_start = (range_start_char >= 0)? TRANSLATE (range_start_char): + range_start_char; + range_end = TRANSLATE (p[0]); + /* Report an error if the range is empty and the syntax prohibits + this. */ + ret = ((syntax & RE_NO_EMPTY_RANGES) + && (range_start > range_end))? REG_ERANGE : REG_NOERROR; + + /* Insert space to the end of the char_ranges. */ + insert_space(2, b - char_set[5] - 2, b - 1); + *(b - char_set[5] - 2) = range_start; + *(b - char_set[5] - 1) = range_end; + char_set[4]++; /* ranges_index */ + } + /* Have to increment the pointer into the pattern string, so the + caller isn't still at the ending character. */ + (*p_ptr)++; + + return ret; +} +#else /* BYTE */ +/* Read the ending character of a range (in a bracket expression) from the + uncompiled pattern *P_PTR (which ends at PEND). We assume the + starting character is in `P[-2]'. (`P[-1]' is the character `-'.) + Then we set the translation of all bits between the starting and + ending characters (inclusive) in the compiled pattern B. + + Return an error code. + + We use these short variable names so we can use the same macros as + `regex_compile' itself. */ + +static reg_errcode_t +byte_compile_range (unsigned int range_start_char, + const char **p_ptr, const char *pend, + RE_TRANSLATE_TYPE translate, reg_syntax_t syntax, + unsigned char *b) +{ + unsigned this_char; + const char *p = *p_ptr; + reg_errcode_t ret; +# if _LIBC + const unsigned char *collseq; + unsigned int start_colseq; + unsigned int end_colseq; +# else + unsigned end_char; +# endif + + if (p == pend) + return REG_ERANGE; + + /* Have to increment the pointer into the pattern string, so the + caller isn't still at the ending character. */ + (*p_ptr)++; + + /* Report an error if the range is empty and the syntax prohibits this. */ + ret = syntax & RE_NO_EMPTY_RANGES ? REG_ERANGE : REG_NOERROR; + +# if _LIBC + collseq = (const unsigned char *) _NL_CURRENT (LC_COLLATE, + _NL_COLLATE_COLLSEQMB); + + start_colseq = collseq[(unsigned char) TRANSLATE (range_start_char)]; + end_colseq = collseq[(unsigned char) TRANSLATE (p[0])]; + for (this_char = 0; this_char <= (unsigned char) -1; ++this_char) + { + unsigned int this_colseq = collseq[(unsigned char) TRANSLATE (this_char)]; + + if (start_colseq <= this_colseq && this_colseq <= end_colseq) + { + SET_LIST_BIT (TRANSLATE (this_char)); + ret = REG_NOERROR; + } + } +# else + /* Here we see why `this_char' has to be larger than an `unsigned + char' -- we would otherwise go into an infinite loop, since all + characters <= 0xff. */ + range_start_char = TRANSLATE (range_start_char); + /* TRANSLATE(p[0]) is casted to char (not unsigned char) in TRANSLATE, + and some compilers cast it to int implicitly, so following for_loop + may fall to (almost) infinite loop. + e.g. If translate[p[0]] = 0xff, end_char may equals to 0xffffffff. + To avoid this, we cast p[0] to unsigned int and truncate it. */ + end_char = ((unsigned)TRANSLATE(p[0]) & ((1 << BYTEWIDTH) - 1)); + + for (this_char = range_start_char; this_char <= end_char; ++this_char) + { + SET_LIST_BIT (TRANSLATE (this_char)); + ret = REG_NOERROR; + } +# endif + + return ret; +} +#endif /* WCHAR */ + +/* re_compile_fastmap computes a ``fastmap'' for the compiled pattern in + BUFP. A fastmap records which of the (1 << BYTEWIDTH) possible + characters can start a string that matches the pattern. This fastmap + is used by re_search to skip quickly over impossible starting points. + + The caller must supply the address of a (1 << BYTEWIDTH)-byte data + area as BUFP->fastmap. + + We set the `fastmap', `fastmap_accurate', and `can_be_null' fields in + the pattern buffer. + + Returns 0 if we succeed, -2 if an internal error. */ + +#ifdef WCHAR +/* local function for re_compile_fastmap. + truncate wchar_t character to char. */ + +static unsigned char +truncate_wchar (CHAR_T c) +{ + unsigned char buf[MB_CUR_MAX]; + mbstate_t state; + int retval; + memset (&state, '\0', sizeof (state)); + retval = wcrtomb (buf, c, &state); + return retval > 0 ? buf[0] : (unsigned char) c; +} +#endif /* WCHAR */ + +static int +PREFIX(re_compile_fastmap) (struct re_pattern_buffer *bufp) +{ + int j, k; +#ifdef MATCH_MAY_ALLOCATE + PREFIX(fail_stack_type) fail_stack; +#endif +#ifndef REGEX_MALLOC + char *destination; +#endif + + register char *fastmap = bufp->fastmap; + +#ifdef WCHAR + /* We need to cast pattern to (wchar_t*), because we casted this compiled + pattern to (char*) in regex_compile. */ + UCHAR_T *pattern = (UCHAR_T*)bufp->buffer; + register UCHAR_T *pend = (UCHAR_T*) (bufp->buffer + bufp->used); +#else /* BYTE */ + UCHAR_T *pattern = bufp->buffer; + register UCHAR_T *pend = pattern + bufp->used; +#endif /* WCHAR */ + UCHAR_T *p = pattern; + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#endif + + /* Assume that each path through the pattern can be null until + proven otherwise. We set this false at the bottom of switch + statement, to which we get only if a particular path doesn't + match the empty string. */ + boolean path_can_be_null = true; + + /* We aren't doing a `succeed_n' to begin with. */ + boolean succeed_n_p = false; + + assert (fastmap != NULL && p != NULL); + + INIT_FAIL_STACK (); + bzero (fastmap, 1 << BYTEWIDTH); /* Assume nothing's valid. */ + bufp->fastmap_accurate = 1; /* It will be when we're done. */ + bufp->can_be_null = 0; + + while (1) + { + if (p == pend || *p == succeed) + { + /* We have reached the (effective) end of pattern. */ + if (!FAIL_STACK_EMPTY ()) + { + bufp->can_be_null |= path_can_be_null; + + /* Reset for next path. */ + path_can_be_null = true; + + p = fail_stack.stack[--fail_stack.avail].pointer; + + continue; + } + else + break; + } + + /* We should never be about to go beyond the end of the pattern. */ + assert (p < pend); + + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { + + /* I guess the idea here is to simply not bother with a fastmap + if a backreference is used, since it's too hard to figure out + the fastmap for the corresponding group. Setting + `can_be_null' stops `re_search_2' from using the fastmap, so + that is all we do. */ + case duplicate: + bufp->can_be_null = 1; + goto done; + + + /* Following are the cases which match a character. These end + with `break'. */ + +#ifdef WCHAR + case exactn: + fastmap[truncate_wchar(p[1])] = 1; + break; +#else /* BYTE */ + case exactn: + fastmap[p[1]] = 1; + break; +#endif /* WCHAR */ +#ifdef MBS_SUPPORT + case exactn_bin: + fastmap[p[1]] = 1; + break; +#endif + +#ifdef WCHAR + /* It is hard to distinguish fastmap from (multi byte) characters + which depends on current locale. */ + case charset: + case charset_not: + case wordchar: + case notwordchar: + bufp->can_be_null = 1; + goto done; +#else /* BYTE */ + case charset: + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH))) + fastmap[j] = 1; + break; + + + case charset_not: + /* Chars beyond end of map must be allowed. */ + for (j = *p * BYTEWIDTH; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + for (j = *p++ * BYTEWIDTH - 1; j >= 0; j--) + if (!(p[j / BYTEWIDTH] & (1 << (j % BYTEWIDTH)))) + fastmap[j] = 1; + break; + + + case wordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == Sword) + fastmap[j] = 1; + break; + + + case notwordchar: + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != Sword) + fastmap[j] = 1; + break; +#endif /* WCHAR */ + + case anychar: + { + int fastmap_newline = fastmap['\n']; + + /* `.' matches anything ... */ + for (j = 0; j < (1 << BYTEWIDTH); j++) + fastmap[j] = 1; + + /* ... except perhaps newline. */ + if (!(bufp->syntax & RE_DOT_NEWLINE)) + fastmap['\n'] = fastmap_newline; + + /* Return if we have already set `can_be_null'; if we have, + then the fastmap is irrelevant. Something's wrong here. */ + else if (bufp->can_be_null) + goto done; + + /* Otherwise, have to check alternative paths. */ + break; + } + +#ifdef emacs + case syntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) == (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + case notsyntaxspec: + k = *p++; + for (j = 0; j < (1 << BYTEWIDTH); j++) + if (SYNTAX (j) != (enum syntaxcode) k) + fastmap[j] = 1; + break; + + + /* All cases after this match the empty string. These end with + `continue'. */ + + + case before_dot: + case at_dot: + case after_dot: + continue; +#endif /* emacs */ + + + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbound: + case notwordbound: + case wordbeg: + case wordend: + case push_dummy_failure: + continue; + + + case jump_n: + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case jump_past_alt: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + if (j > 0) + continue; + + /* Jump backward implies we just went through the body of a + loop and matched nothing. Opcode jumped to should be + `on_failure_jump' or `succeed_n'. Just treat it like an + ordinary jump. For a * loop, it has pushed its failure + point already; if so, discard that as redundant. */ + if ((re_opcode_t) *p != on_failure_jump + && (re_opcode_t) *p != succeed_n) + continue; + + p++; + EXTRACT_NUMBER_AND_INCR (j, p); + p += j; + + /* If what's on the stack is where we are now, pop it. */ + if (!FAIL_STACK_EMPTY () + && fail_stack.stack[fail_stack.avail - 1].pointer == p) + fail_stack.avail--; + + continue; + + + case on_failure_jump: + case on_failure_keep_string_jump: + handle_on_failure_jump: + EXTRACT_NUMBER_AND_INCR (j, p); + + /* For some patterns, e.g., `(a?)?', `p+j' here points to the + end of the pattern. We don't want to push such a point, + since when we restore it above, entering the switch will + increment `p' past the end of the pattern. We don't need + to push such a point since we obviously won't find any more + fastmap entries beyond `pend'. Such a pattern can match + the null string, though. */ + if (p + j < pend) + { + if (!PUSH_PATTERN_OP (p + j, fail_stack)) + { + RESET_FAIL_STACK (); + return -2; + } + } + else + bufp->can_be_null = 1; + + if (succeed_n_p) + { + EXTRACT_NUMBER_AND_INCR (k, p); /* Skip the n. */ + succeed_n_p = false; + } + + continue; + + + case succeed_n: + /* Get to the number of times to succeed. */ + p += OFFSET_ADDRESS_SIZE; + + /* Increment p past the n for when k != 0. */ + EXTRACT_NUMBER_AND_INCR (k, p); + if (k == 0) + { + p -= 2 * OFFSET_ADDRESS_SIZE; + succeed_n_p = true; /* Spaghetti code alert. */ + goto handle_on_failure_jump; + } + continue; + + + case set_number_at: + p += 2 * OFFSET_ADDRESS_SIZE; + continue; + + + case start_memory: + case stop_memory: + p += 2; + continue; + + + default: + abort (); /* We have listed all the cases. */ + } /* switch *p++ */ + + /* Getting here means we have found the possible starting + characters for one path of the pattern -- and that the empty + string does not match. We need not follow this path further. + Instead, look at the next alternative (remembered on the + stack), or quit if no more. The test at the top of the loop + does these things. */ + path_can_be_null = false; + p = pend; + } /* while p */ + + /* Set `can_be_null' for the last path (also the first path, if the + pattern is empty). */ + bufp->can_be_null |= path_can_be_null; + + done: + RESET_FAIL_STACK (); + return 0; +} + +#else /* not INSIDE_RECURSION */ + +int +re_compile_fastmap (struct re_pattern_buffer *bufp) +{ +# ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + return wcs_re_compile_fastmap(bufp); + else +# endif + return byte_re_compile_fastmap(bufp); +} /* re_compile_fastmap */ +#ifdef _LIBC +weak_alias (__re_compile_fastmap, re_compile_fastmap) +#endif + + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use + this memory for recording register information. STARTS and ENDS + must be allocated using the malloc library routine, and must each + be at least NUM_REGS * sizeof (regoff_t) bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ + +void +re_set_registers (struct re_pattern_buffer *bufp, + struct re_registers *regs, + unsigned int num_regs, + regoff_t *starts, regoff_t *ends) +{ + if (num_regs) + { + bufp->regs_allocated = REGS_REALLOCATE; + regs->num_regs = num_regs; + regs->start = starts; + regs->end = ends; + } + else + { + bufp->regs_allocated = REGS_UNALLOCATED; + regs->num_regs = 0; + regs->start = regs->end = (regoff_t *) 0; + } +} +#ifdef _LIBC +weak_alias (__re_set_registers, re_set_registers) +#endif + +/* Searching routines. */ + +/* Like re_search_2, below, but only one string is specified, and + doesn't let you say where to stop matching. */ + +int +re_search (struct re_pattern_buffer *bufp, + const char *string, + int size, int startpos, int range, + struct re_registers *regs) +{ + return re_search_2 (bufp, NULL, 0, string, size, startpos, range, + regs, size); +} +#ifdef _LIBC +weak_alias (__re_search, re_search) +#endif + + +/* Using the compiled pattern in BUFP->buffer, first tries to match the + virtual concatenation of STRING1 and STRING2, starting first at index + STARTPOS, then at STARTPOS + 1, and so on. + + STRING1 and STRING2 have length SIZE1 and SIZE2, respectively. + + RANGE is how far to scan while trying to match. RANGE = 0 means try + only at STARTPOS; in general, the last start tried is STARTPOS + + RANGE. + + In REGS, return the indices of the virtual concatenation of STRING1 + and STRING2 that matched the entire BUFP->buffer and its contained + subexpressions. + + Do not consider matching one past the index STOP in the virtual + concatenation of STRING1 and STRING2. + + We return either the position in the strings at which the match was + found, -1 if no match, or -2 if error (such as failure + stack overflow). */ + +int +re_search_2 (struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int startpos, int range, + struct re_registers *regs, + int stop) +{ +# ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + return wcs_re_search_2 (bufp, string1, size1, string2, size2, startpos, + range, regs, stop); + else +# endif + return byte_re_search_2 (bufp, string1, size1, string2, size2, startpos, + range, regs, stop); +} /* re_search_2 */ +#ifdef _LIBC +weak_alias (__re_search_2, re_search_2) +#endif + +#endif /* not INSIDE_RECURSION */ + +#ifdef INSIDE_RECURSION + +#ifdef MATCH_MAY_ALLOCATE +# define FREE_VAR(var) if (var) REGEX_FREE (var); var = NULL +#else +# define FREE_VAR(var) if (var) free (var); var = NULL +#endif + +#ifdef WCHAR +# define MAX_ALLOCA_SIZE 2000 + +# define FREE_WCS_BUFFERS() \ + do { \ + if (size1 > MAX_ALLOCA_SIZE) \ + { \ + free (wcs_string1); \ + free (mbs_offset1); \ + } \ + else \ + { \ + FREE_VAR (wcs_string1); \ + FREE_VAR (mbs_offset1); \ + } \ + if (size2 > MAX_ALLOCA_SIZE) \ + { \ + free (wcs_string2); \ + free (mbs_offset2); \ + } \ + else \ + { \ + FREE_VAR (wcs_string2); \ + FREE_VAR (mbs_offset2); \ + } \ + } while (0) + +#endif + + +static int +PREFIX(re_search_2) (struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int startpos, int range, + struct re_registers *regs, + int stop) +{ + int val; + register char *fastmap = bufp->fastmap; + register RE_TRANSLATE_TYPE translate = bufp->translate; + int total_size = size1 + size2; + int endpos = startpos + range; +#ifdef WCHAR + /* We need wchar_t* buffers correspond to cstring1, cstring2. */ + wchar_t *wcs_string1 = NULL, *wcs_string2 = NULL; + /* We need the size of wchar_t buffers correspond to csize1, csize2. */ + int wcs_size1 = 0, wcs_size2 = 0; + /* offset buffer for optimizatoin. See convert_mbs_to_wc. */ + int *mbs_offset1 = NULL, *mbs_offset2 = NULL; + /* They hold whether each wchar_t is binary data or not. */ + char *is_binary = NULL; +#endif /* WCHAR */ + + /* Check for out-of-range STARTPOS. */ + if (startpos < 0 || startpos > total_size) + return -1; + + /* Fix up RANGE if it might eventually take us outside + the virtual concatenation of STRING1 and STRING2. + Make sure we won't move STARTPOS below 0 or above TOTAL_SIZE. */ + if (endpos < 0) + range = 0 - startpos; + else if (endpos > total_size) + range = total_size - startpos; + + /* If the search isn't to be a backwards one, don't waste time in a + search for a pattern that must be anchored. */ + if (bufp->used > 0 && range > 0 + && ((re_opcode_t) bufp->buffer[0] == begbuf + /* `begline' is like `begbuf' if it cannot match at newlines. */ + || ((re_opcode_t) bufp->buffer[0] == begline + && !bufp->newline_anchor))) + { + if (startpos > 0) + return -1; + else + range = 1; + } + +#ifdef emacs + /* In a forward search for something that starts with \=. + don't keep searching past point. */ + if (bufp->used > 0 && (re_opcode_t) bufp->buffer[0] == at_dot && range > 0) + { + range = PT - startpos; + if (range <= 0) + return -1; + } +#endif /* emacs */ + + /* Update the fastmap now if not correct already. */ + if (fastmap && !bufp->fastmap_accurate) + if (re_compile_fastmap (bufp) == -2) + return -2; + +#ifdef WCHAR + /* Allocate wchar_t array for wcs_string1 and wcs_string2 and + fill them with converted string. */ + if (size1 != 0) + { + if (size1 > MAX_ALLOCA_SIZE) + { + wcs_string1 = TALLOC (size1 + 1, CHAR_T); + mbs_offset1 = TALLOC (size1 + 1, int); + is_binary = TALLOC (size1 + 1, char); + } + else + { + wcs_string1 = REGEX_TALLOC (size1 + 1, CHAR_T); + mbs_offset1 = REGEX_TALLOC (size1 + 1, int); + is_binary = REGEX_TALLOC (size1 + 1, char); + } + if (!wcs_string1 || !mbs_offset1 || !is_binary) + { + if (size1 > MAX_ALLOCA_SIZE) + { + free (wcs_string1); + free (mbs_offset1); + free (is_binary); + } + else + { + FREE_VAR (wcs_string1); + FREE_VAR (mbs_offset1); + FREE_VAR (is_binary); + } + return -2; + } + wcs_size1 = convert_mbs_to_wcs(wcs_string1, string1, size1, + mbs_offset1, is_binary); + wcs_string1[wcs_size1] = L'\0'; /* for a sentinel */ + if (size1 > MAX_ALLOCA_SIZE) + free (is_binary); + else + FREE_VAR (is_binary); + } + if (size2 != 0) + { + if (size2 > MAX_ALLOCA_SIZE) + { + wcs_string2 = TALLOC (size2 + 1, CHAR_T); + mbs_offset2 = TALLOC (size2 + 1, int); + is_binary = TALLOC (size2 + 1, char); + } + else + { + wcs_string2 = REGEX_TALLOC (size2 + 1, CHAR_T); + mbs_offset2 = REGEX_TALLOC (size2 + 1, int); + is_binary = REGEX_TALLOC (size2 + 1, char); + } + if (!wcs_string2 || !mbs_offset2 || !is_binary) + { + FREE_WCS_BUFFERS (); + if (size2 > MAX_ALLOCA_SIZE) + free (is_binary); + else + FREE_VAR (is_binary); + return -2; + } + wcs_size2 = convert_mbs_to_wcs(wcs_string2, string2, size2, + mbs_offset2, is_binary); + wcs_string2[wcs_size2] = L'\0'; /* for a sentinel */ + if (size2 > MAX_ALLOCA_SIZE) + free (is_binary); + else + FREE_VAR (is_binary); + } +#endif /* WCHAR */ + + + /* Loop through the string, looking for a place to start matching. */ + for (;;) + { + /* If a fastmap is supplied, skip quickly over characters that + cannot be the start of a match. If the pattern can match the + null string, however, we don't need to skip characters; we want + the first null string. */ + if (fastmap && startpos < total_size && !bufp->can_be_null) + { + if (range > 0) /* Searching forwards. */ + { + register const char *d; + register int lim = 0; + int irange = range; + + if (startpos < size1 && startpos + range >= size1) + lim = range - (size1 - startpos); + + d = (startpos >= size1 ? string2 - size1 : string1) + startpos; + + /* Written out as an if-else to avoid testing `translate' + inside the loop. */ + if (translate) + while (range > lim + && !fastmap[(unsigned char) + translate[(unsigned char) *d++]]) + range--; + else + while (range > lim && !fastmap[(unsigned char) *d++]) + range--; + + startpos += irange - range; + } + else /* Searching backwards. */ + { + register CHAR_T c = (size1 == 0 || startpos >= size1 + ? string2[startpos - size1] + : string1[startpos]); + + if (!fastmap[(unsigned char) TRANSLATE (c)]) + goto advance; + } + } + + /* If can't match the null string, and that's all we have left, fail. */ + if (range >= 0 && startpos == total_size && fastmap + && !bufp->can_be_null) + { +#ifdef WCHAR + FREE_WCS_BUFFERS (); +#endif + return -1; + } + +#ifdef WCHAR + val = wcs_re_match_2_internal (bufp, string1, size1, string2, + size2, startpos, regs, stop, + wcs_string1, wcs_size1, + wcs_string2, wcs_size2, + mbs_offset1, mbs_offset2); +#else /* BYTE */ + val = byte_re_match_2_internal (bufp, string1, size1, string2, + size2, startpos, regs, stop); +#endif /* BYTE */ + +#ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +#endif + + if (val >= 0) + { +#ifdef WCHAR + FREE_WCS_BUFFERS (); +#endif + return startpos; + } + + if (val == -2) + { +#ifdef WCHAR + FREE_WCS_BUFFERS (); +#endif + return -2; + } + + advance: + if (!range) + break; + else if (range > 0) + { + range--; + startpos++; + } + else + { + range++; + startpos--; + } + } +#ifdef WCHAR + FREE_WCS_BUFFERS (); +#endif + return -1; +} + +#ifdef WCHAR +/* This converts PTR, a pointer into one of the search wchar_t strings + `string1' and `string2' into an multibyte string offset from the + beginning of that string. We use mbs_offset to optimize. + See convert_mbs_to_wcs. */ +# define POINTER_TO_OFFSET(ptr) \ + (FIRST_STRING_P (ptr) \ + ? ((regoff_t)(mbs_offset1 != NULL? mbs_offset1[(ptr)-string1] : 0)) \ + : ((regoff_t)((mbs_offset2 != NULL? mbs_offset2[(ptr)-string2] : 0) \ + + csize1))) +#else /* BYTE */ +/* This converts PTR, a pointer into one of the search strings `string1' + and `string2' into an offset from the beginning of that string. */ +# define POINTER_TO_OFFSET(ptr) \ + (FIRST_STRING_P (ptr) \ + ? ((regoff_t) ((ptr) - string1)) \ + : ((regoff_t) ((ptr) - string2 + size1))) +#endif /* WCHAR */ + +/* Macros for dealing with the split strings in re_match_2. */ + +#define MATCHING_IN_FIRST_STRING (dend == end_match_1) + +/* Call before fetching a character with *d. This switches over to + string2 if necessary. */ +#define PREFETCH() \ + while (d == dend) \ + { \ + /* End of string2 => fail. */ \ + if (dend == end_match_2) \ + goto fail; \ + /* End of string1 => advance to string2. */ \ + d = string2; \ + dend = end_match_2; \ + } + +/* Test if at very beginning or at very end of the virtual concatenation + of `string1' and `string2'. If only one string, it's `string2'. */ +#define AT_STRINGS_BEG(d) ((d) == (size1 ? string1 : string2) || !size2) +#define AT_STRINGS_END(d) ((d) == end2) + + +/* Test if D points to a character which is word-constituent. We have + two special cases to check for: if past the end of string1, look at + the first character in string2; and if before the beginning of + string2, look at the last character in string1. */ +#ifdef WCHAR +/* Use internationalized API instead of SYNTAX. */ +# define WORDCHAR_P(d) \ + (iswalnum ((wint_t)((d) == end1 ? *string2 \ + : (d) == string2 - 1 ? *(end1 - 1) : *(d))) != 0 \ + || ((d) == end1 ? *string2 \ + : (d) == string2 - 1 ? *(end1 - 1) : *(d)) == L'_') +#else /* BYTE */ +# define WORDCHAR_P(d) \ + (SYNTAX ((d) == end1 ? *string2 \ + : (d) == string2 - 1 ? *(end1 - 1) : *(d)) \ + == Sword) +#endif /* WCHAR */ + +/* Disabled due to a compiler bug -- see comment at case wordbound */ +#if 0 +/* Test if the character before D and the one at D differ with respect + to being word-constituent. */ +#define AT_WORD_BOUNDARY(d) \ + (AT_STRINGS_BEG (d) || AT_STRINGS_END (d) \ + || WORDCHAR_P (d - 1) != WORDCHAR_P (d)) +#endif + +/* Free everything we malloc. */ +#ifdef MATCH_MAY_ALLOCATE +# ifdef WCHAR +# define FREE_VARIABLES() \ + do { \ + REGEX_FREE_STACK (fail_stack.stack); \ + FREE_VAR (regstart); \ + FREE_VAR (regend); \ + FREE_VAR (old_regstart); \ + FREE_VAR (old_regend); \ + FREE_VAR (best_regstart); \ + FREE_VAR (best_regend); \ + FREE_VAR (reg_info); \ + FREE_VAR (reg_dummy); \ + FREE_VAR (reg_info_dummy); \ + if (!cant_free_wcs_buf) \ + { \ + FREE_VAR (string1); \ + FREE_VAR (string2); \ + FREE_VAR (mbs_offset1); \ + FREE_VAR (mbs_offset2); \ + } \ + } while (0) +# else /* BYTE */ +# define FREE_VARIABLES() \ + do { \ + REGEX_FREE_STACK (fail_stack.stack); \ + FREE_VAR (regstart); \ + FREE_VAR (regend); \ + FREE_VAR (old_regstart); \ + FREE_VAR (old_regend); \ + FREE_VAR (best_regstart); \ + FREE_VAR (best_regend); \ + FREE_VAR (reg_info); \ + FREE_VAR (reg_dummy); \ + FREE_VAR (reg_info_dummy); \ + } while (0) +# endif /* WCHAR */ +#else +# ifdef WCHAR +# define FREE_VARIABLES() \ + do { \ + if (!cant_free_wcs_buf) \ + { \ + FREE_VAR (string1); \ + FREE_VAR (string2); \ + FREE_VAR (mbs_offset1); \ + FREE_VAR (mbs_offset2); \ + } \ + } while (0) +# else /* BYTE */ +# define FREE_VARIABLES() ((void)0) /* Do nothing! But inhibit gcc warning. */ +# endif /* WCHAR */ +#endif /* not MATCH_MAY_ALLOCATE */ + +/* These values must meet several constraints. They must not be valid + register values; since we have a limit of 255 registers (because + we use only one byte in the pattern for the register number), we can + use numbers larger than 255. They must differ by 1, because of + NUM_FAILURE_ITEMS above. And the value for the lowest register must + be larger than the value for the highest register, so we do not try + to actually save any registers when none are active. */ +#define NO_HIGHEST_ACTIVE_REG (1 << BYTEWIDTH) +#define NO_LOWEST_ACTIVE_REG (NO_HIGHEST_ACTIVE_REG + 1) + +#else /* not INSIDE_RECURSION */ +/* Matching routines. */ + +#ifndef emacs /* Emacs never uses this. */ +/* re_match is like re_match_2 except it takes only a single string. */ + +int +re_match (struct re_pattern_buffer *bufp, + const char *string, + int size, int pos, + struct re_registers *regs) +{ + int result; +# ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + result = wcs_re_match_2_internal (bufp, NULL, 0, string, size, + pos, regs, size, + NULL, 0, NULL, 0, NULL, NULL); + else +# endif + result = byte_re_match_2_internal (bufp, NULL, 0, string, size, + pos, regs, size); +# ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +# endif + return result; +} +# ifdef _LIBC +weak_alias (__re_match, re_match) +# endif +#endif /* not emacs */ + +#endif /* not INSIDE_RECURSION */ + +#ifdef INSIDE_RECURSION +static boolean PREFIX(group_match_null_string_p) (UCHAR_T **p, + UCHAR_T *end, + PREFIX(register_info_type) *reg_info); +static boolean PREFIX(alt_match_null_string_p) (UCHAR_T *p, + UCHAR_T *end, + PREFIX(register_info_type) *reg_info); +static boolean PREFIX(common_op_match_null_string_p) (UCHAR_T **p, + UCHAR_T *end, + PREFIX(register_info_type) *reg_info); +static int PREFIX(bcmp_translate) (const CHAR_T *s1, const CHAR_T *s2, + int len, char *translate); +#else /* not INSIDE_RECURSION */ + +/* re_match_2 matches the compiled pattern in BUFP against the + the (virtual) concatenation of STRING1 and STRING2 (of length SIZE1 + and SIZE2, respectively). We start matching at POS, and stop + matching at STOP. + + If REGS is non-null and the `no_sub' field of BUFP is nonzero, we + store offsets for the substring each group matched in REGS. See the + documentation for exactly how many groups we fill. + + We return -1 if no match, -2 if an internal error (such as the + failure stack overflowing). Otherwise, we return the length of the + matched substring. */ + +int +re_match_2 (struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int pos, struct re_registers *regs, + int stop) +{ + int result; +# ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + result = wcs_re_match_2_internal (bufp, string1, size1, string2, size2, + pos, regs, stop, + NULL, 0, NULL, 0, NULL, NULL); + else +# endif + result = byte_re_match_2_internal (bufp, string1, size1, string2, size2, + pos, regs, stop); + +#ifndef REGEX_MALLOC +# ifdef C_ALLOCA + alloca (0); +# endif +#endif + return result; +} +#ifdef _LIBC +weak_alias (__re_match_2, re_match_2) +#endif + +#endif /* not INSIDE_RECURSION */ + +#ifdef INSIDE_RECURSION + +#ifdef WCHAR + +/* This check the substring (from 0, to length) of the multibyte string, + to which offset_buffer correspond. And count how many wchar_t_characters + the substring occupy. We use offset_buffer to optimization. + See convert_mbs_to_wcs. */ + +static int +count_mbs_length (int *offset_buffer, int length) +{ + int upper, lower; + + /* Check whether the size is valid. */ + if (length < 0) + return -1; + + if (offset_buffer == NULL) + return 0; + + /* If there are no multibyte character, offset_buffer[i] == i. + Optmize for this case. */ + if (offset_buffer[length] == length) + return length; + + /* Set up upper with length. (because for all i, offset_buffer[i] >= i) */ + upper = length; + lower = 0; + + while (true) + { + int middle = (lower + upper) / 2; + if (middle == lower || middle == upper) + break; + if (offset_buffer[middle] > length) + upper = middle; + else if (offset_buffer[middle] < length) + lower = middle; + else + return middle; + } + + return -1; +} +#endif /* WCHAR */ + +/* This is a separate function so that we can force an alloca cleanup + afterwards. */ +#ifdef WCHAR +static int +wcs_re_match_2_internal (struct re_pattern_buffer *bufp, + const char *cstring1, int csize1, + const char *cstring2, int csize2, + int pos, + struct re_registers *regs, + int stop, + /* string1 == string2 == NULL means + string1/2, size1/2 and mbs_offset1/2 need + setting up in this function. */ + /* We need wchar_t * buffers corresponding to + cstring1, cstring2. */ + wchar_t *string1, int size1, + wchar_t *string2, int size2, + /* Offset buffer for optimization. See + convert_mbs_to_wc. */ + int *mbs_offset1, + int *mbs_offset2) +#else /* BYTE */ +static int +byte_re_match_2_internal (struct re_pattern_buffer *bufp, + const char *string1, int size1, + const char *string2, int size2, + int pos, + struct re_registers *regs, + int stop) +#endif /* BYTE */ +{ + /* General temporaries. */ + int mcnt; + UCHAR_T *p1; +#ifdef WCHAR + /* They hold whether each wchar_t is binary data or not. */ + char *is_binary = NULL; + /* If true, we can't free string1/2, mbs_offset1/2. */ + int cant_free_wcs_buf = 1; +#endif /* WCHAR */ + + /* Just past the end of the corresponding string. */ + const CHAR_T *end1, *end2; + + /* Pointers into string1 and string2, just past the last characters in + each to consider matching. */ + const CHAR_T *end_match_1, *end_match_2; + + /* Where we are in the data, and the end of the current string. */ + const CHAR_T *d, *dend; + + /* Where we are in the pattern, and the end of the pattern. */ +#ifdef WCHAR + UCHAR_T *pattern, *p; + register UCHAR_T *pend; +#else /* BYTE */ + UCHAR_T *p = bufp->buffer; + register UCHAR_T *pend = p + bufp->used; +#endif /* WCHAR */ + + /* Mark the opcode just after a start_memory, so we can test for an + empty subpattern when we get to the stop_memory. */ + UCHAR_T *just_past_start_mem = 0; + + /* We use this to map every character in the string. */ + RE_TRANSLATE_TYPE translate = bufp->translate; + + /* Failure point stack. Each place that can handle a failure further + down the line pushes a failure point on this stack. It consists of + restart, regend, and reg_info for all registers corresponding to + the subexpressions we're currently inside, plus the number of such + registers, and, finally, two char *'s. The first char * is where + to resume scanning the pattern; the second one is where to resume + scanning the strings. If the latter is zero, the failure point is + a ``dummy''; if a failure happens and the failure point is a dummy, + it gets discarded and the next next one is tried. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + PREFIX(fail_stack_type) fail_stack; +#endif +#ifdef DEBUG + static unsigned failure_id; + unsigned nfailure_points_pushed = 0, nfailure_points_popped = 0; +#endif + +#ifdef REL_ALLOC + /* This holds the pointer to the failure stack, when + it is allocated relocatably. */ + fail_stack_elt_t *failure_stack_ptr; +#endif + + /* We fill all the registers internally, independent of what we + return, for use in backreferences. The number here includes + an element for register zero. */ + size_t num_regs = bufp->re_nsub + 1; + + /* The currently active registers. */ + active_reg_t lowest_active_reg = NO_LOWEST_ACTIVE_REG; + active_reg_t highest_active_reg = NO_HIGHEST_ACTIVE_REG; + + /* Information on the contents of registers. These are pointers into + the input strings; they record just what was matched (on this + attempt) by a subexpression part of the pattern, that is, the + regnum-th regstart pointer points to where in the pattern we began + matching and the regnum-th regend points to right after where we + stopped matching the regnum-th subexpression. (The zeroth register + keeps track of what the whole pattern matches.) */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_T **regstart, **regend; +#endif + + /* If a group that's operated upon by a repetition operator fails to + match anything, then the register for its start will need to be + restored because it will have been set to wherever in the string we + are when we last see its open-group operator. Similarly for a + register's end. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_T **old_regstart, **old_regend; +#endif + + /* The is_active field of reg_info helps us keep track of which (possibly + nested) subexpressions we are currently in. The matched_something + field of reg_info[reg_num] helps us tell whether or not we have + matched any of the pattern so far this time through the reg_num-th + subexpression. These two fields get reset each time through any + loop their register is in. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, this is global. */ + PREFIX(register_info_type) *reg_info; +#endif + + /* The following record the register info as found in the above + variables when we find a match better than any we've seen before. + This happens as we backtrack through the failure points, which in + turn happens only if we have not yet matched the entire string. */ + unsigned best_regs_set = false; +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_T **best_regstart, **best_regend; +#endif + + /* Logically, this is `best_regend[0]'. But we don't want to have to + allocate space for that if we're not allocating space for anything + else (see below). Also, we never need info about register 0 for + any of the other register vectors, and it seems rather a kludge to + treat `best_regend' differently than the rest. So we keep track of + the end of the best match so far in a separate variable. We + initialize this to NULL so that when we backtrack the first time + and need to test it, it's not garbage. */ + const CHAR_T *match_end = NULL; + + /* This helps SET_REGS_MATCHED avoid doing redundant work. */ + int set_regs_matched_done = 0; + + /* Used when we pop values we don't care about. */ +#ifdef MATCH_MAY_ALLOCATE /* otherwise, these are global. */ + const CHAR_T **reg_dummy; + PREFIX(register_info_type) *reg_info_dummy; +#endif + +#ifdef DEBUG + /* Counts the total number of registers pushed. */ + unsigned num_regs_pushed = 0; +#endif + + /* Definitions for state transitions. More efficiently for gcc. */ +#ifdef __GNUC__ +# if defined HAVE_SUBTRACT_LOCAL_LABELS && defined SHARED +# define NEXT \ + do \ + { \ + int offset; \ + const void *__unbounded ptr; \ + offset = (p == pend \ + ? 0 : jmptable[SWITCH_ENUM_CAST ((re_opcode_t) *p++)]); \ + ptr = &&end_of_pattern + offset; \ + goto *ptr; \ + } \ + while (0) +# define REF(x) \ + &&label_##x - &&end_of_pattern +# define JUMP_TABLE_TYPE const int +# else +# define NEXT \ + do \ + { \ + const void *__unbounded ptr; \ + ptr = (p == pend ? &&end_of_pattern \ + : jmptable[SWITCH_ENUM_CAST ((re_opcode_t) *p++)]); \ + goto *ptr; \ + } \ + while (0) +# define REF(x) \ + &&label_##x +# define JUMP_TABLE_TYPE const void *const +# endif +# define CASE(x) label_##x + static JUMP_TABLE_TYPE jmptable[] = + { + REF (no_op), + REF (succeed), + REF (exactn), +# ifdef MBS_SUPPORT + REF (exactn_bin), +# endif + REF (anychar), + REF (charset), + REF (charset_not), + REF (start_memory), + REF (stop_memory), + REF (duplicate), + REF (begline), + REF (endline), + REF (begbuf), + REF (endbuf), + REF (jump), + REF (jump_past_alt), + REF (on_failure_jump), + REF (on_failure_keep_string_jump), + REF (pop_failure_jump), + REF (maybe_pop_jump), + REF (dummy_failure_jump), + REF (push_dummy_failure), + REF (succeed_n), + REF (jump_n), + REF (set_number_at), + REF (wordchar), + REF (notwordchar), + REF (wordbeg), + REF (wordend), + REF (wordbound), + REF (notwordbound) +# ifdef emacs + ,REF (before_dot), + REF (at_dot), + REF (after_dot), + REF (syntaxspec), + REF (notsyntaxspec) +# endif + }; +#else +# define NEXT \ + break +# define CASE(x) \ + case x +#endif + + DEBUG_PRINT1 ("\n\nEntering re_match_2.\n"); + + INIT_FAIL_STACK (); + +#ifdef MATCH_MAY_ALLOCATE + /* Do not bother to initialize all the register variables if there are + no groups in the pattern, as it takes a fair amount of time. If + there are groups, we include space for register 0 (the whole + pattern), even though we never use it, since it simplifies the + array indexing. We should fix this. */ + if (bufp->re_nsub) + { + regstart = REGEX_TALLOC (num_regs, const CHAR_T *); + regend = REGEX_TALLOC (num_regs, const CHAR_T *); + old_regstart = REGEX_TALLOC (num_regs, const CHAR_T *); + old_regend = REGEX_TALLOC (num_regs, const CHAR_T *); + best_regstart = REGEX_TALLOC (num_regs, const CHAR_T *); + best_regend = REGEX_TALLOC (num_regs, const CHAR_T *); + reg_info = REGEX_TALLOC (num_regs, PREFIX(register_info_type)); + reg_dummy = REGEX_TALLOC (num_regs, const CHAR_T *); + reg_info_dummy = REGEX_TALLOC (num_regs, PREFIX(register_info_type)); + + if (!(regstart && regend && old_regstart && old_regend && reg_info + && best_regstart && best_regend && reg_dummy && reg_info_dummy)) + { + FREE_VARIABLES (); + return -2; + } + } + else + { + /* We must initialize all our variables to NULL, so that + `FREE_VARIABLES' doesn't try to free them. */ + regstart = regend = old_regstart = old_regend = best_regstart + = best_regend = reg_dummy = NULL; + reg_info = reg_info_dummy = (PREFIX(register_info_type) *) NULL; + } +#endif /* MATCH_MAY_ALLOCATE */ + + /* The starting position is bogus. */ +#ifdef WCHAR + if (pos < 0 || pos > csize1 + csize2) +#else /* BYTE */ + if (pos < 0 || pos > size1 + size2) +#endif + { + FREE_VARIABLES (); + return -1; + } + +#ifdef WCHAR + /* Allocate wchar_t array for string1 and string2 and + fill them with converted string. */ + if (string1 == NULL && string2 == NULL) + { + /* We need seting up buffers here. */ + + /* We must free wcs buffers in this function. */ + cant_free_wcs_buf = 0; + + if (csize1 != 0) + { + string1 = REGEX_TALLOC (csize1 + 1, CHAR_T); + mbs_offset1 = REGEX_TALLOC (csize1 + 1, int); + is_binary = REGEX_TALLOC (csize1 + 1, char); + if (!string1 || !mbs_offset1 || !is_binary) + { + FREE_VAR (string1); + FREE_VAR (mbs_offset1); + FREE_VAR (is_binary); + return -2; + } + } + if (csize2 != 0) + { + string2 = REGEX_TALLOC (csize2 + 1, CHAR_T); + mbs_offset2 = REGEX_TALLOC (csize2 + 1, int); + is_binary = REGEX_TALLOC (csize2 + 1, char); + if (!string2 || !mbs_offset2 || !is_binary) + { + FREE_VAR (string1); + FREE_VAR (mbs_offset1); + FREE_VAR (string2); + FREE_VAR (mbs_offset2); + FREE_VAR (is_binary); + return -2; + } + size2 = convert_mbs_to_wcs(string2, cstring2, csize2, + mbs_offset2, is_binary); + string2[size2] = L'\0'; /* for a sentinel */ + FREE_VAR (is_binary); + } + } + + /* We need to cast pattern to (wchar_t*), because we casted this compiled + pattern to (char*) in regex_compile. */ + p = pattern = (CHAR_T*)bufp->buffer; + pend = (CHAR_T*)(bufp->buffer + bufp->used); + +#endif /* WCHAR */ + + /* Initialize subexpression text positions to -1 to mark ones that no + start_memory/stop_memory has been seen for. Also initialize the + register information struct. */ + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + regstart[mcnt] = regend[mcnt] + = old_regstart[mcnt] = old_regend[mcnt] = REG_UNSET_VALUE; + + REG_MATCH_NULL_STRING_P (reg_info[mcnt]) = MATCH_NULL_UNSET_VALUE; + IS_ACTIVE (reg_info[mcnt]) = 0; + MATCHED_SOMETHING (reg_info[mcnt]) = 0; + EVER_MATCHED_SOMETHING (reg_info[mcnt]) = 0; + } + + /* We move `string1' into `string2' if the latter's empty -- but not if + `string1' is null. */ + if (size2 == 0 && string1 != NULL) + { + string2 = string1; + size2 = size1; + string1 = 0; + size1 = 0; +#ifdef WCHAR + mbs_offset2 = mbs_offset1; + csize2 = csize1; + mbs_offset1 = NULL; + csize1 = 0; +#endif + } + end1 = string1 + size1; + end2 = string2 + size2; + + /* Compute where to stop matching, within the two strings. */ +#ifdef WCHAR + if (stop <= csize1) + { + mcnt = count_mbs_length(mbs_offset1, stop); + end_match_1 = string1 + mcnt; + end_match_2 = string2; + } + else + { + if (stop > csize1 + csize2) + stop = csize1 + csize2; + end_match_1 = end1; + mcnt = count_mbs_length(mbs_offset2, stop-csize1); + end_match_2 = string2 + mcnt; + } + if (mcnt < 0) + { /* count_mbs_length return error. */ + FREE_VARIABLES (); + return -1; + } +#else + if (stop <= size1) + { + end_match_1 = string1 + stop; + end_match_2 = string2; + } + else + { + end_match_1 = end1; + end_match_2 = string2 + stop - size1; + } +#endif /* WCHAR */ + + /* `p' scans through the pattern as `d' scans through the data. + `dend' is the end of the input string that `d' points within. `d' + is advanced into the following input string whenever necessary, but + this happens before fetching; therefore, at the beginning of the + loop, `d' can be pointing at the end of a string, but it cannot + equal `string2'. */ +#ifdef WCHAR + if (size1 > 0 && pos <= csize1) + { + mcnt = count_mbs_length(mbs_offset1, pos); + d = string1 + mcnt; + dend = end_match_1; + } + else + { + mcnt = count_mbs_length(mbs_offset2, pos-csize1); + d = string2 + mcnt; + dend = end_match_2; + } + + if (mcnt < 0) + { /* count_mbs_length return error. */ + FREE_VARIABLES (); + return -1; + } +#else + if (size1 > 0 && pos <= size1) + { + d = string1 + pos; + dend = end_match_1; + } + else + { + d = string2 + pos - size1; + dend = end_match_2; + } +#endif /* WCHAR */ + + DEBUG_PRINT1 ("The compiled pattern is:\n"); + DEBUG_PRINT_COMPILED_PATTERN (bufp, p, pend); + DEBUG_PRINT1 ("The string to match is: `"); + DEBUG_PRINT_DOUBLE_STRING (d, string1, size1, string2, size2); + DEBUG_PRINT1 ("'\n"); + + /* This loops over pattern commands. It exits by returning from the + function if the match is complete, or it drops through if the match + fails at this starting point in the input data. */ + for (;;) + { +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + +#ifdef __GNUC__ + NEXT; +#else + if (p == pend) +#endif + { +#ifdef __GNUC__ + end_of_pattern: +#endif + /* End of pattern means we might have succeeded. */ + DEBUG_PRINT1 ("end of pattern ... "); + + /* If we haven't matched the entire string, and we want the + longest match, try backtracking. */ + if (d != end_match_2) + { + /* 1 if this match ends in the same string (string1 or string2) + as the best previous match. */ + boolean same_str_p = (FIRST_STRING_P (match_end) + == MATCHING_IN_FIRST_STRING); + /* 1 if this match is the best seen so far. */ + boolean best_match_p; + + /* AIX compiler got confused when this was combined + with the previous declaration. */ + if (same_str_p) + best_match_p = d > match_end; + else + best_match_p = !MATCHING_IN_FIRST_STRING; + + DEBUG_PRINT1 ("backtracking.\n"); + + if (!FAIL_STACK_EMPTY ()) + { /* More failure points to try. */ + + /* If exceeds best match so far, save it. */ + if (!best_regs_set || best_match_p) + { + best_regs_set = true; + match_end = d; + + DEBUG_PRINT1 ("\nSAVING match as best so far.\n"); + + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + best_regstart[mcnt] = regstart[mcnt]; + best_regend[mcnt] = regend[mcnt]; + } + } + goto fail; + } + + /* If no failure points, don't restore garbage. And if + last match is real best match, don't restore second + best one. */ + else if (best_regs_set && !best_match_p) + { + restore_best_regs: + /* Restore best match. It may happen that `dend == + end_match_1' while the restored d is in string2. + For example, the pattern `x.*y.*z' against the + strings `x-' and `y-z-', if the two strings are + not consecutive in memory. */ + DEBUG_PRINT1 ("Restoring best registers.\n"); + + d = match_end; + dend = ((d >= string1 && d <= end1) + ? end_match_1 : end_match_2); + + for (mcnt = 1; (unsigned) mcnt < num_regs; mcnt++) + { + regstart[mcnt] = best_regstart[mcnt]; + regend[mcnt] = best_regend[mcnt]; + } + } + } /* d != end_match_2 */ + + succeed_label: + DEBUG_PRINT1 ("Accepting match.\n"); + /* If caller wants register contents data back, do it. */ + if (regs && !bufp->no_sub) + { + /* Have the register data arrays been allocated? */ + if (bufp->regs_allocated == REGS_UNALLOCATED) + { /* No. So allocate them with malloc. We need one + extra element beyond `num_regs' for the `-1' marker + GNU code uses. */ + regs->num_regs = MAX (RE_NREGS, num_regs + 1); + regs->start = TALLOC (regs->num_regs, regoff_t); + regs->end = TALLOC (regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + { + FREE_VARIABLES (); + return -2; + } + bufp->regs_allocated = REGS_REALLOCATE; + } + else if (bufp->regs_allocated == REGS_REALLOCATE) + { /* Yes. If we need more elements than were already + allocated, reallocate them. If we need fewer, just + leave it alone. */ + if (regs->num_regs < num_regs + 1) + { + regs->num_regs = num_regs + 1; + RETALLOC (regs->start, regs->num_regs, regoff_t); + RETALLOC (regs->end, regs->num_regs, regoff_t); + if (regs->start == NULL || regs->end == NULL) + { + FREE_VARIABLES (); + return -2; + } + } + } + else + { + /* These braces fend off a "empty body in an else-statement" + warning under GCC when assert expands to nothing. */ + assert (bufp->regs_allocated == REGS_FIXED); + } + + /* Convert the pointer data in `regstart' and `regend' to + indices. Register zero has to be set differently, + since we haven't kept track of any info for it. */ + if (regs->num_regs > 0) + { + regs->start[0] = pos; +#ifdef WCHAR + if (MATCHING_IN_FIRST_STRING) + regs->end[0] = (mbs_offset1 != NULL ? + mbs_offset1[d-string1] : 0); + else + regs->end[0] = csize1 + (mbs_offset2 != NULL + ? mbs_offset2[d-string2] : 0); +#else + regs->end[0] = (MATCHING_IN_FIRST_STRING + ? ((regoff_t) (d - string1)) + : ((regoff_t) (d - string2 + size1))); +#endif /* WCHAR */ + } + + /* Go through the first `min (num_regs, regs->num_regs)' + registers, since that is all we initialized. */ + for (mcnt = 1; (unsigned) mcnt < MIN (num_regs, regs->num_regs); + mcnt++) + { + if (REG_UNSET (regstart[mcnt]) || REG_UNSET (regend[mcnt])) + regs->start[mcnt] = regs->end[mcnt] = -1; + else + { + regs->start[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regstart[mcnt]); + regs->end[mcnt] + = (regoff_t) POINTER_TO_OFFSET (regend[mcnt]); + } + } + + /* If the regs structure we return has more elements than + were in the pattern, set the extra elements to -1. If + we (re)allocated the registers, this is the case, + because we always allocate enough to have at least one + -1 at the end. */ + for (mcnt = num_regs; (unsigned) mcnt < regs->num_regs; mcnt++) + regs->start[mcnt] = regs->end[mcnt] = -1; + } /* regs && !bufp->no_sub */ + + DEBUG_PRINT4 ("%u failure points pushed, %u popped (%u remain).\n", + nfailure_points_pushed, nfailure_points_popped, + nfailure_points_pushed - nfailure_points_popped); + DEBUG_PRINT2 ("%u registers pushed.\n", num_regs_pushed); + +#ifdef WCHAR + if (MATCHING_IN_FIRST_STRING) + mcnt = mbs_offset1 != NULL ? mbs_offset1[d-string1] : 0; + else + mcnt = (mbs_offset2 != NULL ? mbs_offset2[d-string2] : 0) + + csize1; + mcnt -= pos; +#else + mcnt = d - pos - (MATCHING_IN_FIRST_STRING + ? string1 : string2 - size1); +#endif /* WCHAR */ + + DEBUG_PRINT2 ("Returning %d from re_match_2.\n", mcnt); + + FREE_VARIABLES (); + return mcnt; + } + +#ifndef __GNUC__ + /* Otherwise match next pattern command. */ + switch (SWITCH_ENUM_CAST ((re_opcode_t) *p++)) + { +#endif + /* Ignore these. Used to ignore the n of succeed_n's which + currently have n == 0. */ + CASE (no_op): + DEBUG_PRINT1 ("EXECUTING no_op.\n"); + NEXT; + + CASE (succeed): + DEBUG_PRINT1 ("EXECUTING succeed.\n"); + goto succeed_label; + + /* Match the next n pattern characters exactly. The following + byte in the pattern defines n, and the n bytes after that + are the characters to match. */ + CASE (exactn): +#ifdef MBS_SUPPORT + CASE (exactn_bin): +#endif + mcnt = *p++; + DEBUG_PRINT2 ("EXECUTING exactn %d.\n", mcnt); + + /* This is written out as an if-else so we don't waste time + testing `translate' inside the loop. */ + if (translate) + { + do + { + PREFETCH (); +#ifdef WCHAR + if (*d <= 0xff) + { + if ((UCHAR_T) translate[(unsigned char) *d++] + != (UCHAR_T) *p++) + goto fail; + } + else + { + if (*d++ != (CHAR_T) *p++) + goto fail; + } +#else + if ((UCHAR_T) translate[(unsigned char) *d++] + != (UCHAR_T) *p++) + goto fail; +#endif /* WCHAR */ + } + while (--mcnt); + } + else + { + do + { + PREFETCH (); + if (*d++ != (CHAR_T) *p++) goto fail; + } + while (--mcnt); + } + SET_REGS_MATCHED (); + NEXT; + + + /* Match any character except possibly a newline or a null. */ + CASE (anychar): + DEBUG_PRINT1 ("EXECUTING anychar.\n"); + + PREFETCH (); + + if ((!(bufp->syntax & RE_DOT_NEWLINE) && TRANSLATE (*d) == '\n') + || (bufp->syntax & RE_DOT_NOT_NULL && TRANSLATE (*d) == '\000')) + goto fail; + + SET_REGS_MATCHED (); + DEBUG_PRINT2 (" Matched `%ld'.\n", (long int) *d); + d++; + NEXT; + + + CASE (charset): + CASE (charset_not): + { + register UCHAR_T c; +#ifdef WCHAR + unsigned int i, char_class_length, coll_symbol_length, + equiv_class_length, ranges_length, chars_length, length; + CHAR_T *workp, *workp2, *charset_top; +#define WORK_BUFFER_SIZE 128 + CHAR_T str_buf[WORK_BUFFER_SIZE]; +# ifdef _LIBC + uint32_t nrules; +# endif /* _LIBC */ +#endif /* WCHAR */ + boolean not = (re_opcode_t) *(p - 1) == charset_not; + + DEBUG_PRINT2 ("EXECUTING charset%s.\n", not ? "_not" : ""); + PREFETCH (); + c = TRANSLATE (*d); /* The character to match. */ +#ifdef WCHAR +# ifdef _LIBC + nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); +# endif /* _LIBC */ + charset_top = p - 1; + char_class_length = *p++; + coll_symbol_length = *p++; + equiv_class_length = *p++; + ranges_length = *p++; + chars_length = *p++; + /* p points charset[6], so the address of the next instruction + (charset[l+m+n+2o+k+p']) equals p[l+m+n+2*o+p'], + where l=length of char_classes, m=length of collating_symbol, + n=equivalence_class, o=length of char_range, + p'=length of character. */ + workp = p; + /* Update p to indicate the next instruction. */ + p += char_class_length + coll_symbol_length+ equiv_class_length + + 2*ranges_length + chars_length; + + /* match with char_class? */ + for (i = 0; i < char_class_length ; i += CHAR_CLASS_SIZE) + { + wctype_t wctype; + uintptr_t alignedp = ((uintptr_t)workp + + __alignof__(wctype_t) - 1) + & ~(uintptr_t)(__alignof__(wctype_t) - 1); + wctype = *((wctype_t*)alignedp); + workp += CHAR_CLASS_SIZE; + if (iswctype((wint_t)c, wctype)) + goto char_set_matched; + } + + /* match with collating_symbol? */ +# ifdef _LIBC + if (nrules != 0) + { + const unsigned char *extra = (const unsigned char *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_SYMB_EXTRAMB); + + for (workp2 = workp + coll_symbol_length ; workp < workp2 ; + workp++) + { + int32_t *wextra; + wextra = (int32_t*)(extra + *workp++); + for (i = 0; i < *wextra; ++i) + if (TRANSLATE(d[i]) != wextra[1 + i]) + break; + + if (i == *wextra) + { + /* Update d, however d will be incremented at + char_set_matched:, we decrement d here. */ + d += i - 1; + goto char_set_matched; + } + } + } + else /* (nrules == 0) */ +# endif + /* If we can't look up collation data, we use wcscoll + instead. */ + { + for (workp2 = workp + coll_symbol_length ; workp < workp2 ;) + { + const CHAR_T *backup_d = d, *backup_dend = dend; + length = wcslen (workp); + + /* If wcscoll(the collating symbol, whole string) > 0, + any substring of the string never match with the + collating symbol. */ + if (wcscoll (workp, d) > 0) + { + workp += length + 1; + continue; + } + + /* First, we compare the collating symbol with + the first character of the string. + If it don't match, we add the next character to + the compare buffer in turn. */ + for (i = 0 ; i < WORK_BUFFER_SIZE-1 ; i++, d++) + { + int match; + if (d == dend) + { + if (dend == end_match_2) + break; + d = string2; + dend = end_match_2; + } + + /* add next character to the compare buffer. */ + str_buf[i] = TRANSLATE(*d); + str_buf[i+1] = '\0'; + + match = wcscoll (workp, str_buf); + if (match == 0) + goto char_set_matched; + + if (match < 0) + /* (str_buf > workp) indicate (str_buf + X > workp), + because for all X (str_buf + X > str_buf). + So we don't need continue this loop. */ + break; + + /* Otherwise(str_buf < workp), + (str_buf+next_character) may equals (workp). + So we continue this loop. */ + } + /* not matched */ + d = backup_d; + dend = backup_dend; + workp += length + 1; + } + } + /* match with equivalence_class? */ +# ifdef _LIBC + if (nrules != 0) + { + const CHAR_T *backup_d = d, *backup_dend = dend; + /* Try to match the equivalence class against + those known to the collate implementation. */ + const int32_t *table; + const int32_t *weights; + const int32_t *extra; + const int32_t *indirect; + int32_t idx, idx2; + wint_t *cp; + size_t len; + + /* This #include defines a local function! */ +# include + + table = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEWC); + weights = (const wint_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_WEIGHTWC); + extra = (const wint_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAWC); + indirect = (const int32_t *) + _NL_CURRENT (LC_COLLATE, _NL_COLLATE_INDIRECTWC); + + /* Write 1 collating element to str_buf, and + get its index. */ + idx2 = 0; + + for (i = 0 ; idx2 == 0 && i < WORK_BUFFER_SIZE - 1; i++) + { + cp = (wint_t*)str_buf; + if (d == dend) + { + if (dend == end_match_2) + break; + d = string2; + dend = end_match_2; + } + str_buf[i] = TRANSLATE(*(d+i)); + str_buf[i+1] = '\0'; /* sentinel */ + idx2 = findidx ((const wint_t**)&cp); + } + + /* Update d, however d will be incremented at + char_set_matched:, we decrement d here. */ + d = backup_d + ((wchar_t*)cp - (wchar_t*)str_buf - 1); + if (d >= dend) + { + if (dend == end_match_2) + d = dend; + else + { + d = string2; + dend = end_match_2; + } + } + + len = weights[idx2]; + + for (workp2 = workp + equiv_class_length ; workp < workp2 ; + workp++) + { + idx = (int32_t)*workp; + /* We already checked idx != 0 in regex_compile. */ + + if (idx2 != 0 && len == weights[idx]) + { + int cnt = 0; + while (cnt < len && (weights[idx + 1 + cnt] + == weights[idx2 + 1 + cnt])) + ++cnt; + + if (cnt == len) + goto char_set_matched; + } + } + /* not matched */ + d = backup_d; + dend = backup_dend; + } + else /* (nrules == 0) */ +# endif + /* If we can't look up collation data, we use wcscoll + instead. */ + { + for (workp2 = workp + equiv_class_length ; workp < workp2 ;) + { + const CHAR_T *backup_d = d, *backup_dend = dend; + length = wcslen (workp); + + /* If wcscoll(the collating symbol, whole string) > 0, + any substring of the string never match with the + collating symbol. */ + if (wcscoll (workp, d) > 0) + { + workp += length + 1; + break; + } + + /* First, we compare the equivalence class with + the first character of the string. + If it don't match, we add the next character to + the compare buffer in turn. */ + for (i = 0 ; i < WORK_BUFFER_SIZE - 1 ; i++, d++) + { + int match; + if (d == dend) + { + if (dend == end_match_2) + break; + d = string2; + dend = end_match_2; + } + + /* add next character to the compare buffer. */ + str_buf[i] = TRANSLATE(*d); + str_buf[i+1] = '\0'; + + match = wcscoll (workp, str_buf); + + if (match == 0) + goto char_set_matched; + + if (match < 0) + /* (str_buf > workp) indicate (str_buf + X > workp), + because for all X (str_buf + X > str_buf). + So we don't need continue this loop. */ + break; + + /* Otherwise(str_buf < workp), + (str_buf+next_character) may equals (workp). + So we continue this loop. */ + } + /* not matched */ + d = backup_d; + dend = backup_dend; + workp += length + 1; + } + } + + /* match with char_range? */ +# ifdef _LIBC + if (nrules != 0) + { + uint32_t collseqval; + const char *collseq = (const char *) + _NL_CURRENT(LC_COLLATE, _NL_COLLATE_COLLSEQWC); + + collseqval = collseq_table_lookup (collseq, c); + + for (; workp < p - chars_length ;) + { + uint32_t start_val, end_val; + + /* We already compute the collation sequence value + of the characters (or collating symbols). */ + start_val = (uint32_t) *workp++; /* range_start */ + end_val = (uint32_t) *workp++; /* range_end */ + + if (start_val <= collseqval && collseqval <= end_val) + goto char_set_matched; + } + } + else +# endif + { + /* We set range_start_char at str_buf[0], range_end_char + at str_buf[4], and compared char at str_buf[2]. */ + str_buf[1] = 0; + str_buf[2] = c; + str_buf[3] = 0; + str_buf[5] = 0; + for (; workp < p - chars_length ;) + { + wchar_t *range_start_char, *range_end_char; + + /* match if (range_start_char <= c <= range_end_char). */ + + /* If range_start(or end) < 0, we assume -range_start(end) + is the offset of the collating symbol which is specified + as the character of the range start(end). */ + + /* range_start */ + if (*workp < 0) + range_start_char = charset_top - (*workp++); + else + { + str_buf[0] = *workp++; + range_start_char = str_buf; + } + + /* range_end */ + if (*workp < 0) + range_end_char = charset_top - (*workp++); + else + { + str_buf[4] = *workp++; + range_end_char = str_buf + 4; + } + + if (wcscoll (range_start_char, str_buf+2) <= 0 + && wcscoll (str_buf+2, range_end_char) <= 0) + goto char_set_matched; + } + } + + /* match with char? */ + for (; workp < p ; workp++) + if (c == *workp) + goto char_set_matched; + + not = !not; + + char_set_matched: + if (not) goto fail; +#else + /* Cast to `unsigned' instead of `unsigned char' in case the + bit list is a full 32 bytes long. */ + if (c < (unsigned) (*p * BYTEWIDTH) + && p[1 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + p += 1 + *p; + + if (!not) goto fail; +#undef WORK_BUFFER_SIZE +#endif /* WCHAR */ + SET_REGS_MATCHED (); + d++; + NEXT; + } + + + /* The beginning of a group is represented by start_memory. + The arguments are the register number in the next byte, and the + number of groups inner to this one in the next. The text + matched within the group is recorded (in the internal + registers data structure) under the register number. */ + CASE (start_memory): + DEBUG_PRINT3 ("EXECUTING start_memory %ld (%ld):\n", + (long int) *p, (long int) p[1]); + + /* Find out if this group can match the empty string. */ + p1 = p; /* To send to group_match_null_string_p. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[*p]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[*p]) + = PREFIX(group_match_null_string_p) (&p1, pend, reg_info); + + /* Save the position in the string where we were the last time + we were at this open-group operator in case the group is + operated upon by a repetition operator, e.g., with `(a*)*b' + against `ab'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regstart[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regstart[*p]) ? d : regstart[*p] + : regstart[*p]; + DEBUG_PRINT2 (" old_regstart: %d\n", + POINTER_TO_OFFSET (old_regstart[*p])); + + regstart[*p] = d; + DEBUG_PRINT2 (" regstart: %d\n", POINTER_TO_OFFSET (regstart[*p])); + + IS_ACTIVE (reg_info[*p]) = 1; + MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 0; + + /* This is the new highest active register. */ + highest_active_reg = *p; + + /* If nothing was active before, this is the new lowest active + register. */ + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *p; + + /* Move past the register number and inner group count. */ + p += 2; + just_past_start_mem = p; + + NEXT; + + + /* The stop_memory opcode represents the end of a group. Its + arguments are the same as start_memory's: the register + number, and the number of inner groups. */ + CASE (stop_memory): + DEBUG_PRINT3 ("EXECUTING stop_memory %ld (%ld):\n", + (long int) *p, (long int) p[1]); + + /* We need to save the string position the last time we were at + this close-group operator in case the group is operated + upon by a repetition operator, e.g., with `((a*)*(b*)*)*' + against `aba'; then we want to ignore where we are now in + the string in case this attempt to match fails. */ + old_regend[*p] = REG_MATCH_NULL_STRING_P (reg_info[*p]) + ? REG_UNSET (regend[*p]) ? d : regend[*p] + : regend[*p]; + DEBUG_PRINT2 (" old_regend: %d\n", + POINTER_TO_OFFSET (old_regend[*p])); + + regend[*p] = d; + DEBUG_PRINT2 (" regend: %d\n", POINTER_TO_OFFSET (regend[*p])); + + /* This register isn't active anymore. */ + IS_ACTIVE (reg_info[*p]) = 0; + + /* Clear this whenever we change the register activity status. */ + set_regs_matched_done = 0; + + /* If this was the only register active, nothing is active + anymore. */ + if (lowest_active_reg == highest_active_reg) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + { /* We must scan for the new highest active register, since + it isn't necessarily one less than now: consider + (a(b)c(d(e)f)g). When group 3 ends, after the f), the + new highest active register is 1. */ + UCHAR_T r = *p - 1; + while (r > 0 && !IS_ACTIVE (reg_info[r])) + r--; + + /* If we end up at register zero, that means that we saved + the registers as the result of an `on_failure_jump', not + a `start_memory', and we jumped to past the innermost + `stop_memory'. For example, in ((.)*) we save + registers 1 and 2 as a result of the *, but when we pop + back to the second ), we are at the stop_memory 1. + Thus, nothing is active. */ + if (r == 0) + { + lowest_active_reg = NO_LOWEST_ACTIVE_REG; + highest_active_reg = NO_HIGHEST_ACTIVE_REG; + } + else + highest_active_reg = r; + } + + /* If just failed to match something this time around with a + group that's operated on by a repetition operator, try to + force exit from the ``loop'', and restore the register + information for this group that we had before trying this + last match. */ + if ((!MATCHED_SOMETHING (reg_info[*p]) + || just_past_start_mem == p - 1) + && (p + 2) < pend) + { + boolean is_a_jump_n = false; + + p1 = p + 2; + mcnt = 0; + switch ((re_opcode_t) *p1++) + { + case jump_n: + is_a_jump_n = true; + case pop_failure_jump: + case maybe_pop_jump: + case jump: + case dummy_failure_jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (is_a_jump_n) + p1 += OFFSET_ADDRESS_SIZE; + break; + + default: + /* do nothing */ ; + } + p1 += mcnt; + + /* If the next operation is a jump backwards in the pattern + to an on_failure_jump right before the start_memory + corresponding to this stop_memory, exit from the loop + by forcing a failure after pushing on the stack the + on_failure_jump's jump in the pattern, and d. */ + if (mcnt < 0 && (re_opcode_t) *p1 == on_failure_jump + && (re_opcode_t) p1[1+OFFSET_ADDRESS_SIZE] == start_memory + && p1[2+OFFSET_ADDRESS_SIZE] == *p) + { + /* If this group ever matched anything, then restore + what its registers were before trying this last + failed match, e.g., with `(a*)*b' against `ab' for + regstart[1], and, e.g., with `((a*)*(b*)*)*' + against `aba' for regend[3]. + + Also restore the registers for inner groups for, + e.g., `((a*)(b*))*' against `aba' (register 3 would + otherwise get trashed). */ + + if (EVER_MATCHED_SOMETHING (reg_info[*p])) + { + unsigned r; + + EVER_MATCHED_SOMETHING (reg_info[*p]) = 0; + + /* Restore this and inner groups' (if any) registers. */ + for (r = *p; r < (unsigned) *p + (unsigned) *(p + 1); + r++) + { + regstart[r] = old_regstart[r]; + + /* xx why this test? */ + if (old_regend[r] >= regstart[r]) + regend[r] = old_regend[r]; + } + } + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + PUSH_FAILURE_POINT (p1 + mcnt, d, -2); + + goto fail; + } + } + + /* Move past the register number and the inner group count. */ + p += 2; + NEXT; + + + /* \ has been turned into a `duplicate' command which is + followed by the numeric value of as the register number. */ + CASE (duplicate): + { + register const CHAR_T *d2, *dend2; + int regno = *p++; /* Get which register to match against. */ + DEBUG_PRINT2 ("EXECUTING duplicate %d.\n", regno); + + /* Can't back reference a group which we've never matched. */ + if (REG_UNSET (regstart[regno]) || REG_UNSET (regend[regno])) + goto fail; + + /* Where in input to try to start matching. */ + d2 = regstart[regno]; + + /* Where to stop matching; if both the place to start and + the place to stop matching are in the same string, then + set to the place to stop, otherwise, for now have to use + the end of the first string. */ + + dend2 = ((FIRST_STRING_P (regstart[regno]) + == FIRST_STRING_P (regend[regno])) + ? regend[regno] : end_match_1); + for (;;) + { + /* If necessary, advance to next segment in register + contents. */ + while (d2 == dend2) + { + if (dend2 == end_match_2) break; + if (dend2 == regend[regno]) break; + + /* End of string1 => advance to string2. */ + d2 = string2; + dend2 = regend[regno]; + } + /* At end of register contents => success */ + if (d2 == dend2) break; + + /* If necessary, advance to next segment in data. */ + PREFETCH (); + + /* How many characters left in this segment to match. */ + mcnt = dend - d; + + /* Want how many consecutive characters we can match in + one shot, so, if necessary, adjust the count. */ + if (mcnt > dend2 - d2) + mcnt = dend2 - d2; + + /* Compare that many; failure if mismatch, else move + past them. */ + if (translate + ? PREFIX(bcmp_translate) (d, d2, mcnt, translate) + : memcmp (d, d2, mcnt*sizeof(UCHAR_T))) + goto fail; + d += mcnt, d2 += mcnt; + + /* Do this because we've match some characters. */ + SET_REGS_MATCHED (); + } + } + NEXT; + + + /* begline matches the empty string at the beginning of the string + (unless `not_bol' is set in `bufp'), and, if + `newline_anchor' is set, after newlines. */ + CASE (begline): + DEBUG_PRINT1 ("EXECUTING begline.\n"); + + if (AT_STRINGS_BEG (d)) + { + if (!bufp->not_bol) + { + NEXT; + } + } + else if (d[-1] == '\n' && bufp->newline_anchor) + { + NEXT; + } + /* In all other cases, we fail. */ + goto fail; + + + /* endline is the dual of begline. */ + CASE (endline): + DEBUG_PRINT1 ("EXECUTING endline.\n"); + + if (AT_STRINGS_END (d)) + { + if (!bufp->not_eol) + { + NEXT; + } + } + + /* We have to ``prefetch'' the next character. */ + else if ((d == end1 ? *string2 : *d) == '\n' + && bufp->newline_anchor) + { + NEXT; + } + goto fail; + + + /* Match at the very beginning of the data. */ + CASE (begbuf): + DEBUG_PRINT1 ("EXECUTING begbuf.\n"); + if (AT_STRINGS_BEG (d)) + { + NEXT; + } + goto fail; + + + /* Match at the very end of the data. */ + CASE (endbuf): + DEBUG_PRINT1 ("EXECUTING endbuf.\n"); + if (AT_STRINGS_END (d)) + { + NEXT; + } + goto fail; + + + /* on_failure_keep_string_jump is used to optimize `.*\n'. It + pushes NULL as the value for the string on the stack. Then + `pop_failure_point' will keep the current value for the + string, instead of restoring it. To see why, consider + matching `foo\nbar' against `.*\n'. The .* matches the foo; + then the . fails against the \n. But the next thing we want + to do is match the \n against the \n; if we restored the + string value, we would be back at the foo. + + Because this is used only in specific cases, we don't need to + check all the things that `on_failure_jump' does, to make + sure the right things get saved on the stack. Hence we don't + share its code. The only reason to push anything on the + stack at all is that otherwise we would have to change + `anychar's code to do something besides goto fail in this + case; that seems worse than this. */ + CASE (on_failure_keep_string_jump): + DEBUG_PRINT1 ("EXECUTING on_failure_keep_string_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p):\n", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x):\n", mcnt, p + mcnt); +#endif + + PUSH_FAILURE_POINT (p + mcnt, NULL, -2); + NEXT; + + + /* Uses of on_failure_jump: + + Each alternative starts with an on_failure_jump that points + to the beginning of the next alternative. Each alternative + except the last ends with a jump that in effect jumps past + the rest of the alternatives. (They really jump to the + ending jump of the following alternative, because tensioning + these jumps is a hassle.) + + Repeats start with an on_failure_jump that points past both + the repetition text and either the following jump or + pop_failure_jump back to this on_failure_jump. */ + CASE (on_failure_jump): + on_failure: + DEBUG_PRINT1 ("EXECUTING on_failure_jump"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" %d (to %p)", mcnt, p + mcnt); +#else + DEBUG_PRINT3 (" %d (to 0x%x)", mcnt, p + mcnt); +#endif + + /* If this on_failure_jump comes right before a group (i.e., + the original * applied to a group), save the information + for that group and all inner ones, so that if we fail back + to this point, the group's information will be correct. + For example, in \(a*\)*\1, we need the preceding group, + and in \(zz\(a*\)b*\)\2, we need the inner group. */ + + /* We can't use `p' to check ahead because we push + a failure point to `p + mcnt' after we do this. */ + p1 = p; + + /* We need to skip no_op's before we look for the + start_memory in case this on_failure_jump is happening as + the result of a completed succeed_n, as in \(a\)\{1,3\}b\1 + against aba. */ + while (p1 < pend && (re_opcode_t) *p1 == no_op) + p1++; + + if (p1 < pend && (re_opcode_t) *p1 == start_memory) + { + /* We have a new highest active register now. This will + get reset at the start_memory we are about to get to, + but we will have saved all the registers relevant to + this repetition op, as described above. */ + highest_active_reg = *(p1 + 1) + *(p1 + 2); + if (lowest_active_reg == NO_LOWEST_ACTIVE_REG) + lowest_active_reg = *(p1 + 1); + } + + DEBUG_PRINT1 (":\n"); + PUSH_FAILURE_POINT (p + mcnt, d, -2); + NEXT; + + + /* A smart repeat ends with `maybe_pop_jump'. + We change it to either `pop_failure_jump' or `jump'. */ + CASE (maybe_pop_jump): + EXTRACT_NUMBER_AND_INCR (mcnt, p); + DEBUG_PRINT2 ("EXECUTING maybe_pop_jump %d.\n", mcnt); + { + register UCHAR_T *p2 = p; + + /* Compare the beginning of the repeat with what in the + pattern follows its end. If we can establish that there + is nothing that they would both match, i.e., that we + would have to backtrack because of (as in, e.g., `a*a') + then we can change to pop_failure_jump, because we'll + never have to backtrack. + + This is not true in the case of alternatives: in + `(a|ab)*' we do need to backtrack to the `ab' alternative + (e.g., if the string was `ab'). But instead of trying to + detect that here, the alternative has put on a dummy + failure point which is what we will end up popping. */ + + /* Skip over open/close-group commands. + If what follows this loop is a ...+ construct, + look at what begins its body, since we will have to + match at least one of that. */ + while (1) + { + if (p2 + 2 < pend + && ((re_opcode_t) *p2 == stop_memory + || (re_opcode_t) *p2 == start_memory)) + p2 += 3; + else if (p2 + 2 + 2 * OFFSET_ADDRESS_SIZE < pend + && (re_opcode_t) *p2 == dummy_failure_jump) + p2 += 2 + 2 * OFFSET_ADDRESS_SIZE; + else + break; + } + + p1 = p + mcnt; + /* p1[0] ... p1[2] are the `on_failure_jump' corresponding + to the `maybe_finalize_jump' of this case. Examine what + follows. */ + + /* If we're at the end of the pattern, we can change. */ + if (p2 == pend) + { + /* Consider what happens when matching ":\(.*\)" + against ":/". I don't really understand this code + yet. */ + p[-(1+OFFSET_ADDRESS_SIZE)] = (UCHAR_T) + pop_failure_jump; + DEBUG_PRINT1 + (" End of pattern: change to `pop_failure_jump'.\n"); + } + + else if ((re_opcode_t) *p2 == exactn +#ifdef MBS_SUPPORT + || (re_opcode_t) *p2 == exactn_bin +#endif + || (bufp->newline_anchor && (re_opcode_t) *p2 == endline)) + { + register UCHAR_T c + = *p2 == (UCHAR_T) endline ? '\n' : p2[2]; + + if (((re_opcode_t) p1[1+OFFSET_ADDRESS_SIZE] == exactn +#ifdef MBS_SUPPORT + || (re_opcode_t) p1[1+OFFSET_ADDRESS_SIZE] == exactn_bin +#endif + ) && p1[3+OFFSET_ADDRESS_SIZE] != c) + { + p[-(1+OFFSET_ADDRESS_SIZE)] = (UCHAR_T) + pop_failure_jump; +#ifdef WCHAR + DEBUG_PRINT3 (" %C != %C => pop_failure_jump.\n", + (wint_t) c, + (wint_t) p1[3+OFFSET_ADDRESS_SIZE]); +#else + DEBUG_PRINT3 (" %c != %c => pop_failure_jump.\n", + (char) c, + (char) p1[3+OFFSET_ADDRESS_SIZE]); +#endif + } + +#ifndef WCHAR + else if ((re_opcode_t) p1[3] == charset + || (re_opcode_t) p1[3] == charset_not) + { + int not = (re_opcode_t) p1[3] == charset_not; + + if (c < (unsigned) (p1[4] * BYTEWIDTH) + && p1[5 + c / BYTEWIDTH] & (1 << (c % BYTEWIDTH))) + not = !not; + + /* `not' is equal to 1 if c would match, which means + that we can't change to pop_failure_jump. */ + if (!not) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } +#endif /* not WCHAR */ + } +#ifndef WCHAR + else if ((re_opcode_t) *p2 == charset) + { + /* We win if the first character of the loop is not part + of the charset. */ + if ((re_opcode_t) p1[3] == exactn + && ! ((int) p2[1] * BYTEWIDTH > (int) p1[5] + && (p2[2 + p1[5] / BYTEWIDTH] + & (1 << (p1[5] % BYTEWIDTH))))) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + + else if ((re_opcode_t) p1[3] == charset_not) + { + int idx; + /* We win if the charset_not inside the loop + lists every character listed in the charset after. */ + for (idx = 0; idx < (int) p2[1]; idx++) + if (! (p2[2 + idx] == 0 + || (idx < (int) p1[4] + && ((p2[2 + idx] & ~ p1[5 + idx]) == 0)))) + break; + + if (idx == p2[1]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + else if ((re_opcode_t) p1[3] == charset) + { + int idx; + /* We win if the charset inside the loop + has no overlap with the one after the loop. */ + for (idx = 0; + idx < (int) p2[1] && idx < (int) p1[4]; + idx++) + if ((p2[2 + idx] & p1[5 + idx]) != 0) + break; + + if (idx == p2[1] || idx == p1[4]) + { + p[-3] = (unsigned char) pop_failure_jump; + DEBUG_PRINT1 (" No match => pop_failure_jump.\n"); + } + } + } +#endif /* not WCHAR */ + } + p -= OFFSET_ADDRESS_SIZE; /* Point at relative address again. */ + if ((re_opcode_t) p[-1] != pop_failure_jump) + { + p[-1] = (UCHAR_T) jump; + DEBUG_PRINT1 (" Match => jump.\n"); + goto unconditional_jump; + } + /* Note fall through. */ + + + /* The end of a simple repeat has a pop_failure_jump back to + its matching on_failure_jump, where the latter will push a + failure point. The pop_failure_jump takes off failure + points put on by this pop_failure_jump's matching + on_failure_jump; we got through the pattern to here from the + matching on_failure_jump, so didn't fail. */ + CASE (pop_failure_jump): + { + /* We need to pass separate storage for the lowest and + highest registers, even though we don't care about the + actual values. Otherwise, we will restore only one + register from the stack, since lowest will == highest in + `pop_failure_point'. */ + active_reg_t dummy_low_reg, dummy_high_reg; + UCHAR_T *pdummy = NULL; + const CHAR_T *sdummy = NULL; + + DEBUG_PRINT1 ("EXECUTING pop_failure_jump.\n"); + POP_FAILURE_POINT (sdummy, pdummy, + dummy_low_reg, dummy_high_reg, + reg_dummy, reg_dummy, reg_info_dummy); + } + /* Note fall through. */ + + unconditional_jump: +#ifdef _LIBC + DEBUG_PRINT2 ("\n%p: ", p); +#else + DEBUG_PRINT2 ("\n0x%x: ", p); +#endif + /* Note fall through. */ + + /* Unconditionally jump (without popping any failure points). */ + CASE (jump): + EXTRACT_NUMBER_AND_INCR (mcnt, p); /* Get the amount to jump. */ + DEBUG_PRINT2 ("EXECUTING jump %d ", mcnt); + p += mcnt; /* Do the jump. */ +#ifdef _LIBC + DEBUG_PRINT2 ("(to %p).\n", p); +#else + DEBUG_PRINT2 ("(to 0x%x).\n", p); +#endif + NEXT; + + + /* We need this opcode so we can detect where alternatives end + in `group_match_null_string_p' et al. */ + CASE (jump_past_alt): + DEBUG_PRINT1 ("EXECUTING jump_past_alt.\n"); + goto unconditional_jump; + + + /* Normally, the on_failure_jump pushes a failure point, which + then gets popped at pop_failure_jump. We will end up at + pop_failure_jump, also, and with a pattern of, say, `a+', we + are skipping over the on_failure_jump, so we have to push + something meaningless for pop_failure_jump to pop. */ + CASE (dummy_failure_jump): + DEBUG_PRINT1 ("EXECUTING dummy_failure_jump.\n"); + /* It doesn't matter what we push for the string here. What + the code at `fail' tests is the value for the pattern. */ + PUSH_FAILURE_POINT (NULL, NULL, -2); + goto unconditional_jump; + + + /* At the end of an alternative, we need to push a dummy failure + point in case we are followed by a `pop_failure_jump', because + we don't want the failure point for the alternative to be + popped. For example, matching `(a|ab)*' against `aab' + requires that we match the `ab' alternative. */ + CASE (push_dummy_failure): + DEBUG_PRINT1 ("EXECUTING push_dummy_failure.\n"); + /* See comments just above at `dummy_failure_jump' about the + two zeroes. */ + PUSH_FAILURE_POINT (NULL, NULL, -2); + NEXT; + + /* Have to succeed matching what follows at least n times. + After that, handle like `on_failure_jump'. */ + CASE (succeed_n): + EXTRACT_NUMBER (mcnt, p + OFFSET_ADDRESS_SIZE); + DEBUG_PRINT2 ("EXECUTING succeed_n %d.\n", mcnt); + + assert (mcnt >= 0); + /* Originally, this is how many times we HAVE to succeed. */ + if (mcnt > 0) + { + mcnt--; + p += OFFSET_ADDRESS_SIZE; + STORE_NUMBER_AND_INCR (p, mcnt); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p - OFFSET_ADDRESS_SIZE + , mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p - OFFSET_ADDRESS_SIZE + , mcnt); +#endif + } + else if (mcnt == 0) + { +#ifdef _LIBC + DEBUG_PRINT2 (" Setting two bytes from %p to no_op.\n", + p + OFFSET_ADDRESS_SIZE); +#else + DEBUG_PRINT2 (" Setting two bytes from 0x%x to no_op.\n", + p + OFFSET_ADDRESS_SIZE); +#endif /* _LIBC */ + +#ifdef WCHAR + p[1] = (UCHAR_T) no_op; +#else + p[2] = (UCHAR_T) no_op; + p[3] = (UCHAR_T) no_op; +#endif /* WCHAR */ + goto on_failure; + } + NEXT; + + CASE (jump_n): + EXTRACT_NUMBER (mcnt, p + OFFSET_ADDRESS_SIZE); + DEBUG_PRINT2 ("EXECUTING jump_n %d.\n", mcnt); + + /* Originally, this is how many times we CAN jump. */ + if (mcnt) + { + mcnt--; + STORE_NUMBER (p + OFFSET_ADDRESS_SIZE, mcnt); + +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p + OFFSET_ADDRESS_SIZE, + mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p + OFFSET_ADDRESS_SIZE, + mcnt); +#endif /* _LIBC */ + goto unconditional_jump; + } + /* If don't have to jump any more, skip over the rest of command. */ + else + p += 2 * OFFSET_ADDRESS_SIZE; + NEXT; + + CASE (set_number_at): + { + DEBUG_PRINT1 ("EXECUTING set_number_at.\n"); + + EXTRACT_NUMBER_AND_INCR (mcnt, p); + p1 = p + mcnt; + EXTRACT_NUMBER_AND_INCR (mcnt, p); +#ifdef _LIBC + DEBUG_PRINT3 (" Setting %p to %d.\n", p1, mcnt); +#else + DEBUG_PRINT3 (" Setting 0x%x to %d.\n", p1, mcnt); +#endif + STORE_NUMBER (p1, mcnt); + NEXT; + } + +#if 0 + /* The DEC Alpha C compiler 3.x generates incorrect code for the + test WORDCHAR_P (d - 1) != WORDCHAR_P (d) in the expansion of + AT_WORD_BOUNDARY, so this code is disabled. Expanding the + macro and introducing temporary variables works around the bug. */ + + CASE (wordbound): + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + { + NEXT; + } + goto fail; + + CASE (notwordbound): + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_WORD_BOUNDARY (d)) + goto fail; + NEXT; +#else + CASE (wordbound): + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING wordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + { + NEXT; + } + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + { + NEXT; + } + goto fail; + } + + CASE (notwordbound): + { + boolean prevchar, thischar; + + DEBUG_PRINT1 ("EXECUTING notwordbound.\n"); + if (AT_STRINGS_BEG (d) || AT_STRINGS_END (d)) + goto fail; + + prevchar = WORDCHAR_P (d - 1); + thischar = WORDCHAR_P (d); + if (prevchar != thischar) + goto fail; + NEXT; + } +#endif + + CASE (wordbeg): + DEBUG_PRINT1 ("EXECUTING wordbeg.\n"); + if (!AT_STRINGS_END (d) && WORDCHAR_P (d) + && (AT_STRINGS_BEG (d) || !WORDCHAR_P (d - 1))) + { + NEXT; + } + goto fail; + + CASE (wordend): + DEBUG_PRINT1 ("EXECUTING wordend.\n"); + if (!AT_STRINGS_BEG (d) && WORDCHAR_P (d - 1) + && (AT_STRINGS_END (d) || !WORDCHAR_P (d))) + { + NEXT; + } + goto fail; + +#ifdef emacs + CASE (before_dot): + DEBUG_PRINT1 ("EXECUTING before_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) >= point) + goto fail; + NEXT; + + CASE (at_dot): + DEBUG_PRINT1 ("EXECUTING at_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) != point) + goto fail; + NEXT; + + CASE (after_dot): + DEBUG_PRINT1 ("EXECUTING after_dot.\n"); + if (PTR_CHAR_POS ((unsigned char *) d) <= point) + goto fail; + NEXT; + + CASE (syntaxspec): + DEBUG_PRINT2 ("EXECUTING syntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchsyntax; + + CASE (wordchar): + DEBUG_PRINT1 ("EXECUTING Emacs wordchar.\n"); + mcnt = (int) Sword; + matchsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) != (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + NEXT; + + CASE (notsyntaxspec): + DEBUG_PRINT2 ("EXECUTING notsyntaxspec %d.\n", mcnt); + mcnt = *p++; + goto matchnotsyntax; + + CASE (notwordchar): + DEBUG_PRINT1 ("EXECUTING Emacs notwordchar.\n"); + mcnt = (int) Sword; + matchnotsyntax: + PREFETCH (); + /* Can't use *d++ here; SYNTAX may be an unsafe macro. */ + d++; + if (SYNTAX (d[-1]) == (enum syntaxcode) mcnt) + goto fail; + SET_REGS_MATCHED (); + NEXT; + +#else /* not emacs */ + CASE (wordchar): + DEBUG_PRINT1 ("EXECUTING non-Emacs wordchar.\n"); + PREFETCH (); + if (!WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + NEXT; + + CASE (notwordchar): + DEBUG_PRINT1 ("EXECUTING non-Emacs notwordchar.\n"); + PREFETCH (); + if (WORDCHAR_P (d)) + goto fail; + SET_REGS_MATCHED (); + d++; + NEXT; +#endif /* not emacs */ + +#ifndef __GNUC__ + default: + abort (); + } + continue; /* Successfully executed one pattern command; keep going. */ +#endif + + + /* We goto here if a matching operation fails. */ + fail: + if (!FAIL_STACK_EMPTY ()) + { /* A restart point is known. Restore to that state. */ + DEBUG_PRINT1 ("\nFAIL:\n"); + POP_FAILURE_POINT (d, p, + lowest_active_reg, highest_active_reg, + regstart, regend, reg_info); + + /* If this failure point is a dummy, try the next one. */ + if (!p) + goto fail; + + /* If we failed to the end of the pattern, don't examine *p. */ + assert (p <= pend); + if (p < pend) + { + boolean is_a_jump_n = false; + + /* If failed to a backwards jump that's part of a repetition + loop, need to pop this failure point and use the next one. */ + switch ((re_opcode_t) *p) + { + case jump_n: + is_a_jump_n = true; + case maybe_pop_jump: + case pop_failure_jump: + case jump: + p1 = p + 1; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + + if ((is_a_jump_n && (re_opcode_t) *p1 == succeed_n) + || (!is_a_jump_n + && (re_opcode_t) *p1 == on_failure_jump)) + goto fail; + break; + default: + /* do nothing */ ; + } + } + + if (d >= string1 && d <= end1) + dend = end_match_1; + } + else + break; /* Matching at this starting point really fails. */ + } /* for (;;) */ + + if (best_regs_set) + goto restore_best_regs; + + FREE_VARIABLES (); + + return -1; /* Failure to match. */ +} /* re_match_2 */ + +/* Subroutine definitions for re_match_2. */ + + +/* We are passed P pointing to a register number after a start_memory. + + Return true if the pattern up to the corresponding stop_memory can + match the empty string, and false otherwise. + + If we find the matching stop_memory, sets P to point to one past its number. + Otherwise, sets P to an undefined byte less than or equal to END. + + We don't handle duplicates properly (yet). */ + +static boolean +PREFIX(group_match_null_string_p) (UCHAR_T **p, UCHAR_T *end, + PREFIX(register_info_type) *reg_info) +{ + int mcnt; + /* Point to after the args to the start_memory. */ + UCHAR_T *p1 = *p + 2; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and return true or + false, as appropriate, when we get to one that can't, or to the + matching stop_memory. */ + + switch ((re_opcode_t) *p1) + { + /* Could be either a loop or a series of alternatives. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + /* If the next operation is not a jump backwards in the + pattern. */ + + if (mcnt >= 0) + { + /* Go through the on_failure_jumps of the alternatives, + seeing if any of the alternatives cannot match nothing. + The last alternative starts with only a jump, + whereas the rest start with on_failure_jump and end + with a jump, e.g., here is the pattern for `a|b|c': + + /on_failure_jump/0/6/exactn/1/a/jump_past_alt/0/6 + /on_failure_jump/0/6/exactn/1/b/jump_past_alt/0/3 + /exactn/1/c + + So, we have to first go through the first (n-1) + alternatives and then deal with the last one separately. */ + + + /* Deal with the first (n-1) alternatives, which start + with an on_failure_jump (see above) that jumps to right + past a jump_past_alt. */ + + while ((re_opcode_t) p1[mcnt-(1+OFFSET_ADDRESS_SIZE)] == + jump_past_alt) + { + /* `mcnt' holds how many bytes long the alternative + is, including the ending `jump_past_alt' and + its number. */ + + if (!PREFIX(alt_match_null_string_p) (p1, p1 + mcnt - + (1 + OFFSET_ADDRESS_SIZE), + reg_info)) + return false; + + /* Move to right after this alternative, including the + jump_past_alt. */ + p1 += mcnt; + + /* Break if it's the beginning of an n-th alternative + that doesn't begin with an on_failure_jump. */ + if ((re_opcode_t) *p1 != on_failure_jump) + break; + + /* Still have to check that it's not an n-th + alternative that starts with an on_failure_jump. */ + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if ((re_opcode_t) p1[mcnt-(1+OFFSET_ADDRESS_SIZE)] != + jump_past_alt) + { + /* Get to the beginning of the n-th alternative. */ + p1 -= 1 + OFFSET_ADDRESS_SIZE; + break; + } + } + + /* Deal with the last alternative: go back and get number + of the `jump_past_alt' just before it. `mcnt' contains + the length of the alternative. */ + EXTRACT_NUMBER (mcnt, p1 - OFFSET_ADDRESS_SIZE); + + if (!PREFIX(alt_match_null_string_p) (p1, p1 + mcnt, reg_info)) + return false; + + p1 += mcnt; /* Get past the n-th alternative. */ + } /* if mcnt > 0 */ + break; + + + case stop_memory: + assert (p1[1] == **p); + *p = p1 + 2; + return true; + + + default: + if (!PREFIX(common_op_match_null_string_p) (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return false; +} /* group_match_null_string_p */ + + +/* Similar to group_match_null_string_p, but doesn't deal with alternatives: + It expects P to be the first byte of a single alternative and END one + byte past the last. The alternative can contain groups. */ + +static boolean +PREFIX(alt_match_null_string_p) (UCHAR_T *p, UCHAR_T *end, + PREFIX(register_info_type) *reg_info) +{ + int mcnt; + UCHAR_T *p1 = p; + + while (p1 < end) + { + /* Skip over opcodes that can match nothing, and break when we get + to one that can't. */ + + switch ((re_opcode_t) *p1) + { + /* It's a loop. */ + case on_failure_jump: + p1++; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + break; + + default: + if (!PREFIX(common_op_match_null_string_p) (&p1, end, reg_info)) + return false; + } + } /* while p1 < end */ + + return true; +} /* alt_match_null_string_p */ + + +/* Deals with the ops common to group_match_null_string_p and + alt_match_null_string_p. + + Sets P to one after the op and its arguments, if any. */ + +static boolean +PREFIX(common_op_match_null_string_p) (UCHAR_T **p, UCHAR_T *end, + PREFIX(register_info_type) *reg_info) +{ + int mcnt; + boolean ret; + int reg_no; + UCHAR_T *p1 = *p; + + switch ((re_opcode_t) *p1++) + { + case no_op: + case begline: + case endline: + case begbuf: + case endbuf: + case wordbeg: + case wordend: + case wordbound: + case notwordbound: +#ifdef emacs + case before_dot: + case at_dot: + case after_dot: +#endif + break; + + case start_memory: + reg_no = *p1; + assert (reg_no > 0 && reg_no <= MAX_REGNUM); + ret = PREFIX(group_match_null_string_p) (&p1, end, reg_info); + + /* Have to set this here in case we're checking a group which + contains a group and a back reference to it. */ + + if (REG_MATCH_NULL_STRING_P (reg_info[reg_no]) == MATCH_NULL_UNSET_VALUE) + REG_MATCH_NULL_STRING_P (reg_info[reg_no]) = ret; + + if (!ret) + return false; + break; + + /* If this is an optimized succeed_n for zero times, make the jump. */ + case jump: + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + if (mcnt >= 0) + p1 += mcnt; + else + return false; + break; + + case succeed_n: + /* Get to the number of times to succeed. */ + p1 += OFFSET_ADDRESS_SIZE; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + + if (mcnt == 0) + { + p1 -= 2 * OFFSET_ADDRESS_SIZE; + EXTRACT_NUMBER_AND_INCR (mcnt, p1); + p1 += mcnt; + } + else + return false; + break; + + case duplicate: + if (!REG_MATCH_NULL_STRING_P (reg_info[*p1])) + return false; + break; + + case set_number_at: + p1 += 2 * OFFSET_ADDRESS_SIZE; + + default: + /* All other opcodes mean we cannot match the empty string. */ + return false; + } + + *p = p1; + return true; +} /* common_op_match_null_string_p */ + + +/* Return zero if TRANSLATE[S1] and TRANSLATE[S2] are identical for LEN + bytes; nonzero otherwise. */ + +static int +PREFIX(bcmp_translate) (const CHAR_T *s1, const CHAR_T *s2, + register int len, + RE_TRANSLATE_TYPE translate) +{ + register const UCHAR_T *p1 = (const UCHAR_T *) s1; + register const UCHAR_T *p2 = (const UCHAR_T *) s2; + while (len) + { +#ifdef WCHAR + if (((*p1<=0xff)?translate[*p1++]:*p1++) + != ((*p2<=0xff)?translate[*p2++]:*p2++)) + return 1; +#else /* BYTE */ + if (translate[*p1++] != translate[*p2++]) return 1; +#endif /* WCHAR */ + len--; + } + return 0; +} + + +#else /* not INSIDE_RECURSION */ + +/* Entry points for GNU code. */ + +/* re_compile_pattern is the GNU regular expression compiler: it + compiles PATTERN (of length SIZE) and puts the result in BUFP. + Returns 0 if the pattern was valid, otherwise an error string. + + Assumes the `allocated' (and perhaps `buffer') and `translate' fields + are set in BUFP on entry. + + We call regex_compile to do the actual compilation. */ + +const char * +re_compile_pattern (const char *pattern, + size_t length, + struct re_pattern_buffer *bufp) +{ + reg_errcode_t ret; + + /* GNU code is written to assume at least RE_NREGS registers will be set + (and at least one extra will be -1). */ + bufp->regs_allocated = REGS_UNALLOCATED; + + /* And GNU code determines whether or not to get register information + by passing null for the REGS argument to re_match, etc., not by + setting no_sub. */ + bufp->no_sub = 0; + + /* Match anchors at newline. */ + bufp->newline_anchor = 1; + +# ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + ret = wcs_regex_compile (pattern, length, re_syntax_options, bufp); + else +# endif + ret = byte_regex_compile (pattern, length, re_syntax_options, bufp); + + if (!ret) + return NULL; + return gettext (re_error_msgid + re_error_msgid_idx[(int) ret]); +} +#ifdef _LIBC +weak_alias (__re_compile_pattern, re_compile_pattern) +#endif + +/* Entry points compatible with 4.2 BSD regex library. We don't define + them unless specifically requested. */ + +#if defined _REGEX_RE_COMP || defined _LIBC + +/* BSD has one and only one pattern buffer. */ +static struct re_pattern_buffer re_comp_buf; + +char * +#ifdef _LIBC +/* Make these definitions weak in libc, so POSIX programs can redefine + these names if they don't use our functions, and still use + regcomp/regexec below without link errors. */ +weak_function +#endif +re_comp (const char *s) +{ + reg_errcode_t ret; + + if (!s) + { + if (!re_comp_buf.buffer) + return gettext ("No previous regular expression"); + return 0; + } + + if (!re_comp_buf.buffer) + { + re_comp_buf.buffer = malloc (200); + if (re_comp_buf.buffer == NULL) + return (char *) gettext (re_error_msgid + + re_error_msgid_idx[(int) REG_ESPACE]); + re_comp_buf.allocated = 200; + + re_comp_buf.fastmap = malloc (1 << BYTEWIDTH); + if (re_comp_buf.fastmap == NULL) + return (char *) gettext (re_error_msgid + + re_error_msgid_idx[(int) REG_ESPACE]); + } + + /* Since `re_exec' always passes NULL for the `regs' argument, we + don't need to initialize the pattern buffer fields which affect it. */ + + /* Match anchors at newlines. */ + re_comp_buf.newline_anchor = 1; + +# ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + ret = wcs_regex_compile (s, strlen (s), re_syntax_options, &re_comp_buf); + else +# endif + ret = byte_regex_compile (s, strlen (s), re_syntax_options, &re_comp_buf); + + if (!ret) + return NULL; + + /* Yes, we're discarding `const' here if !HAVE_LIBINTL. */ + return (char *) gettext (re_error_msgid + re_error_msgid_idx[(int) ret]); +} + + +int +#ifdef _LIBC +weak_function +#endif +re_exec (const char *s) +{ + const int len = strlen (s); + return + 0 <= re_search (&re_comp_buf, s, len, 0, len, 0); +} + +#endif /* _REGEX_RE_COMP */ + +/* POSIX.2 functions. Don't define these for Emacs. */ + +#ifndef emacs + +/* regcomp takes a regular expression as a string and compiles it. + + PREG is a regex_t *. We do not expect any fields to be initialized, + since POSIX says we shouldn't. Thus, we set + + `buffer' to the compiled pattern; + `used' to the length of the compiled pattern; + `syntax' to RE_SYNTAX_POSIX_EXTENDED if the + REG_EXTENDED bit in CFLAGS is set; otherwise, to + RE_SYNTAX_POSIX_BASIC; + `newline_anchor' to REG_NEWLINE being set in CFLAGS; + `fastmap' to an allocated space for the fastmap; + `fastmap_accurate' to zero; + `re_nsub' to the number of subexpressions in PATTERN. + + PATTERN is the address of the pattern string. + + CFLAGS is a series of bits which affect compilation. + + If REG_EXTENDED is set, we use POSIX extended syntax; otherwise, we + use POSIX basic syntax. + + If REG_NEWLINE is set, then . and [^...] don't match newline. + Also, regexec will try a match beginning after every newline. + + If REG_ICASE is set, then we considers upper- and lowercase + versions of letters to be equivalent when matching. + + If REG_NOSUB is set, then when PREG is passed to regexec, that + routine will report only success or failure, and nothing about the + registers. + + It returns 0 if it succeeds, nonzero if it doesn't. (See regex.h for + the return codes and their meanings.) */ + +int +regcomp (regex_t *preg, const char *pattern, int cflags) +{ + reg_errcode_t ret; + reg_syntax_t syntax + = (cflags & REG_EXTENDED) ? + RE_SYNTAX_POSIX_EXTENDED : RE_SYNTAX_POSIX_BASIC; + + /* regex_compile will allocate the space for the compiled pattern. */ + preg->buffer = 0; + preg->allocated = 0; + preg->used = 0; + + /* Try to allocate space for the fastmap. */ + preg->fastmap = malloc (1 << BYTEWIDTH); + + if (cflags & REG_ICASE) + { + unsigned i; + + preg->translate = malloc (CHAR_SET_SIZE + * sizeof (*(RE_TRANSLATE_TYPE)0)); + if (preg->translate == NULL) + return (int) REG_ESPACE; + + /* Map uppercase characters to corresponding lowercase ones. */ + for (i = 0; i < CHAR_SET_SIZE; i++) + preg->translate[i] = ISUPPER (i) ? TOLOWER (i) : i; + } + else + preg->translate = NULL; + + /* If REG_NEWLINE is set, newlines are treated differently. */ + if (cflags & REG_NEWLINE) + { /* REG_NEWLINE implies neither . nor [^...] match newline. */ + syntax &= ~RE_DOT_NEWLINE; + syntax |= RE_HAT_LISTS_NOT_NEWLINE; + /* It also changes the matching behavior. */ + preg->newline_anchor = 1; + } + else + preg->newline_anchor = 0; + + preg->no_sub = !!(cflags & REG_NOSUB); + + /* POSIX says a null character in the pattern terminates it, so we + can use strlen here in compiling the pattern. */ +# ifdef MBS_SUPPORT + if (MB_CUR_MAX != 1) + ret = wcs_regex_compile (pattern, strlen (pattern), syntax, preg); + else +# endif + ret = byte_regex_compile (pattern, strlen (pattern), syntax, preg); + + /* POSIX doesn't distinguish between an unmatched open-group and an + unmatched close-group: both are REG_EPAREN. */ + if (ret == REG_ERPAREN) ret = REG_EPAREN; + + if (ret == REG_NOERROR && preg->fastmap) + { + /* Compute the fastmap now, since regexec cannot modify the pattern + buffer. */ + if (re_compile_fastmap (preg) == -2) + { + /* Some error occurred while computing the fastmap, just forget + about it. */ + free (preg->fastmap); + preg->fastmap = NULL; + } + } + + return (int) ret; +} +#ifdef _LIBC +weak_alias (__regcomp, regcomp) +#endif + + +/* regexec searches for a given pattern, specified by PREG, in the + string STRING. + + If NMATCH is zero or REG_NOSUB was set in the cflags argument to + `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at + least NMATCH elements, and we set them to the offsets of the + corresponding matched substrings. + + EFLAGS specifies `execution flags' which affect matching: if + REG_NOTBOL is set, then ^ does not match at the beginning of the + string; if REG_NOTEOL is set, then $ does not match at the end. + + We return 0 if we find a match and REG_NOMATCH if not. */ + +int +regexec (const regex_t *preg, const char *string, + size_t nmatch, regmatch_t pmatch[], int eflags) +{ + int ret; + struct re_registers regs; + regex_t private_preg; + int len = strlen (string); + boolean want_reg_info = !preg->no_sub && nmatch > 0; + + private_preg = *preg; + + private_preg.not_bol = !!(eflags & REG_NOTBOL); + private_preg.not_eol = !!(eflags & REG_NOTEOL); + + /* The user has told us exactly how many registers to return + information about, via `nmatch'. We have to pass that on to the + matching routines. */ + private_preg.regs_allocated = REGS_FIXED; + + if (want_reg_info) + { + regs.num_regs = nmatch; + regs.start = TALLOC (nmatch * 2, regoff_t); + if (regs.start == NULL) + return (int) REG_NOMATCH; + regs.end = regs.start + nmatch; + } + + /* Perform the searching operation. */ + ret = re_search (&private_preg, string, len, + /* start: */ 0, /* range: */ len, + want_reg_info ? ®s : 0); + + /* Copy the register information to the POSIX structure. */ + if (want_reg_info) + { + if (ret >= 0) + { + unsigned r; + + for (r = 0; r < nmatch; r++) + { + pmatch[r].rm_so = regs.start[r]; + pmatch[r].rm_eo = regs.end[r]; + } + } + + /* If we needed the temporary register info, free the space now. */ + free (regs.start); + } + + /* We want zero return to mean success, unlike `re_search'. */ + return ret >= 0 ? (int) REG_NOERROR : (int) REG_NOMATCH; +} +#ifdef _LIBC +weak_alias (__regexec, regexec) +#endif + + +/* Returns a message corresponding to an error code, ERRCODE, returned + from either regcomp or regexec. We don't use PREG here. */ + +size_t +regerror (int errcode, const regex_t *preg, char *errbuf, size_t errbuf_size) +{ + const char *msg; + size_t msg_size; + + if (errcode < 0 + || errcode >= (int) (sizeof (re_error_msgid_idx) + / sizeof (re_error_msgid_idx[0]))) + /* Only error codes returned by the rest of the code should be passed + to this routine. If we are given anything else, or if other regex + code generates an invalid error code, then the program has a bug. + Dump core so we can fix it. */ + abort (); + + msg = gettext (re_error_msgid + re_error_msgid_idx[errcode]); + + msg_size = strlen (msg) + 1; /* Includes the null. */ + + if (errbuf_size != 0) + { + if (msg_size > errbuf_size) + { +#if defined HAVE_MEMPCPY || defined _LIBC + *((char *) __mempcpy (errbuf, msg, errbuf_size - 1)) = '\0'; +#else + memcpy (errbuf, msg, errbuf_size - 1); + errbuf[errbuf_size - 1] = 0; +#endif + } + else + memcpy (errbuf, msg, msg_size); + } + + return msg_size; +} +#ifdef _LIBC +weak_alias (__regerror, regerror) +#endif + + +/* Free dynamically allocated space used by PREG. */ + +void +regfree (regex_t *preg) +{ + if (preg->buffer != NULL) + free (preg->buffer); + preg->buffer = NULL; + + preg->allocated = 0; + preg->used = 0; + + if (preg->fastmap != NULL) + free (preg->fastmap); + preg->fastmap = NULL; + preg->fastmap_accurate = 0; + + if (preg->translate != NULL) + free (preg->translate); + preg->translate = NULL; +} +#ifdef _LIBC +weak_alias (__regfree, regfree) +#endif + +#endif /* not emacs */ + +#endif /* not INSIDE_RECURSION */ + + +#undef STORE_NUMBER +#undef STORE_NUMBER_AND_INCR +#undef EXTRACT_NUMBER +#undef EXTRACT_NUMBER_AND_INCR + +#undef DEBUG_PRINT_COMPILED_PATTERN +#undef DEBUG_PRINT_DOUBLE_STRING + +#undef INIT_FAIL_STACK +#undef RESET_FAIL_STACK +#undef DOUBLE_FAIL_STACK +#undef PUSH_PATTERN_OP +#undef PUSH_FAILURE_POINTER +#undef PUSH_FAILURE_INT +#undef PUSH_FAILURE_ELT +#undef POP_FAILURE_POINTER +#undef POP_FAILURE_INT +#undef POP_FAILURE_ELT +#undef DEBUG_PUSH +#undef DEBUG_POP +#undef PUSH_FAILURE_POINT +#undef POP_FAILURE_POINT + +#undef REG_UNSET_VALUE +#undef REG_UNSET + +#undef PATFETCH +#undef PATFETCH_RAW +#undef PATUNFETCH +#undef TRANSLATE + +#undef INIT_BUF_SIZE +#undef GET_BUFFER_SPACE +#undef BUF_PUSH +#undef BUF_PUSH_2 +#undef BUF_PUSH_3 +#undef STORE_JUMP +#undef STORE_JUMP2 +#undef INSERT_JUMP +#undef INSERT_JUMP2 +#undef EXTEND_BUFFER +#undef GET_UNSIGNED_NUMBER +#undef FREE_STACK_RETURN + +# undef POINTER_TO_OFFSET +# undef MATCHING_IN_FRST_STRING +# undef PREFETCH +# undef AT_STRINGS_BEG +# undef AT_STRINGS_END +# undef WORDCHAR_P +# undef FREE_VAR +# undef FREE_VARIABLES +# undef NO_HIGHEST_ACTIVE_REG +# undef NO_LOWEST_ACTIVE_REG + +# undef CHAR_T +# undef UCHAR_T +# undef COMPILED_BUFFER_VAR +# undef OFFSET_ADDRESS_SIZE +# undef CHAR_CLASS_SIZE +# undef PREFIX +# undef ARG_PREFIX +# undef PUT_CHAR +# undef BYTE +# undef WCHAR + +# define DEFINED_ONCE diff --git a/contrib/diff/lib/regex.h b/contrib/diff/lib/regex.h new file mode 100644 index 0000000..b4bbc01 --- /dev/null +++ b/contrib/diff/lib/regex.h @@ -0,0 +1,556 @@ +/* Definitions for data structures and routines for the regular + expression library. + Copyright (C) 1985,1989-93,1995-98,2000,2001,2002,2003 + Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _REGEX_H +#define _REGEX_H 1 + +#include + +/* Allow the use in C++ code. */ +#ifdef __cplusplus +extern "C" { +#endif + +/* POSIX says that must be included (by the caller) before + . */ + +#if !defined _POSIX_C_SOURCE && !defined _POSIX_SOURCE && defined VMS +/* VMS doesn't have `size_t' in , even though POSIX says it + should be there. */ +# include +#endif + +/* The following two types have to be signed and unsigned integer type + wide enough to hold a value of a pointer. For most ANSI compilers + ptrdiff_t and size_t should be likely OK. Still size of these two + types is 2 for Microsoft C. Ugh... */ +typedef long int s_reg_t; +typedef unsigned long int active_reg_t; + +/* The following bits are used to determine the regexp syntax we + recognize. The set/not-set meanings are chosen so that Emacs syntax + remains the value 0. The bits are given in alphabetical order, and + the definitions shifted by one from the previous bit; thus, when we + add or remove a bit, only one other definition need change. */ +typedef unsigned long int reg_syntax_t; + +/* If this bit is not set, then \ inside a bracket expression is literal. + If set, then such a \ quotes the following character. */ +#define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1) + +/* If this bit is not set, then + and ? are operators, and \+ and \? are + literals. + If set, then \+ and \? are operators and + and ? are literals. */ +#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1) + +/* If this bit is set, then character classes are supported. They are: + [:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:], + [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:]. + If not set, then character classes are not supported. */ +#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1) + +/* If this bit is set, then ^ and $ are always anchors (outside bracket + expressions, of course). + If this bit is not set, then it depends: + ^ is an anchor if it is at the beginning of a regular + expression or after an open-group or an alternation operator; + $ is an anchor if it is at the end of a regular expression, or + before a close-group or an alternation operator. + + This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because + POSIX draft 11.2 says that * etc. in leading positions is undefined. + We already implemented a previous draft which made those constructs + invalid, though, so we haven't changed the code back. */ +#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1) + +/* If this bit is set, then special characters are always special + regardless of where they are in the pattern. + If this bit is not set, then special characters are special only in + some contexts; otherwise they are ordinary. Specifically, + * + ? and intervals are only special when not after the beginning, + open-group, or alternation operator. */ +#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1) + +/* If this bit is set, then *, +, ?, and { cannot be first in an re or + immediately after an alternation or begin-group operator. */ +#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1) + +/* If this bit is set, then . matches newline. + If not set, then it doesn't. */ +#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1) + +/* If this bit is set, then . doesn't match NUL. + If not set, then it does. */ +#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1) + +/* If this bit is set, nonmatching lists [^...] do not match newline. + If not set, they do. */ +#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1) + +/* If this bit is set, either \{...\} or {...} defines an + interval, depending on RE_NO_BK_BRACES. + If not set, \{, \}, {, and } are literals. */ +#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1) + +/* If this bit is set, +, ? and | aren't recognized as operators. + If not set, they are. */ +#define RE_LIMITED_OPS (RE_INTERVALS << 1) + +/* If this bit is set, newline is an alternation operator. + If not set, newline is literal. */ +#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1) + +/* If this bit is set, then `{...}' defines an interval, and \{ and \} + are literals. + If not set, then `\{...\}' defines an interval. */ +#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1) + +/* If this bit is set, (...) defines a group, and \( and \) are literals. + If not set, \(...\) defines a group, and ( and ) are literals. */ +#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1) + +/* If this bit is set, then \ matches . + If not set, then \ is a back-reference. */ +#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1) + +/* If this bit is set, then | is an alternation operator, and \| is literal. + If not set, then \| is an alternation operator, and | is literal. */ +#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1) + +/* If this bit is set, then an ending range point collating higher + than the starting range point, as in [z-a], is invalid. + If not set, then when ending range point collates higher than the + starting range point, the range is ignored. */ +#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1) + +/* If this bit is set, then an unmatched ) is ordinary. + If not set, then an unmatched ) is invalid. */ +#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1) + +/* If this bit is set, succeed as soon as we match the whole pattern, + without further backtracking. */ +#define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1) + +/* If this bit is set, do not process the GNU regex operators. + If not set, then the GNU regex operators are recognized. */ +#define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1) + +/* If this bit is set, turn on internal regex debugging. + If not set, and debugging was on, turn it off. + This only works if regex.c is compiled -DDEBUG. + We define this bit always, so that all that's needed to turn on + debugging is to recompile regex.c; the calling code can always have + this bit set, and it won't affect anything in the normal case. */ +#define RE_DEBUG (RE_NO_GNU_OPS << 1) + +/* If this bit is set, a syntactically invalid interval is treated as + a string of ordinary characters. For example, the ERE 'a{1' is + treated as 'a\{1'. */ +#define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1) + +/* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +#define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1) + +/* This global variable defines the particular regexp syntax to use (for + some interfaces). When a regexp is compiled, the syntax used is + stored in the pattern buffer, so changing this does not affect + already-compiled regexps. */ +extern reg_syntax_t re_syntax_options; + +/* Define combinations of the above bits for the standard possibilities. + (The [[[ comments delimit what gets put into the Texinfo file, so + don't delete them!) */ +/* [[[begin syntaxes]]] */ +#define RE_SYNTAX_EMACS 0 + +#define RE_SYNTAX_AWK \ + (RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \ + | RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \ + | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GNU_AWK \ + ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ + & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS \ + | RE_CONTEXT_INVALID_OPS )) + +#define RE_SYNTAX_POSIX_AWK \ + (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ + | RE_INTERVALS | RE_NO_GNU_OPS) + +#define RE_SYNTAX_GREP \ + (RE_BK_PLUS_QM | RE_CHAR_CLASSES \ + | RE_HAT_LISTS_NOT_NEWLINE | RE_INTERVALS \ + | RE_NEWLINE_ALT) + +#define RE_SYNTAX_EGREP \ + (RE_CHAR_CLASSES | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_HAT_LISTS_NOT_NEWLINE \ + | RE_NEWLINE_ALT | RE_NO_BK_PARENS \ + | RE_NO_BK_VBAR) + +#define RE_SYNTAX_POSIX_EGREP \ + (RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \ + | RE_INVALID_INTERVAL_ORD) + +/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */ +#define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC + +#define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC + +/* Syntax bits common to both basic and extended POSIX regex syntax. */ +#define _RE_SYNTAX_POSIX_COMMON \ + (RE_CHAR_CLASSES | RE_DOT_NEWLINE | RE_DOT_NOT_NULL \ + | RE_INTERVALS | RE_NO_EMPTY_RANGES) + +#define RE_SYNTAX_POSIX_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM) + +/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes + RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this + isn't minimal, since other operators, such as \`, aren't disabled. */ +#define RE_SYNTAX_POSIX_MINIMAL_BASIC \ + (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS) + +#define RE_SYNTAX_POSIX_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_VBAR \ + | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD) + +/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is + removed and RE_NO_BK_REFS is added. */ +#define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \ + (_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \ + | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \ + | RE_NO_BK_PARENS | RE_NO_BK_REFS \ + | RE_NO_BK_VBAR | RE_UNMATCHED_RIGHT_PAREN_ORD) +/* [[[end syntaxes]]] */ + +/* Maximum number of duplicates an interval can allow. Some systems + (erroneously) define this in other header files, but we want our + value, so remove any previous define. */ +#ifdef RE_DUP_MAX +# undef RE_DUP_MAX +#endif +/* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */ +#define RE_DUP_MAX (0x7fff) + + +/* POSIX `cflags' bits (i.e., information for `regcomp'). */ + +/* If this bit is set, then use extended regular expression syntax. + If not set, then use basic regular expression syntax. */ +#define REG_EXTENDED 1 + +/* If this bit is set, then ignore case when matching. + If not set, then case is significant. */ +#define REG_ICASE (REG_EXTENDED << 1) + +/* If this bit is set, then anchors do not match at newline + characters in the string. + If not set, then anchors do match at newlines. */ +#define REG_NEWLINE (REG_ICASE << 1) + +/* If this bit is set, then report only success or fail in regexec. + If not set, then returns differ between not matching and errors. */ +#define REG_NOSUB (REG_NEWLINE << 1) + + +/* POSIX `eflags' bits (i.e., information for regexec). */ + +/* If this bit is set, then the beginning-of-line operator doesn't match + the beginning of the string (presumably because it's not the + beginning of a line). + If not set, then the beginning-of-line operator does match the + beginning of the string. */ +#define REG_NOTBOL 1 + +/* Like REG_NOTBOL, except for the end-of-line. */ +#define REG_NOTEOL (1 << 1) + + +/* If any error codes are removed, changed, or added, update the + `re_error_msg' table in regex.c. */ +typedef enum +{ +#ifdef _XOPEN_SOURCE + REG_ENOSYS = -1, /* This will never happen for this implementation. */ +#endif + + REG_NOERROR = 0, /* Success. */ + REG_NOMATCH, /* Didn't find a match (for regexec). */ + + /* POSIX regcomp return error codes. (In the order listed in the + standard.) */ + REG_BADPAT, /* Invalid pattern. */ + REG_ECOLLATE, /* Not implemented. */ + REG_ECTYPE, /* Invalid character class name. */ + REG_EESCAPE, /* Trailing backslash. */ + REG_ESUBREG, /* Invalid back reference. */ + REG_EBRACK, /* Unmatched left bracket. */ + REG_EPAREN, /* Parenthesis imbalance. */ + REG_EBRACE, /* Unmatched \{. */ + REG_BADBR, /* Invalid contents of \{\}. */ + REG_ERANGE, /* Invalid range end. */ + REG_ESPACE, /* Ran out of memory. */ + REG_BADRPT, /* No preceding re for repetition op. */ + + /* Error codes we've added. */ + REG_EEND, /* Premature end. */ + REG_ESIZE, /* Compiled pattern bigger than 2^16 bytes. */ + REG_ERPAREN /* Unmatched ) or \); not returned from regcomp. */ +} reg_errcode_t; + +/* This data structure represents a compiled pattern. Before calling + the pattern compiler, the fields `buffer', `allocated', `fastmap', + `translate', and `no_sub' can be set. After the pattern has been + compiled, the `re_nsub' field is available. All other fields are + private to the regex routines. */ + +#ifndef RE_TRANSLATE_TYPE +# define RE_TRANSLATE_TYPE char * +#endif + +struct re_pattern_buffer +{ +/* [[[begin pattern_buffer]]] */ + /* Space that holds the compiled pattern. It is declared as + `unsigned char *' because its elements are + sometimes used as array indexes. */ + unsigned char *buffer; + + /* Number of bytes to which `buffer' points. */ + unsigned long int allocated; + + /* Number of bytes actually used in `buffer'. */ + unsigned long int used; + + /* Syntax setting with which the pattern was compiled. */ + reg_syntax_t syntax; + + /* Pointer to a fastmap, if any, otherwise zero. re_search uses + the fastmap, if there is one, to skip over impossible + starting points for matches. */ + char *fastmap; + + /* Either a translate table to apply to all characters before + comparing them, or zero for no translation. The translation + is applied to a pattern when it is compiled and to a string + when it is matched. */ + RE_TRANSLATE_TYPE translate; + + /* Number of subexpressions found by the compiler. */ + size_t re_nsub; + + /* Zero if this pattern cannot match the empty string, one else. + Well, in truth it's used only in `re_search_2', to see + whether or not we should use the fastmap, so we don't set + this absolutely perfectly; see `re_compile_fastmap' (the + `duplicate' case). */ + unsigned can_be_null : 1; + + /* If REGS_UNALLOCATED, allocate space in the `regs' structure + for `max (RE_NREGS, re_nsub + 1)' groups. + If REGS_REALLOCATE, reallocate space if necessary. + If REGS_FIXED, use what's there. */ +#define REGS_UNALLOCATED 0 +#define REGS_REALLOCATE 1 +#define REGS_FIXED 2 + unsigned regs_allocated : 2; + + /* Set to zero when `regex_compile' compiles a pattern; set to one + by `re_compile_fastmap' if it updates the fastmap. */ + unsigned fastmap_accurate : 1; + + /* If set, `re_match_2' does not return information about + subexpressions. */ + unsigned no_sub : 1; + + /* If set, a beginning-of-line anchor doesn't match at the + beginning of the string. */ + unsigned not_bol : 1; + + /* Similarly for an end-of-line anchor. */ + unsigned not_eol : 1; + + /* If true, an anchor at a newline matches. */ + unsigned newline_anchor : 1; + +/* [[[end pattern_buffer]]] */ +}; + +typedef struct re_pattern_buffer regex_t; + +/* Type for byte offsets within the string. POSIX mandates this. */ +typedef int regoff_t; + + +/* This is the structure we store register match data in. See + regex.texinfo for a full description of what registers match. */ +struct re_registers +{ + unsigned num_regs; + regoff_t *start; + regoff_t *end; +}; + + +/* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer, + `re_match_2' returns information about at least this many registers + the first time a `regs' structure is passed. */ +#ifndef RE_NREGS +# define RE_NREGS 30 +#endif + + +/* POSIX specification for registers. Aside from the different names than + `re_registers', POSIX uses an array of structures, instead of a + structure of arrays. */ +typedef struct +{ + regoff_t rm_so; /* Byte offset from string's start to substring's start. */ + regoff_t rm_eo; /* Byte offset from string's start to substring's end. */ +} regmatch_t; + +/* Declarations for routines. */ + +/* Sets the current default syntax to SYNTAX, and return the old syntax. + You can also simply assign to the `re_syntax_options' variable. */ +extern reg_syntax_t re_set_syntax (reg_syntax_t syntax); + +/* Compile the regular expression PATTERN, with length LENGTH + and syntax given by the global `re_syntax_options', into the buffer + BUFFER. Return NULL if successful, and an error string if not. */ +extern const char *re_compile_pattern (const char *pattern, size_t length, + struct re_pattern_buffer *buffer); + + +/* Compile a fastmap for the compiled pattern in BUFFER; used to + accelerate searches. Return 0 if successful and -2 if was an + internal error. */ +extern int re_compile_fastmap (struct re_pattern_buffer *buffer); + + +/* Search in the string STRING (with length LENGTH) for the pattern + compiled into BUFFER. Start searching at position START, for RANGE + characters. Return the starting position of the match, -1 for no + match, or -2 for an internal error. Also return register + information in REGS (if REGS and BUFFER->no_sub are nonzero). */ +extern int re_search (struct re_pattern_buffer *buffer, const char *string, + int length, int start, int range, + struct re_registers *regs); + + +/* Like `re_search', but search in the concatenation of STRING1 and + STRING2. Also, stop searching at index START + STOP. */ +extern int re_search_2 (struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, int range, struct re_registers *regs, + int stop); + + +/* Like `re_search', but return how many characters in STRING the regexp + in BUFFER matched, starting at position START. */ +extern int re_match (struct re_pattern_buffer *buffer, const char *string, + int length, int start, struct re_registers *regs); + + +/* Relates to `re_match' as `re_search_2' relates to `re_search'. */ +extern int re_match_2 (struct re_pattern_buffer *buffer, const char *string1, + int length1, const char *string2, int length2, + int start, struct re_registers *regs, int stop); + + +/* Set REGS to hold NUM_REGS registers, storing them in STARTS and + ENDS. Subsequent matches using BUFFER and REGS will use this memory + for recording register information. STARTS and ENDS must be + allocated with malloc, and must each be at least `NUM_REGS * sizeof + (regoff_t)' bytes long. + + If NUM_REGS == 0, then subsequent matches should allocate their own + register data. + + Unless this function is called, the first search or match using + PATTERN_BUFFER will allocate its own register data, without + freeing the old data. */ +extern void re_set_registers (struct re_pattern_buffer *buffer, + struct re_registers *regs, unsigned num_regs, + regoff_t *starts, regoff_t *ends); + +#if defined _REGEX_RE_COMP || defined _LIBC +# ifndef _CRAY +/* 4.2 bsd compatibility. */ +extern char *re_comp (const char *); +extern int re_exec (const char *); +# endif +#endif + +/* GCC 2.95 and later have "__restrict"; C99 compilers have + "restrict", and "configure" may have defined "restrict". */ +#ifndef __restrict +# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)) +# if defined restrict || 199901L <= __STDC_VERSION__ +# define __restrict restrict +# else +# define __restrict +# endif +# endif +#endif +/* gcc 3.1 and up support the [restrict] syntax. */ +#ifndef __restrict_arr +# if __GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1) +# define __restrict_arr __restrict +# else +# define __restrict_arr +# endif +#endif + +/* POSIX compatibility. */ +extern int regcomp (regex_t *__restrict __preg, + const char *__restrict __pattern, + int __cflags); + +extern int regexec (const regex_t *__restrict __preg, + const char *__restrict __string, size_t __nmatch, + regmatch_t __pmatch[__restrict_arr], + int __eflags); + +extern size_t regerror (int __errcode, const regex_t *__preg, + char *__errbuf, size_t __errbuf_size); + +extern void regfree (regex_t *__preg); + + +#ifdef __cplusplus +} +#endif /* C++ */ + +#endif /* regex.h */ + +/* +Local variables: +make-backup-files: t +version-control: t +trim-versions-without-asking: nil +End: +*/ diff --git a/contrib/diff/lib/setmode.c b/contrib/diff/lib/setmode.c new file mode 100644 index 0000000..b2b93a6 --- /dev/null +++ b/contrib/diff/lib/setmode.c @@ -0,0 +1,67 @@ +/* Set a file descriptor's mode to binary or to text. + + Copyright (C) 2001, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert */ + +#if HAVE_CONFIG_H +# include +#endif + +#if HAVE_STDBOOL_H +# include +#else +typedef enum {false = 0, true = 1} bool; +#endif + +#if HAVE_SETMODE_DOS +# include +# if HAVE_FCNTL_H +# include +# endif +# if HAVE_UNISTD_H +# include +# endif +#endif + +#include "setmode.h" +#undef set_binary_mode + +#ifndef __attribute__ +# if __GNUC__ < 3 || __STRICT_ANSI__ +# define __attribute__(x) +# endif +#endif + +/* Set the binary mode of FD to MODE, returning its previous mode. + MODE is 1 for binary and 0 for text. If setting the mode might + cause problems, ignore the request and return MODE. Always return + 1 on POSIX platforms, which do not distinguish between text and + binary. */ + +#if HAVE_SETMODE_DOS +bool +set_binary_mode (int fd, bool mode) +{ + if (isatty (fd)) + return mode; + return setmode (fd, mode ? O_BINARY : O_TEXT) != O_TEXT; +} +#else +static char dummy; +#endif diff --git a/contrib/diff/lib/setmode.h b/contrib/diff/lib/setmode.h new file mode 100644 index 0000000..8381d12 --- /dev/null +++ b/contrib/diff/lib/setmode.h @@ -0,0 +1,27 @@ +/* Set a file descriptor's mode to binary or to text. + + Copyright (C) 2001, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert */ + +#ifndef set_binary_mode +bool set_binary_mode (int, bool); +# if ! HAVE_SETMODE_DOS +# define set_binary_mode(fd, mode) true +# endif +#endif diff --git a/contrib/diff/lib/stdbool_.h b/contrib/diff/lib/stdbool_.h new file mode 100644 index 0000000..e33715a --- /dev/null +++ b/contrib/diff/lib/stdbool_.h @@ -0,0 +1,93 @@ +/* Copyright (C) 2001, 2002, 2003 Free Software Foundation, Inc. + Written by Bruno Haible , 2001. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _STDBOOL_H +#define _STDBOOL_H + +/* ISO C 99 for platforms that lack it. */ + +/* Usage suggestions: + + Programs that use should be aware of some limitations + and standards compliance issues. + + Standards compliance: + + - must be #included before 'bool', 'false', 'true' + can be used. + + - You cannot assume that sizeof (bool) == 1. + + - Programs should not undefine the macros bool, true, and false, + as C99 lists that as an "obsolescent feature". + + Limitations of this substitute, when used in a C89 environment: + + - must be #included before the '_Bool' type can be used. + + - You cannot assume that _Bool is a typedef; it might be a macro. + + - In C99, casts and automatic conversions to '_Bool' or 'bool' are + performed in such a way that every nonzero value gets converted + to 'true', and zero gets converted to 'false'. This doesn't work + with this substitute. With this substitute, only the values 0 and 1 + give the expected result when converted to _Bool' or 'bool'. + + Also, it is suggested that programs use 'bool' rather than '_Bool'; + this isn't required, but 'bool' is more common. */ + + +/* 7.16. Boolean type and values */ + +/* BeOS already #defines false 0, true 1. We use the same + definitions below, but temporarily we have to #undef them. */ +#ifdef __BEOS__ +# include /* defines bool but not _Bool */ +# undef false +# undef true +#endif + +/* For the sake of symbolic names in gdb, we define true and false as + enum constants, not only as macros. + It is tempting to write + typedef enum { false = 0, true = 1 } _Bool; + so that gdb prints values of type 'bool' symbolically. But if we do + this, values of type '_Bool' may promote to 'int' or 'unsigned int' + (see ISO C 99 6.7.2.2.(4)); however, '_Bool' must promote to 'int' + (see ISO C 99 6.3.1.1.(2)). So we add a negative value to the + enum; this ensures that '_Bool' promotes to 'int'. */ +#if !(defined __cplusplus || defined __BEOS__) +# if !@HAVE__BOOL@ +# if defined __SUNPRO_C && (__SUNPRO_C < 0x550 || __STDC__ == 1) + /* Avoid stupid "warning: _Bool is a keyword in ISO C99". */ +# define _Bool signed char +enum { false = 0, true = 1 }; +# else +typedef enum { _Bool_must_promote_to_int = -1, false = 0, true = 1 } _Bool; +# endif +# endif +#else +typedef bool _Bool; +#endif +#define bool _Bool + +/* The other macros must be usable in preprocessor directives. */ +#define false 0 +#define true 1 +#define __bool_true_false_are_defined 1 + +#endif /* _STDBOOL_H */ diff --git a/contrib/diff/lib/strcase.h b/contrib/diff/lib/strcase.h new file mode 100644 index 0000000..f17e648 --- /dev/null +++ b/contrib/diff/lib/strcase.h @@ -0,0 +1,35 @@ +/* Case-insensitive string comparison functions. + Copyright (C) 1995-1996, 2001, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef _STRCASE_H +#define _STRCASE_H + +#include + +/* Compare strings S1 and S2, ignoring case, returning less than, equal to or + greater than zero if S1 is lexicographically less than, equal to or greater + than S2. + Note: This function does not work correctly in multibyte locales. */ +extern int strcasecmp (const char *s1, const char *s2); + +/* Compare no more than N characters of strings S1 and S2, ignoring case, + returning less than, equal to or greater than zero if S1 is + lexicographically less than, equal to or greater than S2. + Note: This function can not work correctly in multibyte locales. */ +extern int strncasecmp (const char *s1, const char *s2, size_t n); + +#endif /* _STRCASE_H */ diff --git a/contrib/diff/lib/strcasecmp.c b/contrib/diff/lib/strcasecmp.c new file mode 100644 index 0000000..cf4ab88 --- /dev/null +++ b/contrib/diff/lib/strcasecmp.c @@ -0,0 +1,66 @@ +/* strcasecmp.c -- case insensitive string comparator + Copyright (C) 1998, 1999 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#ifdef LENGTH_LIMIT +# define STRXCASECMP_FUNCTION strncasecmp +# define STRXCASECMP_DECLARE_N , size_t n +# define LENGTH_LIMIT_EXPR(Expr) Expr +#else +# define STRXCASECMP_FUNCTION strcasecmp +# define STRXCASECMP_DECLARE_N /* empty */ +# define LENGTH_LIMIT_EXPR(Expr) 0 +#endif + +#include +#include + +#define TOLOWER(Ch) (isupper (Ch) ? tolower (Ch) : (Ch)) + +/* Compare {{no more than N characters of }}strings S1 and S2, + ignoring case, returning less than, equal to or + greater than zero if S1 is lexicographically less + than, equal to or greater than S2. */ + +int +STRXCASECMP_FUNCTION (const char *s1, const char *s2 STRXCASECMP_DECLARE_N) +{ + register const unsigned char *p1 = (const unsigned char *) s1; + register const unsigned char *p2 = (const unsigned char *) s2; + unsigned char c1, c2; + + if (p1 == p2 || LENGTH_LIMIT_EXPR (n == 0)) + return 0; + + do + { + c1 = TOLOWER (*p1); + c2 = TOLOWER (*p2); + + if (LENGTH_LIMIT_EXPR (--n == 0) || c1 == '\0') + break; + + ++p1; + ++p2; + } + while (c1 == c2); + + return c1 - c2; +} diff --git a/contrib/diff/lib/strftime.c b/contrib/diff/lib/strftime.c new file mode 100644 index 0000000..dccbaf7 --- /dev/null +++ b/contrib/diff/lib/strftime.c @@ -0,0 +1,1330 @@ +/* Copyright (C) 1991-1999, 2000, 2001, 2003 Free Software Foundation, Inc. + + NOTE: The canonical source of this file is maintained with the GNU C Library. + Bugs can be reported to bug-glibc@prep.ai.mit.edu. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifdef HAVE_CONFIG_H +# include +#endif + +#ifdef _LIBC +# define HAVE_MBLEN 1 +# define HAVE_MBRLEN 1 +# define HAVE_STRUCT_ERA_ENTRY 1 +# define HAVE_TM_GMTOFF 1 +# define HAVE_TM_ZONE 1 +# define HAVE_TZNAME 1 +# define HAVE_TZSET 1 +# define MULTIBYTE_IS_FORMAT_SAFE 1 +# include "../locale/localeinfo.h" +#endif + +#include +#include /* Some systems define `time_t' here. */ + +#ifdef TIME_WITH_SYS_TIME +# include +# include +#else +# ifdef HAVE_SYS_TIME_H +# include +# else +# include +# endif +#endif +#if HAVE_TZNAME +extern char *tzname[]; +#endif + +/* Do multibyte processing if multibytes are supported, unless + multibyte sequences are safe in formats. Multibyte sequences are + safe if they cannot contain byte sequences that look like format + conversion specifications. The GNU C Library uses UTF8 multibyte + encoding, which is safe for formats, but strftime.c can be used + with other C libraries that use unsafe encodings. */ +#define DO_MULTIBYTE (HAVE_MBLEN && ! MULTIBYTE_IS_FORMAT_SAFE) + +#if DO_MULTIBYTE +# if HAVE_MBRLEN +# include +# else + /* Simulate mbrlen with mblen as best we can. */ +# define mbstate_t int +# define mbrlen(s, n, ps) mblen (s, n) +# define mbsinit(ps) (*(ps) == 0) +# endif + static const mbstate_t mbstate_zero; +#endif + +#include +#include +#include +#include + +#ifdef COMPILE_WIDE +# include +# define CHAR_T wchar_t +# define UCHAR_T unsigned int +# define L_(Str) L##Str +# define NLW(Sym) _NL_W##Sym + +# define MEMCPY(d, s, n) __wmemcpy (d, s, n) +# define STRLEN(s) __wcslen (s) + +#else +# define CHAR_T char +# define UCHAR_T unsigned char +# define L_(Str) Str +# define NLW(Sym) Sym + +# define MEMCPY(d, s, n) memcpy (d, s, n) +# define STRLEN(s) strlen (s) + +# ifdef _LIBC +# define MEMPCPY(d, s, n) __mempcpy (d, s, n) +# else +# ifndef HAVE_MEMPCPY +# define MEMPCPY(d, s, n) ((void *) ((char *) memcpy (d, s, n) + (n))) +# endif +# endif +#endif + +#define TYPE_SIGNED(t) ((t) -1 < 0) + +/* Bound on length of the string representing an integer value of type t. + Subtract one for the sign bit if t is signed; + 302 / 1000 is log10 (2) rounded up; + add one for integer division truncation; + add one more for a minus sign if t is signed. */ +#define INT_STRLEN_BOUND(t) \ + ((sizeof (t) * CHAR_BIT - TYPE_SIGNED (t)) * 302 / 1000 + 1 + TYPE_SIGNED (t)) + +#define TM_YEAR_BASE 1900 + +#ifndef __isleap +/* Nonzero if YEAR is a leap year (every 4 years, + except every 100th isn't, and every 400th is). */ +# define __isleap(year) \ + ((year) % 4 == 0 && ((year) % 100 != 0 || (year) % 400 == 0)) +#endif + + +#ifdef _LIBC +# define tzname __tzname +# define tzset __tzset +#endif + +#if !HAVE_TM_GMTOFF +/* Portable standalone applications should supply a "time_r.h" that + declares a POSIX-compliant localtime_r, for the benefit of older + implementations that lack localtime_r or have a nonstandard one. + See the gnulib time_r module for one way to implement this. */ +# include "time_r.h" +# undef __gmtime_r +# undef __localtime_r +# define __gmtime_r gmtime_r +# define __localtime_r localtime_r +#endif + + +#ifdef COMPILE_WIDE +# define memset_space(P, Len) (wmemset (P, L' ', Len), (P) += (Len)) +# define memset_zero(P, Len) (wmemset (P, L'0', Len), (P) += (Len)) +#else +# define memset_space(P, Len) (memset (P, ' ', Len), (P) += (Len)) +# define memset_zero(P, Len) (memset (P, '0', Len), (P) += (Len)) +#endif + +#define add(n, f) \ + do \ + { \ + int _n = (n); \ + int _delta = width - _n; \ + int _incr = _n + (_delta > 0 ? _delta : 0); \ + if ((size_t) _incr >= maxsize - i) \ + return 0; \ + if (p) \ + { \ + if (_delta > 0) \ + { \ + if (pad == L_('0')) \ + memset_zero (p, _delta); \ + else \ + memset_space (p, _delta); \ + } \ + f; \ + p += _n; \ + } \ + i += _incr; \ + } while (0) + +#define cpy(n, s) \ + add ((n), \ + if (to_lowcase) \ + memcpy_lowcase (p, (s), _n LOCALE_ARG); \ + else if (to_uppcase) \ + memcpy_uppcase (p, (s), _n LOCALE_ARG); \ + else \ + MEMCPY ((void *) p, (void const *) (s), _n)) + +#ifdef COMPILE_WIDE +# ifndef USE_IN_EXTENDED_LOCALE_MODEL +# undef __mbsrtowcs_l +# define __mbsrtowcs_l(d, s, l, st, loc) __mbsrtowcs (d, s, l, st) +# endif +# define widen(os, ws, l) \ + { \ + mbstate_t __st; \ + const char *__s = os; \ + memset (&__st, '\0', sizeof (__st)); \ + l = __mbsrtowcs_l (NULL, &__s, 0, &__st, loc); \ + ws = (wchar_t *) alloca ((l + 1) * sizeof (wchar_t)); \ + (void) __mbsrtowcs_l (ws, &__s, l, &__st, loc); \ + } +#endif + + +#if defined _LIBC && defined USE_IN_EXTENDED_LOCALE_MODEL +/* We use this code also for the extended locale handling where the + function gets as an additional argument the locale which has to be + used. To access the values we have to redefine the _NL_CURRENT + macro. */ +# define strftime __strftime_l +# define wcsftime __wcsftime_l +# undef _NL_CURRENT +# define _NL_CURRENT(category, item) \ + (current->values[_NL_ITEM_INDEX (item)].string) +# define LOCALE_ARG , loc +# define LOCALE_PARAM_PROTO , __locale_t loc +# define HELPER_LOCALE_ARG , current +#else +# define LOCALE_PARAM_PROTO +# define LOCALE_ARG +# ifdef _LIBC +# define HELPER_LOCALE_ARG , _NL_CURRENT_DATA (LC_TIME) +# else +# define HELPER_LOCALE_ARG +# endif +#endif + +#ifdef COMPILE_WIDE +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define TOUPPER(Ch, L) __towupper_l (Ch, L) +# define TOLOWER(Ch, L) __towlower_l (Ch, L) +# else +# define TOUPPER(Ch, L) towupper (Ch) +# define TOLOWER(Ch, L) towlower (Ch) +# endif +#else +# ifdef _LIBC +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define TOUPPER(Ch, L) __toupper_l (Ch, L) +# define TOLOWER(Ch, L) __tolower_l (Ch, L) +# else +# define TOUPPER(Ch, L) toupper (Ch) +# define TOLOWER(Ch, L) tolower (Ch) +# endif +# else +# define TOUPPER(Ch, L) (islower (Ch) ? toupper (Ch) : (Ch)) +# define TOLOWER(Ch, L) (isupper (Ch) ? tolower (Ch) : (Ch)) +# endif +#endif +/* We don't use `isdigit' here since the locale dependent + interpretation is not what we want here. We only need to accept + the arabic digits in the ASCII range. One day there is perhaps a + more reliable way to accept other sets of digits. */ +#define ISDIGIT(Ch) ((unsigned int) (Ch) - L_('0') <= 9) + +static CHAR_T * +memcpy_lowcase (CHAR_T *dest, const CHAR_T *src, + size_t len LOCALE_PARAM_PROTO) +{ + while (len-- > 0) + dest[len] = TOLOWER ((UCHAR_T) src[len], loc); + return dest; +} + +static CHAR_T * +memcpy_uppcase (CHAR_T *dest, const CHAR_T *src, + size_t len LOCALE_PARAM_PROTO) +{ + while (len-- > 0) + dest[len] = TOUPPER ((UCHAR_T) src[len], loc); + return dest; +} + + +#if ! HAVE_TM_GMTOFF +/* Yield the difference between *A and *B, + measured in seconds, ignoring leap seconds. */ +# define tm_diff ftime_tm_diff +static int +tm_diff (const struct tm *a, const struct tm *b) +{ + /* Compute intervening leap days correctly even if year is negative. + Take care to avoid int overflow in leap day calculations, + but it's OK to assume that A and B are close to each other. */ + int a4 = (a->tm_year >> 2) + (TM_YEAR_BASE >> 2) - ! (a->tm_year & 3); + int b4 = (b->tm_year >> 2) + (TM_YEAR_BASE >> 2) - ! (b->tm_year & 3); + int a100 = a4 / 25 - (a4 % 25 < 0); + int b100 = b4 / 25 - (b4 % 25 < 0); + int a400 = a100 >> 2; + int b400 = b100 >> 2; + int intervening_leap_days = (a4 - b4) - (a100 - b100) + (a400 - b400); + int years = a->tm_year - b->tm_year; + int days = (365 * years + intervening_leap_days + + (a->tm_yday - b->tm_yday)); + return (60 * (60 * (24 * days + (a->tm_hour - b->tm_hour)) + + (a->tm_min - b->tm_min)) + + (a->tm_sec - b->tm_sec)); +} +#endif /* ! HAVE_TM_GMTOFF */ + + + +/* The number of days from the first day of the first ISO week of this + year to the year day YDAY with week day WDAY. ISO weeks start on + Monday; the first ISO week has the year's first Thursday. YDAY may + be as small as YDAY_MINIMUM. */ +#define ISO_WEEK_START_WDAY 1 /* Monday */ +#define ISO_WEEK1_WDAY 4 /* Thursday */ +#define YDAY_MINIMUM (-366) +#ifdef __GNUC__ +__inline__ +#endif +static int +iso_week_days (int yday, int wday) +{ + /* Add enough to the first operand of % to make it nonnegative. */ + int big_enough_multiple_of_7 = (-YDAY_MINIMUM / 7 + 2) * 7; + return (yday + - (yday - wday + ISO_WEEK1_WDAY + big_enough_multiple_of_7) % 7 + + ISO_WEEK1_WDAY - ISO_WEEK_START_WDAY); +} + + +#if !(defined _NL_CURRENT || HAVE_STRFTIME) +static CHAR_T const weekday_name[][10] = + { + L_("Sunday"), L_("Monday"), L_("Tuesday"), L_("Wednesday"), + L_("Thursday"), L_("Friday"), L_("Saturday") + }; +static CHAR_T const month_name[][10] = + { + L_("January"), L_("February"), L_("March"), L_("April"), L_("May"), + L_("June"), L_("July"), L_("August"), L_("September"), L_("October"), + L_("November"), L_("December") + }; +#endif + + +/* When compiling this file, GNU applications can #define my_strftime + to a symbol (typically nstrftime) to get an extended strftime with + extra arguments UT and NS. Emacs is a special case for now, but + this Emacs-specific code can be removed once Emacs's config.h + defines my_strftime. */ +#if defined emacs && !defined my_strftime +# define my_strftime nstrftime +#endif + +#ifdef my_strftime +# define extra_args , ut, ns +# define extra_args_spec , int ut, int ns +#else +# ifdef COMPILE_WIDE +# define my_strftime wcsftime +# define nl_get_alt_digit _nl_get_walt_digit +# else +# define my_strftime strftime +# define nl_get_alt_digit _nl_get_alt_digit +# endif +# define extra_args +# define extra_args_spec +/* We don't have this information in general. */ +# define ut 0 +# define ns 0 +#endif + +#if ! defined _LIBC && ! HAVE_RUN_TZSET_TEST +/* Solaris 2.5.x and 2.6 tzset sometimes modify the storage returned + by localtime. On such systems, we must use the tzset and localtime + wrappers to work around the bug. */ +"you must run the autoconf test for a working tzset function" +#endif + + +/* Write information from TP into S according to the format + string FORMAT, writing no more that MAXSIZE characters + (including the terminating '\0') and returning number of + characters written. If S is NULL, nothing will be written + anywhere, so to determine how many characters would be + written, use NULL for S and (size_t) UINT_MAX for MAXSIZE. */ +size_t +my_strftime (CHAR_T *s, size_t maxsize, const CHAR_T *format, + const struct tm *tp extra_args_spec LOCALE_PARAM_PROTO) +{ +#if defined _LIBC && defined USE_IN_EXTENDED_LOCALE_MODEL + struct locale_data *const current = loc->__locales[LC_TIME]; +#endif + + int hour12 = tp->tm_hour; +#ifdef _NL_CURRENT + /* We cannot make the following values variables since we must delay + the evaluation of these values until really needed since some + expressions might not be valid in every situation. The `struct tm' + might be generated by a strptime() call that initialized + only a few elements. Dereference the pointers only if the format + requires this. Then it is ok to fail if the pointers are invalid. */ +# define a_wkday \ + ((const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(ABDAY_1) + tp->tm_wday)) +# define f_wkday \ + ((const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(DAY_1) + tp->tm_wday)) +# define a_month \ + ((const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(ABMON_1) + tp->tm_mon)) +# define f_month \ + ((const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(MON_1) + tp->tm_mon)) +# define ampm \ + ((const CHAR_T *) _NL_CURRENT (LC_TIME, tp->tm_hour > 11 \ + ? NLW(PM_STR) : NLW(AM_STR))) + +# define aw_len STRLEN (a_wkday) +# define am_len STRLEN (a_month) +# define ap_len STRLEN (ampm) +#else +# if !HAVE_STRFTIME +# define f_wkday (weekday_name[tp->tm_wday]) +# define f_month (month_name[tp->tm_mon]) +# define a_wkday f_wkday +# define a_month f_month +# define ampm (L_("AMPM") + 2 * (tp->tm_hour > 11)) + + size_t aw_len = 3; + size_t am_len = 3; + size_t ap_len = 2; +# endif +#endif + const char *zone; + size_t i = 0; + CHAR_T *p = s; + const CHAR_T *f; +#if DO_MULTIBYTE && !defined COMPILE_WIDE + const char *format_end = NULL; +#endif + + zone = NULL; +#if HAVE_TM_ZONE + /* The POSIX test suite assumes that setting + the environment variable TZ to a new value before calling strftime() + will influence the result (the %Z format) even if the information in + TP is computed with a totally different time zone. + This is bogus: though POSIX allows bad behavior like this, + POSIX does not require it. Do the right thing instead. */ + zone = (const char *) tp->tm_zone; +#endif +#if HAVE_TZNAME + if (ut) + { + if (! (zone && *zone)) + zone = "GMT"; + } + else + { + /* POSIX.1 requires that local time zone information be used as + though strftime called tzset. */ +# if HAVE_TZSET + tzset (); +# endif + } +#endif + + if (hour12 > 12) + hour12 -= 12; + else + if (hour12 == 0) + hour12 = 12; + + for (f = format; *f != '\0'; ++f) + { + int pad = 0; /* Padding for number ('-', '_', or 0). */ + int modifier; /* Field modifier ('E', 'O', or 0). */ + int digits; /* Max digits for numeric format. */ + int number_value; /* Numeric value to be printed. */ + int negative_number; /* 1 if the number is negative. */ + const CHAR_T *subfmt; + CHAR_T *bufp; + CHAR_T buf[1 + (sizeof (int) < sizeof (time_t) + ? INT_STRLEN_BOUND (time_t) + : INT_STRLEN_BOUND (int))]; + int width = -1; + int to_lowcase = 0; + int to_uppcase = 0; + int change_case = 0; + int format_char; + +#if DO_MULTIBYTE && !defined COMPILE_WIDE + switch (*f) + { + case L_('%'): + break; + + case L_('\b'): case L_('\t'): case L_('\n'): + case L_('\v'): case L_('\f'): case L_('\r'): + case L_(' '): case L_('!'): case L_('"'): case L_('#'): case L_('&'): + case L_('\''): case L_('('): case L_(')'): case L_('*'): case L_('+'): + case L_(','): case L_('-'): case L_('.'): case L_('/'): case L_('0'): + case L_('1'): case L_('2'): case L_('3'): case L_('4'): case L_('5'): + case L_('6'): case L_('7'): case L_('8'): case L_('9'): case L_(':'): + case L_(';'): case L_('<'): case L_('='): case L_('>'): case L_('?'): + case L_('A'): case L_('B'): case L_('C'): case L_('D'): case L_('E'): + case L_('F'): case L_('G'): case L_('H'): case L_('I'): case L_('J'): + case L_('K'): case L_('L'): case L_('M'): case L_('N'): case L_('O'): + case L_('P'): case L_('Q'): case L_('R'): case L_('S'): case L_('T'): + case L_('U'): case L_('V'): case L_('W'): case L_('X'): case L_('Y'): + case L_('Z'): case L_('['): case L_('\\'): case L_(']'): case L_('^'): + case L_('_'): case L_('a'): case L_('b'): case L_('c'): case L_('d'): + case L_('e'): case L_('f'): case L_('g'): case L_('h'): case L_('i'): + case L_('j'): case L_('k'): case L_('l'): case L_('m'): case L_('n'): + case L_('o'): case L_('p'): case L_('q'): case L_('r'): case L_('s'): + case L_('t'): case L_('u'): case L_('v'): case L_('w'): case L_('x'): + case L_('y'): case L_('z'): case L_('{'): case L_('|'): case L_('}'): + case L_('~'): + /* The C Standard requires these 98 characters (plus '%') to + be in the basic execution character set. None of these + characters can start a multibyte sequence, so they need + not be analyzed further. */ + add (1, *p = *f); + continue; + + default: + /* Copy this multibyte sequence until we reach its end, find + an error, or come back to the initial shift state. */ + { + mbstate_t mbstate = mbstate_zero; + size_t len = 0; + size_t fsize; + + if (! format_end) + format_end = f + strlen (f) + 1; + fsize = format_end - f; + + do + { + size_t bytes = mbrlen (f + len, fsize - len, &mbstate); + + if (bytes == 0) + break; + + if (bytes == (size_t) -2) + { + len += strlen (f + len); + break; + } + + if (bytes == (size_t) -1) + { + len++; + break; + } + + len += bytes; + } + while (! mbsinit (&mbstate)); + + cpy (len, f); + f += len - 1; + continue; + } + } + +#else /* ! DO_MULTIBYTE */ + + /* Either multibyte encodings are not supported, they are + safe for formats, so any non-'%' byte can be copied through, + or this is the wide character version. */ + if (*f != L_('%')) + { + add (1, *p = *f); + continue; + } + +#endif /* ! DO_MULTIBYTE */ + + /* Check for flags that can modify a format. */ + while (1) + { + switch (*++f) + { + /* This influences the number formats. */ + case L_('_'): + case L_('-'): + case L_('0'): + pad = *f; + continue; + + /* This changes textual output. */ + case L_('^'): + to_uppcase = 1; + continue; + case L_('#'): + change_case = 1; + continue; + + default: + break; + } + break; + } + + /* As a GNU extension we allow to specify the field width. */ + if (ISDIGIT (*f)) + { + width = 0; + do + { + if (width > INT_MAX / 10 + || (width == INT_MAX / 10 && *f - L_('0') > INT_MAX % 10)) + /* Avoid overflow. */ + width = INT_MAX; + else + { + width *= 10; + width += *f - L_('0'); + } + ++f; + } + while (ISDIGIT (*f)); + } + + /* Check for modifiers. */ + switch (*f) + { + case L_('E'): + case L_('O'): + modifier = *f++; + break; + + default: + modifier = 0; + break; + } + + /* Now do the specified format. */ + format_char = *f; + switch (format_char) + { +#define DO_NUMBER(d, v) \ + digits = d > width ? d : width; \ + number_value = v; goto do_number +#define DO_NUMBER_SPACEPAD(d, v) \ + digits = d > width ? d : width; \ + number_value = v; goto do_number_spacepad + + case L_('%'): + if (modifier != 0) + goto bad_format; + add (1, *p = *f); + break; + + case L_('a'): + if (modifier != 0) + goto bad_format; + if (change_case) + { + to_uppcase = 1; + to_lowcase = 0; + } +#if defined _NL_CURRENT || !HAVE_STRFTIME + cpy (aw_len, a_wkday); + break; +#else + goto underlying_strftime; +#endif + + case 'A': + if (modifier != 0) + goto bad_format; + if (change_case) + { + to_uppcase = 1; + to_lowcase = 0; + } +#if defined _NL_CURRENT || !HAVE_STRFTIME + cpy (STRLEN (f_wkday), f_wkday); + break; +#else + goto underlying_strftime; +#endif + + case L_('b'): + case L_('h'): + if (change_case) + { + to_uppcase = 1; + to_lowcase = 0; + } + if (modifier != 0) + goto bad_format; +#if defined _NL_CURRENT || !HAVE_STRFTIME + cpy (am_len, a_month); + break; +#else + goto underlying_strftime; +#endif + + case L_('B'): + if (modifier != 0) + goto bad_format; + if (change_case) + { + to_uppcase = 1; + to_lowcase = 0; + } +#if defined _NL_CURRENT || !HAVE_STRFTIME + cpy (STRLEN (f_month), f_month); + break; +#else + goto underlying_strftime; +#endif + + case L_('c'): + if (modifier == L_('O')) + goto bad_format; +#ifdef _NL_CURRENT + if (! (modifier == 'E' + && (*(subfmt = + (const CHAR_T *) _NL_CURRENT (LC_TIME, + NLW(ERA_D_T_FMT))) + != '\0'))) + subfmt = (const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(D_T_FMT)); +#else +# if HAVE_STRFTIME + goto underlying_strftime; +# else + subfmt = L_("%a %b %e %H:%M:%S %Y"); +# endif +#endif + + subformat: + { + CHAR_T *old_start = p; + size_t len = my_strftime (NULL, (size_t) -1, subfmt, + tp extra_args LOCALE_ARG); + add (len, my_strftime (p, maxsize - i, subfmt, + tp extra_args LOCALE_ARG)); + + if (to_uppcase) + while (old_start < p) + { + *old_start = TOUPPER ((UCHAR_T) *old_start, loc); + ++old_start; + } + } + break; + +#if HAVE_STRFTIME && ! (defined _NL_CURRENT && HAVE_STRUCT_ERA_ENTRY) + underlying_strftime: + { + /* The relevant information is available only via the + underlying strftime implementation, so use that. */ + char ufmt[4]; + char *u = ufmt; + char ubuf[1024]; /* enough for any single format in practice */ + size_t len; + /* Make sure we're calling the actual underlying strftime. + In some cases, config.h contains something like + "#define strftime rpl_strftime". */ +# ifdef strftime +# undef strftime + size_t strftime (); +# endif + + *u++ = '%'; + if (modifier != 0) + *u++ = modifier; + *u++ = format_char; + *u = '\0'; + len = strftime (ubuf, sizeof ubuf, ufmt, tp); + if (len == 0 && ubuf[0] != '\0') + return 0; + cpy (len, ubuf); + } + break; +#endif + + case L_('C'): + if (modifier == L_('O')) + goto bad_format; + if (modifier == L_('E')) + { +#if HAVE_STRUCT_ERA_ENTRY + struct era_entry *era = _nl_get_era_entry (tp HELPER_LOCALE_ARG); + if (era) + { +# ifdef COMPILE_WIDE + size_t len = __wcslen (era->era_wname); + cpy (len, era->era_wname); +# else + size_t len = strlen (era->era_name); + cpy (len, era->era_name); +# endif + break; + } +#else +# if HAVE_STRFTIME + goto underlying_strftime; +# endif +#endif + } + + { + int year = tp->tm_year + TM_YEAR_BASE; + DO_NUMBER (1, year / 100 - (year % 100 < 0)); + } + + case L_('x'): + if (modifier == L_('O')) + goto bad_format; +#ifdef _NL_CURRENT + if (! (modifier == L_('E') + && (*(subfmt = + (const CHAR_T *)_NL_CURRENT (LC_TIME, NLW(ERA_D_FMT))) + != L_('\0')))) + subfmt = (const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(D_FMT)); + goto subformat; +#else +# if HAVE_STRFTIME + goto underlying_strftime; +# else + /* Fall through. */ +# endif +#endif + case L_('D'): + if (modifier != 0) + goto bad_format; + subfmt = L_("%m/%d/%y"); + goto subformat; + + case L_('d'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, tp->tm_mday); + + case L_('e'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER_SPACEPAD (2, tp->tm_mday); + + /* All numeric formats set DIGITS and NUMBER_VALUE and then + jump to one of these two labels. */ + + do_number_spacepad: + /* Force `_' flag unless overridden by `0' or `-' flag. */ + if (pad != L_('0') && pad != L_('-')) + pad = L_('_'); + + do_number: + /* Format the number according to the MODIFIER flag. */ + + if (modifier == L_('O') && 0 <= number_value) + { +#ifdef _NL_CURRENT + /* Get the locale specific alternate representation of + the number NUMBER_VALUE. If none exist NULL is returned. */ + const CHAR_T *cp = nl_get_alt_digit (number_value + HELPER_LOCALE_ARG); + + if (cp != NULL) + { + size_t digitlen = STRLEN (cp); + if (digitlen != 0) + { + cpy (digitlen, cp); + break; + } + } +#else +# if HAVE_STRFTIME + goto underlying_strftime; +# endif +#endif + } + { + unsigned int u = number_value; + + bufp = buf + sizeof (buf) / sizeof (buf[0]); + negative_number = number_value < 0; + + if (negative_number) + u = -u; + + do + *--bufp = u % 10 + L_('0'); + while ((u /= 10) != 0); + } + + do_number_sign_and_padding: + if (negative_number) + *--bufp = L_('-'); + + if (pad != L_('-')) + { + int padding = digits - (buf + (sizeof (buf) / sizeof (buf[0])) + - bufp); + + if (padding > 0) + { + if (pad == L_('_')) + { + if ((size_t) padding >= maxsize - i) + return 0; + + if (p) + memset_space (p, padding); + i += padding; + width = width > padding ? width - padding : 0; + } + else + { + if ((size_t) digits >= maxsize - i) + return 0; + + if (negative_number) + { + ++bufp; + + if (p) + *p++ = L_('-'); + ++i; + } + + if (p) + memset_zero (p, padding); + i += padding; + width = 0; + } + } + } + + cpy (buf + sizeof (buf) / sizeof (buf[0]) - bufp, bufp); + break; + + case L_('F'): + if (modifier != 0) + goto bad_format; + subfmt = L_("%Y-%m-%d"); + goto subformat; + + case L_('H'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, tp->tm_hour); + + case L_('I'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, hour12); + + case L_('k'): /* GNU extension. */ + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER_SPACEPAD (2, tp->tm_hour); + + case L_('l'): /* GNU extension. */ + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER_SPACEPAD (2, hour12); + + case L_('j'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (3, 1 + tp->tm_yday); + + case L_('M'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, tp->tm_min); + + case L_('m'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, tp->tm_mon + 1); + +#ifndef _LIBC + case L_('N'): /* GNU extension. */ + if (modifier == L_('E')) + goto bad_format; + + number_value = ns; + if (width != -1) + { + /* Take an explicit width less than 9 as a precision. */ + int j; + for (j = width; j < 9; j++) + number_value /= 10; + } + + DO_NUMBER (9, number_value); +#endif + + case L_('n'): + add (1, *p = L_('\n')); + break; + + case L_('P'): + to_lowcase = 1; +#if !defined _NL_CURRENT && HAVE_STRFTIME + format_char = L_('p'); +#endif + /* FALLTHROUGH */ + + case L_('p'): + if (change_case) + { + to_uppcase = 0; + to_lowcase = 1; + } +#if defined _NL_CURRENT || !HAVE_STRFTIME + cpy (ap_len, ampm); + break; +#else + goto underlying_strftime; +#endif + + case L_('R'): + subfmt = L_("%H:%M"); + goto subformat; + + case L_('r'): +#if !defined _NL_CURRENT && HAVE_STRFTIME + goto underlying_strftime; +#else +# ifdef _NL_CURRENT + if (*(subfmt = (const CHAR_T *) _NL_CURRENT (LC_TIME, + NLW(T_FMT_AMPM))) + == L_('\0')) +# endif + subfmt = L_("%I:%M:%S %p"); + goto subformat; +#endif + + case L_('S'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, tp->tm_sec); + + case L_('s'): /* GNU extension. */ + { + struct tm ltm; + time_t t; + + ltm = *tp; + t = mktime (<m); + + /* Generate string value for T using time_t arithmetic; + this works even if sizeof (long) < sizeof (time_t). */ + + bufp = buf + sizeof (buf) / sizeof (buf[0]); + negative_number = t < 0; + + do + { + int d = t % 10; + t /= 10; + + if (negative_number) + { + d = -d; + + /* Adjust if division truncates to minus infinity. */ + if (0 < -1 % 10 && d < 0) + { + t++; + d += 10; + } + } + + *--bufp = d + L_('0'); + } + while (t != 0); + + digits = 1; + goto do_number_sign_and_padding; + } + + case L_('X'): + if (modifier == L_('O')) + goto bad_format; +#ifdef _NL_CURRENT + if (! (modifier == L_('E') + && (*(subfmt = + (const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(ERA_T_FMT))) + != L_('\0')))) + subfmt = (const CHAR_T *) _NL_CURRENT (LC_TIME, NLW(T_FMT)); + goto subformat; +#else +# if HAVE_STRFTIME + goto underlying_strftime; +# else + /* Fall through. */ +# endif +#endif + case L_('T'): + subfmt = L_("%H:%M:%S"); + goto subformat; + + case L_('t'): + add (1, *p = L_('\t')); + break; + + case L_('u'): + DO_NUMBER (1, (tp->tm_wday - 1 + 7) % 7 + 1); + + case L_('U'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, (tp->tm_yday - tp->tm_wday + 7) / 7); + + case L_('V'): + case L_('g'): + case L_('G'): + if (modifier == L_('E')) + goto bad_format; + { + int year = tp->tm_year + TM_YEAR_BASE; + int days = iso_week_days (tp->tm_yday, tp->tm_wday); + + if (days < 0) + { + /* This ISO week belongs to the previous year. */ + year--; + days = iso_week_days (tp->tm_yday + (365 + __isleap (year)), + tp->tm_wday); + } + else + { + int d = iso_week_days (tp->tm_yday - (365 + __isleap (year)), + tp->tm_wday); + if (0 <= d) + { + /* This ISO week belongs to the next year. */ + year++; + days = d; + } + } + + switch (*f) + { + case L_('g'): + DO_NUMBER (2, (year % 100 + 100) % 100); + + case L_('G'): + DO_NUMBER (1, year); + + default: + DO_NUMBER (2, days / 7 + 1); + } + } + + case L_('W'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (2, (tp->tm_yday - (tp->tm_wday - 1 + 7) % 7 + 7) / 7); + + case L_('w'): + if (modifier == L_('E')) + goto bad_format; + + DO_NUMBER (1, tp->tm_wday); + + case L_('Y'): + if (modifier == 'E') + { +#if HAVE_STRUCT_ERA_ENTRY + struct era_entry *era = _nl_get_era_entry (tp HELPER_LOCALE_ARG); + if (era) + { +# ifdef COMPILE_WIDE + subfmt = era->era_wformat; +# else + subfmt = era->era_format; +# endif + goto subformat; + } +#else +# if HAVE_STRFTIME + goto underlying_strftime; +# endif +#endif + } + if (modifier == L_('O')) + goto bad_format; + else + DO_NUMBER (1, tp->tm_year + TM_YEAR_BASE); + + case L_('y'): + if (modifier == L_('E')) + { +#if HAVE_STRUCT_ERA_ENTRY + struct era_entry *era = _nl_get_era_entry (tp HELPER_LOCALE_ARG); + if (era) + { + int delta = tp->tm_year - era->start_date[0]; + DO_NUMBER (1, (era->offset + + delta * era->absolute_direction)); + } +#else +# if HAVE_STRFTIME + goto underlying_strftime; +# endif +#endif + } + DO_NUMBER (2, (tp->tm_year % 100 + 100) % 100); + + case L_('Z'): + if (change_case) + { + to_uppcase = 0; + to_lowcase = 1; + } + +#if HAVE_TZNAME + /* The tzset() call might have changed the value. */ + if (!(zone && *zone) && tp->tm_isdst >= 0) + zone = tzname[tp->tm_isdst]; +#endif + if (! zone) + zone = ""; + +#ifdef COMPILE_WIDE + { + /* The zone string is always given in multibyte form. We have + to transform it first. */ + wchar_t *wczone; + size_t len; + widen (zone, wczone, len); + cpy (len, wczone); + } +#else + cpy (strlen (zone), zone); +#endif + break; + + case L_('z'): + if (tp->tm_isdst < 0) + break; + + { + int diff; +#if HAVE_TM_GMTOFF + diff = tp->tm_gmtoff; +#else + if (ut) + diff = 0; + else + { + struct tm gtm; + struct tm ltm; + time_t lt; + + ltm = *tp; + lt = mktime (<m); + + if (lt == (time_t) -1) + { + /* mktime returns -1 for errors, but -1 is also a + valid time_t value. Check whether an error really + occurred. */ + struct tm tm; + + if (! __localtime_r (<, &tm) + || ((ltm.tm_sec ^ tm.tm_sec) + | (ltm.tm_min ^ tm.tm_min) + | (ltm.tm_hour ^ tm.tm_hour) + | (ltm.tm_mday ^ tm.tm_mday) + | (ltm.tm_mon ^ tm.tm_mon) + | (ltm.tm_year ^ tm.tm_year))) + break; + } + + if (! __gmtime_r (<, >m)) + break; + + diff = tm_diff (<m, >m); + } +#endif + + if (diff < 0) + { + add (1, *p = L_('-')); + diff = -diff; + } + else + add (1, *p = L_('+')); + + diff /= 60; + DO_NUMBER (4, (diff / 60) * 100 + diff % 60); + } + + case L_('\0'): /* GNU extension: % at end of format. */ + --f; + /* Fall through. */ + default: + /* Unknown format; output the format, including the '%', + since this is most likely the right thing to do if a + multibyte string has been misparsed. */ + bad_format: + { + int flen; + for (flen = 1; f[1 - flen] != L_('%'); flen++) + continue; + cpy (flen, &f[1 - flen]); + } + break; + } + } + + if (p && maxsize != 0) + *p = L_('\0'); + return i; +} +#ifdef _LIBC +libc_hidden_def (my_strftime) +#endif + + +#ifdef emacs +/* For Emacs we have a separate interface which corresponds to the normal + strftime function plus the ut argument, but without the ns argument. */ +size_t +emacs_strftimeu (char *s, size_t maxsize, const char *format, + const struct tm *tp, int ut) +{ + return my_strftime (s, maxsize, format, tp, ut, 0); +} +#endif diff --git a/contrib/diff/lib/stripslash.c b/contrib/diff/lib/stripslash.c new file mode 100644 index 0000000..c6b319e --- /dev/null +++ b/contrib/diff/lib/stripslash.c @@ -0,0 +1,39 @@ +/* stripslash.c -- remove redundant trailing slashes from a file name + Copyright (C) 1990, 2001, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "dirname.h" + +/* Remove trailing slashes from PATH. + Return nonzero if a trailing slash was removed. + This is useful when using filename completion from a shell that + adds a "/" after directory names (such as tcsh and bash), because + the Unix rename and rmdir system calls return an "Invalid argument" error + when given a path that ends in "/" (except for the root directory). */ + +int +strip_trailing_slashes (char *path) +{ + char *base = base_name (path); + char *base_lim = base + base_len (base); + int had_slash = *base_lim; + *base_lim = '\0'; + return had_slash; +} diff --git a/contrib/diff/lib/strncasecmp.c b/contrib/diff/lib/strncasecmp.c new file mode 100644 index 0000000..68d95aa --- /dev/null +++ b/contrib/diff/lib/strncasecmp.c @@ -0,0 +1,2 @@ +#define LENGTH_LIMIT +#include "strcasecmp.c" diff --git a/contrib/diff/lib/strtoimax.c b/contrib/diff/lib/strtoimax.c new file mode 100644 index 0000000..4ce741c --- /dev/null +++ b/contrib/diff/lib/strtoimax.c @@ -0,0 +1,80 @@ +/* Convert string representation of a number into an intmax_t value. + Copyright (C) 1999, 2001, 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#if HAVE_CONFIG_H +# include +#endif + +#if HAVE_INTTYPES_H +# include +#elif HAVE_STDINT_H +# include +#endif + +#include + +/* Verify a requirement at compile-time (unlike assert, which is runtime). */ +#define verify(name, assertion) struct name { char a[(assertion) ? 1 : -1]; } + +#ifdef UNSIGNED +# ifndef HAVE_DECL_STRTOULL +"this configure-time declaration test was not run" +# endif +# if !HAVE_DECL_STRTOULL && HAVE_UNSIGNED_LONG_LONG +unsigned long long strtoull (char const *, char **, int); +# endif + +#else + +# ifndef HAVE_DECL_STRTOLL +"this configure-time declaration test was not run" +# endif +# if !HAVE_DECL_STRTOLL && HAVE_UNSIGNED_LONG_LONG +long long strtoll (char const *, char **, int); +# endif +#endif + +#ifdef UNSIGNED +# undef HAVE_LONG_LONG +# define HAVE_LONG_LONG HAVE_UNSIGNED_LONG_LONG +# define INT uintmax_t +# define strtoimax strtoumax +# define strtol strtoul +# define strtoll strtoull +#else +# define INT intmax_t +#endif + +INT +strtoimax (char const *ptr, char **endptr, int base) +{ +#if HAVE_LONG_LONG + verify (size_is_that_of_long_or_long_long, + (sizeof (INT) == sizeof (long) + || sizeof (INT) == sizeof (long long))); + + if (sizeof (INT) != sizeof (long)) + return strtoll (ptr, endptr, base); +#else + verify (size_is_that_of_long, + sizeof (INT) == sizeof (long)); +#endif + + return strtol (ptr, endptr, base); +} diff --git a/contrib/diff/lib/strtol.c b/contrib/diff/lib/strtol.c new file mode 100644 index 0000000..e2b1589 --- /dev/null +++ b/contrib/diff/lib/strtol.c @@ -0,0 +1,432 @@ +/* Convert string representation of a number into an integer value. + + Copyright (C) 1991, 1992, 1994, 1995, 1996, 1997, 1998, 1999, 2003 + Free Software Foundation, Inc. + + NOTE: The canonical source of this file is maintained with the GNU C + Library. Bugs can be reported to bug-glibc@gnu.org. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#ifdef _LIBC +# define USE_NUMBER_GROUPING +#endif + +#include +#include +#ifndef errno +extern int errno; +#endif +#ifndef __set_errno +# define __set_errno(Val) errno = (Val) +#endif + +#include +#include +#include +#include + +#ifdef USE_NUMBER_GROUPING +# include "../locale/localeinfo.h" +#endif + +/* Nonzero if we are defining `strtoul' or `strtoull', operating on + unsigned integers. */ +#ifndef UNSIGNED +# define UNSIGNED 0 +# define INT LONG int +#else +# define INT unsigned LONG int +#endif + +/* Determine the name. */ +#ifdef USE_IN_EXTENDED_LOCALE_MODEL +# if UNSIGNED +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol __wcstoull_l +# else +# define strtol __wcstoul_l +# endif +# else +# ifdef QUAD +# define strtol __strtoull_l +# else +# define strtol __strtoul_l +# endif +# endif +# else +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol __wcstoll_l +# else +# define strtol __wcstol_l +# endif +# else +# ifdef QUAD +# define strtol __strtoll_l +# else +# define strtol __strtol_l +# endif +# endif +# endif +#else +# if UNSIGNED +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol wcstoull +# else +# define strtol wcstoul +# endif +# else +# ifdef QUAD +# define strtol strtoull +# else +# define strtol strtoul +# endif +# endif +# else +# ifdef USE_WIDE_CHAR +# ifdef QUAD +# define strtol wcstoll +# else +# define strtol wcstol +# endif +# else +# ifdef QUAD +# define strtol strtoll +# endif +# endif +# endif +#endif + +/* If QUAD is defined, we are defining `strtoll' or `strtoull', + operating on `long long int's. */ +#ifdef QUAD +# define LONG long long +# define STRTOL_LONG_MIN LONG_LONG_MIN +# define STRTOL_LONG_MAX LONG_LONG_MAX +# define STRTOL_ULONG_MAX ULONG_LONG_MAX + +/* The extra casts work around common compiler bugs, + e.g. Cray C 5.0.3.0 when t == time_t. */ +# ifndef TYPE_SIGNED +# define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) +# endif +# ifndef TYPE_MINIMUM +# define TYPE_MINIMUM(t) ((t) (TYPE_SIGNED (t) \ + ? ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1) \ + : (t) 0)) +# endif +# ifndef TYPE_MAXIMUM +# define TYPE_MAXIMUM(t) ((t) (~ (t) 0 - TYPE_MINIMUM (t))) +# endif + +# ifndef ULONG_LONG_MAX +# define ULONG_LONG_MAX TYPE_MAXIMUM (unsigned long long) +# endif +# ifndef LONG_LONG_MAX +# define LONG_LONG_MAX TYPE_MAXIMUM (long long int) +# endif +# ifndef LONG_LONG_MIN +# define LONG_LONG_MIN TYPE_MINIMUM (long long int) +# endif + +# if __GNUC__ == 2 && __GNUC_MINOR__ < 7 + /* Work around gcc bug with using this constant. */ + static const unsigned long long int maxquad = ULONG_LONG_MAX; +# undef STRTOL_ULONG_MAX +# define STRTOL_ULONG_MAX maxquad +# endif +#else +# define LONG long +# define STRTOL_LONG_MIN LONG_MIN +# define STRTOL_LONG_MAX LONG_MAX +# define STRTOL_ULONG_MAX ULONG_MAX +#endif + + +/* We use this code also for the extended locale handling where the + function gets as an additional argument the locale which has to be + used. To access the values we have to redefine the _NL_CURRENT + macro. */ +#ifdef USE_IN_EXTENDED_LOCALE_MODEL +# undef _NL_CURRENT +# define _NL_CURRENT(category, item) \ + (current->values[_NL_ITEM_INDEX (item)].string) +# define LOCALE_PARAM , loc +# define LOCALE_PARAM_PROTO , __locale_t loc +#else +# define LOCALE_PARAM +# define LOCALE_PARAM_PROTO +#endif + +#if defined _LIBC || defined HAVE_WCHAR_H +# include +#endif + +#ifdef USE_WIDE_CHAR +# include +# define L_(Ch) L##Ch +# define UCHAR_TYPE wint_t +# define STRING_TYPE wchar_t +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define ISSPACE(Ch) __iswspace_l ((Ch), loc) +# define ISALPHA(Ch) __iswalpha_l ((Ch), loc) +# define TOUPPER(Ch) __towupper_l ((Ch), loc) +# else +# define ISSPACE(Ch) iswspace (Ch) +# define ISALPHA(Ch) iswalpha (Ch) +# define TOUPPER(Ch) towupper (Ch) +# endif +#else +# if defined STDC_HEADERS || (!defined isascii && !defined HAVE_ISASCII) +# define IN_CTYPE_DOMAIN(c) 1 +# else +# define IN_CTYPE_DOMAIN(c) isascii(c) +# endif +# define L_(Ch) Ch +# define UCHAR_TYPE unsigned char +# define STRING_TYPE char +# ifdef USE_IN_EXTENDED_LOCALE_MODEL +# define ISSPACE(Ch) __isspace_l ((Ch), loc) +# define ISALPHA(Ch) __isalpha_l ((Ch), loc) +# define TOUPPER(Ch) __toupper_l ((Ch), loc) +# else +# define ISSPACE(Ch) (IN_CTYPE_DOMAIN (Ch) && isspace (Ch)) +# define ISALPHA(Ch) (IN_CTYPE_DOMAIN (Ch) && isalpha (Ch)) +# define TOUPPER(Ch) (IN_CTYPE_DOMAIN (Ch) ? toupper (Ch) : (Ch)) +# endif +#endif + +#define INTERNAL(X) INTERNAL1(X) +#define INTERNAL1(X) __##X##_internal +#define WEAKNAME(X) WEAKNAME1(X) + +#ifdef USE_NUMBER_GROUPING +/* This file defines a function to check for correct grouping. */ +# include "grouping.h" +#endif + + + +/* Convert NPTR to an `unsigned long int' or `long int' in base BASE. + If BASE is 0 the base is determined by the presence of a leading + zero, indicating octal or a leading "0x" or "0X", indicating hexadecimal. + If BASE is < 2 or > 36, it is reset to 10. + If ENDPTR is not NULL, a pointer to the character after the last + one converted is stored in *ENDPTR. */ + +INT +INTERNAL (strtol) (const STRING_TYPE *nptr, STRING_TYPE **endptr, + int base, int group LOCALE_PARAM_PROTO) +{ + int negative; + register unsigned LONG int cutoff; + register unsigned int cutlim; + register unsigned LONG int i; + register const STRING_TYPE *s; + register UCHAR_TYPE c; + const STRING_TYPE *save, *end; + int overflow; + +#ifdef USE_NUMBER_GROUPING +# ifdef USE_IN_EXTENDED_LOCALE_MODEL + struct locale_data *current = loc->__locales[LC_NUMERIC]; +# endif + /* The thousands character of the current locale. */ + wchar_t thousands = L'\0'; + /* The numeric grouping specification of the current locale, + in the format described in . */ + const char *grouping; + + if (group) + { + grouping = _NL_CURRENT (LC_NUMERIC, GROUPING); + if (*grouping <= 0 || *grouping == CHAR_MAX) + grouping = NULL; + else + { + /* Figure out the thousands separator character. */ +# if defined _LIBC || defined _HAVE_BTOWC + thousands = __btowc (*_NL_CURRENT (LC_NUMERIC, THOUSANDS_SEP)); + if (thousands == WEOF) + thousands = L'\0'; +# endif + if (thousands == L'\0') + grouping = NULL; + } + } + else + grouping = NULL; +#endif + + if (base < 0 || base == 1 || base > 36) + { + __set_errno (EINVAL); + return 0; + } + + save = s = nptr; + + /* Skip white space. */ + while (ISSPACE (*s)) + ++s; + if (*s == L_('\0')) + goto noconv; + + /* Check for a sign. */ + if (*s == L_('-')) + { + negative = 1; + ++s; + } + else if (*s == L_('+')) + { + negative = 0; + ++s; + } + else + negative = 0; + + /* Recognize number prefix and if BASE is zero, figure it out ourselves. */ + if (*s == L_('0')) + { + if ((base == 0 || base == 16) && TOUPPER (s[1]) == L_('X')) + { + s += 2; + base = 16; + } + else if (base == 0) + base = 8; + } + else if (base == 0) + base = 10; + + /* Save the pointer so we can check later if anything happened. */ + save = s; + +#ifdef USE_NUMBER_GROUPING + if (group) + { + /* Find the end of the digit string and check its grouping. */ + end = s; + for (c = *end; c != L_('\0'); c = *++end) + if ((wchar_t) c != thousands + && ((wchar_t) c < L_('0') || (wchar_t) c > L_('9')) + && (!ISALPHA (c) || (int) (TOUPPER (c) - L_('A') + 10) >= base)) + break; + if (*s == thousands) + end = s; + else + end = correctly_grouped_prefix (s, end, thousands, grouping); + } + else +#endif + end = NULL; + + cutoff = STRTOL_ULONG_MAX / (unsigned LONG int) base; + cutlim = STRTOL_ULONG_MAX % (unsigned LONG int) base; + + overflow = 0; + i = 0; + for (c = *s; c != L_('\0'); c = *++s) + { + if (s == end) + break; + if (c >= L_('0') && c <= L_('9')) + c -= L_('0'); + else if (ISALPHA (c)) + c = TOUPPER (c) - L_('A') + 10; + else + break; + if ((int) c >= base) + break; + /* Check for overflow. */ + if (i > cutoff || (i == cutoff && c > cutlim)) + overflow = 1; + else + { + i *= (unsigned LONG int) base; + i += c; + } + } + + /* Check if anything actually happened. */ + if (s == save) + goto noconv; + + /* Store in ENDPTR the address of one character + past the last character we converted. */ + if (endptr != NULL) + *endptr = (STRING_TYPE *) s; + +#if !UNSIGNED + /* Check for a value that is within the range of + `unsigned LONG int', but outside the range of `LONG int'. */ + if (overflow == 0 + && i > (negative + ? -((unsigned LONG int) (STRTOL_LONG_MIN + 1)) + 1 + : (unsigned LONG int) STRTOL_LONG_MAX)) + overflow = 1; +#endif + + if (overflow) + { + __set_errno (ERANGE); +#if UNSIGNED + return STRTOL_ULONG_MAX; +#else + return negative ? STRTOL_LONG_MIN : STRTOL_LONG_MAX; +#endif + } + + /* Return the result of the appropriate sign. */ + return negative ? -i : i; + +noconv: + /* We must handle a special case here: the base is 0 or 16 and the + first two characters are '0' and 'x', but the rest are no + hexadecimal digits. This is no error case. We return 0 and + ENDPTR points to the `x`. */ + if (endptr != NULL) + { + if (save - nptr >= 2 && TOUPPER (save[-1]) == L_('X') + && save[-2] == L_('0')) + *endptr = (STRING_TYPE *) &save[-1]; + else + /* There was no number to convert. */ + *endptr = (STRING_TYPE *) nptr; + } + + return 0L; +} + +/* External user entry point. */ + + +INT +#ifdef weak_function +weak_function +#endif +strtol (const STRING_TYPE *nptr, STRING_TYPE **endptr, + int base LOCALE_PARAM_PROTO) +{ + return INTERNAL (strtol) (nptr, endptr, base, 0 LOCALE_PARAM); +} diff --git a/contrib/diff/lib/strtoll.c b/contrib/diff/lib/strtoll.c new file mode 100644 index 0000000..8e9b261 --- /dev/null +++ b/contrib/diff/lib/strtoll.c @@ -0,0 +1,33 @@ +/* Function to parse a `long long int' from text. + Copyright (C) 1995, 1996, 1997, 1999, 2001 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#define QUAD 1 + +#include + +#ifdef _LIBC +# ifdef SHARED +# include + +# if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_2) +compat_symbol (libc, __strtoll_internal, __strtoq_internal, GLIBC_2_0); +# endif + +# endif +weak_alias (strtoll, strtoq) +#endif diff --git a/contrib/diff/lib/strtoul.c b/contrib/diff/lib/strtoul.c new file mode 100644 index 0000000..bb62f9b --- /dev/null +++ b/contrib/diff/lib/strtoul.c @@ -0,0 +1,20 @@ +/* Copyright (C) 1991, 1997 Free Software Foundation, Inc. + This file is part of the GNU C Library. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#define UNSIGNED 1 + +#include "strtol.c" diff --git a/contrib/diff/lib/strtoull.c b/contrib/diff/lib/strtoull.c new file mode 100644 index 0000000..d6aa1f8 --- /dev/null +++ b/contrib/diff/lib/strtoull.c @@ -0,0 +1,27 @@ +/* Function to parse an `unsigned long long int' from text. + Copyright (C) 1995, 1996, 1997, 1999 Free Software Foundation, Inc. + NOTE: The canonical source of this file is maintained with the GNU C + Library. Bugs can be reported to bug-glibc@gnu.org. + + This program is free software; you can redistribute it and/or modify it + under the terms of the GNU General Public License as published by the + Free Software Foundation; either version 2, or (at your option) any + later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#define QUAD 1 + +#include "strtoul.c" + +#ifdef _LIBC +strong_alias (__strtoull_internal, __strtouq_internal) +weak_alias (strtoull, strtouq) +#endif diff --git a/contrib/diff/lib/strtoumax.c b/contrib/diff/lib/strtoumax.c new file mode 100644 index 0000000..dc395d6 --- /dev/null +++ b/contrib/diff/lib/strtoumax.c @@ -0,0 +1,2 @@ +#define UNSIGNED 1 +#include "strtoimax.c" diff --git a/contrib/diff/lib/tempname.c b/contrib/diff/lib/tempname.c new file mode 100644 index 0000000..b3d0874 --- /dev/null +++ b/contrib/diff/lib/tempname.c @@ -0,0 +1,344 @@ +/* tempname.c - generate the name of a temporary file. + + Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, + 2000, 2001, 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include +#include + +#include +#ifndef __set_errno +# define __set_errno(Val) errno = (Val) +#endif + +#include +#ifndef P_tmpdir +# define P_tmpdir "/tmp" +#endif +#ifndef TMP_MAX +# define TMP_MAX 238328 +#endif +#ifndef __GT_FILE +# define __GT_FILE 0 +# define __GT_BIGFILE 1 +# define __GT_DIR 2 +# define __GT_NOCREATE 3 +#endif + +#include +#include +#include + +#if HAVE_FCNTL_H || _LIBC +# include +#endif + +#if HAVE_SYS_TIME_H || _LIBC +# include +#endif + +#if HAVE_STDINT_H || _LIBC +# include +#endif +#if HAVE_INTTYPES_H +# include +#endif + +#if HAVE_UNISTD_H || _LIBC +# include +#endif + +#include +#if STAT_MACROS_BROKEN +# undef S_ISDIR +#endif +#if !defined S_ISDIR && defined S_IFDIR +# define S_ISDIR(mode) (((mode) & S_IFMT) == S_IFDIR) +#endif +#if !S_IRUSR && S_IREAD +# define S_IRUSR S_IREAD +#endif +#if !S_IRUSR +# define S_IRUSR 00400 +#endif +#if !S_IWUSR && S_IWRITE +# define S_IWUSR S_IWRITE +#endif +#if !S_IWUSR +# define S_IWUSR 00200 +#endif +#if !S_IXUSR && S_IEXEC +# define S_IXUSR S_IEXEC +#endif +#if !S_IXUSR +# define S_IXUSR 00100 +#endif + +#if _LIBC +# define struct_stat64 struct stat64 +#else +# define struct_stat64 struct stat +# define __getpid getpid +# define __gettimeofday gettimeofday +# define __mkdir mkdir +# define __open open +# define __open64 open +# define __lxstat64(version, path, buf) lstat (path, buf) +# define __xstat64(version, path, buf) stat (path, buf) +#endif + +#if ! (HAVE___SECURE_GETENV || _LIBC) +# define __secure_getenv getenv +#endif + +#ifdef _LIBC +# include +# if HP_TIMING_AVAIL +# define RANDOM_BITS(Var) \ + if (__builtin_expect (value == UINT64_C (0), 0)) \ + { \ + /* If this is the first time this function is used initialize \ + the variable we accumulate the value in to some somewhat \ + random value. If we'd not do this programs at startup time \ + might have a reduced set of possible names, at least on slow \ + machines. */ \ + struct timeval tv; \ + __gettimeofday (&tv, NULL); \ + value = ((uint64_t) tv.tv_usec << 16) ^ tv.tv_sec; \ + } \ + HP_TIMING_NOW (Var) +# endif +#endif + +/* Use the widest available unsigned type if uint64_t is not + available. The algorithm below extracts a number less than 62**6 + (approximately 2**35.725) from uint64_t, so ancient hosts where + uintmax_t is only 32 bits lose about 3.725 bits of randomness, + which is better than not having mkstemp at all. */ +#if !defined UINT64_MAX && !defined uint64_t +# define uint64_t uintmax_t +#endif + +/* Return nonzero if DIR is an existent directory. */ +static int +direxists (const char *dir) +{ + struct_stat64 buf; + return __xstat64 (_STAT_VER, dir, &buf) == 0 && S_ISDIR (buf.st_mode); +} + +/* Path search algorithm, for tmpnam, tmpfile, etc. If DIR is + non-null and exists, uses it; otherwise uses the first of $TMPDIR, + P_tmpdir, /tmp that exists. Copies into TMPL a template suitable + for use with mk[s]temp. Will fail (-1) if DIR is non-null and + doesn't exist, none of the searched dirs exists, or there's not + enough space in TMPL. */ +int +__path_search (char *tmpl, size_t tmpl_len, const char *dir, const char *pfx, + int try_tmpdir) +{ + const char *d; + size_t dlen, plen; + + if (!pfx || !pfx[0]) + { + pfx = "file"; + plen = 4; + } + else + { + plen = strlen (pfx); + if (plen > 5) + plen = 5; + } + + if (try_tmpdir) + { + d = __secure_getenv ("TMPDIR"); + if (d != NULL && direxists (d)) + dir = d; + else if (dir != NULL && direxists (dir)) + /* nothing */ ; + else + dir = NULL; + } + if (dir == NULL) + { + if (direxists (P_tmpdir)) + dir = P_tmpdir; + else if (strcmp (P_tmpdir, "/tmp") != 0 && direxists ("/tmp")) + dir = "/tmp"; + else + { + __set_errno (ENOENT); + return -1; + } + } + + dlen = strlen (dir); + while (dlen > 1 && dir[dlen - 1] == '/') + dlen--; /* remove trailing slashes */ + + /* check we have room for "${dir}/${pfx}XXXXXX\0" */ + if (tmpl_len < dlen + 1 + plen + 6 + 1) + { + __set_errno (EINVAL); + return -1; + } + + sprintf (tmpl, "%.*s/%.*sXXXXXX", (int) dlen, dir, (int) plen, pfx); + return 0; +} + +/* These are the characters used in temporary filenames. */ +static const char letters[] = +"abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789"; + +/* Generate a temporary file name based on TMPL. TMPL must match the + rules for mk[s]temp (i.e. end in "XXXXXX"). The name constructed + does not exist at the time of the call to __gen_tempname. TMPL is + overwritten with the result. + + KIND may be one of: + __GT_NOCREATE: simply verify that the name does not exist + at the time of the call. + __GT_FILE: create the file using open(O_CREAT|O_EXCL) + and return a read-write fd. The file is mode 0600. + __GT_BIGFILE: same as __GT_FILE but use open64(). + __GT_DIR: create a directory, which will be mode 0700. + + We use a clever algorithm to get hard-to-predict names. */ +int +__gen_tempname (char *tmpl, int kind) +{ + int len; + char *XXXXXX; + static uint64_t value; + uint64_t random_time_bits; + unsigned int count; + int fd = -1; + int save_errno = errno; + struct_stat64 st; + + /* A lower bound on the number of temporary files to attempt to + generate. The maximum total number of temporary file names that + can exist for a given template is 62**6. It should never be + necessary to try all these combinations. Instead if a reasonable + number of names is tried (we define reasonable as 62**3) fail to + give the system administrator the chance to remove the problems. */ + unsigned int attempts_min = 62 * 62 * 62; + + /* The number of times to attempt to generate a temporary file. To + conform to POSIX, this must be no smaller than TMP_MAX. */ + unsigned int attempts = attempts_min < TMP_MAX ? TMP_MAX : attempts_min; + + len = strlen (tmpl); + if (len < 6 || strcmp (&tmpl[len - 6], "XXXXXX")) + { + __set_errno (EINVAL); + return -1; + } + + /* This is where the Xs start. */ + XXXXXX = &tmpl[len - 6]; + + /* Get some more or less random data. */ +#ifdef RANDOM_BITS + RANDOM_BITS (random_time_bits); +#else +# if HAVE_GETTIMEOFDAY || _LIBC + { + struct timeval tv; + __gettimeofday (&tv, NULL); + random_time_bits = ((uint64_t) tv.tv_usec << 16) ^ tv.tv_sec; + } +# else + random_time_bits = time (NULL); +# endif +#endif + value += random_time_bits ^ __getpid (); + + for (count = 0; count < attempts; value += 7777, ++count) + { + uint64_t v = value; + + /* Fill in the random bits. */ + XXXXXX[0] = letters[v % 62]; + v /= 62; + XXXXXX[1] = letters[v % 62]; + v /= 62; + XXXXXX[2] = letters[v % 62]; + v /= 62; + XXXXXX[3] = letters[v % 62]; + v /= 62; + XXXXXX[4] = letters[v % 62]; + v /= 62; + XXXXXX[5] = letters[v % 62]; + + switch (kind) + { + case __GT_FILE: + fd = __open (tmpl, O_RDWR | O_CREAT | O_EXCL, S_IRUSR | S_IWUSR); + break; + + case __GT_BIGFILE: + fd = __open64 (tmpl, O_RDWR | O_CREAT | O_EXCL, S_IRUSR | S_IWUSR); + break; + + case __GT_DIR: + fd = __mkdir (tmpl, S_IRUSR | S_IWUSR | S_IXUSR); + break; + + case __GT_NOCREATE: + /* This case is backward from the other three. __gen_tempname + succeeds if __xstat fails because the name does not exist. + Note the continue to bypass the common logic at the bottom + of the loop. */ + if (__lxstat64 (_STAT_VER, tmpl, &st) < 0) + { + if (errno == ENOENT) + { + __set_errno (save_errno); + return 0; + } + else + /* Give up now. */ + return -1; + } + continue; + + default: + assert (! "invalid KIND in __gen_tempname"); + } + + if (fd >= 0) + { + __set_errno (save_errno); + return fd; + } + else if (errno != EEXIST) + return -1; + } + + /* We got out of the loop because we ran out of combinations to try. */ + __set_errno (EEXIST); + return -1; +} diff --git a/contrib/diff/lib/time_r.c b/contrib/diff/lib/time_r.c new file mode 100644 index 0000000..6c42504 --- /dev/null +++ b/contrib/diff/lib/time_r.c @@ -0,0 +1,69 @@ +/* Reentrant time functions like localtime_r. + + Copyright (C) 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "time_r.h" + +#include + +static char * +copy_string_result (char *dest, char const *src) +{ + if (! src) + return 0; + return strcpy (dest, src); +} + +static struct tm * +copy_tm_result (struct tm *dest, struct tm const *src) +{ + if (! src) + return 0; + *dest = *src; + return dest; +} + + +char * +asctime_r (struct tm const * restrict tm, char * restrict buf) +{ + return copy_string_result (buf, asctime (tm)); +} + +char * +ctime_r (time_t const *t, char *buf) +{ + return copy_string_result (buf, ctime (t)); +} + +struct tm * +gmtime_r (time_t const * restrict t, struct tm * restrict tp) +{ + return copy_tm_result (tp, gmtime (t)); +} + +struct tm * +localtime_r (time_t const * restrict t, struct tm * restrict tp) +{ + return copy_tm_result (tp, localtime (t)); +} diff --git a/contrib/diff/lib/time_r.h b/contrib/diff/lib/time_r.h new file mode 100644 index 0000000..fbd3375 --- /dev/null +++ b/contrib/diff/lib/time_r.h @@ -0,0 +1,46 @@ +/* Reentrant time functions like localtime_r. + + Copyright (C) 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#ifndef _TIME_R_H +#define _TIME_R_H + +/* Include first, since it may declare these functions with + signatures that disagree with POSIX, and we don't want to rename + those declarations. */ +#include + +#if !HAVE_TIME_R_POSIX +# undef asctime_r +# undef ctime_r +# undef gmtime_r +# undef localtime_r + +# define asctime_r rpl_asctime_r +# define ctime_r rpl_ctime_r +# define gmtime_r rpl_gmtime_r +# define localtime_r rpl_localtime_r + +char *asctime_r (struct tm const * restrict, char * restrict); +char *ctime_r (time_t const *, char *); +struct tm *gmtime_r (time_t const * restrict, struct tm * restrict); +struct tm *localtime_r (time_t const * restrict, struct tm * restrict); +#endif + +#endif diff --git a/contrib/diff/lib/umaxtostr.c b/contrib/diff/lib/umaxtostr.c new file mode 100644 index 0000000..4f49a7f --- /dev/null +++ b/contrib/diff/lib/umaxtostr.c @@ -0,0 +1,3 @@ +#define inttostr umaxtostr +#define inttype uintmax_t +#include "inttostr.c" diff --git a/contrib/diff/lib/unlocked-io.h b/contrib/diff/lib/unlocked-io.h new file mode 100644 index 0000000..36a7a48 --- /dev/null +++ b/contrib/diff/lib/unlocked-io.h @@ -0,0 +1,132 @@ +/* Prefer faster, non-thread-safe stdio functions if available. + + Copyright (C) 2001, 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License along + with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Jim Meyering. */ + +#ifndef UNLOCKED_IO_H +# define UNLOCKED_IO_H 1 + +# ifndef USE_UNLOCKED_IO +# define USE_UNLOCKED_IO 1 +# endif + +# if USE_UNLOCKED_IO + +/* These are wrappers for functions/macros from the GNU C library, and + from other C libraries supporting POSIX's optional thread-safe functions. + + The standard I/O functions are thread-safe. These *_unlocked ones are + more efficient but not thread-safe. That they're not thread-safe is + fine since all of the applications in this package are single threaded. + + Also, some code that is shared with the GNU C library may invoke + the *_unlocked functions directly. On hosts that lack those + functions, invoke the non-thread-safe versions instead. */ + +# include + +# if HAVE_DECL_CLEARERR_UNLOCKED +# undef clearerr +# define clearerr(x) clearerr_unlocked (x) +# else +# define clearerr_unlocked(x) clearerr (x) +# endif +# if HAVE_DECL_FEOF_UNLOCKED +# undef feof +# define feof(x) feof_unlocked (x) +# else +# define feof_unlocked(x) feof (x) +# endif +# if HAVE_DECL_FERROR_UNLOCKED +# undef ferror +# define ferror(x) ferror_unlocked (x) +# else +# define ferror_unlocked(x) ferror (x) +# endif +# if HAVE_DECL_FFLUSH_UNLOCKED +# undef fflush +# define fflush(x) fflush_unlocked (x) +# else +# define fflush_unlocked(x) fflush (x) +# endif +# if HAVE_DECL_FGETS_UNLOCKED +# undef fgets +# define fgets(x,y,z) fgets_unlocked (x,y,z) +# else +# define fgets_unlocked(x,y,z) fgets (x,y,z) +# endif +# if HAVE_DECL_FPUTC_UNLOCKED +# undef fputc +# define fputc(x,y) fputc_unlocked (x,y) +# else +# define fputc_unlocked(x,y) fputc (x,y) +# endif +# if HAVE_DECL_FPUTS_UNLOCKED +# undef fputs +# define fputs(x,y) fputs_unlocked (x,y) +# else +# define fputs_unlocked(x,y) fputs (x,y) +# endif +# if HAVE_DECL_FREAD_UNLOCKED +# undef fread +# define fread(w,x,y,z) fread_unlocked (w,x,y,z) +# else +# define fread_unlocked(w,x,y,z) fread (w,x,y,z) +# endif +# if HAVE_DECL_FWRITE_UNLOCKED +# undef fwrite +# define fwrite(w,x,y,z) fwrite_unlocked (w,x,y,z) +# else +# define fwrite_unlocked(w,x,y,z) fwrite (w,x,y,z) +# endif +# if HAVE_DECL_GETC_UNLOCKED +# undef getc +# define getc(x) getc_unlocked (x) +# else +# define getc_unlocked(x) getc (x) +# endif +# if HAVE_DECL_GETCHAR_UNLOCKED +# undef getchar +# define getchar() getchar_unlocked () +# else +# define getchar_unlocked() getchar () +# endif +# if HAVE_DECL_PUTC_UNLOCKED +# undef putc +# define putc(x,y) putc_unlocked (x,y) +# else +# define putc_unlocked(x,y) putc (x,y) +# endif +# if HAVE_DECL_PUTCHAR_UNLOCKED +# undef putchar +# define putchar(x) putchar_unlocked (x) +# else +# define putchar_unlocked(x) putchar (x) +# endif + +# undef flockfile +# define flockfile(x) ((void) 0) + +# undef ftrylockfile +# define ftrylockfile(x) 0 + +# undef funlockfile +# define funlockfile(x) ((void) 0) + +# endif /* USE_UNLOCKED_IO */ +#endif /* UNLOCKED_IO_H */ diff --git a/contrib/diff/lib/version-etc.c b/contrib/diff/lib/version-etc.c new file mode 100644 index 0000000..7523b08 --- /dev/null +++ b/contrib/diff/lib/version-etc.c @@ -0,0 +1,176 @@ +/* Utility to help print --version output in a consistent format. + Copyright (C) 1999-2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Jim Meyering. */ + +#if HAVE_CONFIG_H +# include +#endif + +/* Specification. */ +#include "version-etc.h" + +#include +#include +#include +#include "unlocked-io.h" + +#include "gettext.h" +#define _(msgid) gettext (msgid) + +/* Default copyright goes to the FSF. */ + +const char* version_etc_copyright = + /* Do *not* mark this string for translation. */ + "Copyright (C) 2004 Free Software Foundation, Inc."; + + +/* Like version_etc, below, but with the NULL-terminated author list + provided via a variable of type va_list. */ +void +version_etc_va (FILE *stream, + const char *command_name, const char *package, + const char *version, va_list authors) +{ + unsigned int n_authors; + + /* Count the number of authors. */ + { + va_list tmp_authors; + +#ifdef __va_copy + __va_copy (tmp_authors, authors); +#else + tmp_authors = authors; +#endif + + n_authors = 0; + while (va_arg (tmp_authors, const char *) != NULL) + ++n_authors; + } + + if (command_name) + fprintf (stream, "%s (%s) %s\n", command_name, package, version); + else + fprintf (stream, "%s %s\n", package, version); + + switch (n_authors) + { + case 0: + /* The caller must provide at least one author name. */ + abort (); + case 1: + /* TRANSLATORS: %s denotes an author name. */ + vfprintf (stream, _("Written by %s.\n"), authors); + break; + case 2: + /* TRANSLATORS: Each %s denotes an author name. */ + vfprintf (stream, _("Written by %s and %s.\n"), authors); + break; + case 3: + /* TRANSLATORS: Each %s denotes an author name. */ + vfprintf (stream, _("Written by %s, %s, and %s.\n"), authors); + break; + case 4: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\nand %s.\n"), authors); + break; + case 5: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\n%s, and %s.\n"), authors); + break; + case 6: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\n%s, %s, and %s.\n"), + authors); + break; + case 7: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("Written by %s, %s, %s,\n%s, %s, %s, and %s.\n"), + authors); + break; + case 8: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("\ +Written by %s, %s, %s,\n%s, %s, %s, %s,\nand %s.\n"), + authors); + break; + case 9: + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("\ +Written by %s, %s, %s,\n%s, %s, %s, %s,\n%s, and %s.\n"), + authors); + break; + default: + /* 10 or more authors. Use an abbreviation, since the human reader + will probably not want to read the entire list anyway. */ + /* TRANSLATORS: Each %s denotes an author name. + You can use line breaks, estimating that each author name occupies + ca. 16 screen columns and that a screen line has ca. 80 columns. */ + vfprintf (stream, _("\ +Written by %s, %s, %s,\n%s, %s, %s, %s,\n%s, %s, and others.\n"), + authors); + break; + } + va_end (authors); + putc ('\n', stream); + + fputs (version_etc_copyright, stream); + putc ('\n', stream); + + fputs (_("\ +This is free software; see the source for copying conditions. There is NO\n\ +warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.\n"), + stream); +} + + +/* Display the --version information the standard way. + + If COMMAND_NAME is NULL, the PACKAGE is asumed to be the name of + the program. The formats are therefore: + + PACKAGE VERSION + + or + + COMMAND_NAME (PACKAGE) VERSION. + + The author names are passed as separate arguments, with an additional + NULL argument at the end. */ +void +version_etc (FILE *stream, + const char *command_name, const char *package, + const char *version, /* const char *author1, ...*/ ...) +{ + va_list authors; + + va_start (authors, version); + version_etc_va (stream, command_name, package, version, authors); +} diff --git a/contrib/diff/lib/version-etc.h b/contrib/diff/lib/version-etc.h new file mode 100644 index 0000000..d505e75 --- /dev/null +++ b/contrib/diff/lib/version-etc.h @@ -0,0 +1,37 @@ +/* Utility to help print --version output in a consistent format. + Copyright (C) 1999, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Jim Meyering. */ + +#ifndef VERSION_ETC_H +# define VERSION_ETC_H 1 + +# include +# include + +extern const char *version_etc_copyright; + +extern void version_etc_va (FILE *stream, + const char *command_name, const char *package, + const char *version, va_list authors); + +extern void version_etc (FILE *stream, + const char *command_name, const char *package, + const char *version, + /* const char *author1, ...*/ ...); + +#endif /* VERSION_ETC_H */ diff --git a/contrib/diff/lib/waitpid.c b/contrib/diff/lib/waitpid.c new file mode 100644 index 0000000..bc07898 --- /dev/null +++ b/contrib/diff/lib/waitpid.c @@ -0,0 +1,72 @@ +/* Emulate waitpid on systems that just have wait. + Copyright (C) 1994, 1995, 1998, 1999 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include +#ifndef errno +extern int errno; +#endif + +#define WAITPID_CHILDREN 8 +static pid_t waited_pid[WAITPID_CHILDREN]; +static int waited_status[WAITPID_CHILDREN]; + +pid_t +waitpid (pid_t pid, int *stat_loc, int options) +{ + int i; + pid_t p; + + if (!options && (pid == -1 || 0 < pid)) + { + /* If we have already waited for this child, return it immediately. */ + for (i = 0; i < WAITPID_CHILDREN; i++) + { + p = waited_pid[i]; + if (p && (p == pid || pid == -1)) + { + waited_pid[i] = 0; + goto success; + } + } + + /* The child has not returned yet; wait for it, accumulating status. */ + for (i = 0; i < WAITPID_CHILDREN; i++) + if (! waited_pid[i]) + { + p = wait (&waited_status[i]); + if (p < 0) + return p; + if (p == pid || pid == -1) + goto success; + waited_pid[i] = p; + } + } + + /* We cannot emulate this wait call, e.g. because of too many children. */ + errno = EINVAL; + return -1; + +success: + if (stat_loc) + *stat_loc = waited_status[i]; + return p; +} diff --git a/contrib/diff/lib/xalloc.h b/contrib/diff/lib/xalloc.h new file mode 100644 index 0000000..4b65858 --- /dev/null +++ b/contrib/diff/lib/xalloc.h @@ -0,0 +1,87 @@ +/* xalloc.h -- malloc with out-of-memory checking + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, + 1999, 2000, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef XALLOC_H_ +# define XALLOC_H_ + +# include + +# ifndef __attribute__ +# if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 8) || __STRICT_ANSI__ +# define __attribute__(x) +# endif +# endif + +# ifndef ATTRIBUTE_NORETURN +# define ATTRIBUTE_NORETURN __attribute__ ((__noreturn__)) +# endif + +/* If this pointer is non-zero, run the specified function upon each + allocation failure. It is initialized to zero. */ +extern void (*xalloc_fail_func) (void); + +/* If XALLOC_FAIL_FUNC is undefined or a function that returns, this + message is output. It is translated via gettext. + Its value is "memory exhausted". */ +extern char const xalloc_msg_memory_exhausted[]; + +/* This function is always triggered when memory is exhausted. It is + in charge of honoring the two previous items. It exits with status + exit_failure (defined in exitfail.h). This is the + function to call when one wants the program to die because of a + memory allocation failure. */ +extern void xalloc_die (void) ATTRIBUTE_NORETURN; + +void *xmalloc (size_t s); +void *xnmalloc (size_t n, size_t s); +void *xzalloc (size_t s); +void *xcalloc (size_t n, size_t s); +void *xrealloc (void *p, size_t s); +void *xnrealloc (void *p, size_t n, size_t s); +void *x2realloc (void *p, size_t *pn); +void *x2nrealloc (void *p, size_t *pn, size_t s); +void *xclone (void const *p, size_t s); +char *xstrdup (const char *str); + +/* Return 1 if an array of N objects, each of size S, cannot exist due + to size arithmetic overflow. S must be positive and N must be + nonnegative. This is a macro, not an inline function, so that it + works correctly even when SIZE_MAX < N. + + By gnulib convention, SIZE_MAX represents overflow in size + calculations, so the conservative dividend to use here is + SIZE_MAX - 1, since SIZE_MAX might represent an overflowed value. + However, malloc (SIZE_MAX) fails on all known hosts where + sizeof (ptrdiff_t) <= sizeof (size_t), so do not bother to test for + exactly-SIZE_MAX allocations on such hosts; this avoids a test and + branch when S is known to be 1. */ +# define xalloc_oversized(n, s) \ + ((size_t) (sizeof (ptrdiff_t) <= sizeof (size_t) ? -1 : -2) / (s) < (n)) + +/* These macros are deprecated; they will go away soon, and are retained + temporarily only to ease conversion to the functions described above. */ +# define CCLONE(p, n) xclone (p, (n) * sizeof *(p)) +# define CLONE(p) xclone (p, sizeof *(p)) +# define NEW(type, var) type *var = xmalloc (sizeof (type)) +# define XCALLOC(type, n) xcalloc (n, sizeof (type)) +# define XMALLOC(type, n) xnmalloc (n, sizeof (type)) +# define XREALLOC(p, type, n) xnrealloc (p, n, sizeof (type)) +# define XFREE(p) free (p) + +#endif /* !XALLOC_H_ */ diff --git a/contrib/diff/lib/xmalloc.c b/contrib/diff/lib/xmalloc.c new file mode 100644 index 0000000..181006b --- /dev/null +++ b/contrib/diff/lib/xmalloc.c @@ -0,0 +1,255 @@ +/* xmalloc.c -- malloc with out of memory checking + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 2003, + 1999, 2000, 2002, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +#include "xalloc.h" + +#include +#include + +#include "gettext.h" +#define _(msgid) gettext (msgid) +#define N_(msgid) msgid + +#include "error.h" +#include "exitfail.h" + +#ifndef SIZE_MAX +# define SIZE_MAX ((size_t) -1) +#endif + +#ifndef HAVE_MALLOC +"you must run the autoconf test for a GNU libc compatible malloc" +#endif + +#ifndef HAVE_REALLOC +"you must run the autoconf test for a GNU libc compatible realloc" +#endif + +/* If non NULL, call this function when memory is exhausted. */ +void (*xalloc_fail_func) (void) = 0; + +/* If XALLOC_FAIL_FUNC is NULL, or does return, display this message + before exiting when memory is exhausted. Goes through gettext. */ +char const xalloc_msg_memory_exhausted[] = N_("memory exhausted"); + +void +xalloc_die (void) +{ + if (xalloc_fail_func) + (*xalloc_fail_func) (); + error (exit_failure, 0, "%s", _(xalloc_msg_memory_exhausted)); + /* The `noreturn' cannot be given to error, since it may return if + its first argument is 0. To help compilers understand the + xalloc_die does terminate, call abort. */ + abort (); +} + +/* Allocate an array of N objects, each with S bytes of memory, + dynamically, with error checking. S must be nonzero. */ + +static inline void * +xnmalloc_inline (size_t n, size_t s) +{ + void *p; + if (xalloc_oversized (n, s) || ! (p = malloc (n * s))) + xalloc_die (); + return p; +} + +void * +xnmalloc (size_t n, size_t s) +{ + return xnmalloc_inline (n, s); +} + +/* Allocate N bytes of memory dynamically, with error checking. */ + +void * +xmalloc (size_t n) +{ + return xnmalloc_inline (n, 1); +} + +/* Change the size of an allocated block of memory P to an array of N + objects each of S bytes, with error checking. S must be nonzero. */ + +static inline void * +xnrealloc_inline (void *p, size_t n, size_t s) +{ + if (xalloc_oversized (n, s) || ! (p = realloc (p, n * s))) + xalloc_die (); + return p; +} + +void * +xnrealloc (void *p, size_t n, size_t s) +{ + return xnrealloc_inline (p, n, s); +} + +/* Change the size of an allocated block of memory P to N bytes, + with error checking. */ + +void * +xrealloc (void *p, size_t n) +{ + return xnrealloc_inline (p, n, 1); +} + + +/* If P is null, allocate a block of at least *PN such objects; + otherwise, reallocate P so that it contains more than *PN objects + each of S bytes. *PN must be nonzero unless P is null, and S must + be nonzero. Set *PN to the new number of objects, and return the + pointer to the new block. *PN is never set to zero, and the + returned pointer is never null. + + Repeated reallocations are guaranteed to make progress, either by + allocating an initial block with a nonzero size, or by allocating a + larger block. + + In the following implementation, nonzero sizes are doubled so that + repeated reallocations have O(N log N) overall cost rather than + O(N**2) cost, but the specification for this function does not + guarantee that sizes are doubled. + + Here is an example of use: + + int *p = NULL; + size_t used = 0; + size_t allocated = 0; + + void + append_int (int value) + { + if (used == allocated) + p = x2nrealloc (p, &allocated, sizeof *p); + p[used++] = value; + } + + This causes x2nrealloc to allocate a block of some nonzero size the + first time it is called. + + To have finer-grained control over the initial size, set *PN to a + nonzero value before calling this function with P == NULL. For + example: + + int *p = NULL; + size_t used = 0; + size_t allocated = 0; + size_t allocated1 = 1000; + + void + append_int (int value) + { + if (used == allocated) + { + p = x2nrealloc (p, &allocated1, sizeof *p); + allocated = allocated1; + } + p[used++] = value; + } + + */ + +static inline void * +x2nrealloc_inline (void *p, size_t *pn, size_t s) +{ + size_t n = *pn; + + if (! p) + { + if (! n) + { + /* The approximate size to use for initial small allocation + requests, when the invoking code specifies an old size of + zero. 64 bytes is the largest "small" request for the + GNU C library malloc. */ + enum { DEFAULT_MXFAST = 64 }; + + n = DEFAULT_MXFAST / s; + n += !n; + } + } + else + { + if (SIZE_MAX / 2 / s < n) + xalloc_die (); + n *= 2; + } + + *pn = n; + return xrealloc (p, n * s); +} + +void * +x2nrealloc (void *p, size_t *pn, size_t s) +{ + return x2nrealloc_inline (p, pn, s); +} + +/* If P is null, allocate a block of at least *PN bytes; otherwise, + reallocate P so that it contains more than *PN bytes. *PN must be + nonzero unless P is null. Set *PN to the new block's size, and + return the pointer to the new block. *PN is never set to zero, and + the returned pointer is never null. */ + +void * +x2realloc (void *p, size_t *pn) +{ + return x2nrealloc_inline (p, pn, 1); +} + +/* Allocate S bytes of zeroed memory dynamically, with error checking. + There's no need for xnzalloc (N, S), since it would be equivalent + to xcalloc (N, S). */ + +void * +xzalloc (size_t s) +{ + return memset (xmalloc (s), 0, s); +} + +/* Allocate zeroed memory for N elements of S bytes, with error + checking. S must be nonzero. */ + +void * +xcalloc (size_t n, size_t s) +{ + void *p; + /* Test for overflow, since some calloc implementations don't have + proper overflow checks. */ + if (xalloc_oversized (n, s) || ! (p = calloc (n, s))) + xalloc_die (); + return p; +} + +/* Clone an object P of size S, with error checking. There's no need + for xnclone (P, N, S), since xclone (P, N * S) works without any + need for an arithmetic overflow check. */ + +void * +xclone (void const *p, size_t s) +{ + return memcpy (xmalloc (s), p, s); +} diff --git a/contrib/diff/lib/xstrdup.c b/contrib/diff/lib/xstrdup.c new file mode 100644 index 0000000..58f18be --- /dev/null +++ b/contrib/diff/lib/xstrdup.c @@ -0,0 +1,33 @@ +/* xstrdup.c -- copy a string with out of memory checking + Copyright (C) 1990, 1996, 1998, 2001, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#if HAVE_CONFIG_H +# include +#endif + +/* Specification. */ +#include "xalloc.h" + +#include + +/* Return a newly allocated copy of STRING. */ + +char * +xstrdup (const char *string) +{ + return xclone (string, strlen (string) + 1); +} diff --git a/contrib/diff/lib/xstrtol.c b/contrib/diff/lib/xstrtol.c new file mode 100644 index 0000000..d0aa0a9 --- /dev/null +++ b/contrib/diff/lib/xstrtol.c @@ -0,0 +1,295 @@ +/* A more useful interface to strtol. + + Copyright (C) 1995, 1996, 1998, 1999, 2000, 2001, 2003 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Jim Meyering. */ + +#if HAVE_CONFIG_H +# include +#endif + +#ifndef __strtol +# define __strtol strtol +# define __strtol_t long int +# define __xstrtol xstrtol +# define STRTOL_T_MINIMUM LONG_MIN +# define STRTOL_T_MAXIMUM LONG_MAX +#endif + +/* Some pre-ANSI implementations (e.g. SunOS 4) + need stderr defined if assertion checking is enabled. */ +#include + +#include +#include +#include +#include + +#include +#ifndef errno +extern int errno; +#endif + +#include + +/* The extra casts work around common compiler bugs. */ +#define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) +#define TYPE_MINIMUM(t) ((t) (TYPE_SIGNED (t) \ + ? ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1) \ + : (t) 0)) +#define TYPE_MAXIMUM(t) ((t) (~ (t) 0 - TYPE_MINIMUM (t))) + +#ifndef STRTOL_T_MINIMUM +# define STRTOL_T_MINIMUM TYPE_MINIMUM (__strtol_t) +# define STRTOL_T_MAXIMUM TYPE_MAXIMUM (__strtol_t) +#endif + +#if defined (STDC_HEADERS) || (!defined (isascii) && !defined (HAVE_ISASCII)) +# define IN_CTYPE_DOMAIN(c) 1 +#else +# define IN_CTYPE_DOMAIN(c) isascii(c) +#endif + +#define ISSPACE(c) (IN_CTYPE_DOMAIN (c) && isspace (c)) + +#include "xstrtol.h" + +#if !HAVE_DECL_STRTOIMAX && !defined strtoimax +intmax_t strtoimax (); +#endif + +#if !HAVE_DECL_STRTOUMAX && !defined strtoumax +uintmax_t strtoumax (); +#endif + +static strtol_error +bkm_scale (__strtol_t *x, int scale_factor) +{ + if (TYPE_SIGNED (__strtol_t) && *x < STRTOL_T_MINIMUM / scale_factor) + { + *x = STRTOL_T_MINIMUM; + return LONGINT_OVERFLOW; + } + if (STRTOL_T_MAXIMUM / scale_factor < *x) + { + *x = STRTOL_T_MAXIMUM; + return LONGINT_OVERFLOW; + } + *x *= scale_factor; + return LONGINT_OK; +} + +static strtol_error +bkm_scale_by_power (__strtol_t *x, int base, int power) +{ + strtol_error err = LONGINT_OK; + while (power--) + err |= bkm_scale (x, base); + return err; +} + +/* FIXME: comment. */ + +strtol_error +__xstrtol (const char *s, char **ptr, int strtol_base, + __strtol_t *val, const char *valid_suffixes) +{ + char *t_ptr; + char **p; + __strtol_t tmp; + strtol_error err = LONGINT_OK; + + assert (0 <= strtol_base && strtol_base <= 36); + + p = (ptr ? ptr : &t_ptr); + + if (! TYPE_SIGNED (__strtol_t)) + { + const char *q = s; + while (ISSPACE ((unsigned char) *q)) + ++q; + if (*q == '-') + return LONGINT_INVALID; + } + + errno = 0; + tmp = __strtol (s, p, strtol_base); + + if (*p == s) + { + /* If there is no number but there is a valid suffix, assume the + number is 1. The string is invalid otherwise. */ + if (valid_suffixes && **p && strchr (valid_suffixes, **p)) + tmp = 1; + else + return LONGINT_INVALID; + } + else if (errno != 0) + { + if (errno != ERANGE) + return LONGINT_INVALID; + err = LONGINT_OVERFLOW; + } + + /* Let valid_suffixes == NULL mean `allow any suffix'. */ + /* FIXME: update all callers except the ones that allow suffixes + after the number, changing last parameter NULL to `""'. */ + if (!valid_suffixes) + { + *val = tmp; + return err; + } + + if (**p != '\0') + { + int base = 1024; + int suffixes = 1; + strtol_error overflow; + + if (!strchr (valid_suffixes, **p)) + { + *val = tmp; + return err | LONGINT_INVALID_SUFFIX_CHAR; + } + + if (strchr (valid_suffixes, '0')) + { + /* The ``valid suffix'' '0' is a special flag meaning that + an optional second suffix is allowed, which can change + the base. A suffix "B" (e.g. "100MB") stands for a power + of 1000, whereas a suffix "iB" (e.g. "100MiB") stands for + a power of 1024. If no suffix (e.g. "100M"), assume + power-of-1024. */ + + switch (p[0][1]) + { + case 'i': + if (p[0][2] == 'B') + suffixes += 2; + break; + + case 'B': + case 'D': /* 'D' is obsolescent */ + base = 1000; + suffixes++; + break; + } + } + + switch (**p) + { + case 'b': + overflow = bkm_scale (&tmp, 512); + break; + + case 'B': + overflow = bkm_scale (&tmp, 1024); + break; + + case 'c': + overflow = 0; + break; + + case 'E': /* exa or exbi */ + overflow = bkm_scale_by_power (&tmp, base, 6); + break; + + case 'G': /* giga or gibi */ + case 'g': /* 'g' is undocumented; for compatibility only */ + overflow = bkm_scale_by_power (&tmp, base, 3); + break; + + case 'k': /* kilo */ + case 'K': /* kibi */ + overflow = bkm_scale_by_power (&tmp, base, 1); + break; + + case 'M': /* mega or mebi */ + case 'm': /* 'm' is undocumented; for compatibility only */ + overflow = bkm_scale_by_power (&tmp, base, 2); + break; + + case 'P': /* peta or pebi */ + overflow = bkm_scale_by_power (&tmp, base, 5); + break; + + case 'T': /* tera or tebi */ + case 't': /* 't' is undocumented; for compatibility only */ + overflow = bkm_scale_by_power (&tmp, base, 4); + break; + + case 'w': + overflow = bkm_scale (&tmp, 2); + break; + + case 'Y': /* yotta or 2**80 */ + overflow = bkm_scale_by_power (&tmp, base, 8); + break; + + case 'Z': /* zetta or 2**70 */ + overflow = bkm_scale_by_power (&tmp, base, 7); + break; + + default: + *val = tmp; + return err | LONGINT_INVALID_SUFFIX_CHAR; + } + + err |= overflow; + *p += suffixes; + if (**p) + err |= LONGINT_INVALID_SUFFIX_CHAR; + } + + *val = tmp; + return err; +} + +#ifdef TESTING_XSTRTO + +# include +# include "error.h" + +char *program_name; + +int +main (int argc, char **argv) +{ + strtol_error s_err; + int i; + + program_name = argv[0]; + for (i=1; i%lu (%s)\n", argv[i], val, p); + } + else + { + STRTOL_FATAL_ERROR (argv[i], "arg", s_err); + } + } + exit (0); +} + +#endif /* TESTING_XSTRTO */ diff --git a/contrib/diff/lib/xstrtol.h b/contrib/diff/lib/xstrtol.h new file mode 100644 index 0000000..5765c88 --- /dev/null +++ b/contrib/diff/lib/xstrtol.h @@ -0,0 +1,91 @@ +/* A more useful interface to strtol. + + Copyright (C) 1995, 1996, 1998, 1999, 2001, 2002, 2003, 2004 Free + Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#ifndef XSTRTOL_H_ +# define XSTRTOL_H_ 1 + +# include "exitfail.h" + +/* Get uintmax_t. */ +# if HAVE_INTTYPES_H +# include +# else +# if HAVE_STDINT_H +# include +# endif +# endif + +# ifndef _STRTOL_ERROR +enum strtol_error + { + LONGINT_OK = 0, + + /* These two values can be ORed together, to indicate that both + errors occurred. */ + LONGINT_OVERFLOW = 1, + LONGINT_INVALID_SUFFIX_CHAR = 2, + + LONGINT_INVALID_SUFFIX_CHAR_WITH_OVERFLOW = (LONGINT_INVALID_SUFFIX_CHAR + | LONGINT_OVERFLOW), + LONGINT_INVALID = 4 + }; +typedef enum strtol_error strtol_error; +# endif + +# define _DECLARE_XSTRTOL(name, type) \ + strtol_error name (const char *, char **, int, type *, const char *); +_DECLARE_XSTRTOL (xstrtol, long int) +_DECLARE_XSTRTOL (xstrtoul, unsigned long int) +_DECLARE_XSTRTOL (xstrtoimax, intmax_t) +_DECLARE_XSTRTOL (xstrtoumax, uintmax_t) + +# define _STRTOL_ERROR(Exit_code, Str, Argument_type_string, Err) \ + do \ + { \ + switch ((Err)) \ + { \ + default: \ + abort (); \ + \ + case LONGINT_INVALID: \ + error ((Exit_code), 0, "invalid %s `%s'", \ + (Argument_type_string), (Str)); \ + break; \ + \ + case LONGINT_INVALID_SUFFIX_CHAR: \ + case LONGINT_INVALID_SUFFIX_CHAR | LONGINT_OVERFLOW: \ + error ((Exit_code), 0, "invalid character following %s in `%s'", \ + (Argument_type_string), (Str)); \ + break; \ + \ + case LONGINT_OVERFLOW: \ + error ((Exit_code), 0, "%s `%s' too large", \ + (Argument_type_string), (Str)); \ + break; \ + } \ + } \ + while (0) + +# define STRTOL_FATAL_ERROR(Str, Argument_type_string, Err) \ + _STRTOL_ERROR (exit_failure, Str, Argument_type_string, Err) + +# define STRTOL_FAIL_WARN(Str, Argument_type_string, Err) \ + _STRTOL_ERROR (0, Str, Argument_type_string, Err) + +#endif /* not XSTRTOL_H_ */ diff --git a/contrib/diff/lib/xstrtoul.c b/contrib/diff/lib/xstrtoul.c new file mode 100644 index 0000000..285f7b9 --- /dev/null +++ b/contrib/diff/lib/xstrtoul.c @@ -0,0 +1,6 @@ +#define __strtol strtoul +#define __strtol_t unsigned long int +#define __xstrtol xstrtoul +#define STRTOL_T_MINIMUM 0 +#define STRTOL_T_MAXIMUM ULONG_MAX +#include "xstrtol.c" diff --git a/contrib/diff/lib/xstrtoumax.c b/contrib/diff/lib/xstrtoumax.c new file mode 100644 index 0000000..a8858bf --- /dev/null +++ b/contrib/diff/lib/xstrtoumax.c @@ -0,0 +1,37 @@ +/* xstrtoumax.c -- A more useful interface to strtoumax. + Copyright (C) 1999, 2003 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; if not, write to the Free Software Foundation, + Inc., 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* Written by Paul Eggert. */ + +#if HAVE_CONFIG_H +# include +#endif + +#if HAVE_INTTYPES_H +# include +#elif HAVE_STDINT_H +# include +#endif + +#define __strtol strtoumax +#define __strtol_t uintmax_t +#define __xstrtol xstrtoumax +#ifdef UINTMAX_MAX +# define STRTOL_T_MINIMUM 0 +# define STRTOL_T_MAXIMUM UINTMAX_MAX +#endif +#include "xstrtol.c" diff --git a/contrib/diff/man/Makefile.am b/contrib/diff/man/Makefile.am new file mode 100644 index 0000000..d3a1ee2 --- /dev/null +++ b/contrib/diff/man/Makefile.am @@ -0,0 +1,34 @@ +# Automakefile for GNU diffutils man pages + +# Copyright (C) 2002 Free Software Foundation, Inc. + +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; either version 2, or (at your option) +## any later version. + +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. + +## You should have received a copy of the GNU General Public License +## along with this program; if not, write to the Free Software +## Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA +## 02111-1307, USA. + +dist_man_MANS = cmp.1 diff.1 diff3.1 sdiff.1 +MAINTAINERCLEANFILES = $(dist_man_MANS) + +S = $(top_srcdir)/src +cmp.1: $S/cmp.c +diff.1: $S/diff.c +diff3.1: $S/diff3.c +sdiff.1: $S/sdiff.c + +# Depend on configure.ac to get version number changes. +$(dist_man_MANS): $(top_srcdir)/configure.ac + base=`expr $@ : '\(.*\).1'` && \ + (echo '[NAME]' && sed 's@/\* *@@; s/-/\\-/; q' $S/$$base.c) | \ + $(HELP2MAN) -i - -S '$(PACKAGE) $(VERSION)' ../src/$$base | \ + sed 's/^\.B info .*/.B info diff/' >$@ diff --git a/contrib/diff/man/cmp.1 b/contrib/diff/man/cmp.1 new file mode 100644 index 0000000..ffc664e --- /dev/null +++ b/contrib/diff/man/cmp.1 @@ -0,0 +1,63 @@ +.\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.33. +.TH CMP "1" "April 2004" "diffutils 2.8.7" "User Commands" +.SH NAME +cmp \- compare two files byte by byte +.SH SYNOPSIS +.B cmp +[\fIOPTION\fR]... \fIFILE1 \fR[\fIFILE2 \fR[\fISKIP1 \fR[\fISKIP2\fR]]] +.SH DESCRIPTION +Compare two files byte by byte. +.TP +\fB\-b\fR \fB\-\-print\-bytes\fR +Print differing bytes. +.TP +\fB\-i\fR SKIP \fB\-\-ignore\-initial\fR=\fISKIP\fR +Skip the first SKIP bytes of input. +.HP +\fB\-i\fR SKIP1:SKIP2 \fB\-\-ignore\-initial\fR=\fISKIP1\fR:SKIP2 +.IP +Skip the first SKIP1 bytes of FILE1 and the first SKIP2 bytes of FILE2. +.TP +\fB\-l\fR \fB\-\-verbose\fR +Output byte numbers and values of all differing bytes. +.TP +\fB\-n\fR LIMIT \fB\-\-bytes\fR=\fILIMIT\fR +Compare at most LIMIT bytes. +.TP +\fB\-s\fR \fB\-\-quiet\fR \fB\-\-silent\fR +Output nothing; yield exit status only. +.TP +\fB\-v\fR \fB\-\-version\fR +Output version info. +.TP +\fB\-\-help\fR +Output this help. +.PP +SKIP1 and SKIP2 are the number of bytes to skip in each file. +SKIP values may be followed by the following multiplicative suffixes: +kB 1000, K 1024, MB 1,000,000, M 1,048,576, +GB 1,000,000,000, G 1,073,741,824, and so on for T, P, E, Z, Y. +.PP +If a FILE is `-' or missing, read standard input. +Exit status is 0 if inputs are the same, 1 if different, 2 if trouble. +.SH AUTHOR +Written by Torbjorn Granlund and David MacKenzie. +.SH "REPORTING BUGS" +Report bugs to . +.SH COPYRIGHT +Copyright \(co 2004 Free Software Foundation, Inc. +.br +This is free software; see the source for copying conditions. There is NO +warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +.SH "SEE ALSO" +The full documentation for +.B cmp +is maintained as a Texinfo manual. If the +.B info +and +.B cmp +programs are properly installed at your site, the command +.IP +.B info diff +.PP +should give you access to the complete manual. diff --git a/contrib/diff/man/diff.1 b/contrib/diff/man/diff.1 new file mode 100644 index 0000000..cef8e1a --- /dev/null +++ b/contrib/diff/man/diff.1 @@ -0,0 +1,227 @@ +.\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.33. +.TH DIFF "1" "April 2004" "diffutils 2.8.7" "User Commands" +.SH NAME +diff \- compare files line by line +.SH SYNOPSIS +.B diff +[\fIOPTION\fR]... \fIFILES\fR +.SH DESCRIPTION +Compare files line by line. +.TP +\fB\-i\fR \fB\-\-ignore\-case\fR +Ignore case differences in file contents. +.TP +\fB\-\-ignore\-file\-name\-case\fR +Ignore case when comparing file names. +.TP +\fB\-\-no\-ignore\-file\-name\-case\fR +Consider case when comparing file names. +.TP +\fB\-E\fR \fB\-\-ignore\-tab\-expansion\fR +Ignore changes due to tab expansion. +.TP +\fB\-b\fR \fB\-\-ignore\-space\-change\fR +Ignore changes in the amount of white space. +.TP +\fB\-w\fR \fB\-\-ignore\-all\-space\fR +Ignore all white space. +.TP +\fB\-B\fR \fB\-\-ignore\-blank\-lines\fR +Ignore changes whose lines are all blank. +.TP +\fB\-I\fR RE \fB\-\-ignore\-matching\-lines\fR=\fIRE\fR +Ignore changes whose lines all match RE. +.TP +\fB\-\-strip\-trailing\-cr\fR +Strip trailing carriage return on input. +.TP +\fB\-a\fR \fB\-\-text\fR +Treat all files as text. +.TP +\fB\-c\fR \fB\-C\fR NUM \fB\-\-context\fR[=\fINUM\fR] +Output NUM (default 3) lines of copied context. +.TP +\fB\-u\fR \fB\-U\fR NUM \fB\-\-unified\fR[=\fINUM\fR] +Output NUM (default 3) lines of unified context. +.TP +\fB\-\-label\fR LABEL +Use LABEL instead of file name. +.TP +\fB\-p\fR \fB\-\-show\-c\-function\fR +Show which C function each change is in. +.TP +\fB\-F\fR RE \fB\-\-show\-function\-line\fR=\fIRE\fR +Show the most recent line matching RE. +.TP +\fB\-q\fR \fB\-\-brief\fR +Output only whether files differ. +.TP +\fB\-e\fR \fB\-\-ed\fR +Output an ed script. +.TP +\fB\-\-normal\fR +Output a normal diff. +.TP +\fB\-n\fR \fB\-\-rcs\fR +Output an RCS format diff. +.TP +\fB\-y\fR \fB\-\-side\-by\-side\fR +Output in two columns. +.TP +\fB\-W\fR NUM \fB\-\-width\fR=\fINUM\fR +Output at most NUM (default 130) print columns. +.TP +\fB\-\-left\-column\fR +Output only the left column of common lines. +.TP +\fB\-\-suppress\-common\-lines\fR +Do not output common lines. +.TP +\fB\-D\fR NAME \fB\-\-ifdef\fR=\fINAME\fR +Output merged file to show `#ifdef NAME' diffs. +.TP +\fB\-\-GTYPE\-group\-format\fR=\fIGFMT\fR +Similar, but format GTYPE input groups with GFMT. +.TP +\fB\-\-line\-format\fR=\fILFMT\fR +Similar, but format all input lines with LFMT. +.TP +\fB\-\-LTYPE\-line\-format\fR=\fILFMT\fR +Similar, but format LTYPE input lines with LFMT. +.TP +LTYPE is `old', `new', or `unchanged'. +GTYPE is LTYPE or `changed'. +.IP +GFMT may contain: +.TP +%< +lines from FILE1 +.TP +%> +lines from FILE2 +.TP +%= +lines common to FILE1 and FILE2 +.TP +%[-][WIDTH][.[PREC]]{doxX}LETTER +printf-style spec for LETTER +.IP +LETTERs are as follows for new group, lower case for old group: +.TP +F +first line number +.TP +L +last line number +.TP +N +number of lines = L-F+1 +.TP +E +F-1 +.TP +M +L+1 +.IP +LFMT may contain: +.TP +%L +contents of line +.TP +%l +contents of line, excluding any trailing newline +.TP +%[-][WIDTH][.[PREC]]{doxX}n +printf-style spec for input line number +.IP +Either GFMT or LFMT may contain: +.TP +%% +% +.TP +%c'C' +the single character C +.TP +%c'\eOOO' +the character with octal code OOO +.TP +\fB\-l\fR \fB\-\-paginate\fR +Pass the output through `pr' to paginate it. +.TP +\fB\-t\fR \fB\-\-expand\-tabs\fR +Expand tabs to spaces in output. +.TP +\fB\-T\fR \fB\-\-initial\-tab\fR +Make tabs line up by prepending a tab. +.TP +\fB\-\-tabsize\fR=\fINUM\fR +Tab stops are every NUM (default 8) print columns. +.TP +\fB\-r\fR \fB\-\-recursive\fR +Recursively compare any subdirectories found. +.TP +\fB\-N\fR \fB\-\-new\-file\fR +Treat absent files as empty. +.TP +\fB\-\-unidirectional\-new\-file\fR +Treat absent first files as empty. +.TP +\fB\-s\fR \fB\-\-report\-identical\-files\fR +Report when two files are the same. +.TP +\fB\-x\fR PAT \fB\-\-exclude\fR=\fIPAT\fR +Exclude files that match PAT. +.TP +\fB\-X\fR FILE \fB\-\-exclude\-from\fR=\fIFILE\fR +Exclude files that match any pattern in FILE. +.TP +\fB\-S\fR FILE \fB\-\-starting\-file\fR=\fIFILE\fR +Start with FILE when comparing directories. +.TP +\fB\-\-from\-file\fR=\fIFILE1\fR +Compare FILE1 to all operands. FILE1 can be a directory. +.TP +\fB\-\-to\-file\fR=\fIFILE2\fR +Compare all operands to FILE2. FILE2 can be a directory. +.TP +\fB\-\-horizon\-lines\fR=\fINUM\fR +Keep NUM lines of the common prefix and suffix. +.TP +\fB\-d\fR \fB\-\-minimal\fR +Try hard to find a smaller set of changes. +.TP +\fB\-\-speed\-large\-files\fR +Assume large files and many scattered small changes. +.TP +\fB\-v\fR \fB\-\-version\fR +Output version info. +.TP +\fB\-\-help\fR +Output this help. +.PP +FILES are `FILE1 FILE2' or `DIR1 DIR2' or `DIR FILE...' or `FILE... DIR'. +If \fB\-\-from\-file\fR or \fB\-\-to\-file\fR is given, there are no restrictions on FILES. +If a FILE is `-', read standard input. +Exit status is 0 if inputs are the same, 1 if different, 2 if trouble. +.SH AUTHOR +Written by Paul Eggert, Mike Haertel, David Hayes, +Richard Stallman, and Len Tower. +.SH "REPORTING BUGS" +Report bugs to . +.SH COPYRIGHT +Copyright \(co 2004 Free Software Foundation, Inc. +.br +This is free software; see the source for copying conditions. There is NO +warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +.SH "SEE ALSO" +The full documentation for +.B diff +is maintained as a Texinfo manual. If the +.B info +and +.B diff +programs are properly installed at your site, the command +.IP +.B info diff +.PP +should give you access to the complete manual. diff --git a/contrib/diff/man/diff3.1 b/contrib/diff/man/diff3.1 new file mode 100644 index 0000000..f7daee2 --- /dev/null +++ b/contrib/diff/man/diff3.1 @@ -0,0 +1,78 @@ +.\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.33. +.TH DIFF3 "1" "April 2004" "diffutils 2.8.7" "User Commands" +.SH NAME +diff3 \- compare three files line by line +.SH SYNOPSIS +.B diff3 +[\fIOPTION\fR]... \fIMYFILE OLDFILE YOURFILE\fR +.SH DESCRIPTION +Compare three files line by line. +.TP +\fB\-e\fR \fB\-\-ed\fR +Output unmerged changes from OLDFILE to YOURFILE into MYFILE. +.TP +\fB\-E\fR \fB\-\-show\-overlap\fR +Output unmerged changes, bracketing conflicts. +.TP +\fB\-A\fR \fB\-\-show\-all\fR +Output all changes, bracketing conflicts. +.TP +\fB\-x\fR \fB\-\-overlap\-only\fR +Output overlapping changes. +.TP +\fB\-X\fR +Output overlapping changes, bracketing them. +.TP +\fB\-3\fR \fB\-\-easy\-only\fR +Output unmerged nonoverlapping changes. +.TP +\fB\-m\fR \fB\-\-merge\fR +Output merged file instead of ed script (default \fB\-A\fR). +.TP +\fB\-L\fR LABEL \fB\-\-label\fR=\fILABEL\fR +Use LABEL instead of file name. +.TP +\fB\-i\fR +Append `w' and `q' commands to ed scripts. +.TP +\fB\-a\fR \fB\-\-text\fR +Treat all files as text. +.TP +\fB\-\-strip\-trailing\-cr\fR +Strip trailing carriage return on input. +.TP +\fB\-T\fR \fB\-\-initial\-tab\fR +Make tabs line up by prepending a tab. +.TP +\fB\-\-diff\-program\fR=\fIPROGRAM\fR +Use PROGRAM to compare files. +.TP +\fB\-v\fR \fB\-\-version\fR +Output version info. +.TP +\fB\-\-help\fR +Output this help. +.PP +If a FILE is `-', read standard input. +Exit status is 0 if successful, 1 if conflicts, 2 if trouble. +.SH AUTHOR +Written by Randy Smith. +.SH "REPORTING BUGS" +Report bugs to . +.SH COPYRIGHT +Copyright \(co 2004 Free Software Foundation, Inc. +.br +This is free software; see the source for copying conditions. There is NO +warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +.SH "SEE ALSO" +The full documentation for +.B diff3 +is maintained as a Texinfo manual. If the +.B info +and +.B diff3 +programs are properly installed at your site, the command +.IP +.B info diff +.PP +should give you access to the complete manual. diff --git a/contrib/diff/man/sdiff.1 b/contrib/diff/man/sdiff.1 new file mode 100644 index 0000000..ab64c7e --- /dev/null +++ b/contrib/diff/man/sdiff.1 @@ -0,0 +1,90 @@ +.\" DO NOT MODIFY THIS FILE! It was generated by help2man 1.33. +.TH SDIFF "1" "April 2004" "diffutils 2.8.7" "User Commands" +.SH NAME +sdiff \- side-by-side merge of file differences +.SH SYNOPSIS +.B sdiff +[\fIOPTION\fR]... \fIFILE1 FILE2\fR +.SH DESCRIPTION +Side-by-side merge of file differences. +.TP +\fB\-o\fR FILE \fB\-\-output\fR=\fIFILE\fR +Operate interactively, sending output to FILE. +.TP +\fB\-i\fR \fB\-\-ignore\-case\fR +Consider upper- and lower-case to be the same. +.TP +\fB\-E\fR \fB\-\-ignore\-tab\-expansion\fR +Ignore changes due to tab expansion. +.TP +\fB\-b\fR \fB\-\-ignore\-space\-change\fR +Ignore changes in the amount of white space. +.TP +\fB\-W\fR \fB\-\-ignore\-all\-space\fR +Ignore all white space. +.TP +\fB\-B\fR \fB\-\-ignore\-blank\-lines\fR +Ignore changes whose lines are all blank. +.TP +\fB\-I\fR RE \fB\-\-ignore\-matching\-lines\fR=\fIRE\fR +Ignore changes whose lines all match RE. +.TP +\fB\-\-strip\-trailing\-cr\fR +Strip trailing carriage return on input. +.TP +\fB\-a\fR \fB\-\-text\fR +Treat all files as text. +.TP +\fB\-w\fR NUM \fB\-\-width\fR=\fINUM\fR +Output at most NUM (default 130) print columns. +.TP +\fB\-l\fR \fB\-\-left\-column\fR +Output only the left column of common lines. +.TP +\fB\-s\fR \fB\-\-suppress\-common\-lines\fR +Do not output common lines. +.TP +\fB\-t\fR \fB\-\-expand\-tabs\fR +Expand tabs to spaces in output. +.TP +\fB\-\-tabsize\fR=\fINUM\fR +Tab stops are every NUM (default 8) print columns. +.TP +\fB\-d\fR \fB\-\-minimal\fR +Try hard to find a smaller set of changes. +.TP +\fB\-H\fR \fB\-\-speed\-large\-files\fR +Assume large files and many scattered small changes. +.TP +\fB\-\-diff\-program\fR=\fIPROGRAM\fR +Use PROGRAM to compare files. +.TP +\fB\-v\fR \fB\-\-version\fR +Output version info. +.TP +\fB\-\-help\fR +Output this help. +.PP +If a FILE is `-', read standard input. +Exit status is 0 if inputs are the same, 1 if different, 2 if trouble. +.SH AUTHOR +Written by Thomas Lord. +.SH "REPORTING BUGS" +Report bugs to . +.SH COPYRIGHT +Copyright \(co 2004 Free Software Foundation, Inc. +.br +This is free software; see the source for copying conditions. There is NO +warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. +.SH "SEE ALSO" +The full documentation for +.B sdiff +is maintained as a Texinfo manual. If the +.B info +and +.B sdiff +programs are properly installed at your site, the command +.IP +.B info diff +.PP +should give you access to the complete manual. diff --git a/contrib/diff/src/Makefile.am b/contrib/diff/src/Makefile.am new file mode 100644 index 0000000..14ff6ec --- /dev/null +++ b/contrib/diff/src/Makefile.am @@ -0,0 +1,44 @@ +# Automakefile for GNU diffutils programs. + +# Copyright (C) 2001, 2002 Free Software Foundation, Inc. + +## This program is free software; you can redistribute it and/or modify +## it under the terms of the GNU General Public License as published by +## the Free Software Foundation; either version 2, or (at your option) +## any later version. + +## This program is distributed in the hope that it will be useful, +## but WITHOUT ANY WARRANTY; without even the implied warranty of +## MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the +## GNU General Public License for more details. + +## You should have received a copy of the GNU General Public License +## along with this program; if not, write to the Free Software +## Foundation, Inc., 59 Temple Place - Suite 330, Boston, MA +## 02111-1307, USA. + +bin_PROGRAMS = cmp diff diff3 sdiff + +noinst_HEADERS = system.h + +localedir = $(datadir)/locale + +AM_CPPFLAGS = -I../lib -I$(top_srcdir)/lib + +LDADD = ../lib/libdiffutils.a @LIBINTL@ +diff_LDADD = $(LDADD) @LIB_CLOCK_GETTIME@ + +cmp_SOURCES = cmp.c +diff3_SOURCES = diff3.c +sdiff_SOURCES = sdiff.c +diff_SOURCES = \ + analyze.c context.c diff.c diff.h dir.c ed.c ifdef.c io.c \ + normal.c side.c util.c + +MOSTLYCLEANFILES = paths.h paths.ht + +cmp.$(OBJEXT) diff3.$(OBJEXT) diff.$(OBJEXT) sdiff.$(OBJEXT): paths.h +paths.h: + (echo '#define DEFAULT_DIFF_PROGRAM "$(bindir)/'`echo diff|sed '$(transform)'`'"' && \ + echo '#define LOCALEDIR "$(localedir)"') >$@t + mv $@t $@ diff --git a/contrib/diff/src/analyze.c b/contrib/diff/src/analyze.c new file mode 100644 index 0000000..7d6750a --- /dev/null +++ b/contrib/diff/src/analyze.c @@ -0,0 +1,1038 @@ +/* Analyze file differences for GNU DIFF. + + Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1998, 2001, 2002, + 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +/* The basic algorithm is described in: + "An O(ND) Difference Algorithm and its Variations", Eugene Myers, + Algorithmica Vol. 1 No. 2, 1986, pp. 251-266; + see especially section 4.2, which describes the variation used below. + Unless the --minimal option is specified, this code uses the TOO_EXPENSIVE + heuristic, by Paul Eggert, to limit the cost to O(N**1.5 log N) + at the price of producing suboptimal output for large inputs with + many differences. + + The basic algorithm was independently discovered as described in: + "Algorithms for Approximate String Matching", E. Ukkonen, + Information and Control Vol. 64, 1985, pp. 100-118. */ + +#include "diff.h" +#include +#include +#include +#include + +static lin *xvec, *yvec; /* Vectors being compared. */ +static lin *fdiag; /* Vector, indexed by diagonal, containing + 1 + the X coordinate of the point furthest + along the given diagonal in the forward + search of the edit matrix. */ +static lin *bdiag; /* Vector, indexed by diagonal, containing + the X coordinate of the point furthest + along the given diagonal in the backward + search of the edit matrix. */ +static lin too_expensive; /* Edit scripts longer than this are too + expensive to compute. */ + +#define SNAKE_LIMIT 20 /* Snakes bigger than this are considered `big'. */ + +struct partition +{ + lin xmid, ymid; /* Midpoints of this partition. */ + bool lo_minimal; /* Nonzero if low half will be analyzed minimally. */ + bool hi_minimal; /* Likewise for high half. */ +}; + +/* Find the midpoint of the shortest edit script for a specified + portion of the two files. + + Scan from the beginnings of the files, and simultaneously from the ends, + doing a breadth-first search through the space of edit-sequence. + When the two searches meet, we have found the midpoint of the shortest + edit sequence. + + If FIND_MINIMAL is nonzero, find the minimal edit script regardless + of expense. Otherwise, if the search is too expensive, use + heuristics to stop the search and report a suboptimal answer. + + Set PART->(xmid,ymid) to the midpoint (XMID,YMID). The diagonal number + XMID - YMID equals the number of inserted lines minus the number + of deleted lines (counting only lines before the midpoint). + + Set PART->lo_minimal to true iff the minimal edit script for the + left half of the partition is known; similarly for PART->hi_minimal. + + This function assumes that the first lines of the specified portions + of the two files do not match, and likewise that the last lines do not + match. The caller must trim matching lines from the beginning and end + of the portions it is going to specify. + + If we return the "wrong" partitions, + the worst this can do is cause suboptimal diff output. + It cannot cause incorrect diff output. */ + +static void +diag (lin xoff, lin xlim, lin yoff, lin ylim, bool find_minimal, + struct partition *part) +{ + lin *const fd = fdiag; /* Give the compiler a chance. */ + lin *const bd = bdiag; /* Additional help for the compiler. */ + lin const *const xv = xvec; /* Still more help for the compiler. */ + lin const *const yv = yvec; /* And more and more . . . */ + lin const dmin = xoff - ylim; /* Minimum valid diagonal. */ + lin const dmax = xlim - yoff; /* Maximum valid diagonal. */ + lin const fmid = xoff - yoff; /* Center diagonal of top-down search. */ + lin const bmid = xlim - ylim; /* Center diagonal of bottom-up search. */ + lin fmin = fmid, fmax = fmid; /* Limits of top-down search. */ + lin bmin = bmid, bmax = bmid; /* Limits of bottom-up search. */ + lin c; /* Cost. */ + bool odd = (fmid - bmid) & 1; /* True if southeast corner is on an odd + diagonal with respect to the northwest. */ + + fd[fmid] = xoff; + bd[bmid] = xlim; + + for (c = 1;; ++c) + { + lin d; /* Active diagonal. */ + bool big_snake = false; + + /* Extend the top-down search by an edit step in each diagonal. */ + fmin > dmin ? fd[--fmin - 1] = -1 : ++fmin; + fmax < dmax ? fd[++fmax + 1] = -1 : --fmax; + for (d = fmax; d >= fmin; d -= 2) + { + lin x, y, oldx, tlo = fd[d - 1], thi = fd[d + 1]; + + if (tlo >= thi) + x = tlo + 1; + else + x = thi; + oldx = x; + y = x - d; + while (x < xlim && y < ylim && xv[x] == yv[y]) + ++x, ++y; + if (x - oldx > SNAKE_LIMIT) + big_snake = true; + fd[d] = x; + if (odd && bmin <= d && d <= bmax && bd[d] <= x) + { + part->xmid = x; + part->ymid = y; + part->lo_minimal = part->hi_minimal = true; + return; + } + } + + /* Similarly extend the bottom-up search. */ + bmin > dmin ? bd[--bmin - 1] = LIN_MAX : ++bmin; + bmax < dmax ? bd[++bmax + 1] = LIN_MAX : --bmax; + for (d = bmax; d >= bmin; d -= 2) + { + lin x, y, oldx, tlo = bd[d - 1], thi = bd[d + 1]; + + if (tlo < thi) + x = tlo; + else + x = thi - 1; + oldx = x; + y = x - d; + while (x > xoff && y > yoff && xv[x - 1] == yv[y - 1]) + --x, --y; + if (oldx - x > SNAKE_LIMIT) + big_snake = true; + bd[d] = x; + if (!odd && fmin <= d && d <= fmax && x <= fd[d]) + { + part->xmid = x; + part->ymid = y; + part->lo_minimal = part->hi_minimal = true; + return; + } + } + + if (find_minimal) + continue; + + /* Heuristic: check occasionally for a diagonal that has made + lots of progress compared with the edit distance. + If we have any such, find the one that has made the most + progress and return it as if it had succeeded. + + With this heuristic, for files with a constant small density + of changes, the algorithm is linear in the file size. */ + + if (200 < c && big_snake && speed_large_files) + { + lin best = 0; + + for (d = fmax; d >= fmin; d -= 2) + { + lin dd = d - fmid; + lin x = fd[d]; + lin y = x - d; + lin v = (x - xoff) * 2 - dd; + if (v > 12 * (c + (dd < 0 ? -dd : dd))) + { + if (v > best + && xoff + SNAKE_LIMIT <= x && x < xlim + && yoff + SNAKE_LIMIT <= y && y < ylim) + { + /* We have a good enough best diagonal; + now insist that it end with a significant snake. */ + int k; + + for (k = 1; xv[x - k] == yv[y - k]; k++) + if (k == SNAKE_LIMIT) + { + best = v; + part->xmid = x; + part->ymid = y; + break; + } + } + } + } + if (best > 0) + { + part->lo_minimal = true; + part->hi_minimal = false; + return; + } + + best = 0; + for (d = bmax; d >= bmin; d -= 2) + { + lin dd = d - bmid; + lin x = bd[d]; + lin y = x - d; + lin v = (xlim - x) * 2 + dd; + if (v > 12 * (c + (dd < 0 ? -dd : dd))) + { + if (v > best + && xoff < x && x <= xlim - SNAKE_LIMIT + && yoff < y && y <= ylim - SNAKE_LIMIT) + { + /* We have a good enough best diagonal; + now insist that it end with a significant snake. */ + int k; + + for (k = 0; xv[x + k] == yv[y + k]; k++) + if (k == SNAKE_LIMIT - 1) + { + best = v; + part->xmid = x; + part->ymid = y; + break; + } + } + } + } + if (best > 0) + { + part->lo_minimal = false; + part->hi_minimal = true; + return; + } + } + + /* Heuristic: if we've gone well beyond the call of duty, + give up and report halfway between our best results so far. */ + if (c >= too_expensive) + { + lin fxybest, fxbest; + lin bxybest, bxbest; + + fxbest = bxbest = 0; /* Pacify `gcc -Wall'. */ + + /* Find forward diagonal that maximizes X + Y. */ + fxybest = -1; + for (d = fmax; d >= fmin; d -= 2) + { + lin x = MIN (fd[d], xlim); + lin y = x - d; + if (ylim < y) + x = ylim + d, y = ylim; + if (fxybest < x + y) + { + fxybest = x + y; + fxbest = x; + } + } + + /* Find backward diagonal that minimizes X + Y. */ + bxybest = LIN_MAX; + for (d = bmax; d >= bmin; d -= 2) + { + lin x = MAX (xoff, bd[d]); + lin y = x - d; + if (y < yoff) + x = yoff + d, y = yoff; + if (x + y < bxybest) + { + bxybest = x + y; + bxbest = x; + } + } + + /* Use the better of the two diagonals. */ + if ((xlim + ylim) - bxybest < fxybest - (xoff + yoff)) + { + part->xmid = fxbest; + part->ymid = fxybest - fxbest; + part->lo_minimal = true; + part->hi_minimal = false; + } + else + { + part->xmid = bxbest; + part->ymid = bxybest - bxbest; + part->lo_minimal = false; + part->hi_minimal = true; + } + return; + } + } +} + +/* Compare in detail contiguous subsequences of the two files + which are known, as a whole, to match each other. + + The results are recorded in the vectors files[N].changed, by + storing 1 in the element for each line that is an insertion or deletion. + + The subsequence of file 0 is [XOFF, XLIM) and likewise for file 1. + + Note that XLIM, YLIM are exclusive bounds. + All line numbers are origin-0 and discarded lines are not counted. + + If FIND_MINIMAL, find a minimal difference no matter how + expensive it is. */ + +static void +compareseq (lin xoff, lin xlim, lin yoff, lin ylim, bool find_minimal) +{ + lin const *xv = xvec; /* Help the compiler. */ + lin const *yv = yvec; + + /* Slide down the bottom initial diagonal. */ + while (xoff < xlim && yoff < ylim && xv[xoff] == yv[yoff]) + ++xoff, ++yoff; + /* Slide up the top initial diagonal. */ + while (xlim > xoff && ylim > yoff && xv[xlim - 1] == yv[ylim - 1]) + --xlim, --ylim; + + /* Handle simple cases. */ + if (xoff == xlim) + while (yoff < ylim) + files[1].changed[files[1].realindexes[yoff++]] = 1; + else if (yoff == ylim) + while (xoff < xlim) + files[0].changed[files[0].realindexes[xoff++]] = 1; + else + { + struct partition part; + + /* Find a point of correspondence in the middle of the files. */ + diag (xoff, xlim, yoff, ylim, find_minimal, &part); + + /* Use the partitions to split this problem into subproblems. */ + compareseq (xoff, part.xmid, yoff, part.ymid, part.lo_minimal); + compareseq (part.xmid, xlim, part.ymid, ylim, part.hi_minimal); + } +} + +/* Discard lines from one file that have no matches in the other file. + + A line which is discarded will not be considered by the actual + comparison algorithm; it will be as if that line were not in the file. + The file's `realindexes' table maps virtual line numbers + (which don't count the discarded lines) into real line numbers; + this is how the actual comparison algorithm produces results + that are comprehensible when the discarded lines are counted. + + When we discard a line, we also mark it as a deletion or insertion + so that it will be printed in the output. */ + +static void +discard_confusing_lines (struct file_data filevec[]) +{ + int f; + lin i; + char *discarded[2]; + lin *equiv_count[2]; + lin *p; + + /* Allocate our results. */ + p = xmalloc ((filevec[0].buffered_lines + filevec[1].buffered_lines) + * (2 * sizeof *p)); + for (f = 0; f < 2; f++) + { + filevec[f].undiscarded = p; p += filevec[f].buffered_lines; + filevec[f].realindexes = p; p += filevec[f].buffered_lines; + } + + /* Set up equiv_count[F][I] as the number of lines in file F + that fall in equivalence class I. */ + + p = zalloc (filevec[0].equiv_max * (2 * sizeof *p)); + equiv_count[0] = p; + equiv_count[1] = p + filevec[0].equiv_max; + + for (i = 0; i < filevec[0].buffered_lines; ++i) + ++equiv_count[0][filevec[0].equivs[i]]; + for (i = 0; i < filevec[1].buffered_lines; ++i) + ++equiv_count[1][filevec[1].equivs[i]]; + + /* Set up tables of which lines are going to be discarded. */ + + discarded[0] = zalloc (filevec[0].buffered_lines + + filevec[1].buffered_lines); + discarded[1] = discarded[0] + filevec[0].buffered_lines; + + /* Mark to be discarded each line that matches no line of the other file. + If a line matches many lines, mark it as provisionally discardable. */ + + for (f = 0; f < 2; f++) + { + size_t end = filevec[f].buffered_lines; + char *discards = discarded[f]; + lin *counts = equiv_count[1 - f]; + lin *equivs = filevec[f].equivs; + size_t many = 5; + size_t tem = end / 64; + + /* Multiply MANY by approximate square root of number of lines. + That is the threshold for provisionally discardable lines. */ + while ((tem = tem >> 2) > 0) + many *= 2; + + for (i = 0; i < end; i++) + { + lin nmatch; + if (equivs[i] == 0) + continue; + nmatch = counts[equivs[i]]; + if (nmatch == 0) + discards[i] = 1; + else if (nmatch > many) + discards[i] = 2; + } + } + + /* Don't really discard the provisional lines except when they occur + in a run of discardables, with nonprovisionals at the beginning + and end. */ + + for (f = 0; f < 2; f++) + { + lin end = filevec[f].buffered_lines; + register char *discards = discarded[f]; + + for (i = 0; i < end; i++) + { + /* Cancel provisional discards not in middle of run of discards. */ + if (discards[i] == 2) + discards[i] = 0; + else if (discards[i] != 0) + { + /* We have found a nonprovisional discard. */ + register lin j; + lin length; + lin provisional = 0; + + /* Find end of this run of discardable lines. + Count how many are provisionally discardable. */ + for (j = i; j < end; j++) + { + if (discards[j] == 0) + break; + if (discards[j] == 2) + ++provisional; + } + + /* Cancel provisional discards at end, and shrink the run. */ + while (j > i && discards[j - 1] == 2) + discards[--j] = 0, --provisional; + + /* Now we have the length of a run of discardable lines + whose first and last are not provisional. */ + length = j - i; + + /* If 1/4 of the lines in the run are provisional, + cancel discarding of all provisional lines in the run. */ + if (provisional * 4 > length) + { + while (j > i) + if (discards[--j] == 2) + discards[j] = 0; + } + else + { + register lin consec; + lin minimum = 1; + lin tem = length >> 2; + + /* MINIMUM is approximate square root of LENGTH/4. + A subrun of two or more provisionals can stand + when LENGTH is at least 16. + A subrun of 4 or more can stand when LENGTH >= 64. */ + while (0 < (tem >>= 2)) + minimum <<= 1; + minimum++; + + /* Cancel any subrun of MINIMUM or more provisionals + within the larger run. */ + for (j = 0, consec = 0; j < length; j++) + if (discards[i + j] != 2) + consec = 0; + else if (minimum == ++consec) + /* Back up to start of subrun, to cancel it all. */ + j -= consec; + else if (minimum < consec) + discards[i + j] = 0; + + /* Scan from beginning of run + until we find 3 or more nonprovisionals in a row + or until the first nonprovisional at least 8 lines in. + Until that point, cancel any provisionals. */ + for (j = 0, consec = 0; j < length; j++) + { + if (j >= 8 && discards[i + j] == 1) + break; + if (discards[i + j] == 2) + consec = 0, discards[i + j] = 0; + else if (discards[i + j] == 0) + consec = 0; + else + consec++; + if (consec == 3) + break; + } + + /* I advances to the last line of the run. */ + i += length - 1; + + /* Same thing, from end. */ + for (j = 0, consec = 0; j < length; j++) + { + if (j >= 8 && discards[i - j] == 1) + break; + if (discards[i - j] == 2) + consec = 0, discards[i - j] = 0; + else if (discards[i - j] == 0) + consec = 0; + else + consec++; + if (consec == 3) + break; + } + } + } + } + } + + /* Actually discard the lines. */ + for (f = 0; f < 2; f++) + { + char *discards = discarded[f]; + lin end = filevec[f].buffered_lines; + lin j = 0; + for (i = 0; i < end; ++i) + if (minimal || discards[i] == 0) + { + filevec[f].undiscarded[j] = filevec[f].equivs[i]; + filevec[f].realindexes[j++] = i; + } + else + filevec[f].changed[i] = 1; + filevec[f].nondiscarded_lines = j; + } + + free (discarded[0]); + free (equiv_count[0]); +} + +/* Adjust inserts/deletes of identical lines to join changes + as much as possible. + + We do something when a run of changed lines include a + line at one end and have an excluded, identical line at the other. + We are free to choose which identical line is included. + `compareseq' usually chooses the one at the beginning, + but usually it is cleaner to consider the following identical line + to be the "change". */ + +static void +shift_boundaries (struct file_data filevec[]) +{ + int f; + + for (f = 0; f < 2; f++) + { + char *changed = filevec[f].changed; + char *other_changed = filevec[1 - f].changed; + lin const *equivs = filevec[f].equivs; + lin i = 0; + lin j = 0; + lin i_end = filevec[f].buffered_lines; + + while (1) + { + lin runlength, start, corresponding; + + /* Scan forwards to find beginning of another run of changes. + Also keep track of the corresponding point in the other file. */ + + while (i < i_end && !changed[i]) + { + while (other_changed[j++]) + continue; + i++; + } + + if (i == i_end) + break; + + start = i; + + /* Find the end of this run of changes. */ + + while (changed[++i]) + continue; + while (other_changed[j]) + j++; + + do + { + /* Record the length of this run of changes, so that + we can later determine whether the run has grown. */ + runlength = i - start; + + /* Move the changed region back, so long as the + previous unchanged line matches the last changed one. + This merges with previous changed regions. */ + + while (start && equivs[start - 1] == equivs[i - 1]) + { + changed[--start] = 1; + changed[--i] = 0; + while (changed[start - 1]) + start--; + while (other_changed[--j]) + continue; + } + + /* Set CORRESPONDING to the end of the changed run, at the last + point where it corresponds to a changed run in the other file. + CORRESPONDING == I_END means no such point has been found. */ + corresponding = other_changed[j - 1] ? i : i_end; + + /* Move the changed region forward, so long as the + first changed line matches the following unchanged one. + This merges with following changed regions. + Do this second, so that if there are no merges, + the changed region is moved forward as far as possible. */ + + while (i != i_end && equivs[start] == equivs[i]) + { + changed[start++] = 0; + changed[i++] = 1; + while (changed[i]) + i++; + while (other_changed[++j]) + corresponding = i; + } + } + while (runlength != i - start); + + /* If possible, move the fully-merged run of changes + back to a corresponding run in the other file. */ + + while (corresponding < i) + { + changed[--start] = 1; + changed[--i] = 0; + while (other_changed[--j]) + continue; + } + } + } +} + +/* Cons an additional entry onto the front of an edit script OLD. + LINE0 and LINE1 are the first affected lines in the two files (origin 0). + DELETED is the number of lines deleted here from file 0. + INSERTED is the number of lines inserted here in file 1. + + If DELETED is 0 then LINE0 is the number of the line before + which the insertion was done; vice versa for INSERTED and LINE1. */ + +static struct change * +add_change (lin line0, lin line1, lin deleted, lin inserted, + struct change *old) +{ + struct change *new = xmalloc (sizeof *new); + + new->line0 = line0; + new->line1 = line1; + new->inserted = inserted; + new->deleted = deleted; + new->link = old; + return new; +} + +/* Scan the tables of which lines are inserted and deleted, + producing an edit script in reverse order. */ + +static struct change * +build_reverse_script (struct file_data const filevec[]) +{ + struct change *script = 0; + char *changed0 = filevec[0].changed; + char *changed1 = filevec[1].changed; + lin len0 = filevec[0].buffered_lines; + lin len1 = filevec[1].buffered_lines; + + /* Note that changedN[len0] does exist, and is 0. */ + + lin i0 = 0, i1 = 0; + + while (i0 < len0 || i1 < len1) + { + if (changed0[i0] | changed1[i1]) + { + lin line0 = i0, line1 = i1; + + /* Find # lines changed here in each file. */ + while (changed0[i0]) ++i0; + while (changed1[i1]) ++i1; + + /* Record this change. */ + script = add_change (line0, line1, i0 - line0, i1 - line1, script); + } + + /* We have reached lines in the two files that match each other. */ + i0++, i1++; + } + + return script; +} + +/* Scan the tables of which lines are inserted and deleted, + producing an edit script in forward order. */ + +static struct change * +build_script (struct file_data const filevec[]) +{ + struct change *script = 0; + char *changed0 = filevec[0].changed; + char *changed1 = filevec[1].changed; + lin i0 = filevec[0].buffered_lines, i1 = filevec[1].buffered_lines; + + /* Note that changedN[-1] does exist, and is 0. */ + + while (i0 >= 0 || i1 >= 0) + { + if (changed0[i0 - 1] | changed1[i1 - 1]) + { + lin line0 = i0, line1 = i1; + + /* Find # lines changed here in each file. */ + while (changed0[i0 - 1]) --i0; + while (changed1[i1 - 1]) --i1; + + /* Record this change. */ + script = add_change (i0, i1, line0 - i0, line1 - i1, script); + } + + /* We have reached lines in the two files that match each other. */ + i0--, i1--; + } + + return script; +} + +/* If CHANGES, briefly report that two files differed. + Return 2 if trouble, CHANGES otherwise. */ +static int +briefly_report (int changes, struct file_data const filevec[]) +{ + if (changes) + { + char const *label0 = file_label[0] ? file_label[0] : filevec[0].name; + char const *label1 = file_label[1] ? file_label[1] : filevec[1].name; + message ("Files %s and %s differ\n", label0, label1); + if (! brief) + changes = 2; + } + + return changes; +} + +/* Report the differences of two files. */ +int +diff_2_files (struct comparison *cmp) +{ + lin diags; + int f; + struct change *e, *p; + struct change *script; + int changes; + + + /* If we have detected that either file is binary, + compare the two files as binary. This can happen + only when the first chunk is read. + Also, --brief without any --ignore-* options means + we can speed things up by treating the files as binary. */ + + if (read_files (cmp->file, files_can_be_treated_as_binary)) + { + /* Files with different lengths must be different. */ + if (cmp->file[0].stat.st_size != cmp->file[1].stat.st_size + && (cmp->file[0].desc < 0 || S_ISREG (cmp->file[0].stat.st_mode)) + && (cmp->file[1].desc < 0 || S_ISREG (cmp->file[1].stat.st_mode))) + changes = 1; + + /* Standard input equals itself. */ + else if (cmp->file[0].desc == cmp->file[1].desc) + changes = 0; + + else + /* Scan both files, a buffer at a time, looking for a difference. */ + { + /* Allocate same-sized buffers for both files. */ + size_t lcm_max = PTRDIFF_MAX - 1; + size_t buffer_size = + buffer_lcm (sizeof (word), + buffer_lcm (STAT_BLOCKSIZE (cmp->file[0].stat), + STAT_BLOCKSIZE (cmp->file[1].stat), + lcm_max), + lcm_max); + for (f = 0; f < 2; f++) + cmp->file[f].buffer = xrealloc (cmp->file[f].buffer, buffer_size); + + for (;; cmp->file[0].buffered = cmp->file[1].buffered = 0) + { + /* Read a buffer's worth from both files. */ + for (f = 0; f < 2; f++) + if (0 <= cmp->file[f].desc) + file_block_read (&cmp->file[f], + buffer_size - cmp->file[f].buffered); + + /* If the buffers differ, the files differ. */ + if (cmp->file[0].buffered != cmp->file[1].buffered + || memcmp (cmp->file[0].buffer, + cmp->file[1].buffer, + cmp->file[0].buffered)) + { + changes = 1; + break; + } + + /* If we reach end of file, the files are the same. */ + if (cmp->file[0].buffered != buffer_size) + { + changes = 0; + break; + } + } + } + + changes = briefly_report (changes, cmp->file); + } + else + { + /* Allocate vectors for the results of comparison: + a flag for each line of each file, saying whether that line + is an insertion or deletion. + Allocate an extra element, always 0, at each end of each vector. */ + + size_t s = cmp->file[0].buffered_lines + cmp->file[1].buffered_lines + 4; + char *flag_space = zalloc (s); + cmp->file[0].changed = flag_space + 1; + cmp->file[1].changed = flag_space + cmp->file[0].buffered_lines + 3; + + /* Some lines are obviously insertions or deletions + because they don't match anything. Detect them now, and + avoid even thinking about them in the main comparison algorithm. */ + + discard_confusing_lines (cmp->file); + + /* Now do the main comparison algorithm, considering just the + undiscarded lines. */ + + xvec = cmp->file[0].undiscarded; + yvec = cmp->file[1].undiscarded; + diags = (cmp->file[0].nondiscarded_lines + + cmp->file[1].nondiscarded_lines + 3); + fdiag = xmalloc (diags * (2 * sizeof *fdiag)); + bdiag = fdiag + diags; + fdiag += cmp->file[1].nondiscarded_lines + 1; + bdiag += cmp->file[1].nondiscarded_lines + 1; + + /* Set TOO_EXPENSIVE to be approximate square root of input size, + bounded below by 256. */ + too_expensive = 1; + for (; diags != 0; diags >>= 2) + too_expensive <<= 1; + too_expensive = MAX (256, too_expensive); + + files[0] = cmp->file[0]; + files[1] = cmp->file[1]; + + compareseq (0, cmp->file[0].nondiscarded_lines, + 0, cmp->file[1].nondiscarded_lines, minimal); + + free (fdiag - (cmp->file[1].nondiscarded_lines + 1)); + + /* Modify the results slightly to make them prettier + in cases where that can validly be done. */ + + shift_boundaries (cmp->file); + + /* Get the results of comparison in the form of a chain + of `struct change's -- an edit script. */ + + if (output_style == OUTPUT_ED) + script = build_reverse_script (cmp->file); + else + script = build_script (cmp->file); + + /* Set CHANGES if we had any diffs. + If some changes are ignored, we must scan the script to decide. */ + if (ignore_blank_lines || ignore_regexp.fastmap) + { + struct change *next = script; + changes = 0; + + while (next && changes == 0) + { + struct change *this, *end; + lin first0, last0, first1, last1; + + /* Find a set of changes that belong together. */ + this = next; + end = find_change (next); + + /* Disconnect them from the rest of the changes, making them + a hunk, and remember the rest for next iteration. */ + next = end->link; + end->link = 0; + + /* Determine whether this hunk is really a difference. */ + if (analyze_hunk (this, &first0, &last0, &first1, &last1)) + changes = 1; + + /* Reconnect the script so it will all be freed properly. */ + end->link = next; + } + } + else + changes = (script != 0); + + if (brief) + changes = briefly_report (changes, cmp->file); + else + { + if (changes | !no_diff_means_no_output) + { + /* Record info for starting up output, + to be used if and when we have some output to print. */ + setup_output (file_label[0] ? file_label[0] : cmp->file[0].name, + file_label[1] ? file_label[1] : cmp->file[1].name, + cmp->parent != 0); + + switch (output_style) + { + case OUTPUT_CONTEXT: + print_context_script (script, false); + break; + + case OUTPUT_UNIFIED: + print_context_script (script, true); + break; + + case OUTPUT_ED: + print_ed_script (script); + break; + + case OUTPUT_FORWARD_ED: + pr_forward_ed_script (script); + break; + + case OUTPUT_RCS: + print_rcs_script (script); + break; + + case OUTPUT_NORMAL: + print_normal_script (script); + break; + + case OUTPUT_IFDEF: + print_ifdef_script (script); + break; + + case OUTPUT_SDIFF: + print_sdiff_script (script); + break; + + default: + abort (); + } + + finish_output (); + } + } + + free (cmp->file[0].undiscarded); + + free (flag_space); + + for (f = 0; f < 2; f++) + { + free (cmp->file[f].equivs); + free (cmp->file[f].linbuf + cmp->file[f].linbuf_base); + } + + for (e = script; e; e = p) + { + p = e->link; + free (e); + } + + if (! ROBUST_OUTPUT_STYLE (output_style)) + for (f = 0; f < 2; ++f) + if (cmp->file[f].missing_newline) + { + error (0, 0, "%s: %s\n", + file_label[f] ? file_label[f] : cmp->file[f].name, + _("No newline at end of file")); + changes = 2; + } + } + + if (cmp->file[0].buffer != cmp->file[1].buffer) + free (cmp->file[0].buffer); + free (cmp->file[1].buffer); + + return changes; +} diff --git a/contrib/diff/src/cmp.c b/contrib/diff/src/cmp.c new file mode 100644 index 0000000..1f965e5 --- /dev/null +++ b/contrib/diff/src/cmp.c @@ -0,0 +1,677 @@ +/* cmp - compare two files byte by byte + + Copyright (C) 1990, 1991, 1992, 1993, 1994, 1995, 1996, 1998, 2001, + 2002, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + See the GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "system.h" +#include "paths.h" + +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#if defined LC_MESSAGES && ENABLE_NLS +# define hard_locale_LC_MESSAGES hard_locale (LC_MESSAGES) +#else +# define hard_locale_LC_MESSAGES 0 +#endif + +static int cmp (void); +static off_t file_position (int); +static size_t block_compare (word const *, word const *); +static size_t block_compare_and_count (word const *, word const *, off_t *); +static void sprintc (char *, unsigned char); + +/* Name under which this program was invoked. */ +char *program_name; + +/* Filenames of the compared files. */ +static char const *file[2]; + +/* File descriptors of the files. */ +static int file_desc[2]; + +/* Status of the files. */ +static struct stat stat_buf[2]; + +/* Read buffers for the files. */ +static word *buffer[2]; + +/* Optimal block size for the files. */ +static size_t buf_size; + +/* Initial prefix to ignore for each file. */ +static off_t ignore_initial[2]; + +/* Number of bytes to compare. */ +static uintmax_t bytes = UINTMAX_MAX; + +/* Output format. */ +static enum comparison_type + { + type_first_diff, /* Print the first difference. */ + type_all_diffs, /* Print all differences. */ + type_status /* Exit status only. */ + } comparison_type; + +/* If nonzero, print values of bytes quoted like cat -t does. */ +static bool opt_print_bytes; + +/* Values for long options that do not have single-letter equivalents. */ +enum +{ + HELP_OPTION = CHAR_MAX + 1 +}; + +static struct option const long_options[] = +{ + {"print-bytes", 0, 0, 'b'}, + {"print-chars", 0, 0, 'c'}, /* obsolescent as of diffutils 2.7.3 */ + {"ignore-initial", 1, 0, 'i'}, + {"verbose", 0, 0, 'l'}, + {"bytes", 1, 0, 'n'}, + {"silent", 0, 0, 's'}, + {"quiet", 0, 0, 's'}, + {"version", 0, 0, 'v'}, + {"help", 0, 0, HELP_OPTION}, + {0, 0, 0, 0} +}; + +static void try_help (char const *, char const *) __attribute__((noreturn)); +static void +try_help (char const *reason_msgid, char const *operand) +{ + if (reason_msgid) + error (0, 0, _(reason_msgid), operand); + error (EXIT_TROUBLE, 0, + _("Try `%s --help' for more information."), program_name); + abort (); +} + +static char const valid_suffixes[] = "kKMGTPEZY0"; + +/* Update ignore_initial[F] according to the result of parsing an + *operand ARGPTR of --ignore-initial, updating *ARGPTR to point + *after the operand. If DELIMITER is nonzero, the operand may be + *followed by DELIMITER; otherwise it must be null-terminated. */ +static void +specify_ignore_initial (int f, char **argptr, char delimiter) +{ + uintmax_t val; + off_t o; + char const *arg = *argptr; + strtol_error e = xstrtoumax (arg, argptr, 0, &val, valid_suffixes); + if (! (e == LONGINT_OK + || (e == LONGINT_INVALID_SUFFIX_CHAR && **argptr == delimiter)) + || (o = val) < 0 || o != val || val == UINTMAX_MAX) + try_help ("invalid --ignore-initial value `%s'", arg); + if (ignore_initial[f] < o) + ignore_initial[f] = o; +} + +/* Specify the output format. */ +static void +specify_comparison_type (enum comparison_type t) +{ + if (comparison_type && comparison_type != t) + try_help ("options -l and -s are incompatible", 0); + comparison_type = t; +} + +static void +check_stdout (void) +{ + if (ferror (stdout)) + error (EXIT_TROUBLE, 0, "%s", _("write failed")); + else if (fclose (stdout) != 0) + error (EXIT_TROUBLE, errno, "%s", _("standard output")); +} + +static char const * const option_help_msgid[] = { + N_("-b --print-bytes Print differing bytes."), + N_("-i SKIP --ignore-initial=SKIP Skip the first SKIP bytes of input."), + N_("-i SKIP1:SKIP2 --ignore-initial=SKIP1:SKIP2"), + N_(" Skip the first SKIP1 bytes of FILE1 and the first SKIP2 bytes of FILE2."), + N_("-l --verbose Output byte numbers and values of all differing bytes."), + N_("-n LIMIT --bytes=LIMIT Compare at most LIMIT bytes."), + N_("-s --quiet --silent Output nothing; yield exit status only."), + N_("-v --version Output version info."), + N_("--help Output this help."), + 0 +}; + +static void +usage (void) +{ + char const * const *p; + + printf (_("Usage: %s [OPTION]... FILE1 [FILE2 [SKIP1 [SKIP2]]]\n"), + program_name); + printf ("%s\n\n", _("Compare two files byte by byte.")); + for (p = option_help_msgid; *p; p++) + printf (" %s\n", _(*p)); + printf ("\n%s\n%s\n\n%s\n%s\n\n%s\n", + _("SKIP1 and SKIP2 are the number of bytes to skip in each file."), + _("SKIP values may be followed by the following multiplicative suffixes:\n\ +kB 1000, K 1024, MB 1,000,000, M 1,048,576,\n\ +GB 1,000,000,000, G 1,073,741,824, and so on for T, P, E, Z, Y."), + _("If a FILE is `-' or missing, read standard input."), + _("Exit status is 0 if inputs are the same, 1 if different, 2 if trouble."), + _("Report bugs to .")); +} + +int +main (int argc, char **argv) +{ + int c, f, exit_status; + size_t words_per_buffer; + + exit_failure = EXIT_TROUBLE; + initialize_main (&argc, &argv); + program_name = argv[0]; + setlocale (LC_ALL, ""); + bindtextdomain (PACKAGE, LOCALEDIR); + textdomain (PACKAGE); + c_stack_action (0); + + /* Parse command line options. */ + + while ((c = getopt_long (argc, argv, "bci:ln:sv", long_options, 0)) + != -1) + switch (c) + { + case 'b': + case 'c': /* 'c' is obsolescent as of diffutils 2.7.3 */ + opt_print_bytes = true; + break; + + case 'i': + specify_ignore_initial (0, &optarg, ':'); + if (*optarg++ == ':') + specify_ignore_initial (1, &optarg, 0); + else if (ignore_initial[1] < ignore_initial[0]) + ignore_initial[1] = ignore_initial[0]; + break; + + case 'l': + specify_comparison_type (type_all_diffs); + break; + + case 'n': + { + uintmax_t n; + if (xstrtoumax (optarg, 0, 0, &n, valid_suffixes) != LONGINT_OK) + try_help ("invalid --bytes value `%s'", optarg); + if (n < bytes) + bytes = n; + } + break; + + case 's': + specify_comparison_type (type_status); + break; + + case 'v': + /* TRANSLATORS: Please translate the second "o" in "Torbjorn + Granlund" to an o-with-umlaut (U+00F6, LATIN SMALL LETTER O + WITH DIAERESIS) if possible. */ + version_etc (stdout, "cmp", PACKAGE_NAME, PACKAGE_VERSION, + _("Torbjorn Granlund"), "David MacKenzie", (char *) 0); + check_stdout (); + return EXIT_SUCCESS; + + case HELP_OPTION: + usage (); + check_stdout (); + return EXIT_SUCCESS; + + default: + try_help (0, 0); + } + + if (optind == argc) + try_help ("missing operand after `%s'", argv[argc - 1]); + + file[0] = argv[optind++]; + file[1] = optind < argc ? argv[optind++] : "-"; + + for (f = 0; f < 2 && optind < argc; f++) + { + char *arg = argv[optind++]; + specify_ignore_initial (f, &arg, 0); + } + + if (optind < argc) + try_help ("extra operand `%s'", argv[optind]); + + for (f = 0; f < 2; f++) + { + /* If file[1] is "-", treat it first; this avoids a misdiagnostic if + stdin is closed and opening file[0] yields file descriptor 0. */ + int f1 = f ^ (strcmp (file[1], "-") == 0); + + /* Two files with the same name and offset are identical. + But wait until we open the file once, for proper diagnostics. */ + if (f && ignore_initial[0] == ignore_initial[1] + && file_name_cmp (file[0], file[1]) == 0) + return EXIT_SUCCESS; + + file_desc[f1] = (strcmp (file[f1], "-") == 0 + ? STDIN_FILENO + : open (file[f1], O_RDONLY, 0)); + if (file_desc[f1] < 0 || fstat (file_desc[f1], stat_buf + f1) != 0) + { + if (file_desc[f1] < 0 && comparison_type == type_status) + exit (EXIT_TROUBLE); + else + error (EXIT_TROUBLE, errno, "%s", file[f1]); + } + + set_binary_mode (file_desc[f1], true); + } + + /* If the files are links to the same inode and have the same file position, + they are identical. */ + + if (0 < same_file (&stat_buf[0], &stat_buf[1]) + && same_file_attributes (&stat_buf[0], &stat_buf[1]) + && file_position (0) == file_position (1)) + return EXIT_SUCCESS; + + /* If output is redirected to the null device, we may assume `-s'. */ + + if (comparison_type != type_status) + { + struct stat outstat, nullstat; + + if (fstat (STDOUT_FILENO, &outstat) == 0 + && stat (NULL_DEVICE, &nullstat) == 0 + && 0 < same_file (&outstat, &nullstat)) + comparison_type = type_status; + } + + /* If only a return code is needed, + and if both input descriptors are associated with plain files, + conclude that the files differ if they have different sizes + and if more bytes will be compared than are in the smaller file. */ + + if (comparison_type == type_status + && S_ISREG (stat_buf[0].st_mode) + && S_ISREG (stat_buf[1].st_mode)) + { + off_t s0 = stat_buf[0].st_size - file_position (0); + off_t s1 = stat_buf[1].st_size - file_position (1); + if (s0 < 0) + s0 = 0; + if (s1 < 0) + s1 = 0; + if (s0 != s1 && MIN (s0, s1) < bytes) + exit (EXIT_FAILURE); + } + + /* Get the optimal block size of the files. */ + + buf_size = buffer_lcm (STAT_BLOCKSIZE (stat_buf[0]), + STAT_BLOCKSIZE (stat_buf[1]), + PTRDIFF_MAX - sizeof (word)); + + /* Allocate word-aligned buffers, with space for sentinels at the end. */ + + words_per_buffer = (buf_size + 2 * sizeof (word) - 1) / sizeof (word); + buffer[0] = xmalloc (2 * sizeof (word) * words_per_buffer); + buffer[1] = buffer[0] + words_per_buffer; + + exit_status = cmp (); + + for (f = 0; f < 2; f++) + if (close (file_desc[f]) != 0) + error (EXIT_TROUBLE, errno, "%s", file[f]); + if (exit_status != 0 && comparison_type != type_status) + check_stdout (); + exit (exit_status); + return exit_status; +} + +/* Compare the two files already open on `file_desc[0]' and `file_desc[1]', + using `buffer[0]' and `buffer[1]'. + Return EXIT_SUCCESS if identical, EXIT_FAILURE if different, + >1 if error. */ + +static int +cmp (void) +{ + off_t line_number = 1; /* Line number (1...) of difference. */ + off_t byte_number = 1; /* Byte number (1...) of difference. */ + uintmax_t remaining = bytes; /* Remaining number of bytes to compare. */ + size_t read0, read1; /* Number of bytes read from each file. */ + size_t first_diff; /* Offset (0...) in buffers of 1st diff. */ + size_t smaller; /* The lesser of `read0' and `read1'. */ + word *buffer0 = buffer[0]; + word *buffer1 = buffer[1]; + char *buf0 = (char *) buffer0; + char *buf1 = (char *) buffer1; + int ret = EXIT_SUCCESS; + int f; + int offset_width; + + if (comparison_type == type_all_diffs) + { + off_t byte_number_max = MIN (bytes, TYPE_MAXIMUM (off_t)); + + for (f = 0; f < 2; f++) + if (S_ISREG (stat_buf[f].st_mode)) + { + off_t file_bytes = stat_buf[f].st_size - file_position (f); + if (file_bytes < byte_number_max) + byte_number_max = file_bytes; + } + + for (offset_width = 1; (byte_number_max /= 10) != 0; offset_width++) + continue; + } + + for (f = 0; f < 2; f++) + { + off_t ig = ignore_initial[f]; + if (ig && file_position (f) == -1) + { + /* lseek failed; read and discard the ignored initial prefix. */ + do + { + size_t bytes_to_read = MIN (ig, buf_size); + size_t r = block_read (file_desc[f], buf0, bytes_to_read); + if (r != bytes_to_read) + { + if (r == SIZE_MAX) + error (EXIT_TROUBLE, errno, "%s", file[f]); + break; + } + ig -= r; + } + while (ig); + } + } + + do + { + size_t bytes_to_read = buf_size; + + if (remaining != UINTMAX_MAX) + { + if (remaining < bytes_to_read) + bytes_to_read = remaining; + remaining -= bytes_to_read; + } + + read0 = block_read (file_desc[0], buf0, bytes_to_read); + if (read0 == SIZE_MAX) + error (EXIT_TROUBLE, errno, "%s", file[0]); + read1 = block_read (file_desc[1], buf1, bytes_to_read); + if (read1 == SIZE_MAX) + error (EXIT_TROUBLE, errno, "%s", file[1]); + + /* Insert sentinels for the block compare. */ + + buf0[read0] = ~buf1[read0]; + buf1[read1] = ~buf0[read1]; + + /* If the line number should be written for differing files, + compare the blocks and count the number of newlines + simultaneously. */ + first_diff = (comparison_type == type_first_diff + ? block_compare_and_count (buffer0, buffer1, &line_number) + : block_compare (buffer0, buffer1)); + + byte_number += first_diff; + smaller = MIN (read0, read1); + + if (first_diff < smaller) + { + switch (comparison_type) + { + case type_first_diff: + { + char byte_buf[INT_BUFSIZE_BOUND (off_t)]; + char line_buf[INT_BUFSIZE_BOUND (off_t)]; + char const *byte_num = offtostr (byte_number, byte_buf); + char const *line_num = offtostr (line_number, line_buf); + if (!opt_print_bytes) + { + /* See POSIX 1003.1-2001 for this format. This + message is used only in the POSIX locale, so it + need not be translated. */ + static char const char_message[] = + "%s %s differ: char %s, line %s\n"; + + /* The POSIX rationale recommends using the word + "byte" outside the POSIX locale. Some gettext + implementations translate even in the POSIX + locale if certain other environment variables + are set, so use "byte" if a translation is + available, or if outside the POSIX locale. */ + static char const byte_msgid[] = + N_("%s %s differ: byte %s, line %s\n"); + char const *byte_message = _(byte_msgid); + bool use_byte_message = (byte_message != byte_msgid + || hard_locale_LC_MESSAGES); + + printf (use_byte_message ? byte_message : char_message, + file[0], file[1], byte_num, line_num); + } + else + { + unsigned char c0 = buf0[first_diff]; + unsigned char c1 = buf1[first_diff]; + char s0[5]; + char s1[5]; + sprintc (s0, c0); + sprintc (s1, c1); + printf (_("%s %s differ: byte %s, line %s is %3o %s %3o %s\n"), + file[0], file[1], byte_num, line_num, + c0, s0, c1, s1); + } + } + /* Fall through. */ + case type_status: + return EXIT_FAILURE; + + case type_all_diffs: + do + { + unsigned char c0 = buf0[first_diff]; + unsigned char c1 = buf1[first_diff]; + if (c0 != c1) + { + char byte_buf[INT_BUFSIZE_BOUND (off_t)]; + char const *byte_num = offtostr (byte_number, byte_buf); + if (!opt_print_bytes) + { + /* See POSIX 1003.1-2001 for this format. */ + printf ("%*s %3o %3o\n", + offset_width, byte_num, c0, c1); + } + else + { + char s0[5]; + char s1[5]; + sprintc (s0, c0); + sprintc (s1, c1); + printf ("%*s %3o %-4s %3o %s\n", + offset_width, byte_num, c0, s0, c1, s1); + } + } + byte_number++; + first_diff++; + } + while (first_diff < smaller); + ret = EXIT_FAILURE; + break; + } + } + + if (read0 != read1) + { + if (comparison_type != type_status) + { + /* See POSIX 1003.1-2001 for this format. */ + fprintf (stderr, _("cmp: EOF on %s\n"), file[read1 < read0]); + } + + return EXIT_FAILURE; + } + } + while (read0 == buf_size); + + return ret; +} + +/* Compare two blocks of memory P0 and P1 until they differ, + and count the number of '\n' occurrences in the common + part of P0 and P1. + If the blocks are not guaranteed to be different, put sentinels at the ends + of the blocks before calling this function. + + Return the offset of the first byte that differs. + Increment *COUNT by the count of '\n' occurrences. */ + +static size_t +block_compare_and_count (word const *p0, word const *p1, off_t *count) +{ + word l; /* One word from first buffer. */ + word const *l0, *l1; /* Pointers into each buffer. */ + char const *c0, *c1; /* Pointers for finding exact address. */ + size_t cnt = 0; /* Number of '\n' occurrences. */ + word nnnn; /* Newline, sizeof (word) times. */ + int i; + + nnnn = 0; + for (i = 0; i < sizeof nnnn; i++) + nnnn = (nnnn << CHAR_BIT) | '\n'; + + /* Find the rough position of the first difference by reading words, + not bytes. */ + + for (l0 = p0, l1 = p1; (l = *l0) == *l1; l0++, l1++) + { + l ^= nnnn; + for (i = 0; i < sizeof l; i++) + { + unsigned char uc = l; + cnt += ! uc; + l >>= CHAR_BIT; + } + } + + /* Find the exact differing position (endianness independent). */ + + for (c0 = (char const *) l0, c1 = (char const *) l1; + *c0 == *c1; + c0++, c1++) + cnt += *c0 == '\n'; + + *count += cnt; + return c0 - (char const *) p0; +} + +/* Compare two blocks of memory P0 and P1 until they differ. + If the blocks are not guaranteed to be different, put sentinels at the ends + of the blocks before calling this function. + + Return the offset of the first byte that differs. */ + +static size_t +block_compare (word const *p0, word const *p1) +{ + word const *l0, *l1; + char const *c0, *c1; + + /* Find the rough position of the first difference by reading words, + not bytes. */ + + for (l0 = p0, l1 = p1; *l0 == *l1; l0++, l1++) + continue; + + /* Find the exact differing position (endianness independent). */ + + for (c0 = (char const *) l0, c1 = (char const *) l1; + *c0 == *c1; + c0++, c1++) + continue; + + return c0 - (char const *) p0; +} + +/* Put into BUF the unsigned char C, making unprintable bytes + visible by quoting like cat -t does. */ + +static void +sprintc (char *buf, unsigned char c) +{ + if (! isprint (c)) + { + if (c >= 128) + { + *buf++ = 'M'; + *buf++ = '-'; + c -= 128; + } + if (c < 32) + { + *buf++ = '^'; + c += 64; + } + else if (c == 127) + { + *buf++ = '^'; + c = '?'; + } + } + + *buf++ = c; + *buf = 0; +} + +/* Position file F to ignore_initial[F] bytes from its initial position, + and yield its new position. Don't try more than once. */ + +static off_t +file_position (int f) +{ + static bool positioned[2]; + static off_t position[2]; + + if (! positioned[f]) + { + positioned[f] = true; + position[f] = lseek (file_desc[f], ignore_initial[f], SEEK_CUR); + } + return position[f]; +} diff --git a/contrib/diff/src/context.c b/contrib/diff/src/context.c new file mode 100644 index 0000000..2055ff1 --- /dev/null +++ b/contrib/diff/src/context.c @@ -0,0 +1,478 @@ +/* Context-format output routines for GNU DIFF. + + Copyright (C) 1988, 1989, 1991, 1992, 1993, 1994, 1995, 1998, 2001, + 2002, 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "diff.h" +#include + +#ifdef ST_MTIM_NSEC +# define TIMESPEC_NS(timespec) ((timespec).ST_MTIM_NSEC) +#else +# define TIMESPEC_NS(timespec) 0 +#endif + +size_t nstrftime (char *, size_t, char const *, struct tm const *, int, int); + +static char const *find_function (char const * const *, lin); +static struct change *find_hunk (struct change *); +static void mark_ignorable (struct change *); +static void pr_context_hunk (struct change *); +static void pr_unidiff_hunk (struct change *); + +/* Last place find_function started searching from. */ +static lin find_function_last_search; + +/* The value find_function returned when it started searching there. */ +static lin find_function_last_match; + +/* Print a label for a context diff, with a file name and date or a label. */ + +static void +print_context_label (char const *mark, + struct file_data *inf, + char const *label) +{ + if (label) + fprintf (outfile, "%s %s\n", mark, label); + else + { + char buf[MAX (INT_STRLEN_BOUND (int) + 32, + INT_STRLEN_BOUND (time_t) + 11)]; + struct tm const *tm = localtime (&inf->stat.st_mtime); + int nsec = TIMESPEC_NS (inf->stat.st_mtim); + if (! (tm && nstrftime (buf, sizeof buf, time_format, tm, 0, nsec))) + { + long int sec = inf->stat.st_mtime; + verify (info_preserved, sizeof inf->stat.st_mtime <= sizeof sec); + sprintf (buf, "%ld.%.9d", sec, nsec); + } + fprintf (outfile, "%s %s\t%s\n", mark, inf->name, buf); + } +} + +/* Print a header for a context diff, with the file names and dates. */ + +void +print_context_header (struct file_data inf[], bool unidiff) +{ + if (unidiff) + { + print_context_label ("---", &inf[0], file_label[0]); + print_context_label ("+++", &inf[1], file_label[1]); + } + else + { + print_context_label ("***", &inf[0], file_label[0]); + print_context_label ("---", &inf[1], file_label[1]); + } +} + +/* Print an edit script in context format. */ + +void +print_context_script (struct change *script, bool unidiff) +{ + if (ignore_blank_lines || ignore_regexp.fastmap) + mark_ignorable (script); + else + { + struct change *e; + for (e = script; e; e = e->link) + e->ignore = false; + } + + find_function_last_search = - files[0].prefix_lines; + find_function_last_match = LIN_MAX; + + if (unidiff) + print_script (script, find_hunk, pr_unidiff_hunk); + else + print_script (script, find_hunk, pr_context_hunk); +} + +/* Print a pair of line numbers with a comma, translated for file FILE. + If the second number is not greater, use the first in place of it. + + Args A and B are internal line numbers. + We print the translated (real) line numbers. */ + +static void +print_context_number_range (struct file_data const *file, lin a, lin b) +{ + long int trans_a, trans_b; + translate_range (file, a, b, &trans_a, &trans_b); + + /* We can have B <= A in the case of a range of no lines. + In this case, we should print the line number before the range, + which is B. + + POSIX 1003.1-2001 requires two line numbers separated by a comma + even if the line numbers are the same. However, this does not + match existing practice and is surely an error in the + specification. */ + + if (trans_b <= trans_a) + fprintf (outfile, "%ld", trans_b); + else + fprintf (outfile, "%ld,%ld", trans_a, trans_b); +} + +/* Print FUNCTION in a context header. */ +static void +print_context_function (FILE *out, char const *function) +{ + int i; + putc (' ', out); + for (i = 0; i < 40 && function[i] != '\n'; i++) + continue; + fwrite (function, sizeof (char), i, out); +} + +/* Print a portion of an edit script in context format. + HUNK is the beginning of the portion to be printed. + The end is marked by a `link' that has been nulled out. + + Prints out lines from both files, and precedes each + line with the appropriate flag-character. */ + +static void +pr_context_hunk (struct change *hunk) +{ + lin first0, last0, first1, last1, i; + char const *prefix; + char const *function; + FILE *out; + + /* Determine range of line numbers involved in each file. */ + + enum changes changes = analyze_hunk (hunk, &first0, &last0, &first1, &last1); + if (! changes) + return; + + /* Include a context's width before and after. */ + + i = - files[0].prefix_lines; + first0 = MAX (first0 - context, i); + first1 = MAX (first1 - context, i); + if (last0 < files[0].valid_lines - context) + last0 += context; + else + last0 = files[0].valid_lines - 1; + if (last1 < files[1].valid_lines - context) + last1 += context; + else + last1 = files[1].valid_lines - 1; + + /* If desired, find the preceding function definition line in file 0. */ + function = 0; + if (function_regexp.fastmap) + function = find_function (files[0].linbuf, first0); + + begin_output (); + out = outfile; + + fprintf (out, "***************"); + + if (function) + print_context_function (out, function); + + fprintf (out, "\n*** "); + print_context_number_range (&files[0], first0, last0); + fprintf (out, " ****\n"); + + if (changes & OLD) + { + struct change *next = hunk; + + for (i = first0; i <= last0; i++) + { + /* Skip past changes that apply (in file 0) + only to lines before line I. */ + + while (next && next->line0 + next->deleted <= i) + next = next->link; + + /* Compute the marking for line I. */ + + prefix = " "; + if (next && next->line0 <= i) + /* The change NEXT covers this line. + If lines were inserted here in file 1, this is "changed". + Otherwise it is "deleted". */ + prefix = (next->inserted > 0 ? "!" : "-"); + + print_1_line (prefix, &files[0].linbuf[i]); + } + } + + fprintf (out, "--- "); + print_context_number_range (&files[1], first1, last1); + fprintf (out, " ----\n"); + + if (changes & NEW) + { + struct change *next = hunk; + + for (i = first1; i <= last1; i++) + { + /* Skip past changes that apply (in file 1) + only to lines before line I. */ + + while (next && next->line1 + next->inserted <= i) + next = next->link; + + /* Compute the marking for line I. */ + + prefix = " "; + if (next && next->line1 <= i) + /* The change NEXT covers this line. + If lines were deleted here in file 0, this is "changed". + Otherwise it is "inserted". */ + prefix = (next->deleted > 0 ? "!" : "+"); + + print_1_line (prefix, &files[1].linbuf[i]); + } + } +} + +/* Print a pair of line numbers with a comma, translated for file FILE. + If the second number is smaller, use the first in place of it. + If the numbers are equal, print just one number. + + Args A and B are internal line numbers. + We print the translated (real) line numbers. */ + +static void +print_unidiff_number_range (struct file_data const *file, lin a, lin b) +{ + long int trans_a, trans_b; + translate_range (file, a, b, &trans_a, &trans_b); + + /* We can have B < A in the case of a range of no lines. + In this case, we print the line number before the range, + which is B. It would be more logical to print A, but + 'patch' expects B in order to detect diffs against empty files. */ + if (trans_b <= trans_a) + fprintf (outfile, trans_b < trans_a ? "%ld,0" : "%ld", trans_b); + else + fprintf (outfile, "%ld,%ld", trans_a, trans_b - trans_a + 1); +} + +/* Print a portion of an edit script in unidiff format. + HUNK is the beginning of the portion to be printed. + The end is marked by a `link' that has been nulled out. + + Prints out lines from both files, and precedes each + line with the appropriate flag-character. */ + +static void +pr_unidiff_hunk (struct change *hunk) +{ + lin first0, last0, first1, last1; + lin i, j, k; + struct change *next; + char const *function; + FILE *out; + + /* Determine range of line numbers involved in each file. */ + + if (! analyze_hunk (hunk, &first0, &last0, &first1, &last1)) + return; + + /* Include a context's width before and after. */ + + i = - files[0].prefix_lines; + first0 = MAX (first0 - context, i); + first1 = MAX (first1 - context, i); + if (last0 < files[0].valid_lines - context) + last0 += context; + else + last0 = files[0].valid_lines - 1; + if (last1 < files[1].valid_lines - context) + last1 += context; + else + last1 = files[1].valid_lines - 1; + + /* If desired, find the preceding function definition line in file 0. */ + function = 0; + if (function_regexp.fastmap) + function = find_function (files[0].linbuf, first0); + + begin_output (); + out = outfile; + + fprintf (out, "@@ -"); + print_unidiff_number_range (&files[0], first0, last0); + fprintf (out, " +"); + print_unidiff_number_range (&files[1], first1, last1); + fprintf (out, " @@"); + + if (function) + print_context_function (out, function); + + putc ('\n', out); + + next = hunk; + i = first0; + j = first1; + + while (i <= last0 || j <= last1) + { + + /* If the line isn't a difference, output the context from file 0. */ + + if (!next || i < next->line0) + { + putc (initial_tab ? '\t' : ' ', out); + print_1_line (0, &files[0].linbuf[i++]); + j++; + } + else + { + /* For each difference, first output the deleted part. */ + + k = next->deleted; + while (k--) + { + putc ('-', out); + if (initial_tab) + putc ('\t', out); + print_1_line (0, &files[0].linbuf[i++]); + } + + /* Then output the inserted part. */ + + k = next->inserted; + while (k--) + { + putc ('+', out); + if (initial_tab) + putc ('\t', out); + print_1_line (0, &files[1].linbuf[j++]); + } + + /* We're done with this hunk, so on to the next! */ + + next = next->link; + } + } +} + +/* Scan a (forward-ordered) edit script for the first place that more than + 2*CONTEXT unchanged lines appear, and return a pointer + to the `struct change' for the last change before those lines. */ + +static struct change * +find_hunk (struct change *start) +{ + struct change *prev; + lin top0, top1; + lin thresh; + + /* Threshold distance is 2 * CONTEXT + 1 between two non-ignorable + changes, but only CONTEXT if one is ignorable. Watch out for + integer overflow, though. */ + lin non_ignorable_threshold = + (LIN_MAX - 1) / 2 < context ? LIN_MAX : 2 * context + 1; + lin ignorable_threshold = context; + + do + { + /* Compute number of first line in each file beyond this changed. */ + top0 = start->line0 + start->deleted; + top1 = start->line1 + start->inserted; + prev = start; + start = start->link; + thresh = (prev->ignore || (start && start->ignore) + ? ignorable_threshold + : non_ignorable_threshold); + /* It is not supposed to matter which file we check in the end-test. + If it would matter, crash. */ + if (start && start->line0 - top0 != start->line1 - top1) + abort (); + } while (start + /* Keep going if less than THRESH lines + elapse before the affected line. */ + && start->line0 - top0 < thresh); + + return prev; +} + +/* Set the `ignore' flag properly in each change in SCRIPT. + It should be 1 if all the lines inserted or deleted in that change + are ignorable lines. */ + +static void +mark_ignorable (struct change *script) +{ + while (script) + { + struct change *next = script->link; + lin first0, last0, first1, last1; + + /* Turn this change into a hunk: detach it from the others. */ + script->link = 0; + + /* Determine whether this change is ignorable. */ + script->ignore = ! analyze_hunk (script, + &first0, &last0, &first1, &last1); + + /* Reconnect the chain as before. */ + script->link = next; + + /* Advance to the following change. */ + script = next; + } +} + +/* Find the last function-header line in LINBUF prior to line number LINENUM. + This is a line containing a match for the regexp in `function_regexp'. + Return the address of the text, or 0 if no function-header is found. */ + +static char const * +find_function (char const * const *linbuf, lin linenum) +{ + lin i = linenum; + lin last = find_function_last_search; + find_function_last_search = i; + + while (last <= --i) + { + /* See if this line is what we want. */ + char const *line = linbuf[i]; + size_t linelen = linbuf[i + 1] - line - 1; + + /* FIXME: re_search's size args should be size_t, not int. */ + int len = MIN (linelen, INT_MAX); + + if (0 <= re_search (&function_regexp, line, len, 0, len, 0)) + { + find_function_last_match = i; + return line; + } + } + /* If we search back to where we started searching the previous time, + find the line we found last time. */ + if (find_function_last_match != LIN_MAX) + return linbuf[find_function_last_match]; + + return 0; +} diff --git a/contrib/diff/src/diff.c b/contrib/diff/src/diff.c new file mode 100644 index 0000000..a975107 --- /dev/null +++ b/contrib/diff/src/diff.c @@ -0,0 +1,1353 @@ +/* diff - compare files line by line + + Copyright (C) 1988, 1989, 1992, 1993, 1994, 1996, 1998, 2001, 2002, + 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + See the GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with GNU DIFF; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#define GDIFF_MAIN +#include "diff.h" +#include "paths.h" +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +#ifndef GUTTER_WIDTH_MINIMUM +# define GUTTER_WIDTH_MINIMUM 3 +#endif + +struct regexp_list +{ + char *regexps; /* chars representing disjunction of the regexps */ + size_t len; /* chars used in `regexps' */ + size_t size; /* size malloc'ed for `regexps'; 0 if not malloc'ed */ + bool multiple_regexps;/* Does `regexps' represent a disjunction? */ + struct re_pattern_buffer *buf; +}; + +static int compare_files (struct comparison const *, char const *, char const *); +static void add_regexp (struct regexp_list *, char const *); +static void summarize_regexp_list (struct regexp_list *); +static void specify_style (enum output_style); +static void specify_value (char const **, char const *, char const *); +static void try_help (char const *, char const *) __attribute__((noreturn)); +static void check_stdout (void); +static void usage (void); + +/* If comparing directories, compare their common subdirectories + recursively. */ +static bool recursive; + +/* In context diffs, show previous lines that match these regexps. */ +static struct regexp_list function_regexp_list; + +/* Ignore changes affecting only lines that match these regexps. */ +static struct regexp_list ignore_regexp_list; + +#if HAVE_SETMODE_DOS +/* Use binary I/O when reading and writing data (--binary). + On POSIX hosts, this has no effect. */ +static bool binary; +#else +enum { binary = true }; +#endif + +/* When comparing directories, if a file appears only in one + directory, treat it as present but empty in the other (-N). + Then `patch' would create the file with appropriate contents. */ +static bool new_file; + +/* When comparing directories, if a file appears only in the second + directory of the two, treat it as present but empty in the other + (--unidirectional-new-file). + Then `patch' would create the file with appropriate contents. */ +static bool unidirectional_new_file; + +/* Report files compared that are the same (-s). + Normally nothing is output when that happens. */ +static bool report_identical_files; + + +/* Return a string containing the command options with which diff was invoked. + Spaces appear between what were separate ARGV-elements. + There is a space at the beginning but none at the end. + If there were no options, the result is an empty string. + + Arguments: OPTIONVEC, a vector containing separate ARGV-elements, and COUNT, + the length of that vector. */ + +static char * +option_list (char **optionvec, int count) +{ + int i; + size_t size = 1; + char *result; + char *p; + + for (i = 0; i < count; i++) + size += 1 + quote_system_arg ((char *) 0, optionvec[i]); + + p = result = xmalloc (size); + + for (i = 0; i < count; i++) + { + *p++ = ' '; + p += quote_system_arg (p, optionvec[i]); + } + + *p = 0; + return result; +} + + +/* Return an option value suitable for add_exclude. */ + +static int +exclude_options (void) +{ + return EXCLUDE_WILDCARDS | (ignore_file_name_case ? FNM_CASEFOLD : 0); +} + +static char const shortopts[] = +"0123456789abBcC:dD:eEfF:hHiI:lL:nNpPqrsS:tTuU:vwW:x:X:y"; + +/* Values for long options that do not have single-letter equivalents. */ +enum +{ + BINARY_OPTION = CHAR_MAX + 1, + FROM_FILE_OPTION, + HELP_OPTION, + HORIZON_LINES_OPTION, + IGNORE_FILE_NAME_CASE_OPTION, + INHIBIT_HUNK_MERGE_OPTION, + LEFT_COLUMN_OPTION, + LINE_FORMAT_OPTION, + NO_IGNORE_FILE_NAME_CASE_OPTION, + NORMAL_OPTION, + SDIFF_MERGE_ASSIST_OPTION, + STRIP_TRAILING_CR_OPTION, + SUPPRESS_COMMON_LINES_OPTION, + TABSIZE_OPTION, + TO_FILE_OPTION, + + /* These options must be in sequence. */ + UNCHANGED_LINE_FORMAT_OPTION, + OLD_LINE_FORMAT_OPTION, + NEW_LINE_FORMAT_OPTION, + + /* These options must be in sequence. */ + UNCHANGED_GROUP_FORMAT_OPTION, + OLD_GROUP_FORMAT_OPTION, + NEW_GROUP_FORMAT_OPTION, + CHANGED_GROUP_FORMAT_OPTION +}; + +static char const group_format_option[][sizeof "--unchanged-group-format"] = + { + "--unchanged-group-format", + "--old-group-format", + "--new-group-format", + "--changed-group-format" + }; + +static char const line_format_option[][sizeof "--unchanged-line-format"] = + { + "--unchanged-line-format", + "--old-line-format", + "--new-line-format" + }; + +static struct option const longopts[] = +{ + {"binary", 0, 0, BINARY_OPTION}, + {"brief", 0, 0, 'q'}, + {"changed-group-format", 1, 0, CHANGED_GROUP_FORMAT_OPTION}, + {"context", 2, 0, 'C'}, + {"ed", 0, 0, 'e'}, + {"exclude", 1, 0, 'x'}, + {"exclude-from", 1, 0, 'X'}, + {"expand-tabs", 0, 0, 't'}, + {"forward-ed", 0, 0, 'f'}, + {"from-file", 1, 0, FROM_FILE_OPTION}, + {"help", 0, 0, HELP_OPTION}, + {"horizon-lines", 1, 0, HORIZON_LINES_OPTION}, + {"ifdef", 1, 0, 'D'}, + {"ignore-all-space", 0, 0, 'w'}, + {"ignore-blank-lines", 0, 0, 'B'}, + {"ignore-case", 0, 0, 'i'}, + {"ignore-file-name-case", 0, 0, IGNORE_FILE_NAME_CASE_OPTION}, + {"ignore-matching-lines", 1, 0, 'I'}, + {"ignore-space-change", 0, 0, 'b'}, + {"ignore-tab-expansion", 0, 0, 'E'}, + {"inhibit-hunk-merge", 0, 0, INHIBIT_HUNK_MERGE_OPTION}, + {"initial-tab", 0, 0, 'T'}, + {"label", 1, 0, 'L'}, + {"left-column", 0, 0, LEFT_COLUMN_OPTION}, + {"line-format", 1, 0, LINE_FORMAT_OPTION}, + {"minimal", 0, 0, 'd'}, + {"new-file", 0, 0, 'N'}, + {"new-group-format", 1, 0, NEW_GROUP_FORMAT_OPTION}, + {"new-line-format", 1, 0, NEW_LINE_FORMAT_OPTION}, + {"no-ignore-file-name-case", 0, 0, NO_IGNORE_FILE_NAME_CASE_OPTION}, + {"normal", 0, 0, NORMAL_OPTION}, + {"old-group-format", 1, 0, OLD_GROUP_FORMAT_OPTION}, + {"old-line-format", 1, 0, OLD_LINE_FORMAT_OPTION}, + {"paginate", 0, 0, 'l'}, + {"rcs", 0, 0, 'n'}, + {"recursive", 0, 0, 'r'}, + {"report-identical-files", 0, 0, 's'}, + {"sdiff-merge-assist", 0, 0, SDIFF_MERGE_ASSIST_OPTION}, + {"show-c-function", 0, 0, 'p'}, + {"show-function-line", 1, 0, 'F'}, + {"side-by-side", 0, 0, 'y'}, + {"speed-large-files", 0, 0, 'H'}, + {"starting-file", 1, 0, 'S'}, + {"strip-trailing-cr", 0, 0, STRIP_TRAILING_CR_OPTION}, + {"suppress-common-lines", 0, 0, SUPPRESS_COMMON_LINES_OPTION}, + {"tabsize", 1, 0, TABSIZE_OPTION}, + {"text", 0, 0, 'a'}, + {"to-file", 1, 0, TO_FILE_OPTION}, + {"unchanged-group-format", 1, 0, UNCHANGED_GROUP_FORMAT_OPTION}, + {"unchanged-line-format", 1, 0, UNCHANGED_LINE_FORMAT_OPTION}, + {"unidirectional-new-file", 0, 0, 'P'}, + {"unified", 2, 0, 'U'}, + {"version", 0, 0, 'v'}, + {"width", 1, 0, 'W'}, + {0, 0, 0, 0} +}; + +int +main (int argc, char **argv) +{ + int exit_status = EXIT_SUCCESS; + int c; + int i; + int prev = -1; + lin ocontext = -1; + bool explicit_context = false; + size_t width = 0; + bool show_c_function = false; + char const *from_file = 0; + char const *to_file = 0; + uintmax_t numval; + char *numend; + + /* Do our initializations. */ + exit_failure = 2; + initialize_main (&argc, &argv); + program_name = argv[0]; + setlocale (LC_ALL, ""); + bindtextdomain (PACKAGE, LOCALEDIR); + textdomain (PACKAGE); + c_stack_action (0); + function_regexp_list.buf = &function_regexp; + ignore_regexp_list.buf = &ignore_regexp; + re_set_syntax (RE_SYNTAX_GREP | RE_NO_POSIX_BACKTRACKING); + excluded = new_exclude (); + + /* Decode the options. */ + + while ((c = getopt_long (argc, argv, shortopts, longopts, 0)) != -1) + { + switch (c) + { + case 0: + break; + + case '0': + case '1': + case '2': + case '3': + case '4': + case '5': + case '6': + case '7': + case '8': + case '9': + if (! ISDIGIT (prev)) + ocontext = c - '0'; + else if (LIN_MAX / 10 < ocontext + || ((ocontext = 10 * ocontext + c - '0') < 0)) + ocontext = LIN_MAX; + break; + + case 'a': + text = true; + break; + + case 'b': + if (ignore_white_space < IGNORE_SPACE_CHANGE) + ignore_white_space = IGNORE_SPACE_CHANGE; + break; + + case 'B': + ignore_blank_lines = true; + break; + + case 'C': + case 'U': + { + if (optarg) + { + numval = strtoumax (optarg, &numend, 10); + if (*numend) + try_help ("invalid context length `%s'", optarg); + if (LIN_MAX < numval) + numval = LIN_MAX; + } + else + numval = 3; + + specify_style (c == 'U' ? OUTPUT_UNIFIED : OUTPUT_CONTEXT); + if (context < numval) + context = numval; + explicit_context = true; + } + break; + + case 'c': + specify_style (OUTPUT_CONTEXT); + if (context < 3) + context = 3; + break; + + case 'd': + minimal = true; + break; + + case 'D': + specify_style (OUTPUT_IFDEF); + { + static char const C_ifdef_group_formats[] = + "%%=%c#ifndef %s\n%%<#endif /* ! %s */\n%c#ifdef %s\n%%>#endif /* %s */\n%c#ifndef %s\n%%<#else /* %s */\n%%>#endif /* %s */\n"; + char *b = xmalloc (sizeof C_ifdef_group_formats + + 7 * strlen (optarg) - 14 /* 7*"%s" */ + - 8 /* 5*"%%" + 3*"%c" */); + sprintf (b, C_ifdef_group_formats, + 0, + optarg, optarg, 0, + optarg, optarg, 0, + optarg, optarg, optarg); + for (i = 0; i < sizeof group_format / sizeof *group_format; i++) + { + specify_value (&group_format[i], b, "-D"); + b += strlen (b) + 1; + } + } + break; + + case 'e': + specify_style (OUTPUT_ED); + break; + + case 'E': + if (ignore_white_space < IGNORE_TAB_EXPANSION) + ignore_white_space = IGNORE_TAB_EXPANSION; + break; + + case 'f': + specify_style (OUTPUT_FORWARD_ED); + break; + + case 'F': + add_regexp (&function_regexp_list, optarg); + break; + + case 'h': + /* Split the files into chunks for faster processing. + Usually does not change the result. + + This currently has no effect. */ + break; + + case 'H': + speed_large_files = true; + break; + + case 'i': + ignore_case = true; + break; + + case 'I': + add_regexp (&ignore_regexp_list, optarg); + break; + + case 'l': + if (!pr_program[0]) + try_help ("pagination not supported on this host", 0); + paginate = true; +#ifdef SIGCHLD + /* Pagination requires forking and waiting, and + System V fork+wait does not work if SIGCHLD is ignored. */ + signal (SIGCHLD, SIG_DFL); +#endif + break; + + case 'L': + if (!file_label[0]) + file_label[0] = optarg; + else if (!file_label[1]) + file_label[1] = optarg; + else + fatal ("too many file label options"); + break; + + case 'n': + specify_style (OUTPUT_RCS); + break; + + case 'N': + new_file = true; + break; + + case 'p': + show_c_function = true; + add_regexp (&function_regexp_list, "^[[:alpha:]$_]"); + break; + + case 'P': + unidirectional_new_file = true; + break; + + case 'q': + brief = true; + break; + + case 'r': + recursive = true; + break; + + case 's': + report_identical_files = true; + break; + + case 'S': + specify_value (&starting_file, optarg, "-S"); + break; + + case 't': + expand_tabs = true; + break; + + case 'T': + initial_tab = true; + break; + + case 'u': + specify_style (OUTPUT_UNIFIED); + if (context < 3) + context = 3; + break; + + case 'v': + version_etc (stdout, "diff", PACKAGE_NAME, PACKAGE_VERSION, + "Paul Eggert", "Mike Haertel", "David Hayes", + "Richard Stallman", "Len Tower", (char *) 0); + check_stdout (); + return EXIT_SUCCESS; + + case 'w': + ignore_white_space = IGNORE_ALL_SPACE; + break; + + case 'x': + add_exclude (excluded, optarg, exclude_options ()); + break; + + case 'X': + if (add_exclude_file (add_exclude, excluded, optarg, + exclude_options (), '\n')) + pfatal_with_name (optarg); + break; + + case 'y': + specify_style (OUTPUT_SDIFF); + break; + + case 'W': + numval = strtoumax (optarg, &numend, 10); + if (! (0 < numval && numval <= SIZE_MAX) || *numend) + try_help ("invalid width `%s'", optarg); + if (width != numval) + { + if (width) + fatal ("conflicting width options"); + width = numval; + } + break; + + case BINARY_OPTION: +#if HAVE_SETMODE_DOS + binary = true; + set_binary_mode (STDOUT_FILENO, true); +#endif + break; + + case FROM_FILE_OPTION: + specify_value (&from_file, optarg, "--from-file"); + break; + + case HELP_OPTION: + usage (); + check_stdout (); + return EXIT_SUCCESS; + + case HORIZON_LINES_OPTION: + numval = strtoumax (optarg, &numend, 10); + if (*numend) + try_help ("invalid horizon length `%s'", optarg); + horizon_lines = MAX (horizon_lines, MIN (numval, LIN_MAX)); + break; + + case IGNORE_FILE_NAME_CASE_OPTION: + ignore_file_name_case = true; + break; + + case INHIBIT_HUNK_MERGE_OPTION: + /* This option is obsolete, but accept it for backward + compatibility. */ + break; + + case LEFT_COLUMN_OPTION: + left_column = true; + break; + + case LINE_FORMAT_OPTION: + specify_style (OUTPUT_IFDEF); + for (i = 0; i < sizeof line_format / sizeof *line_format; i++) + specify_value (&line_format[i], optarg, "--line-format"); + break; + + case NO_IGNORE_FILE_NAME_CASE_OPTION: + ignore_file_name_case = false; + break; + + case NORMAL_OPTION: + specify_style (OUTPUT_NORMAL); + break; + + case SDIFF_MERGE_ASSIST_OPTION: + specify_style (OUTPUT_SDIFF); + sdiff_merge_assist = true; + break; + + case STRIP_TRAILING_CR_OPTION: + strip_trailing_cr = true; + break; + + case SUPPRESS_COMMON_LINES_OPTION: + suppress_common_lines = true; + break; + + case TABSIZE_OPTION: + numval = strtoumax (optarg, &numend, 10); + if (! (0 < numval && numval <= SIZE_MAX) || *numend) + try_help ("invalid tabsize `%s'", optarg); + if (tabsize != numval) + { + if (tabsize) + fatal ("conflicting tabsize options"); + tabsize = numval; + } + break; + + case TO_FILE_OPTION: + specify_value (&to_file, optarg, "--to-file"); + break; + + case UNCHANGED_LINE_FORMAT_OPTION: + case OLD_LINE_FORMAT_OPTION: + case NEW_LINE_FORMAT_OPTION: + specify_style (OUTPUT_IFDEF); + c -= UNCHANGED_LINE_FORMAT_OPTION; + specify_value (&line_format[c], optarg, line_format_option[c]); + break; + + case UNCHANGED_GROUP_FORMAT_OPTION: + case OLD_GROUP_FORMAT_OPTION: + case NEW_GROUP_FORMAT_OPTION: + case CHANGED_GROUP_FORMAT_OPTION: + specify_style (OUTPUT_IFDEF); + c -= UNCHANGED_GROUP_FORMAT_OPTION; + specify_value (&group_format[c], optarg, group_format_option[c]); + break; + + default: + try_help (0, 0); + } + prev = c; + } + + if (output_style == OUTPUT_UNSPECIFIED) + { + if (show_c_function) + { + specify_style (OUTPUT_CONTEXT); + if (ocontext < 0) + context = 3; + } + else + specify_style (OUTPUT_NORMAL); + } + + if (output_style != OUTPUT_CONTEXT || hard_locale (LC_TIME)) + { +#ifdef ST_MTIM_NSEC + time_format = "%Y-%m-%d %H:%M:%S.%N %z"; +#else + time_format = "%Y-%m-%d %H:%M:%S %z"; +#endif + } + else + { + /* See POSIX 1003.1-2001 for this format. */ + time_format = "%a %b %e %T %Y"; + } + + if (0 <= ocontext) + { + bool modern_usage = 200112 <= posix2_version (); + + if ((output_style == OUTPUT_CONTEXT + || output_style == OUTPUT_UNIFIED) + && (context < ocontext + || (ocontext < context && ! explicit_context))) + { + if (modern_usage) + { + error (0, 0, + _("`-%ld' option is obsolete; use `-%c %ld'"), + (long int) ocontext, + output_style == OUTPUT_CONTEXT ? 'C' : 'U', + (long int) ocontext); + try_help (0, 0); + } + context = ocontext; + } + else + { + if (modern_usage) + { + error (0, 0, _("`-%ld' option is obsolete; omit it"), + (long int) ocontext); + try_help (0, 0); + } + } + } + + if (! tabsize) + tabsize = 8; + if (! width) + width = 130; + + { + /* Maximize first the half line width, and then the gutter width, + according to the following constraints: + + 1. Two half lines plus a gutter must fit in a line. + 2. If the half line width is nonzero: + a. The gutter width is at least GUTTER_WIDTH_MINIMUM. + b. If tabs are not expanded to spaces, + a half line plus a gutter is an integral number of tabs, + so that tabs in the right column line up. */ + + intmax_t t = expand_tabs ? 1 : tabsize; + intmax_t w = width; + intmax_t off = (w + t + GUTTER_WIDTH_MINIMUM) / (2 * t) * t; + sdiff_half_width = MAX (0, MIN (off - GUTTER_WIDTH_MINIMUM, w - off)), + sdiff_column2_offset = sdiff_half_width ? off : w; + } + + /* Make the horizon at least as large as the context, so that + shift_boundaries has more freedom to shift the first and last hunks. */ + if (horizon_lines < context) + horizon_lines = context; + + summarize_regexp_list (&function_regexp_list); + summarize_regexp_list (&ignore_regexp_list); + + if (output_style == OUTPUT_IFDEF) + { + for (i = 0; i < sizeof line_format / sizeof *line_format; i++) + if (!line_format[i]) + line_format[i] = "%l\n"; + if (!group_format[OLD]) + group_format[OLD] + = group_format[CHANGED] ? group_format[CHANGED] : "%<"; + if (!group_format[NEW]) + group_format[NEW] + = group_format[CHANGED] ? group_format[CHANGED] : "%>"; + if (!group_format[UNCHANGED]) + group_format[UNCHANGED] = "%="; + if (!group_format[CHANGED]) + group_format[CHANGED] = concat (group_format[OLD], + group_format[NEW], ""); + } + + no_diff_means_no_output = + (output_style == OUTPUT_IFDEF ? + (!*group_format[UNCHANGED] + || (strcmp (group_format[UNCHANGED], "%=") == 0 + && !*line_format[UNCHANGED])) + : (output_style != OUTPUT_SDIFF) | suppress_common_lines); + + files_can_be_treated_as_binary = + (brief & binary + & ~ (ignore_blank_lines | ignore_case | strip_trailing_cr + | (ignore_regexp_list.regexps || ignore_white_space))); + + switch_string = option_list (argv + 1, optind - 1); + + if (from_file) + { + if (to_file) + fatal ("--from-file and --to-file both specified"); + else + for (; optind < argc; optind++) + { + int status = compare_files ((struct comparison *) 0, + from_file, argv[optind]); + if (exit_status < status) + exit_status = status; + } + } + else + { + if (to_file) + for (; optind < argc; optind++) + { + int status = compare_files ((struct comparison *) 0, + argv[optind], to_file); + if (exit_status < status) + exit_status = status; + } + else + { + if (argc - optind != 2) + { + if (argc - optind < 2) + try_help ("missing operand after `%s'", argv[argc - 1]); + else + try_help ("extra operand `%s'", argv[optind + 2]); + } + + exit_status = compare_files ((struct comparison *) 0, + argv[optind], argv[optind + 1]); + } + } + + /* Print any messages that were saved up for last. */ + print_message_queue (); + + check_stdout (); + exit (exit_status); + return exit_status; +} + +/* Append to REGLIST the regexp PATTERN. */ + +static void +add_regexp (struct regexp_list *reglist, char const *pattern) +{ + size_t patlen = strlen (pattern); + char const *m = re_compile_pattern (pattern, patlen, reglist->buf); + + if (m != 0) + error (0, 0, "%s: %s", pattern, m); + else + { + char *regexps = reglist->regexps; + size_t len = reglist->len; + bool multiple_regexps = reglist->multiple_regexps = regexps != 0; + size_t newlen = reglist->len = len + 2 * multiple_regexps + patlen; + size_t size = reglist->size; + + if (size <= newlen) + { + if (!size) + size = 1; + + do size *= 2; + while (size <= newlen); + + reglist->size = size; + reglist->regexps = regexps = xrealloc (regexps, size); + } + if (multiple_regexps) + { + regexps[len++] = '\\'; + regexps[len++] = '|'; + } + memcpy (regexps + len, pattern, patlen + 1); + } +} + +/* Ensure that REGLIST represents the disjunction of its regexps. + This is done here, rather than earlier, to avoid O(N^2) behavior. */ + +static void +summarize_regexp_list (struct regexp_list *reglist) +{ + if (reglist->regexps) + { + /* At least one regexp was specified. Allocate a fastmap for it. */ + reglist->buf->fastmap = xmalloc (1 << CHAR_BIT); + if (reglist->multiple_regexps) + { + /* Compile the disjunction of the regexps. + (If just one regexp was specified, it is already compiled.) */ + char const *m = re_compile_pattern (reglist->regexps, reglist->len, + reglist->buf); + if (m != 0) + error (EXIT_TROUBLE, 0, "%s: %s", reglist->regexps, m); + } + } +} + +static void +try_help (char const *reason_msgid, char const *operand) +{ + if (reason_msgid) + error (0, 0, _(reason_msgid), operand); + error (EXIT_TROUBLE, 0, _("Try `%s --help' for more information."), + program_name); + abort (); +} + +static void +check_stdout (void) +{ + if (ferror (stdout)) + fatal ("write failed"); + else if (fclose (stdout) != 0) + pfatal_with_name (_("standard output")); +} + +static char const * const option_help_msgid[] = { + N_("Compare files line by line."), + "", + N_("-i --ignore-case Ignore case differences in file contents."), + N_("--ignore-file-name-case Ignore case when comparing file names."), + N_("--no-ignore-file-name-case Consider case when comparing file names."), + N_("-E --ignore-tab-expansion Ignore changes due to tab expansion."), + N_("-b --ignore-space-change Ignore changes in the amount of white space."), + N_("-w --ignore-all-space Ignore all white space."), + N_("-B --ignore-blank-lines Ignore changes whose lines are all blank."), + N_("-I RE --ignore-matching-lines=RE Ignore changes whose lines all match RE."), + N_("--strip-trailing-cr Strip trailing carriage return on input."), +#if HAVE_SETMODE_DOS + N_("--binary Read and write data in binary mode."), +#endif + N_("-a --text Treat all files as text."), + "", + N_("-c -C NUM --context[=NUM] Output NUM (default 3) lines of copied context.\n\ +-u -U NUM --unified[=NUM] Output NUM (default 3) lines of unified context.\n\ + --label LABEL Use LABEL instead of file name.\n\ + -p --show-c-function Show which C function each change is in.\n\ + -F RE --show-function-line=RE Show the most recent line matching RE."), + N_("-q --brief Output only whether files differ."), + N_("-e --ed Output an ed script."), + N_("--normal Output a normal diff."), + N_("-n --rcs Output an RCS format diff."), + N_("-y --side-by-side Output in two columns.\n\ + -W NUM --width=NUM Output at most NUM (default 130) print columns.\n\ + --left-column Output only the left column of common lines.\n\ + --suppress-common-lines Do not output common lines."), + N_("-D NAME --ifdef=NAME Output merged file to show `#ifdef NAME' diffs."), + N_("--GTYPE-group-format=GFMT Similar, but format GTYPE input groups with GFMT."), + N_("--line-format=LFMT Similar, but format all input lines with LFMT."), + N_("--LTYPE-line-format=LFMT Similar, but format LTYPE input lines with LFMT."), + N_(" LTYPE is `old', `new', or `unchanged'. GTYPE is LTYPE or `changed'."), + N_(" GFMT may contain:\n\ + %< lines from FILE1\n\ + %> lines from FILE2\n\ + %= lines common to FILE1 and FILE2\n\ + %[-][WIDTH][.[PREC]]{doxX}LETTER printf-style spec for LETTER\n\ + LETTERs are as follows for new group, lower case for old group:\n\ + F first line number\n\ + L last line number\n\ + N number of lines = L-F+1\n\ + E F-1\n\ + M L+1"), + N_(" LFMT may contain:\n\ + %L contents of line\n\ + %l contents of line, excluding any trailing newline\n\ + %[-][WIDTH][.[PREC]]{doxX}n printf-style spec for input line number"), + N_(" Either GFMT or LFMT may contain:\n\ + %% %\n\ + %c'C' the single character C\n\ + %c'\\OOO' the character with octal code OOO"), + "", + N_("-l --paginate Pass the output through `pr' to paginate it."), + N_("-t --expand-tabs Expand tabs to spaces in output."), + N_("-T --initial-tab Make tabs line up by prepending a tab."), + N_("--tabsize=NUM Tab stops are every NUM (default 8) print columns."), + "", + N_("-r --recursive Recursively compare any subdirectories found."), + N_("-N --new-file Treat absent files as empty."), + N_("--unidirectional-new-file Treat absent first files as empty."), + N_("-s --report-identical-files Report when two files are the same."), + N_("-x PAT --exclude=PAT Exclude files that match PAT."), + N_("-X FILE --exclude-from=FILE Exclude files that match any pattern in FILE."), + N_("-S FILE --starting-file=FILE Start with FILE when comparing directories."), + N_("--from-file=FILE1 Compare FILE1 to all operands. FILE1 can be a directory."), + N_("--to-file=FILE2 Compare all operands to FILE2. FILE2 can be a directory."), + "", + N_("--horizon-lines=NUM Keep NUM lines of the common prefix and suffix."), + N_("-d --minimal Try hard to find a smaller set of changes."), + N_("--speed-large-files Assume large files and many scattered small changes."), + "", + N_("-v --version Output version info."), + N_("--help Output this help."), + "", + N_("FILES are `FILE1 FILE2' or `DIR1 DIR2' or `DIR FILE...' or `FILE... DIR'."), + N_("If --from-file or --to-file is given, there are no restrictions on FILES."), + N_("If a FILE is `-', read standard input."), + N_("Exit status is 0 if inputs are the same, 1 if different, 2 if trouble."), + "", + N_("Report bugs to ."), + 0 +}; + +static void +usage (void) +{ + char const * const *p; + + printf (_("Usage: %s [OPTION]... FILES\n"), program_name); + + for (p = option_help_msgid; *p; p++) + { + if (!**p) + putchar ('\n'); + else + { + char const *msg = _(*p); + char const *nl; + while ((nl = strchr (msg, '\n'))) + { + int msglen = nl + 1 - msg; + printf (" %.*s", msglen, msg); + msg = nl + 1; + } + + printf (" %s\n" + 2 * (*msg != ' ' && *msg != '-'), msg); + } + } +} + +/* Set VAR to VALUE, reporting an OPTION error if this is a + conflict. */ +static void +specify_value (char const **var, char const *value, char const *option) +{ + if (*var && strcmp (*var, value) != 0) + { + error (0, 0, _("conflicting %s option value `%s'"), option, value); + try_help (0, 0); + } + *var = value; +} + +/* Set the output style to STYLE, diagnosing conflicts. */ +static void +specify_style (enum output_style style) +{ + if (output_style != style) + { + if (output_style != OUTPUT_UNSPECIFIED) + try_help ("conflicting output style options", 0); + output_style = style; + } +} + +/* Set the last-modified time of *ST to be the current time. */ + +static void +set_mtime_to_now (struct stat *st) +{ +#ifdef ST_MTIM_NSEC + +# if HAVE_CLOCK_GETTIME && defined CLOCK_REALTIME + if (clock_gettime (CLOCK_REALTIME, &st->st_mtim) == 0) + return; +# endif + +# if HAVE_GETTIMEOFDAY + { + struct timeval timeval; + if (gettimeofday (&timeval, 0) == 0) + { + st->st_mtime = timeval.tv_sec; + st->st_mtim.ST_MTIM_NSEC = timeval.tv_usec * 1000; + return; + } + } +# endif + +#endif /* ST_MTIM_NSEC */ + + time (&st->st_mtime); +} + +/* Compare two files (or dirs) with parent comparison PARENT + and names NAME0 and NAME1. + (If PARENT is 0, then the first name is just NAME0, etc.) + This is self-contained; it opens the files and closes them. + + Value is EXIT_SUCCESS if files are the same, EXIT_FAILURE if + different, EXIT_TROUBLE if there is a problem opening them. */ + +static int +compare_files (struct comparison const *parent, + char const *name0, + char const *name1) +{ + struct comparison cmp; +#define DIR_P(f) (S_ISDIR (cmp.file[f].stat.st_mode) != 0) + register int f; + int status = EXIT_SUCCESS; + bool same_files; + char *free0, *free1; + + /* If this is directory comparison, perhaps we have a file + that exists only in one of the directories. + If so, just print a message to that effect. */ + + if (! ((name0 && name1) + || (unidirectional_new_file && name1) + || new_file)) + { + char const *name = name0 == 0 ? name1 : name0; + char const *dir = parent->file[name0 == 0].name; + + /* See POSIX 1003.1-2001 for this format. */ + message ("Only in %s: %s\n", dir, name); + + /* Return EXIT_FAILURE so that diff_dirs will return + EXIT_FAILURE ("some files differ"). */ + return EXIT_FAILURE; + } + + memset (cmp.file, 0, sizeof cmp.file); + cmp.parent = parent; + + /* cmp.file[f].desc markers */ +#define NONEXISTENT (-1) /* nonexistent file */ +#define UNOPENED (-2) /* unopened file (e.g. directory) */ +#define ERRNO_ENCODE(errno) (-3 - (errno)) /* encoded errno value */ + +#define ERRNO_DECODE(desc) (-3 - (desc)) /* inverse of ERRNO_ENCODE */ + + cmp.file[0].desc = name0 == 0 ? NONEXISTENT : UNOPENED; + cmp.file[1].desc = name1 == 0 ? NONEXISTENT : UNOPENED; + + /* Now record the full name of each file, including nonexistent ones. */ + + if (name0 == 0) + name0 = name1; + if (name1 == 0) + name1 = name0; + + if (!parent) + { + free0 = 0; + free1 = 0; + cmp.file[0].name = name0; + cmp.file[1].name = name1; + } + else + { + cmp.file[0].name = free0 + = dir_file_pathname (parent->file[0].name, name0); + cmp.file[1].name = free1 + = dir_file_pathname (parent->file[1].name, name1); + } + + /* Stat the files. */ + + for (f = 0; f < 2; f++) + { + if (cmp.file[f].desc != NONEXISTENT) + { + if (f && file_name_cmp (cmp.file[f].name, cmp.file[0].name) == 0) + { + cmp.file[f].desc = cmp.file[0].desc; + cmp.file[f].stat = cmp.file[0].stat; + } + else if (strcmp (cmp.file[f].name, "-") == 0) + { + cmp.file[f].desc = STDIN_FILENO; + if (fstat (STDIN_FILENO, &cmp.file[f].stat) != 0) + cmp.file[f].desc = ERRNO_ENCODE (errno); + else + { + if (S_ISREG (cmp.file[f].stat.st_mode)) + { + off_t pos = lseek (STDIN_FILENO, (off_t) 0, SEEK_CUR); + if (pos < 0) + cmp.file[f].desc = ERRNO_ENCODE (errno); + else + cmp.file[f].stat.st_size = + MAX (0, cmp.file[f].stat.st_size - pos); + } + + /* POSIX 1003.1-2001 requires current time for + stdin. */ + set_mtime_to_now (&cmp.file[f].stat); + } + } + else if (stat (cmp.file[f].name, &cmp.file[f].stat) != 0) + cmp.file[f].desc = ERRNO_ENCODE (errno); + } + } + + /* Mark files as nonexistent as needed for -N and -P, if they are + inaccessible empty regular files (the kind of files that 'patch' + creates to indicate nonexistent backups), or if they are + top-level files that do not exist but their counterparts do + exist. */ + for (f = 0; f < 2; f++) + if ((new_file || (f == 0 && unidirectional_new_file)) + && (cmp.file[f].desc == UNOPENED + ? (S_ISREG (cmp.file[f].stat.st_mode) + && ! (cmp.file[f].stat.st_mode & (S_IRWXU | S_IRWXG | S_IRWXO)) + && cmp.file[f].stat.st_size == 0) + : (cmp.file[f].desc == ERRNO_ENCODE (ENOENT) + && ! parent + && cmp.file[1 - f].desc == UNOPENED))) + cmp.file[f].desc = NONEXISTENT; + + for (f = 0; f < 2; f++) + if (cmp.file[f].desc == NONEXISTENT) + { + memset (&cmp.file[f].stat, 0, sizeof cmp.file[f].stat); + cmp.file[f].stat.st_mode = cmp.file[1 - f].stat.st_mode; + } + + for (f = 0; f < 2; f++) + { + int e = ERRNO_DECODE (cmp.file[f].desc); + if (0 <= e) + { + errno = e; + perror_with_name (cmp.file[f].name); + status = EXIT_TROUBLE; + } + } + + if (status == EXIT_SUCCESS && ! parent && DIR_P (0) != DIR_P (1)) + { + /* If one is a directory, and it was specified in the command line, + use the file in that dir with the other file's basename. */ + + int fnm_arg = DIR_P (0); + int dir_arg = 1 - fnm_arg; + char const *fnm = cmp.file[fnm_arg].name; + char const *dir = cmp.file[dir_arg].name; + char const *filename = cmp.file[dir_arg].name = free0 + = dir_file_pathname (dir, base_name (fnm)); + + if (strcmp (fnm, "-") == 0) + fatal ("cannot compare `-' to a directory"); + + if (stat (filename, &cmp.file[dir_arg].stat) != 0) + { + perror_with_name (filename); + status = EXIT_TROUBLE; + } + } + + if (status != EXIT_SUCCESS) + { + /* One of the files should exist but does not. */ + } + else if (cmp.file[0].desc == NONEXISTENT + && cmp.file[1].desc == NONEXISTENT) + { + /* Neither file "exists", so there's nothing to compare. */ + } + else if ((same_files + = (cmp.file[0].desc != NONEXISTENT + && cmp.file[1].desc != NONEXISTENT + && 0 < same_file (&cmp.file[0].stat, &cmp.file[1].stat) + && same_file_attributes (&cmp.file[0].stat, + &cmp.file[1].stat))) + && no_diff_means_no_output) + { + /* The two named files are actually the same physical file. + We know they are identical without actually reading them. */ + } + else if (DIR_P (0) & DIR_P (1)) + { + if (output_style == OUTPUT_IFDEF) + fatal ("-D option not supported with directories"); + + /* If both are directories, compare the files in them. */ + + if (parent && !recursive) + { + /* But don't compare dir contents one level down + unless -r was specified. + See POSIX 1003.1-2001 for this format. */ + message ("Common subdirectories: %s and %s\n", + cmp.file[0].name, cmp.file[1].name); + } + else + status = diff_dirs (&cmp, compare_files); + } + else if ((DIR_P (0) | DIR_P (1)) + || (parent + && (! S_ISREG (cmp.file[0].stat.st_mode) + || ! S_ISREG (cmp.file[1].stat.st_mode)))) + { + if (cmp.file[0].desc == NONEXISTENT || cmp.file[1].desc == NONEXISTENT) + { + /* We have a subdirectory that exists only in one directory. */ + + if ((DIR_P (0) | DIR_P (1)) + && recursive + && (new_file + || (unidirectional_new_file + && cmp.file[0].desc == NONEXISTENT))) + status = diff_dirs (&cmp, compare_files); + else + { + char const *dir + = parent->file[cmp.file[0].desc == NONEXISTENT].name; + + /* See POSIX 1003.1-2001 for this format. */ + message ("Only in %s: %s\n", dir, name0); + + status = EXIT_FAILURE; + } + } + else + { + /* We have two files that are not to be compared. */ + + /* See POSIX 1003.1-2001 for this format. */ + message5 ("File %s is a %s while file %s is a %s\n", + file_label[0] ? file_label[0] : cmp.file[0].name, + file_type (&cmp.file[0].stat), + file_label[1] ? file_label[1] : cmp.file[1].name, + file_type (&cmp.file[1].stat)); + + /* This is a difference. */ + status = EXIT_FAILURE; + } + } + else if (files_can_be_treated_as_binary + && S_ISREG (cmp.file[0].stat.st_mode) + && S_ISREG (cmp.file[1].stat.st_mode) + && cmp.file[0].stat.st_size != cmp.file[1].stat.st_size) + { + message ("Files %s and %s differ\n", + file_label[0] ? file_label[0] : cmp.file[0].name, + file_label[1] ? file_label[1] : cmp.file[1].name); + status = EXIT_FAILURE; + } + else + { + /* Both exist and neither is a directory. */ + + /* Open the files and record their descriptors. */ + + if (cmp.file[0].desc == UNOPENED) + if ((cmp.file[0].desc = open (cmp.file[0].name, O_RDONLY, 0)) < 0) + { + perror_with_name (cmp.file[0].name); + status = EXIT_TROUBLE; + } + if (cmp.file[1].desc == UNOPENED) + { + if (same_files) + cmp.file[1].desc = cmp.file[0].desc; + else if ((cmp.file[1].desc = open (cmp.file[1].name, O_RDONLY, 0)) + < 0) + { + perror_with_name (cmp.file[1].name); + status = EXIT_TROUBLE; + } + } + +#if HAVE_SETMODE_DOS + if (binary) + for (f = 0; f < 2; f++) + if (0 <= cmp.file[f].desc) + set_binary_mode (cmp.file[f].desc, true); +#endif + + /* Compare the files, if no error was found. */ + + if (status == EXIT_SUCCESS) + status = diff_2_files (&cmp); + + /* Close the file descriptors. */ + + if (0 <= cmp.file[0].desc && close (cmp.file[0].desc) != 0) + { + perror_with_name (cmp.file[0].name); + status = EXIT_TROUBLE; + } + if (0 <= cmp.file[1].desc && cmp.file[0].desc != cmp.file[1].desc + && close (cmp.file[1].desc) != 0) + { + perror_with_name (cmp.file[1].name); + status = EXIT_TROUBLE; + } + } + + /* Now the comparison has been done, if no error prevented it, + and STATUS is the value this function will return. */ + + if (status == EXIT_SUCCESS) + { + if (report_identical_files && !DIR_P (0)) + message ("Files %s and %s are identical\n", + file_label[0] ? file_label[0] : cmp.file[0].name, + file_label[1] ? file_label[1] : cmp.file[1].name); + } + else + { + /* Flush stdout so that the user sees differences immediately. + This can hurt performance, unfortunately. */ + if (fflush (stdout) != 0) + pfatal_with_name (_("standard output")); + } + + if (free0) + free (free0); + if (free1) + free (free1); + + return status; +} diff --git a/contrib/diff/src/diff.h b/contrib/diff/src/diff.h new file mode 100644 index 0000000..e7a4b2d --- /dev/null +++ b/contrib/diff/src/diff.h @@ -0,0 +1,375 @@ +/* Shared definitions for GNU DIFF + + Copyright (C) 1988, 1989, 1991, 1992, 1993, 1994, 1995, 1998, 2001, + 2002, 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "system.h" +#include +#include +#include + +/* What kind of changes a hunk contains. */ +enum changes +{ + /* No changes: lines common to both files. */ + UNCHANGED, + + /* Deletes only: lines taken from just the first file. */ + OLD, + + /* Inserts only: lines taken from just the second file. */ + NEW, + + /* Both deletes and inserts: a hunk containing both old and new lines. */ + CHANGED +}; + +/* Variables for command line options */ + +#ifndef GDIFF_MAIN +# define XTERN extern +#else +# define XTERN +#endif + +enum output_style +{ + /* No output style specified. */ + OUTPUT_UNSPECIFIED, + + /* Default output style. */ + OUTPUT_NORMAL, + + /* Output the differences with lines of context before and after (-c). */ + OUTPUT_CONTEXT, + + /* Output the differences in a unified context diff format (-u). */ + OUTPUT_UNIFIED, + + /* Output the differences as commands suitable for `ed' (-e). */ + OUTPUT_ED, + + /* Output the diff as a forward ed script (-f). */ + OUTPUT_FORWARD_ED, + + /* Like -f, but output a count of changed lines in each "command" (-n). */ + OUTPUT_RCS, + + /* Output merged #ifdef'd file (-D). */ + OUTPUT_IFDEF, + + /* Output sdiff style (-y). */ + OUTPUT_SDIFF +}; + +/* True for output styles that are robust, + i.e. can handle a file that ends in a non-newline. */ +#define ROBUST_OUTPUT_STYLE(S) ((S) != OUTPUT_ED && (S) != OUTPUT_FORWARD_ED) + +XTERN enum output_style output_style; + +/* Nonzero if output cannot be generated for identical files. */ +XTERN bool no_diff_means_no_output; + +/* Number of lines of context to show in each set of diffs. + This is zero when context is not to be shown. */ +XTERN lin context; + +/* Consider all files as text files (-a). + Don't interpret codes over 0177 as implying a "binary file". */ +XTERN bool text; + +/* Number of lines to keep in identical prefix and suffix. */ +XTERN lin horizon_lines; + +/* The significance of white space during comparisons. */ +XTERN enum +{ + /* All white space is significant (the default). */ + IGNORE_NO_WHITE_SPACE, + + /* Ignore changes due to tab expansion (-E). */ + IGNORE_TAB_EXPANSION, + + /* Ignore changes in horizontal white space (-b). */ + IGNORE_SPACE_CHANGE, + + /* Ignore all horizontal white space (-w). */ + IGNORE_ALL_SPACE +} ignore_white_space; + +/* Ignore changes that affect only blank lines (-B). */ +XTERN bool ignore_blank_lines; + +/* Files can be compared byte-by-byte, as if they were binary. + This depends on various options. */ +XTERN bool files_can_be_treated_as_binary; + +/* Ignore differences in case of letters (-i). */ +XTERN bool ignore_case; + +/* Ignore differences in case of letters in file names. */ +XTERN bool ignore_file_name_case; + +/* File labels for `-c' output headers (--label). */ +XTERN char *file_label[2]; + +/* Regexp to identify function-header lines (-F). */ +XTERN struct re_pattern_buffer function_regexp; + +/* Ignore changes that affect only lines matching this regexp (-I). */ +XTERN struct re_pattern_buffer ignore_regexp; + +/* Say only whether files differ, not how (-q). */ +XTERN bool brief; + +/* Expand tabs in the output so the text lines up properly + despite the characters added to the front of each line (-t). */ +XTERN bool expand_tabs; + +/* Number of columns between tab stops. */ +XTERN size_t tabsize; + +/* Use a tab in the output, rather than a space, before the text of an + input line, so as to keep the proper alignment in the input line + without changing the characters in it (-T). */ +XTERN bool initial_tab; + +/* Remove trailing carriage returns from input. */ +XTERN bool strip_trailing_cr; + +/* In directory comparison, specify file to start with (-S). + This is used for resuming an aborted comparison. + All file names less than this name are ignored. */ +XTERN char const *starting_file; + +/* Pipe each file's output through pr (-l). */ +XTERN bool paginate; + +/* Line group formats for unchanged, old, new, and changed groups. */ +XTERN char const *group_format[CHANGED + 1]; + +/* Line formats for unchanged, old, and new lines. */ +XTERN char const *line_format[NEW + 1]; + +/* If using OUTPUT_SDIFF print extra information to help the sdiff filter. */ +XTERN bool sdiff_merge_assist; + +/* Tell OUTPUT_SDIFF to show only the left version of common lines. */ +XTERN bool left_column; + +/* Tell OUTPUT_SDIFF to not show common lines. */ +XTERN bool suppress_common_lines; + +/* The half line width and column 2 offset for OUTPUT_SDIFF. */ +XTERN size_t sdiff_half_width; +XTERN size_t sdiff_column2_offset; + +/* String containing all the command options diff received, + with spaces between and at the beginning but none at the end. + If there were no options given, this string is empty. */ +XTERN char *switch_string; + +/* Use heuristics for better speed with large files with a small + density of changes. */ +XTERN bool speed_large_files; + +/* Patterns that match file names to be excluded. */ +XTERN struct exclude *excluded; + +/* Don't discard lines. This makes things slower (sometimes much + slower) but will find a guaranteed minimal set of changes. */ +XTERN bool minimal; + +/* Name of program the user invoked (for error messages). */ +XTERN char *program_name; + +/* The strftime format to use for time strings. */ +XTERN char const *time_format; + +/* The result of comparison is an "edit script": a chain of `struct change'. + Each `struct change' represents one place where some lines are deleted + and some are inserted. + + LINE0 and LINE1 are the first affected lines in the two files (origin 0). + DELETED is the number of lines deleted here from file 0. + INSERTED is the number of lines inserted here in file 1. + + If DELETED is 0 then LINE0 is the number of the line before + which the insertion was done; vice versa for INSERTED and LINE1. */ + +struct change +{ + struct change *link; /* Previous or next edit command */ + lin inserted; /* # lines of file 1 changed here. */ + lin deleted; /* # lines of file 0 changed here. */ + lin line0; /* Line number of 1st deleted line. */ + lin line1; /* Line number of 1st inserted line. */ + bool ignore; /* Flag used in context.c. */ +}; + +/* Structures that describe the input files. */ + +/* Data on one input file being compared. */ + +struct file_data { + int desc; /* File descriptor */ + char const *name; /* File name */ + struct stat stat; /* File status */ + + /* Buffer in which text of file is read. */ + word *buffer; + + /* Allocated size of buffer, in bytes. Always a multiple of + sizeof *buffer. */ + size_t bufsize; + + /* Number of valid bytes now in the buffer. */ + size_t buffered; + + /* Array of pointers to lines in the file. */ + char const **linbuf; + + /* linbuf_base <= buffered_lines <= valid_lines <= alloc_lines. + linebuf[linbuf_base ... buffered_lines - 1] are possibly differing. + linebuf[linbuf_base ... valid_lines - 1] contain valid data. + linebuf[linbuf_base ... alloc_lines - 1] are allocated. */ + lin linbuf_base, buffered_lines, valid_lines, alloc_lines; + + /* Pointer to end of prefix of this file to ignore when hashing. */ + char const *prefix_end; + + /* Count of lines in the prefix. + There are this many lines in the file before linbuf[0]. */ + lin prefix_lines; + + /* Pointer to start of suffix of this file to ignore when hashing. */ + char const *suffix_begin; + + /* Vector, indexed by line number, containing an equivalence code for + each line. It is this vector that is actually compared with that + of another file to generate differences. */ + lin *equivs; + + /* Vector, like the previous one except that + the elements for discarded lines have been squeezed out. */ + lin *undiscarded; + + /* Vector mapping virtual line numbers (not counting discarded lines) + to real ones (counting those lines). Both are origin-0. */ + lin *realindexes; + + /* Total number of nondiscarded lines. */ + lin nondiscarded_lines; + + /* Vector, indexed by real origin-0 line number, + containing 1 for a line that is an insertion or a deletion. + The results of comparison are stored here. */ + char *changed; + + /* 1 if file ends in a line with no final newline. */ + bool missing_newline; + + /* 1 if at end of file. */ + bool eof; + + /* 1 more than the maximum equivalence value used for this or its + sibling file. */ + lin equiv_max; +}; + +/* The file buffer, considered as an array of bytes rather than + as an array of words. */ +#define FILE_BUFFER(f) ((char *) (f)->buffer) + +/* Data on two input files being compared. */ + +struct comparison + { + struct file_data file[2]; + struct comparison const *parent; /* parent, if a recursive comparison */ + }; + +/* Describe the two files currently being compared. */ + +XTERN struct file_data files[2]; + +/* Stdio stream to output diffs to. */ + +XTERN FILE *outfile; + +/* Declare various functions. */ + +/* analyze.c */ +int diff_2_files (struct comparison *); + +/* context.c */ +void print_context_header (struct file_data[], bool); +void print_context_script (struct change *, bool); + +/* dir.c */ +int diff_dirs (struct comparison const *, int (*) (struct comparison const *, char const *, char const *)); + +/* ed.c */ +void print_ed_script (struct change *); +void pr_forward_ed_script (struct change *); + +/* ifdef.c */ +void print_ifdef_script (struct change *); + +/* io.c */ +void file_block_read (struct file_data *, size_t); +bool read_files (struct file_data[], bool); + +/* normal.c */ +void print_normal_script (struct change *); + +/* rcs.c */ +void print_rcs_script (struct change *); + +/* side.c */ +void print_sdiff_script (struct change *); + +/* util.c */ +extern char const change_letter[4]; +extern char const pr_program[]; +char *concat (char const *, char const *, char const *); +char *dir_file_pathname (char const *, char const *); +bool lines_differ (char const *, char const *); +lin translate_line_number (struct file_data const *, lin); +struct change *find_change (struct change *); +struct change *find_reverse_change (struct change *); +void *zalloc (size_t); +enum changes analyze_hunk (struct change *, lin *, lin *, lin *, lin *); +void begin_output (void); +void debug_script (struct change *); +void fatal (char const *) __attribute__((noreturn)); +void finish_output (void); +void message (char const *, char const *, char const *); +void message5 (char const *, char const *, char const *, char const *, char const *); +void output_1_line (char const *, char const *, char const *, char const *); +void perror_with_name (char const *); +void pfatal_with_name (char const *) __attribute__((noreturn)); +void print_1_line (char const *, char const * const *); +void print_message_queue (void); +void print_number_range (char, struct file_data *, lin, lin); +void print_script (struct change *, struct change * (*) (struct change *), void (*) (struct change *)); +void setup_output (char const *, char const *, bool); +void translate_range (struct file_data const *, lin, lin, long int *, long int *); diff --git a/contrib/diff/src/diff3.c b/contrib/diff/src/diff3.c new file mode 100644 index 0000000..261eeab --- /dev/null +++ b/contrib/diff/src/diff3.c @@ -0,0 +1,1744 @@ +/* diff3 - compare three files line by line + + Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1996, 1998, 2001, + 2002, 2004 Free Software Foundation, Inc. + + This program is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + This program is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + See the GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "system.h" +#include "paths.h" + +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* Internal data structures and macros for the diff3 program; includes + data structures for both diff3 diffs and normal diffs. */ + +/* Different files within a three way diff. */ +#define FILE0 0 +#define FILE1 1 +#define FILE2 2 + +/* A three way diff is built from two two-way diffs; the file which + the two two-way diffs share is: */ +#define FILEC FILE2 + +/* Different files within a two way diff. + FC is the common file, FO the other file. */ +#define FO 0 +#define FC 1 + +/* The ranges are indexed by */ +#define RANGE_START 0 +#define RANGE_END 1 + +enum diff_type { + ERROR, /* Should not be used */ + ADD, /* Two way diff add */ + CHANGE, /* Two way diff change */ + DELETE, /* Two way diff delete */ + DIFF_ALL, /* All three are different */ + DIFF_1ST, /* Only the first is different */ + DIFF_2ND, /* Only the second */ + DIFF_3RD /* Only the third */ +}; + +/* Two way diff */ +struct diff_block { + lin ranges[2][2]; /* Ranges are inclusive */ + char **lines[2]; /* The actual lines (may contain nulls) */ + size_t *lengths[2]; /* Line lengths (including newlines, if any) */ + struct diff_block *next; +}; + +/* Three way diff */ + +struct diff3_block { + enum diff_type correspond; /* Type of diff */ + lin ranges[3][2]; /* Ranges are inclusive */ + char **lines[3]; /* The actual lines (may contain nulls) */ + size_t *lengths[3]; /* Line lengths (including newlines, if any) */ + struct diff3_block *next; +}; + +/* Access the ranges on a diff block. */ +#define D_LOWLINE(diff, filenum) \ + ((diff)->ranges[filenum][RANGE_START]) +#define D_HIGHLINE(diff, filenum) \ + ((diff)->ranges[filenum][RANGE_END]) +#define D_NUMLINES(diff, filenum) \ + (D_HIGHLINE (diff, filenum) - D_LOWLINE (diff, filenum) + 1) + +/* Access the line numbers in a file in a diff by relative line + numbers (i.e. line number within the diff itself). Note that these + are lvalues and can be used for assignment. */ +#define D_RELNUM(diff, filenum, linenum) \ + ((diff)->lines[filenum][linenum]) +#define D_RELLEN(diff, filenum, linenum) \ + ((diff)->lengths[filenum][linenum]) + +/* And get at them directly, when that should be necessary. */ +#define D_LINEARRAY(diff, filenum) \ + ((diff)->lines[filenum]) +#define D_LENARRAY(diff, filenum) \ + ((diff)->lengths[filenum]) + +/* Next block. */ +#define D_NEXT(diff) ((diff)->next) + +/* Access the type of a diff3 block. */ +#define D3_TYPE(diff) ((diff)->correspond) + +/* Line mappings based on diffs. The first maps off the top of the + diff, the second off of the bottom. */ +#define D_HIGH_MAPLINE(diff, fromfile, tofile, linenum) \ + ((linenum) \ + - D_HIGHLINE ((diff), (fromfile)) \ + + D_HIGHLINE ((diff), (tofile))) + +#define D_LOW_MAPLINE(diff, fromfile, tofile, linenum) \ + ((linenum) \ + - D_LOWLINE ((diff), (fromfile)) \ + + D_LOWLINE ((diff), (tofile))) + +/* Options variables for flags set on command line. */ + +/* If nonzero, treat all files as text files, never as binary. */ +static bool text; + +/* Remove trailing carriage returns from input. */ +static bool strip_trailing_cr; + +/* If nonzero, write out an ed script instead of the standard diff3 format. */ +static bool edscript; + +/* If nonzero, in the case of overlapping diffs (type DIFF_ALL), + preserve the lines which would normally be deleted from + file 1 with a special flagging mechanism. */ +static bool flagging; + +/* Use a tab to align output lines (-T). */ +static bool initial_tab; + +/* If nonzero, do not output information for overlapping diffs. */ +static bool simple_only; + +/* If nonzero, do not output information for non-overlapping diffs. */ +static bool overlap_only; + +/* If nonzero, show information for DIFF_2ND diffs. */ +static bool show_2nd; + +/* If nonzero, include `:wq' at the end of the script + to write out the file being edited. */ +static bool finalwrite; + +/* If nonzero, output a merged file. */ +static bool merge; + +char *program_name; + +static char *read_diff (char const *, char const *, char **); +static char *scan_diff_line (char *, char **, size_t *, char *, char); +static enum diff_type process_diff_control (char **, struct diff_block *); +static bool compare_line_list (char * const[], size_t const[], char * const[], size_t const[], lin); +static bool copy_stringlist (char * const[], size_t const[], char *[], size_t[], lin); +static bool output_diff3_edscript (FILE *, struct diff3_block *, int const[3], int const[3], char const *, char const *, char const *); +static bool output_diff3_merge (FILE *, FILE *, struct diff3_block *, int const[3], int const[3], char const *, char const *, char const *); +static struct diff3_block *create_diff3_block (lin, lin, lin, lin, lin, lin); +static struct diff3_block *make_3way_diff (struct diff_block *, struct diff_block *); +static struct diff3_block *reverse_diff3_blocklist (struct diff3_block *); +static struct diff3_block *using_to_diff3_block (struct diff_block *[2], struct diff_block *[2], int, int, struct diff3_block const *); +static struct diff_block *process_diff (char const *, char const *, struct diff_block **); +static void check_stdout (void); +static void fatal (char const *) __attribute__((noreturn)); +static void output_diff3 (FILE *, struct diff3_block *, int const[3], int const[3]); +static void perror_with_exit (char const *) __attribute__((noreturn)); +static void try_help (char const *, char const *) __attribute__((noreturn)); +static void usage (void); + +static char const *diff_program = DEFAULT_DIFF_PROGRAM; + +/* Values for long options that do not have single-letter equivalents. */ +enum +{ + DIFF_PROGRAM_OPTION = CHAR_MAX + 1, + HELP_OPTION, + STRIP_TRAILING_CR_OPTION +}; + +static struct option const longopts[] = +{ + {"diff-program", 1, 0, DIFF_PROGRAM_OPTION}, + {"easy-only", 0, 0, '3'}, + {"ed", 0, 0, 'e'}, + {"help", 0, 0, HELP_OPTION}, + {"initial-tab", 0, 0, 'T'}, + {"label", 1, 0, 'L'}, + {"merge", 0, 0, 'm'}, + {"overlap-only", 0, 0, 'x'}, + {"show-all", 0, 0, 'A'}, + {"show-overlap", 0, 0, 'E'}, + {"strip-trailing-cr", 0, 0, STRIP_TRAILING_CR_OPTION}, + {"text", 0, 0, 'a'}, + {"version", 0, 0, 'v'}, + {0, 0, 0, 0} +}; + +int +main (int argc, char **argv) +{ + int c, i; + int common; + int mapping[3]; + int rev_mapping[3]; + int incompat = 0; + bool conflicts_found; + struct diff_block *thread0, *thread1, *last_block; + struct diff3_block *diff3; + int tag_count = 0; + char *tag_strings[3]; + char *commonname; + char **file; + struct stat statb; + + exit_failure = 2; + initialize_main (&argc, &argv); + program_name = argv[0]; + setlocale (LC_ALL, ""); + bindtextdomain (PACKAGE, LOCALEDIR); + textdomain (PACKAGE); + c_stack_action (0); + + while ((c = getopt_long (argc, argv, "aeimvx3AEL:TX", longopts, 0)) != -1) + { + switch (c) + { + case 'a': + text = true; + break; + case 'A': + show_2nd = true; + flagging = true; + incompat++; + break; + case 'x': + overlap_only = true; + incompat++; + break; + case '3': + simple_only = true; + incompat++; + break; + case 'i': + finalwrite = true; + break; + case 'm': + merge = true; + break; + case 'X': + overlap_only = true; + /* Fall through. */ + case 'E': + flagging = true; + /* Fall through. */ + case 'e': + incompat++; + break; + case 'T': + initial_tab = true; + break; + case STRIP_TRAILING_CR_OPTION: + strip_trailing_cr = true; + break; + case 'v': + version_etc (stdout, "diff3", PACKAGE_NAME, PACKAGE_VERSION, + "Randy Smith", (char *) 0); + check_stdout (); + return EXIT_SUCCESS; + case DIFF_PROGRAM_OPTION: + diff_program = optarg; + break; + case HELP_OPTION: + usage (); + check_stdout (); + return EXIT_SUCCESS; + case 'L': + /* Handle up to three -L options. */ + if (tag_count < 3) + { + tag_strings[tag_count++] = optarg; + break; + } + try_help ("too many file label options", 0); + default: + try_help (0, 0); + } + } + + edscript = incompat & ~merge; /* -AeExX3 without -m implies ed script. */ + show_2nd |= ~incompat & merge; /* -m without -AeExX3 implies -A. */ + flagging |= ~incompat & merge; + + if (incompat > 1 /* Ensure at most one of -AeExX3. */ + || finalwrite & merge /* -i -m would rewrite input file. */ + || (tag_count && ! flagging)) /* -L requires one of -AEX. */ + try_help ("incompatible options", 0); + + if (argc - optind != 3) + { + if (argc - optind < 3) + try_help ("missing operand after `%s'", argv[argc - 1]); + else + try_help ("extra operand `%s'", argv[optind + 3]); + } + + file = &argv[optind]; + + for (i = tag_count; i < 3; i++) + tag_strings[i] = file[i]; + + /* Always compare file1 to file2, even if file2 is "-". + This is needed for -mAeExX3. Using the file0 as + the common file would produce wrong results, because if the + file0-file1 diffs didn't line up with the file0-file2 diffs + (which is entirely possible since we don't use diff's -n option), + diff3 might report phantom changes from file1 to file2. + + Also, try to compare file0 to file1, because this is where + changes are expected to come from. Diffing between these pairs + of files is more likely to avoid phantom changes from file0 to file1. + + Historically, the default common file was file2, so some older + applications (e.g. Emacs ediff) used file2 as the ancestor. So, + for compatibility, if this is a 3-way diff (not a merge or + edscript), prefer file2 as the common file. */ + + common = 2 - (edscript | merge); + + if (strcmp (file[common], "-") == 0) + { + /* Sigh. We've got standard input as the common file. We can't + call diff twice on stdin. Use the other arg as the common + file instead. */ + common = 3 - common; + if (strcmp (file[0], "-") == 0 || strcmp (file[common], "-") == 0) + fatal ("`-' specified for more than one input file"); + } + + mapping[0] = 0; + mapping[1] = 3 - common; + mapping[2] = common; + + for (i = 0; i < 3; i++) + rev_mapping[mapping[i]] = i; + + for (i = 0; i < 3; i++) + if (strcmp (file[i], "-") != 0) + { + if (stat (file[i], &statb) < 0) + perror_with_exit (file[i]); + else if (S_ISDIR (statb.st_mode)) + error (EXIT_TROUBLE, EISDIR, "%s", file[i]); + } + +#ifdef SIGCHLD + /* System V fork+wait does not work if SIGCHLD is ignored. */ + signal (SIGCHLD, SIG_DFL); +#endif + + /* Invoke diff twice on two pairs of input files, combine the two + diffs, and output them. */ + + commonname = file[rev_mapping[FILEC]]; + thread1 = process_diff (file[rev_mapping[FILE1]], commonname, &last_block); + thread0 = process_diff (file[rev_mapping[FILE0]], commonname, &last_block); + diff3 = make_3way_diff (thread0, thread1); + if (edscript) + conflicts_found + = output_diff3_edscript (stdout, diff3, mapping, rev_mapping, + tag_strings[0], tag_strings[1], tag_strings[2]); + else if (merge) + { + if (! freopen (file[rev_mapping[FILE0]], "r", stdin)) + perror_with_exit (file[rev_mapping[FILE0]]); + conflicts_found + = output_diff3_merge (stdin, stdout, diff3, mapping, rev_mapping, + tag_strings[0], tag_strings[1], tag_strings[2]); + if (ferror (stdin)) + fatal ("read failed"); + } + else + { + output_diff3 (stdout, diff3, mapping, rev_mapping); + conflicts_found = false; + } + + check_stdout (); + exit (conflicts_found); + return conflicts_found; +} + +static void +try_help (char const *reason_msgid, char const *operand) +{ + if (reason_msgid) + error (0, 0, _(reason_msgid), operand); + error (EXIT_TROUBLE, 0, + _("Try `%s --help' for more information."), program_name); + abort (); +} + +static void +check_stdout (void) +{ + if (ferror (stdout)) + fatal ("write failed"); + else if (fclose (stdout) != 0) + perror_with_exit (_("standard output")); +} + +static char const * const option_help_msgid[] = { + N_("-e --ed Output unmerged changes from OLDFILE to YOURFILE into MYFILE."), + N_("-E --show-overlap Output unmerged changes, bracketing conflicts."), + N_("-A --show-all Output all changes, bracketing conflicts."), + N_("-x --overlap-only Output overlapping changes."), + N_("-X Output overlapping changes, bracketing them."), + N_("-3 --easy-only Output unmerged nonoverlapping changes."), + "", + N_("-m --merge Output merged file instead of ed script (default -A)."), + N_("-L LABEL --label=LABEL Use LABEL instead of file name."), + N_("-i Append `w' and `q' commands to ed scripts."), + N_("-a --text Treat all files as text."), + N_("--strip-trailing-cr Strip trailing carriage return on input."), + N_("-T --initial-tab Make tabs line up by prepending a tab."), + N_("--diff-program=PROGRAM Use PROGRAM to compare files."), + "", + N_("-v --version Output version info."), + N_("--help Output this help."), + 0 +}; + +static void +usage (void) +{ + char const * const *p; + + printf (_("Usage: %s [OPTION]... MYFILE OLDFILE YOURFILE\n"), + program_name); + printf ("%s\n\n", _("Compare three files line by line.")); + for (p = option_help_msgid; *p; p++) + if (**p) + printf (" %s\n", _(*p)); + else + putchar ('\n'); + printf ("\n%s\n%s\n\n%s\n", + _("If a FILE is `-', read standard input."), + _("Exit status is 0 if successful, 1 if conflicts, 2 if trouble."), + _("Report bugs to .")); +} + +/* Combine the two diffs together into one. + Here is the algorithm: + + File2 is shared in common between the two diffs. + Diff02 is the diff between 0 and 2. + Diff12 is the diff between 1 and 2. + + 1) Find the range for the first block in File2. + a) Take the lowest of the two ranges (in File2) in the two + current blocks (one from each diff) as being the low + water mark. Assign the upper end of this block as + being the high water mark and move the current block up + one. Mark the block just moved over as to be used. + b) Check the next block in the diff that the high water + mark is *not* from. + + *If* the high water mark is above + the low end of the range in that block, + + mark that block as to be used and move the current + block up. Set the high water mark to the max of + the high end of this block and the current. Repeat b. + + 2) Find the corresponding ranges in File0 (from the blocks + in diff02; line per line outside of diffs) and in File1. + Create a diff3_block, reserving space as indicated by the ranges. + + 3) Copy all of the pointers for file2 in. At least for now, + do memcmp's between corresponding strings in the two diffs. + + 4) Copy all of the pointers for file0 and 1 in. Get what is + needed from file2 (when there isn't a diff block, it's + identical to file2 within the range between diff blocks). + + 5) If the diff blocks used came from only one of the two + strings of diffs, then that file (i.e. the one other than + the common file in that diff) is the odd person out. If + diff blocks are used from both sets, check to see if files + 0 and 1 match: + + Same number of lines? If so, do a set of memcmp's (if + a memcmp matches; copy the pointer over; it'll be easier + later during comparisons). If they match, 0 & 1 are the + same. If not, all three different. + + Then do it again, until the blocks are exhausted. */ + + +/* Make a three way diff (chain of diff3_block's) from two two way + diffs (chains of diff_block's). Assume that each of the two diffs + passed are onto the same file (i.e. that each of the diffs were + made "to" the same file). Return a three way diff pointer with + numbering FILE0 = the other file in diff02, FILE1 = the other file + in diff12, and FILEC = the common file. */ + +static struct diff3_block * +make_3way_diff (struct diff_block *thread0, struct diff_block *thread1) +{ + /* Work on the two diffs passed to it as threads. Thread number 0 + is diff02, thread number 1 is diff12. USING is the base of the + list of blocks to be used to construct each block of the three + way diff; if no blocks from a particular thread are to be used, + that element of USING is 0. LAST_USING contains the last + elements on each of the using lists. + + HIGH_WATER_MARK is the highest line number in the common file + described in any of the diffs in either of the USING lists. + HIGH_WATER_THREAD names the thread. Similarly BASE_WATER_MARK + and BASE_WATER_THREAD describe the lowest line number in the + common file described in any of the diffs in either of the USING + lists. HIGH_WATER_DIFF is the diff from which the + HIGH_WATER_MARK was taken. + + HIGH_WATER_DIFF should always be equal to + LAST_USING[HIGH_WATER_THREAD]. OTHER_DIFF is the next diff to + check for higher water, and should always be equal to + CURRENT[HIGH_WATER_THREAD ^ 1]. OTHER_THREAD is the thread in + which the OTHER_DIFF is, and hence should always be equal to + HIGH_WATER_THREAD ^ 1. + + LAST_DIFF is the last diff block produced by this routine, for + line correspondence purposes between that diff and the one + currently being worked on. It is ZERO_DIFF before any blocks + have been created. */ + + struct diff_block *using[2]; + struct diff_block *last_using[2]; + struct diff_block *current[2]; + + lin high_water_mark; + + int high_water_thread; + int base_water_thread; + int other_thread; + + struct diff_block *high_water_diff; + struct diff_block *other_diff; + + struct diff3_block *result; + struct diff3_block *tmpblock; + struct diff3_block **result_end; + + struct diff3_block const *last_diff3; + + static struct diff3_block const zero_diff3; + + /* Initialization */ + result = 0; + result_end = &result; + current[0] = thread0; current[1] = thread1; + last_diff3 = &zero_diff3; + + /* Sniff up the threads until we reach the end */ + + while (current[0] || current[1]) + { + using[0] = using[1] = last_using[0] = last_using[1] = 0; + + /* Setup low and high water threads, diffs, and marks. */ + if (!current[0]) + base_water_thread = 1; + else if (!current[1]) + base_water_thread = 0; + else + base_water_thread = + (D_LOWLINE (current[0], FC) > D_LOWLINE (current[1], FC)); + + high_water_thread = base_water_thread; + + high_water_diff = current[high_water_thread]; + + high_water_mark = D_HIGHLINE (high_water_diff, FC); + + /* Make the diff you just got info from into the using class */ + using[high_water_thread] + = last_using[high_water_thread] + = high_water_diff; + current[high_water_thread] = high_water_diff->next; + last_using[high_water_thread]->next = 0; + + /* And mark the other diff */ + other_thread = high_water_thread ^ 0x1; + other_diff = current[other_thread]; + + /* Shuffle up the ladder, checking the other diff to see if it + needs to be incorporated. */ + while (other_diff + && D_LOWLINE (other_diff, FC) <= high_water_mark + 1) + { + + /* Incorporate this diff into the using list. Note that + this doesn't take it off the current list */ + if (using[other_thread]) + last_using[other_thread]->next = other_diff; + else + using[other_thread] = other_diff; + last_using[other_thread] = other_diff; + + /* Take it off the current list. Note that this following + code assumes that other_diff enters it equal to + current[high_water_thread ^ 0x1] */ + current[other_thread] = current[other_thread]->next; + other_diff->next = 0; + + /* Set the high_water stuff + If this comparison is equal, then this is the last pass + through this loop; since diff blocks within a given + thread cannot overlap, the high_water_mark will be + *below* the range_start of either of the next diffs. */ + + if (high_water_mark < D_HIGHLINE (other_diff, FC)) + { + high_water_thread ^= 1; + high_water_diff = other_diff; + high_water_mark = D_HIGHLINE (other_diff, FC); + } + + /* Set the other diff */ + other_thread = high_water_thread ^ 0x1; + other_diff = current[other_thread]; + } + + /* The using lists contain a list of all of the blocks to be + included in this diff3_block. Create it. */ + + tmpblock = using_to_diff3_block (using, last_using, + base_water_thread, high_water_thread, + last_diff3); + + if (!tmpblock) + fatal ("internal error: screwup in format of diff blocks"); + + /* Put it on the list. */ + *result_end = tmpblock; + result_end = &tmpblock->next; + + /* Set up corresponding lines correctly. */ + last_diff3 = tmpblock; + } + return result; +} + +/* Take two lists of blocks (from two separate diff threads) and put + them together into one diff3 block. Return a pointer to this diff3 + block or 0 for failure. + + All arguments besides using are for the convenience of the routine; + they could be derived from the using array. LAST_USING is a pair + of pointers to the last blocks in the using structure. LOW_THREAD + and HIGH_THREAD tell which threads contain the lowest and highest + line numbers for File0. LAST_DIFF3 contains the last diff produced + in the calling routine. This is used for lines mappings that + would still be identical to the state that diff ended in. + + A distinction should be made in this routine between the two diffs + that are part of a normal two diff block, and the three diffs that + are part of a diff3_block. */ + +static struct diff3_block * +using_to_diff3_block (struct diff_block *using[2], + struct diff_block *last_using[2], + int low_thread, int high_thread, + struct diff3_block const *last_diff3) +{ + lin low[2], high[2]; + struct diff3_block *result; + struct diff_block *ptr; + int d; + lin i; + + /* Find the range in the common file. */ + lin lowc = D_LOWLINE (using[low_thread], FC); + lin highc = D_HIGHLINE (last_using[high_thread], FC); + + /* Find the ranges in the other files. + If using[d] is null, that means that the file to which that diff + refers is equivalent to the common file over this range. */ + + for (d = 0; d < 2; d++) + if (using[d]) + { + low[d] = D_LOW_MAPLINE (using[d], FC, FO, lowc); + high[d] = D_HIGH_MAPLINE (last_using[d], FC, FO, highc); + } + else + { + low[d] = D_HIGH_MAPLINE (last_diff3, FILEC, FILE0 + d, lowc); + high[d] = D_HIGH_MAPLINE (last_diff3, FILEC, FILE0 + d, highc); + } + + /* Create a block with the appropriate sizes */ + result = create_diff3_block (low[0], high[0], low[1], high[1], lowc, highc); + + /* Copy information for the common file. + Return with a zero if any of the compares failed. */ + + for (d = 0; d < 2; d++) + for (ptr = using[d]; ptr; ptr = D_NEXT (ptr)) + { + lin result_offset = D_LOWLINE (ptr, FC) - lowc; + + if (!copy_stringlist (D_LINEARRAY (ptr, FC), + D_LENARRAY (ptr, FC), + D_LINEARRAY (result, FILEC) + result_offset, + D_LENARRAY (result, FILEC) + result_offset, + D_NUMLINES (ptr, FC))) + return 0; + } + + /* Copy information for file d. First deal with anything that might be + before the first diff. */ + + for (d = 0; d < 2; d++) + { + struct diff_block *u = using[d]; + lin lo = low[d], hi = high[d]; + + for (i = 0; + i + lo < (u ? D_LOWLINE (u, FO) : hi + 1); + i++) + { + D_RELNUM (result, FILE0 + d, i) = D_RELNUM (result, FILEC, i); + D_RELLEN (result, FILE0 + d, i) = D_RELLEN (result, FILEC, i); + } + + for (ptr = u; ptr; ptr = D_NEXT (ptr)) + { + lin result_offset = D_LOWLINE (ptr, FO) - lo; + lin linec; + + if (!copy_stringlist (D_LINEARRAY (ptr, FO), + D_LENARRAY (ptr, FO), + D_LINEARRAY (result, FILE0 + d) + result_offset, + D_LENARRAY (result, FILE0 + d) + result_offset, + D_NUMLINES (ptr, FO))) + return 0; + + /* Catch the lines between here and the next diff */ + linec = D_HIGHLINE (ptr, FC) + 1 - lowc; + for (i = D_HIGHLINE (ptr, FO) + 1 - lo; + i < (D_NEXT (ptr) ? D_LOWLINE (D_NEXT (ptr), FO) : hi + 1) - lo; + i++) + { + D_RELNUM (result, FILE0 + d, i) = D_RELNUM (result, FILEC, linec); + D_RELLEN (result, FILE0 + d, i) = D_RELLEN (result, FILEC, linec); + linec++; + } + } + } + + /* Set correspond */ + if (!using[0]) + D3_TYPE (result) = DIFF_2ND; + else if (!using[1]) + D3_TYPE (result) = DIFF_1ST; + else + { + lin nl0 = D_NUMLINES (result, FILE0); + lin nl1 = D_NUMLINES (result, FILE1); + + if (nl0 != nl1 + || !compare_line_list (D_LINEARRAY (result, FILE0), + D_LENARRAY (result, FILE0), + D_LINEARRAY (result, FILE1), + D_LENARRAY (result, FILE1), + nl0)) + D3_TYPE (result) = DIFF_ALL; + else + D3_TYPE (result) = DIFF_3RD; + } + + return result; +} + +/* Copy pointers from a list of strings to a different list of + strings. If a spot in the second list is already filled, make sure + that it is filled with the same string; if not, return false, the copy + incomplete. Upon successful completion of the copy, return true. */ + +static bool +copy_stringlist (char * const fromptrs[], size_t const fromlengths[], + char *toptrs[], size_t tolengths[], + lin copynum) +{ + register char * const *f = fromptrs; + register char **t = toptrs; + register size_t const *fl = fromlengths; + register size_t *tl = tolengths; + + while (copynum--) + { + if (*t) + { + if (*fl != *tl || memcmp (*f, *t, *fl) != 0) + return false; + } + else + { + *t = *f; + *tl = *fl; + } + + t++; f++; tl++; fl++; + } + + return true; +} + +/* Create a diff3_block, with ranges as specified in the arguments. + Allocate the arrays for the various pointers (and zero them) based + on the arguments passed. Return the block as a result. */ + +static struct diff3_block * +create_diff3_block (lin low0, lin high0, + lin low1, lin high1, + lin low2, lin high2) +{ + struct diff3_block *result = xmalloc (sizeof *result); + lin numlines; + + D3_TYPE (result) = ERROR; + D_NEXT (result) = 0; + + /* Assign ranges */ + D_LOWLINE (result, FILE0) = low0; + D_HIGHLINE (result, FILE0) = high0; + D_LOWLINE (result, FILE1) = low1; + D_HIGHLINE (result, FILE1) = high1; + D_LOWLINE (result, FILE2) = low2; + D_HIGHLINE (result, FILE2) = high2; + + /* Allocate and zero space */ + numlines = D_NUMLINES (result, FILE0); + if (numlines) + { + D_LINEARRAY (result, FILE0) = xcalloc (numlines, sizeof (char *)); + D_LENARRAY (result, FILE0) = xcalloc (numlines, sizeof (size_t)); + } + else + { + D_LINEARRAY (result, FILE0) = 0; + D_LENARRAY (result, FILE0) = 0; + } + + numlines = D_NUMLINES (result, FILE1); + if (numlines) + { + D_LINEARRAY (result, FILE1) = xcalloc (numlines, sizeof (char *)); + D_LENARRAY (result, FILE1) = xcalloc (numlines, sizeof (size_t)); + } + else + { + D_LINEARRAY (result, FILE1) = 0; + D_LENARRAY (result, FILE1) = 0; + } + + numlines = D_NUMLINES (result, FILE2); + if (numlines) + { + D_LINEARRAY (result, FILE2) = xcalloc (numlines, sizeof (char *)); + D_LENARRAY (result, FILE2) = xcalloc (numlines, sizeof (size_t)); + } + else + { + D_LINEARRAY (result, FILE2) = 0; + D_LENARRAY (result, FILE2) = 0; + } + + /* Return */ + return result; +} + +/* Compare two lists of lines of text. + Return 1 if they are equivalent, 0 if not. */ + +static bool +compare_line_list (char * const list1[], size_t const lengths1[], + char * const list2[], size_t const lengths2[], + lin nl) +{ + char * const *l1 = list1; + char * const *l2 = list2; + size_t const *lgths1 = lengths1; + size_t const *lgths2 = lengths2; + + while (nl--) + if (!*l1 || !*l2 || *lgths1 != *lgths2++ + || memcmp (*l1++, *l2++, *lgths1++) != 0) + return false; + return true; +} + +/* Input and parse two way diffs. */ + +static struct diff_block * +process_diff (char const *filea, + char const *fileb, + struct diff_block **last_block) +{ + char *diff_contents; + char *diff_limit; + char *scan_diff; + enum diff_type dt; + lin i; + struct diff_block *block_list, **block_list_end, *bptr; + size_t too_many_lines = (PTRDIFF_MAX + / MIN (sizeof *bptr->lines[1], + sizeof *bptr->lengths[1])); + + diff_limit = read_diff (filea, fileb, &diff_contents); + scan_diff = diff_contents; + block_list_end = &block_list; + bptr = 0; /* Pacify `gcc -W'. */ + + while (scan_diff < diff_limit) + { + bptr = xmalloc (sizeof *bptr); + bptr->lines[0] = bptr->lines[1] = 0; + bptr->lengths[0] = bptr->lengths[1] = 0; + + dt = process_diff_control (&scan_diff, bptr); + if (dt == ERROR || *scan_diff != '\n') + { + fprintf (stderr, _("%s: diff failed: "), program_name); + do + { + putc (*scan_diff, stderr); + } + while (*scan_diff++ != '\n'); + exit (EXIT_TROUBLE); + } + scan_diff++; + + /* Force appropriate ranges to be null, if necessary */ + switch (dt) + { + case ADD: + bptr->ranges[0][0]++; + break; + case DELETE: + bptr->ranges[1][0]++; + break; + case CHANGE: + break; + default: + fatal ("internal error: invalid diff type in process_diff"); + break; + } + + /* Allocate space for the pointers for the lines from filea, and + parcel them out among these pointers */ + if (dt != ADD) + { + lin numlines = D_NUMLINES (bptr, 0); + if (too_many_lines <= numlines) + xalloc_die (); + bptr->lines[0] = xmalloc (numlines * sizeof *bptr->lines[0]); + bptr->lengths[0] = xmalloc (numlines * sizeof *bptr->lengths[0]); + for (i = 0; i < numlines; i++) + scan_diff = scan_diff_line (scan_diff, + &(bptr->lines[0][i]), + &(bptr->lengths[0][i]), + diff_limit, + '<'); + } + + /* Get past the separator for changes */ + if (dt == CHANGE) + { + if (strncmp (scan_diff, "---\n", 4)) + fatal ("invalid diff format; invalid change separator"); + scan_diff += 4; + } + + /* Allocate space for the pointers for the lines from fileb, and + parcel them out among these pointers */ + if (dt != DELETE) + { + lin numlines = D_NUMLINES (bptr, 1); + if (too_many_lines <= numlines) + xalloc_die (); + bptr->lines[1] = xmalloc (numlines * sizeof *bptr->lines[1]); + bptr->lengths[1] = xmalloc (numlines * sizeof *bptr->lengths[1]); + for (i = 0; i < numlines; i++) + scan_diff = scan_diff_line (scan_diff, + &(bptr->lines[1][i]), + &(bptr->lengths[1][i]), + diff_limit, + '>'); + } + + /* Place this block on the blocklist. */ + *block_list_end = bptr; + block_list_end = &bptr->next; + } + + *block_list_end = 0; + *last_block = bptr; + return block_list; +} + +/* Skip tabs and spaces, and return the first character after them. */ + +static char * +skipwhite (char *s) +{ + while (*s == ' ' || *s == '\t') + s++; + return s; +} + +/* Read a nonnegative line number from S, returning the address of the + first character after the line number, and storing the number into + *PNUM. Return 0 if S does not point to a valid line number. */ + +static char * +readnum (char *s, lin *pnum) +{ + unsigned char c = *s; + lin num = 0; + + if (! ISDIGIT (c)) + return 0; + + do + { + num = c - '0' + num * 10; + c = *++s; + } + while (ISDIGIT (c)); + + *pnum = num; + return s; +} + +/* Parse a normal format diff control string. Return the type of the + diff (ERROR if the format is bad). All of the other important + information is filled into to the structure pointed to by db, and + the string pointer (whose location is passed to this routine) is + updated to point beyond the end of the string parsed. Note that + only the ranges in the diff_block will be set by this routine. + + If some specific pair of numbers has been reduced to a single + number, then both corresponding numbers in the diff block are set + to that number. In general these numbers are interpreted as ranges + inclusive, unless being used by the ADD or DELETE commands. It is + assumed that these will be special cased in a superior routine. */ + +static enum diff_type +process_diff_control (char **string, struct diff_block *db) +{ + char *s = *string; + enum diff_type type; + + /* Read first set of digits */ + s = readnum (skipwhite (s), &db->ranges[0][RANGE_START]); + if (! s) + return ERROR; + + /* Was that the only digit? */ + s = skipwhite (s); + if (*s == ',') + { + s = readnum (s + 1, &db->ranges[0][RANGE_END]); + if (! s) + return ERROR; + } + else + db->ranges[0][RANGE_END] = db->ranges[0][RANGE_START]; + + /* Get the letter */ + s = skipwhite (s); + switch (*s) + { + case 'a': + type = ADD; + break; + case 'c': + type = CHANGE; + break; + case 'd': + type = DELETE; + break; + default: + return ERROR; /* Bad format */ + } + s++; /* Past letter */ + + /* Read second set of digits */ + s = readnum (skipwhite (s), &db->ranges[1][RANGE_START]); + if (! s) + return ERROR; + + /* Was that the only digit? */ + s = skipwhite (s); + if (*s == ',') + { + s = readnum (s + 1, &db->ranges[1][RANGE_END]); + if (! s) + return ERROR; + s = skipwhite (s); /* To move to end */ + } + else + db->ranges[1][RANGE_END] = db->ranges[1][RANGE_START]; + + *string = s; + return type; +} + +static char * +read_diff (char const *filea, + char const *fileb, + char **output_placement) +{ + char *diff_result; + size_t current_chunk_size, total; + int fd, wstatus, status; + int werrno = 0; + struct stat pipestat; + +#if HAVE_WORKING_FORK || HAVE_WORKING_VFORK + + char const *argv[9]; + char const **ap; + int fds[2]; + pid_t pid; + + ap = argv; + *ap++ = diff_program; + if (text) + *ap++ = "-a"; + if (strip_trailing_cr) + *ap++ = "--strip-trailing-cr"; + *ap++ = "--horizon-lines=100"; + *ap++ = "--"; + *ap++ = filea; + *ap++ = fileb; + *ap = 0; + + if (pipe (fds) != 0) + perror_with_exit ("pipe"); + + pid = vfork (); + if (pid == 0) + { + /* Child */ + close (fds[0]); + if (fds[1] != STDOUT_FILENO) + { + dup2 (fds[1], STDOUT_FILENO); + close (fds[1]); + } + + /* The cast to (char **) is needed for portability to older + hosts with a nonstandard prototype for execvp. */ + execvp (diff_program, (char **) argv); + + _exit (errno == ENOENT ? 127 : 126); + } + + if (pid == -1) + perror_with_exit ("fork"); + + close (fds[1]); /* Prevent erroneous lack of EOF */ + fd = fds[0]; + +#else + + FILE *fpipe; + char const args[] = " --horizon-lines=100 -- "; + char *command = xmalloc (quote_system_arg (0, diff_program) + + sizeof "-a" + + sizeof "--strip-trailing-cr" + + sizeof args - 1 + + quote_system_arg (0, filea) + 1 + + quote_system_arg (0, fileb) + 1); + char *p = command; + p += quote_system_arg (p, diff_program); + if (text) + { + strcpy (p, " -a"); + p += 3; + } + if (strip_trailing_cr) + { + strcpy (p, " --strip-trailing-cr"); + p += 20; + } + strcpy (p, args); + p += sizeof args - 1; + p += quote_system_arg (p, filea); + *p++ = ' '; + p += quote_system_arg (p, fileb); + *p = 0; + errno = 0; + fpipe = popen (command, "r"); + if (!fpipe) + perror_with_exit (command); + free (command); + fd = fileno (fpipe); + +#endif + + if (fstat (fd, &pipestat) != 0) + perror_with_exit ("fstat"); + current_chunk_size = MAX (1, STAT_BLOCKSIZE (pipestat)); + diff_result = xmalloc (current_chunk_size); + total = 0; + + for (;;) + { + size_t bytes_to_read = current_chunk_size - total; + size_t bytes = block_read (fd, diff_result + total, bytes_to_read); + total += bytes; + if (bytes != bytes_to_read) + { + if (bytes == SIZE_MAX) + perror_with_exit (_("read failed")); + break; + } + if (PTRDIFF_MAX / 2 <= current_chunk_size) + xalloc_die (); + current_chunk_size *= 2; + diff_result = xrealloc (diff_result, current_chunk_size); + } + + if (total != 0 && diff_result[total-1] != '\n') + fatal ("invalid diff format; incomplete last line"); + + *output_placement = diff_result; + +#if ! (HAVE_WORKING_FORK || HAVE_WORKING_VFORK) + + wstatus = pclose (fpipe); + if (wstatus == -1) + werrno = errno; + +#else + + if (close (fd) != 0) + perror_with_exit ("close"); + if (waitpid (pid, &wstatus, 0) < 0) + perror_with_exit ("waitpid"); + +#endif + + status = ! werrno && WIFEXITED (wstatus) ? WEXITSTATUS (wstatus) : INT_MAX; + + if (EXIT_TROUBLE <= status) + error (EXIT_TROUBLE, werrno, + _(status == 126 + ? "subsidiary program `%s' could not be invoked" + : status == 127 + ? "subsidiary program `%s' not found" + : status == INT_MAX + ? "subsidiary program `%s' failed" + : "subsidiary program `%s' failed (exit status %d)"), + diff_program, status); + + return diff_result + total; +} + + +/* Scan a regular diff line (consisting of > or <, followed by a + space, followed by text (including nulls) up to a newline. + + This next routine began life as a macro and many parameters in it + are used as call-by-reference values. */ +static char * +scan_diff_line (char *scan_ptr, char **set_start, size_t *set_length, + char *limit, char leadingchar) +{ + char *line_ptr; + + if (!(scan_ptr[0] == leadingchar + && scan_ptr[1] == ' ')) + fatal ("invalid diff format; incorrect leading line chars"); + + *set_start = line_ptr = scan_ptr + 2; + while (*line_ptr++ != '\n') + continue; + + /* Include newline if the original line ended in a newline, + or if an edit script is being generated. + Copy any missing newline message to stderr if an edit script is being + generated, because edit scripts cannot handle missing newlines. + Return the beginning of the next line. */ + *set_length = line_ptr - *set_start; + if (line_ptr < limit && *line_ptr == '\\') + { + if (edscript) + fprintf (stderr, "%s:", program_name); + else + --*set_length; + line_ptr++; + do + { + if (edscript) + putc (*line_ptr, stderr); + } + while (*line_ptr++ != '\n'); + } + + return line_ptr; +} + +/* Output a three way diff passed as a list of diff3_block's. The + argument MAPPING is indexed by external file number (in the + argument list) and contains the internal file number (from the diff + passed). This is important because the user expects outputs in + terms of the argument list number, and the diff passed may have + been done slightly differently (if the last argument was "-", for + example). REV_MAPPING is the inverse of MAPPING. */ + +static void +output_diff3 (FILE *outputfile, struct diff3_block *diff, + int const mapping[3], int const rev_mapping[3]) +{ + int i; + int oddoneout; + char *cp; + struct diff3_block *ptr; + lin line; + size_t length; + int dontprint; + static int skew_increment[3] = { 2, 3, 1 }; /* 0==>2==>1==>3 */ + char const *line_prefix = initial_tab ? "\t" : " "; + + for (ptr = diff; ptr; ptr = D_NEXT (ptr)) + { + char x[2]; + + switch (ptr->correspond) + { + case DIFF_ALL: + x[0] = 0; + dontprint = 3; /* Print them all */ + oddoneout = 3; /* Nobody's odder than anyone else */ + break; + case DIFF_1ST: + case DIFF_2ND: + case DIFF_3RD: + oddoneout = rev_mapping[ptr->correspond - DIFF_1ST]; + + x[0] = oddoneout + '1'; + x[1] = 0; + dontprint = oddoneout == 0; + break; + default: + fatal ("internal error: invalid diff type passed to output"); + } + fprintf (outputfile, "====%s\n", x); + + /* Go 0, 2, 1 if the first and third outputs are equivalent. */ + for (i = 0; i < 3; + i = (oddoneout == 1 ? skew_increment[i] : i + 1)) + { + int realfile = mapping[i]; + lin lowt = D_LOWLINE (ptr, realfile); + lin hight = D_HIGHLINE (ptr, realfile); + long int llowt = lowt; + long int lhight = hight; + + fprintf (outputfile, "%d:", i + 1); + switch (lowt - hight) + { + case 1: + fprintf (outputfile, "%lda\n", llowt - 1); + break; + case 0: + fprintf (outputfile, "%ldc\n", llowt); + break; + default: + fprintf (outputfile, "%ld,%ldc\n", llowt, lhight); + break; + } + + if (i == dontprint) continue; + + if (lowt <= hight) + { + line = 0; + do + { + fprintf (outputfile, line_prefix); + cp = D_RELNUM (ptr, realfile, line); + length = D_RELLEN (ptr, realfile, line); + fwrite (cp, sizeof (char), length, outputfile); + } + while (++line < hight - lowt + 1); + if (cp[length - 1] != '\n') + fprintf (outputfile, "\n\\ %s\n", + _("No newline at end of file")); + } + } + } +} + + +/* Output to OUTPUTFILE the lines of B taken from FILENUM. Double any + initial '.'s; yield nonzero if any initial '.'s were doubled. */ + +static bool +dotlines (FILE *outputfile, struct diff3_block *b, int filenum) +{ + lin i; + bool leading_dot = false; + + for (i = 0; + i < D_NUMLINES (b, filenum); + i++) + { + char *line = D_RELNUM (b, filenum, i); + if (line[0] == '.') + { + leading_dot = true; + fprintf (outputfile, "."); + } + fwrite (line, sizeof (char), + D_RELLEN (b, filenum, i), outputfile); + } + + return leading_dot; +} + +/* Output to OUTPUTFILE a '.' line. If LEADING_DOT is true, also + output a command that removes initial '.'s starting with line START + and continuing for NUM lines. (START is long int, not lin, for + convenience with printf %ld formats.) */ + +static void +undotlines (FILE *outputfile, bool leading_dot, long int start, lin num) +{ + fprintf (outputfile, ".\n"); + if (leading_dot) + { + if (num == 1) + fprintf (outputfile, "%lds/^\\.//\n", start); + else + fprintf (outputfile, "%ld,%lds/^\\.//\n", start, start + num - 1); + } +} + +/* Output a diff3 set of blocks as an ed script. This script applies + the changes between file's 2 & 3 to file 1. Take the precise + format of the ed script to be output from global variables set + during options processing. Reverse the order of + the set of diff3 blocks in DIFF; this gets + around the problems involved with changing line numbers in an ed + script. + + As in `output_diff3', the variable MAPPING maps from file number + according to the argument list to file number according to the diff + passed. All files listed below are in terms of the argument list. + REV_MAPPING is the inverse of MAPPING. + + FILE0, FILE1 and FILE2 are the strings to print as the names of the + three files. These may be the actual names, or may be the + arguments specified with -L. + + Return 1 if conflicts were found. */ + +static bool +output_diff3_edscript (FILE *outputfile, struct diff3_block *diff, + int const mapping[3], int const rev_mapping[3], + char const *file0, char const *file1, char const *file2) +{ + bool leading_dot; + bool conflicts_found = false; + bool conflict; + struct diff3_block *b; + + for (b = reverse_diff3_blocklist (diff); b; b = b->next) + { + /* Must do mapping correctly. */ + enum diff_type type + = (b->correspond == DIFF_ALL + ? DIFF_ALL + : DIFF_1ST + rev_mapping[b->correspond - DIFF_1ST]); + + long int low0, high0; + + /* If we aren't supposed to do this output block, skip it. */ + switch (type) + { + default: continue; + case DIFF_2ND: if (!show_2nd) continue; conflict = true; break; + case DIFF_3RD: if (overlap_only) continue; conflict = false; break; + case DIFF_ALL: if (simple_only) continue; conflict = flagging; break; + } + + low0 = D_LOWLINE (b, mapping[FILE0]); + high0 = D_HIGHLINE (b, mapping[FILE0]); + + if (conflict) + { + conflicts_found = true; + + + /* Mark end of conflict. */ + + fprintf (outputfile, "%lda\n", high0); + leading_dot = false; + if (type == DIFF_ALL) + { + if (show_2nd) + { + /* Append lines from FILE1. */ + fprintf (outputfile, "||||||| %s\n", file1); + leading_dot = dotlines (outputfile, b, mapping[FILE1]); + } + /* Append lines from FILE2. */ + fprintf (outputfile, "=======\n"); + leading_dot |= dotlines (outputfile, b, mapping[FILE2]); + } + fprintf (outputfile, ">>>>>>> %s\n", file2); + undotlines (outputfile, leading_dot, high0 + 2, + (D_NUMLINES (b, mapping[FILE1]) + + D_NUMLINES (b, mapping[FILE2]) + 1)); + + + /* Mark start of conflict. */ + + fprintf (outputfile, "%lda\n<<<<<<< %s\n", low0 - 1, + type == DIFF_ALL ? file0 : file1); + leading_dot = false; + if (type == DIFF_2ND) + { + /* Prepend lines from FILE1. */ + leading_dot = dotlines (outputfile, b, mapping[FILE1]); + fprintf (outputfile, "=======\n"); + } + undotlines (outputfile, leading_dot, low0 + 1, + D_NUMLINES (b, mapping[FILE1])); + } + else if (D_NUMLINES (b, mapping[FILE2]) == 0) + /* Write out a delete */ + { + if (low0 == high0) + fprintf (outputfile, "%ldd\n", low0); + else + fprintf (outputfile, "%ld,%ldd\n", low0, high0); + } + else + /* Write out an add or change */ + { + switch (high0 - low0) + { + case -1: + fprintf (outputfile, "%lda\n", high0); + break; + case 0: + fprintf (outputfile, "%ldc\n", high0); + break; + default: + fprintf (outputfile, "%ld,%ldc\n", low0, high0); + break; + } + + undotlines (outputfile, dotlines (outputfile, b, mapping[FILE2]), + low0, D_NUMLINES (b, mapping[FILE2])); + } + } + if (finalwrite) fprintf (outputfile, "w\nq\n"); + return conflicts_found; +} + +/* Read from INFILE and output to OUTPUTFILE a set of diff3_blocks + DIFF as a merged file. This acts like 'ed file0 + <[output_diff3_edscript]', except that it works even for binary + data or incomplete lines. + + As before, MAPPING maps from arg list file number to diff file + number, REV_MAPPING is its inverse, and FILE0, FILE1, and FILE2 are + the names of the files. + + Return 1 if conflicts were found. */ + +static bool +output_diff3_merge (FILE *infile, FILE *outputfile, struct diff3_block *diff, + int const mapping[3], int const rev_mapping[3], + char const *file0, char const *file1, char const *file2) +{ + int c; + lin i; + bool conflicts_found = false; + bool conflict; + struct diff3_block *b; + lin linesread = 0; + + for (b = diff; b; b = b->next) + { + /* Must do mapping correctly. */ + enum diff_type type + = ((b->correspond == DIFF_ALL) + ? DIFF_ALL + : DIFF_1ST + rev_mapping[b->correspond - DIFF_1ST]); + char const *format_2nd = "<<<<<<< %s\n"; + + /* If we aren't supposed to do this output block, skip it. */ + switch (type) + { + default: continue; + case DIFF_2ND: if (!show_2nd) continue; conflict = true; break; + case DIFF_3RD: if (overlap_only) continue; conflict = false; break; + case DIFF_ALL: if (simple_only) continue; conflict = flagging; + format_2nd = "||||||| %s\n"; + break; + } + + /* Copy I lines from file 0. */ + i = D_LOWLINE (b, FILE0) - linesread - 1; + linesread += i; + while (0 <= --i) + do + { + c = getc (infile); + if (c == EOF) + { + if (ferror (infile)) + perror_with_exit (_("read failed")); + else if (feof (infile)) + fatal ("input file shrank"); + } + putc (c, outputfile); + } + while (c != '\n'); + + if (conflict) + { + conflicts_found = true; + + if (type == DIFF_ALL) + { + /* Put in lines from FILE0 with bracket. */ + fprintf (outputfile, "<<<<<<< %s\n", file0); + for (i = 0; + i < D_NUMLINES (b, mapping[FILE0]); + i++) + fwrite (D_RELNUM (b, mapping[FILE0], i), sizeof (char), + D_RELLEN (b, mapping[FILE0], i), outputfile); + } + + if (show_2nd) + { + /* Put in lines from FILE1 with bracket. */ + fprintf (outputfile, format_2nd, file1); + for (i = 0; + i < D_NUMLINES (b, mapping[FILE1]); + i++) + fwrite (D_RELNUM (b, mapping[FILE1], i), sizeof (char), + D_RELLEN (b, mapping[FILE1], i), outputfile); + } + + fprintf (outputfile, "=======\n"); + } + + /* Put in lines from FILE2. */ + for (i = 0; + i < D_NUMLINES (b, mapping[FILE2]); + i++) + fwrite (D_RELNUM (b, mapping[FILE2], i), sizeof (char), + D_RELLEN (b, mapping[FILE2], i), outputfile); + + if (conflict) + fprintf (outputfile, ">>>>>>> %s\n", file2); + + /* Skip I lines in file 0. */ + i = D_NUMLINES (b, FILE0); + linesread += i; + while (0 <= --i) + while ((c = getc (infile)) != '\n') + if (c == EOF) + { + if (ferror (infile)) + perror_with_exit (_("read failed")); + else if (feof (infile)) + { + if (i || b->next) + fatal ("input file shrank"); + return conflicts_found; + } + } + } + /* Copy rest of common file. */ + while ((c = getc (infile)) != EOF || !(ferror (infile) | feof (infile))) + putc (c, outputfile); + return conflicts_found; +} + +/* Reverse the order of the list of diff3 blocks. */ + +static struct diff3_block * +reverse_diff3_blocklist (struct diff3_block *diff) +{ + register struct diff3_block *tmp, *next, *prev; + + for (tmp = diff, prev = 0; tmp; tmp = next) + { + next = tmp->next; + tmp->next = prev; + prev = tmp; + } + + return prev; +} + +static void +fatal (char const *msgid) +{ + error (EXIT_TROUBLE, 0, "%s", _(msgid)); + abort (); +} + +static void +perror_with_exit (char const *string) +{ + error (EXIT_TROUBLE, errno, "%s", string); + abort (); +} diff --git a/contrib/diff/src/dir.c b/contrib/diff/src/dir.c new file mode 100644 index 0000000..cc0c8793 --- /dev/null +++ b/contrib/diff/src/dir.c @@ -0,0 +1,288 @@ +/* Read, sort and compare two directories. Used for GNU DIFF. + + Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1998, 2001, 2002, + 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "diff.h" +#include +#include +#include +#include +#include + +/* Read the directory named by DIR and store into DIRDATA a sorted vector + of filenames for its contents. DIR->desc == -1 means this directory is + known to be nonexistent, so set DIRDATA to an empty vector. + Return -1 (setting errno) if error, 0 otherwise. */ + +struct dirdata +{ + size_t nnames; /* Number of names. */ + char const **names; /* Sorted names of files in dir, followed by 0. */ + char *data; /* Allocated storage for file names. */ +}; + +/* Whether file names in directories should be compared with + locale-specific sorting. */ +static bool locale_specific_sorting; + +/* Where to go if locale-specific sorting fails. */ +static jmp_buf failed_locale_specific_sorting; + +static bool dir_loop (struct comparison const *, int); +static int compare_names_for_qsort (void const *, void const *); + + +/* Read a directory and get its vector of names. */ + +static bool +dir_read (struct file_data const *dir, struct dirdata *dirdata) +{ + register struct dirent *next; + register size_t i; + + /* Address of block containing the files that are described. */ + char const **names; + + /* Number of files in directory. */ + size_t nnames; + + /* Allocated and used storage for file name data. */ + char *data; + size_t data_alloc, data_used; + + dirdata->names = 0; + dirdata->data = 0; + nnames = 0; + data = 0; + + if (dir->desc != -1) + { + /* Open the directory and check for errors. */ + register DIR *reading = opendir (dir->name); + if (!reading) + return false; + + /* Initialize the table of filenames. */ + + data_alloc = 512; + data_used = 0; + dirdata->data = data = xmalloc (data_alloc); + + /* Read the directory entries, and insert the subfiles + into the `data' table. */ + + while ((errno = 0, (next = readdir (reading)) != 0)) + { + char *d_name = next->d_name; + size_t d_size = NAMLEN (next) + 1; + + /* Ignore "." and "..". */ + if (d_name[0] == '.' + && (d_name[1] == 0 || (d_name[1] == '.' && d_name[2] == 0))) + continue; + + if (excluded_filename (excluded, d_name)) + continue; + + while (data_alloc < data_used + d_size) + { + if (PTRDIFF_MAX / 2 <= data_alloc) + xalloc_die (); + dirdata->data = data = xrealloc (data, data_alloc *= 2); + } + + memcpy (data + data_used, d_name, d_size); + data_used += d_size; + nnames++; + } + if (errno) + { + int e = errno; + closedir (reading); + errno = e; + return false; + } +#if CLOSEDIR_VOID + closedir (reading); +#else + if (closedir (reading) != 0) + return false; +#endif + } + + /* Create the `names' table from the `data' table. */ + if (PTRDIFF_MAX / sizeof *names - 1 <= nnames) + xalloc_die (); + dirdata->names = names = xmalloc ((nnames + 1) * sizeof *names); + dirdata->nnames = nnames; + for (i = 0; i < nnames; i++) + { + names[i] = data; + data += strlen (data) + 1; + } + names[nnames] = 0; + return true; +} + +/* Compare file names, returning a value compatible with strcmp. */ + +static int +compare_names (char const *name1, char const *name2) +{ + if (locale_specific_sorting) + { + int r; + errno = 0; + if (ignore_file_name_case) + r = strcasecoll (name1, name2); + else + r = strcoll (name1, name2); + if (errno) + { + error (0, errno, _("cannot compare file names `%s' and `%s'"), + name1, name2); + longjmp (failed_locale_specific_sorting, 1); + } + return r; + } + + return (ignore_file_name_case + ? strcasecmp (name1, name2) + : file_name_cmp (name1, name2)); +} + +/* A wrapper for compare_names suitable as an argument for qsort. */ + +static int +compare_names_for_qsort (void const *file1, void const *file2) +{ + char const *const *f1 = file1; + char const *const *f2 = file2; + return compare_names (*f1, *f2); +} + +/* Compare the contents of two directories named in CMP. + This is a top-level routine; it does everything necessary for diff + on two directories. + + CMP->file[0].desc == -1 says directory CMP->file[0] doesn't exist, + but pretend it is empty. Likewise for CMP->file[1]. + + HANDLE_FILE is a caller-provided subroutine called to handle each file. + It gets three operands: CMP, name of file in dir 0, name of file in dir 1. + These names are relative to the original working directory. + + For a file that appears in only one of the dirs, one of the name-args + to HANDLE_FILE is zero. + + Returns the maximum of all the values returned by HANDLE_FILE, + or EXIT_TROUBLE if trouble is encountered in opening files. */ + +int +diff_dirs (struct comparison const *cmp, + int (*handle_file) (struct comparison const *, + char const *, char const *)) +{ + struct dirdata dirdata[2]; + int volatile val = EXIT_SUCCESS; + int i; + + if ((cmp->file[0].desc == -1 || dir_loop (cmp, 0)) + && (cmp->file[1].desc == -1 || dir_loop (cmp, 1))) + { + error (0, 0, "%s: recursive directory loop", + cmp->file[cmp->file[0].desc == -1].name); + return EXIT_TROUBLE; + } + + /* Get contents of both dirs. */ + for (i = 0; i < 2; i++) + if (! dir_read (&cmp->file[i], &dirdata[i])) + { + perror_with_name (cmp->file[i].name); + val = EXIT_TROUBLE; + } + + if (val == EXIT_SUCCESS) + { + char const **volatile names[2]; + names[0] = dirdata[0].names; + names[1] = dirdata[1].names; + + /* Use locale-specific sorting if possible, else native byte order. */ + locale_specific_sorting = true; + if (setjmp (failed_locale_specific_sorting)) + locale_specific_sorting = false; + + /* Sort the directories. */ + for (i = 0; i < 2; i++) + qsort (names[i], dirdata[i].nnames, sizeof *dirdata[i].names, + compare_names_for_qsort); + + /* If `-S name' was given, and this is the topmost level of comparison, + ignore all file names less than the specified starting name. */ + + if (starting_file && ! cmp->parent) + { + while (*names[0] && compare_names (*names[0], starting_file) < 0) + names[0]++; + while (*names[1] && compare_names (*names[1], starting_file) < 0) + names[1]++; + } + + /* Loop while files remain in one or both dirs. */ + while (*names[0] || *names[1]) + { + /* Compare next name in dir 0 with next name in dir 1. + At the end of a dir, + pretend the "next name" in that dir is very large. */ + int nameorder = (!*names[0] ? 1 : !*names[1] ? -1 + : compare_names (*names[0], *names[1])); + int v1 = (*handle_file) (cmp, + 0 < nameorder ? 0 : *names[0]++, + nameorder < 0 ? 0 : *names[1]++); + if (val < v1) + val = v1; + } + } + + for (i = 0; i < 2; i++) + { + if (dirdata[i].names) + free (dirdata[i].names); + if (dirdata[i].data) + free (dirdata[i].data); + } + + return val; +} + +/* Return nonzero if CMP is looping recursively in argument I. */ + +static bool +dir_loop (struct comparison const *cmp, int i) +{ + struct comparison const *p = cmp; + while ((p = p->parent)) + if (0 < same_file (&p->file[i].stat, &cmp->file[i].stat)) + return true; + return false; +} diff --git a/contrib/diff/src/ed.c b/contrib/diff/src/ed.c new file mode 100644 index 0000000..c90ed78 --- /dev/null +++ b/contrib/diff/src/ed.c @@ -0,0 +1,169 @@ +/* Output routines for ed-script format. + + Copyright (C) 1988, 1989, 1991, 1992, 1993, 1995, 1998, 2001, 2004 + Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "diff.h" + +static void print_ed_hunk (struct change *); +static void print_rcs_hunk (struct change *); +static void pr_forward_ed_hunk (struct change *); + +/* Print our script as ed commands. */ + +void +print_ed_script (struct change *script) +{ + print_script (script, find_reverse_change, print_ed_hunk); +} + +/* Print a hunk of an ed diff */ + +static void +print_ed_hunk (struct change *hunk) +{ + lin f0, l0, f1, l1; + enum changes changes; + +#ifdef DEBUG + debug_script (hunk); +#endif + + /* Determine range of line numbers involved in each file. */ + changes = analyze_hunk (hunk, &f0, &l0, &f1, &l1); + if (!changes) + return; + + begin_output (); + + /* Print out the line number header for this hunk */ + print_number_range (',', &files[0], f0, l0); + fprintf (outfile, "%c\n", change_letter[changes]); + + /* Print new/changed lines from second file, if needed */ + if (changes != OLD) + { + lin i; + for (i = f1; i <= l1; i++) + { + if (files[1].linbuf[i][0] == '.' && files[1].linbuf[i][1] == '\n') + { + /* The file's line is just a dot, and it would exit + insert mode. Precede the dot with another dot, exit + insert mode, remove the extra dot, and then resume + insert mode. */ + fprintf (outfile, "..\n.\ns/.//\na\n"); + } + else + print_1_line ("", &files[1].linbuf[i]); + } + + fprintf (outfile, ".\n"); + } +} + +/* Print change script in the style of ed commands, + but print the changes in the order they appear in the input files, + which means that the commands are not truly useful with ed. */ + +void +pr_forward_ed_script (struct change *script) +{ + print_script (script, find_change, pr_forward_ed_hunk); +} + +static void +pr_forward_ed_hunk (struct change *hunk) +{ + lin i, f0, l0, f1, l1; + + /* Determine range of line numbers involved in each file. */ + enum changes changes = analyze_hunk (hunk, &f0, &l0, &f1, &l1); + if (!changes) + return; + + begin_output (); + + fprintf (outfile, "%c", change_letter[changes]); + print_number_range (' ', files, f0, l0); + fprintf (outfile, "\n"); + + /* If deletion only, print just the number range. */ + + if (changes == OLD) + return; + + /* For insertion (with or without deletion), print the number range + and the lines from file 2. */ + + for (i = f1; i <= l1; i++) + print_1_line ("", &files[1].linbuf[i]); + + fprintf (outfile, ".\n"); +} + +/* Print in a format somewhat like ed commands + except that each insert command states the number of lines it inserts. + This format is used for RCS. */ + +void +print_rcs_script (struct change *script) +{ + print_script (script, find_change, print_rcs_hunk); +} + +/* Print a hunk of an RCS diff */ + +static void +print_rcs_hunk (struct change *hunk) +{ + lin i, f0, l0, f1, l1; + long int tf0, tl0, tf1, tl1; + + /* Determine range of line numbers involved in each file. */ + enum changes changes = analyze_hunk (hunk, &f0, &l0, &f1, &l1); + if (!changes) + return; + + begin_output (); + + translate_range (&files[0], f0, l0, &tf0, &tl0); + + if (changes & OLD) + { + fprintf (outfile, "d"); + /* For deletion, print just the starting line number from file 0 + and the number of lines deleted. */ + fprintf (outfile, "%ld %ld\n", tf0, tf0 <= tl0 ? tl0 - tf0 + 1 : 1); + } + + if (changes & NEW) + { + fprintf (outfile, "a"); + + /* Take last-line-number from file 0 and # lines from file 1. */ + translate_range (&files[1], f1, l1, &tf1, &tl1); + fprintf (outfile, "%ld %ld\n", tl0, tf1 <= tl1 ? tl1 - tf1 + 1 : 1); + + /* Print the inserted lines. */ + for (i = f1; i <= l1; i++) + print_1_line ("", &files[1].linbuf[i]); + } +} diff --git a/contrib/diff/src/ifdef.c b/contrib/diff/src/ifdef.c new file mode 100644 index 0000000..22f1cb1 --- /dev/null +++ b/contrib/diff/src/ifdef.c @@ -0,0 +1,430 @@ +/* #ifdef-format output routines for GNU DIFF. + + Copyright (C) 1989, 1991, 1992, 1993, 1994, 2001, 2002, 2004 Free + Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY. No author or distributor + accepts responsibility to anyone for the consequences of using it + or for whether it serves any particular purpose or works at all, + unless he says so in writing. Refer to the GNU DIFF General Public + License for full details. + + Everyone is granted permission to copy, modify and redistribute + GNU DIFF, but only under the conditions described in the + GNU DIFF General Public License. A copy of this license is + supposed to have been given to you along with GNU DIFF so you + can know your rights and responsibilities. It should be in a + file named COPYING. Among other things, the copyright notice + and this notice must be preserved on all copies. */ + +#include "diff.h" + +#include + +struct group +{ + struct file_data const *file; + lin from, upto; /* start and limit lines for this group of lines */ +}; + +static char const *format_group (FILE *, char const *, char, + struct group const *); +static char const *do_printf_spec (FILE *, char const *, + struct file_data const *, lin, + struct group const *); +static char const *scan_char_literal (char const *, char *); +static lin groups_letter_value (struct group const *, char); +static void format_ifdef (char const *, lin, lin, lin, lin); +static void print_ifdef_hunk (struct change *); +static void print_ifdef_lines (FILE *, char const *, struct group const *); + +static lin next_line0; +static lin next_line1; + +/* Print the edit-script SCRIPT as a merged #ifdef file. */ + +void +print_ifdef_script (struct change *script) +{ + next_line0 = next_line1 = - files[0].prefix_lines; + print_script (script, find_change, print_ifdef_hunk); + if (next_line0 < files[0].valid_lines + || next_line1 < files[1].valid_lines) + { + begin_output (); + format_ifdef (group_format[UNCHANGED], + next_line0, files[0].valid_lines, + next_line1, files[1].valid_lines); + } +} + +/* Print a hunk of an ifdef diff. + This is a contiguous portion of a complete edit script, + describing changes in consecutive lines. */ + +static void +print_ifdef_hunk (struct change *hunk) +{ + lin first0, last0, first1, last1; + + /* Determine range of line numbers involved in each file. */ + enum changes changes = analyze_hunk (hunk, &first0, &last0, &first1, &last1); + if (!changes) + return; + + begin_output (); + + /* Print lines up to this change. */ + if (next_line0 < first0 || next_line1 < first1) + format_ifdef (group_format[UNCHANGED], + next_line0, first0, + next_line1, first1); + + /* Print this change. */ + next_line0 = last0 + 1; + next_line1 = last1 + 1; + format_ifdef (group_format[changes], + first0, next_line0, + first1, next_line1); +} + +/* Print a set of lines according to FORMAT. + Lines BEG0 up to END0 are from the first file; + lines BEG1 up to END1 are from the second file. */ + +static void +format_ifdef (char const *format, lin beg0, lin end0, lin beg1, lin end1) +{ + struct group groups[2]; + + groups[0].file = &files[0]; + groups[0].from = beg0; + groups[0].upto = end0; + groups[1].file = &files[1]; + groups[1].from = beg1; + groups[1].upto = end1; + format_group (outfile, format, 0, groups); +} + +/* Print to file OUT a set of lines according to FORMAT. + The format ends at the first free instance of ENDCHAR. + Yield the address of the terminating character. + GROUPS specifies which lines to print. + If OUT is zero, do not actually print anything; just scan the format. */ + +static char const * +format_group (register FILE *out, char const *format, char endchar, + struct group const *groups) +{ + register char c; + register char const *f = format; + + while ((c = *f) != endchar && c != 0) + { + char const *f1 = ++f; + if (c == '%') + switch ((c = *f++)) + { + case '%': + break; + + case '(': + /* Print if-then-else format e.g. `%(n=1?thenpart:elsepart)'. */ + { + int i; + uintmax_t value[2]; + FILE *thenout, *elseout; + + for (i = 0; i < 2; i++) + { + if (ISDIGIT (*f)) + { + char *fend; + errno = 0; + value[i] = strtoumax (f, &fend, 10); + if (errno) + goto bad_format; + f = fend; + } + else + { + value[i] = groups_letter_value (groups, *f); + if (value[i] == -1) + goto bad_format; + f++; + } + if (*f++ != "=?"[i]) + goto bad_format; + } + if (value[0] == value[1]) + thenout = out, elseout = 0; + else + thenout = 0, elseout = out; + f = format_group (thenout, f, ':', groups); + if (*f) + { + f = format_group (elseout, f + 1, ')', groups); + if (*f) + f++; + } + } + continue; + + case '<': + /* Print lines deleted from first file. */ + print_ifdef_lines (out, line_format[OLD], &groups[0]); + continue; + + case '=': + /* Print common lines. */ + print_ifdef_lines (out, line_format[UNCHANGED], &groups[0]); + continue; + + case '>': + /* Print lines inserted from second file. */ + print_ifdef_lines (out, line_format[NEW], &groups[1]); + continue; + + default: + f = do_printf_spec (out, f - 2, 0, 0, groups); + if (f) + continue; + /* Fall through. */ + bad_format: + c = '%'; + f = f1; + break; + } + + if (out) + putc (c, out); + } + + return f; +} + +/* For the line group pair G, return the number corresponding to LETTER. + Return -1 if LETTER is not a group format letter. */ +static lin +groups_letter_value (struct group const *g, char letter) +{ + switch (letter) + { + case 'E': letter = 'e'; g++; break; + case 'F': letter = 'f'; g++; break; + case 'L': letter = 'l'; g++; break; + case 'M': letter = 'm'; g++; break; + case 'N': letter = 'n'; g++; break; + } + + switch (letter) + { + case 'e': return translate_line_number (g->file, g->from) - 1; + case 'f': return translate_line_number (g->file, g->from); + case 'l': return translate_line_number (g->file, g->upto) - 1; + case 'm': return translate_line_number (g->file, g->upto); + case 'n': return g->upto - g->from; + default: return -1; + } +} + +/* Print to file OUT, using FORMAT to print the line group GROUP. + But do nothing if OUT is zero. */ +static void +print_ifdef_lines (register FILE *out, char const *format, + struct group const *group) +{ + struct file_data const *file = group->file; + char const * const *linbuf = file->linbuf; + lin from = group->from, upto = group->upto; + + if (!out) + return; + + /* If possible, use a single fwrite; it's faster. */ + if (!expand_tabs && format[0] == '%') + { + if (format[1] == 'l' && format[2] == '\n' && !format[3] && from < upto) + { + fwrite (linbuf[from], sizeof (char), + linbuf[upto] + (linbuf[upto][-1] != '\n') - linbuf[from], + out); + return; + } + if (format[1] == 'L' && !format[2]) + { + fwrite (linbuf[from], sizeof (char), + linbuf[upto] - linbuf[from], out); + return; + } + } + + for (; from < upto; from++) + { + register char c; + register char const *f = format; + + while ((c = *f++) != 0) + { + char const *f1 = f; + if (c == '%') + switch ((c = *f++)) + { + case '%': + break; + + case 'l': + output_1_line (linbuf[from], + (linbuf[from + 1] + - (linbuf[from + 1][-1] == '\n')), + 0, 0); + continue; + + case 'L': + output_1_line (linbuf[from], linbuf[from + 1], 0, 0); + continue; + + default: + f = do_printf_spec (out, f - 2, file, from, 0); + if (f) + continue; + c = '%'; + f = f1; + break; + } + + putc (c, out); + } + } +} + +static char const * +do_printf_spec (FILE *out, char const *spec, + struct file_data const *file, lin n, + struct group const *groups) +{ + char const *f = spec; + char c; + char c1; + + /* Scan printf-style SPEC of the form %[-'0]*[0-9]*(.[0-9]*)?[cdoxX]. */ + /* assert (*f == '%'); */ + f++; + while ((c = *f++) == '-' || c == '\'' || c == '0') + continue; + while (ISDIGIT (c)) + c = *f++; + if (c == '.') + while (ISDIGIT (c = *f++)) + continue; + c1 = *f++; + + switch (c) + { + case 'c': + if (c1 != '\'') + return 0; + else + { + char value; + f = scan_char_literal (f, &value); + if (!f) + return 0; + if (out) + putc (value, out); + } + break; + + case 'd': case 'o': case 'x': case 'X': + { + lin value; + + if (file) + { + if (c1 != 'n') + return 0; + value = translate_line_number (file, n); + } + else + { + value = groups_letter_value (groups, c1); + if (value < 0) + return 0; + } + + if (out) + { + /* For example, if the spec is "%3xn", use the printf + format spec "%3lx". Here the spec prefix is "%3". */ + long int long_value = value; + size_t spec_prefix_len = f - spec - 2; +#if HAVE_C_VARARRAYS + char format[spec_prefix_len + 3]; +#else + char *format = xmalloc (spec_prefix_len + 3); +#endif + char *p = format + spec_prefix_len; + memcpy (format, spec, spec_prefix_len); + *p++ = 'l'; + *p++ = c; + *p = '\0'; + fprintf (out, format, long_value); +#if ! HAVE_C_VARARRAYS + free (format); +#endif + } + } + break; + + default: + return 0; + } + + return f; +} + +/* Scan the character literal represented in the string LIT; LIT points just + after the initial apostrophe. Put the literal's value into *VALPTR. + Yield the address of the first character after the closing apostrophe, + or zero if the literal is ill-formed. */ +static char const * +scan_char_literal (char const *lit, char *valptr) +{ + register char const *p = lit; + char value; + ptrdiff_t digits; + char c = *p++; + + switch (c) + { + case 0: + case '\'': + return 0; + + case '\\': + value = 0; + while ((c = *p++) != '\'') + { + unsigned int digit = c - '0'; + if (8 <= digit) + return 0; + value = 8 * value + digit; + } + digits = p - lit - 2; + if (! (1 <= digits && digits <= 3)) + return 0; + break; + + default: + value = c; + if (*p++ != '\'') + return 0; + break; + } + + *valptr = value; + return p; +} diff --git a/contrib/diff/src/io.c b/contrib/diff/src/io.c new file mode 100644 index 0000000..3ddc1f8 --- /dev/null +++ b/contrib/diff/src/io.c @@ -0,0 +1,859 @@ +/* File I/O for GNU DIFF. + + Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1998, 2001, 2002, + 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "diff.h" +#include +#include +#include +#include + +/* Rotate an unsigned value to the left. */ +#define ROL(v, n) ((v) << (n) | (v) >> (sizeof (v) * CHAR_BIT - (n))) + +/* Given a hash value and a new character, return a new hash value. */ +#define HASH(h, c) ((c) + ROL (h, 7)) + +/* The type of a hash value. */ +typedef size_t hash_value; +verify (hash_value_is_unsigned, ! TYPE_SIGNED (hash_value)); + +/* Lines are put into equivalence classes of lines that match in lines_differ. + Each equivalence class is represented by one of these structures, + but only while the classes are being computed. + Afterward, each class is represented by a number. */ +struct equivclass +{ + lin next; /* Next item in this bucket. */ + hash_value hash; /* Hash of lines in this class. */ + char const *line; /* A line that fits this class. */ + size_t length; /* That line's length, not counting its newline. */ +}; + +/* Hash-table: array of buckets, each being a chain of equivalence classes. + buckets[-1] is reserved for incomplete lines. */ +static lin *buckets; + +/* Number of buckets in the hash table array, not counting buckets[-1]. */ +static size_t nbuckets; + +/* Array in which the equivalence classes are allocated. + The bucket-chains go through the elements in this array. + The number of an equivalence class is its index in this array. */ +static struct equivclass *equivs; + +/* Index of first free element in the array `equivs'. */ +static lin equivs_index; + +/* Number of elements allocated in the array `equivs'. */ +static lin equivs_alloc; + +/* Read a block of data into a file buffer, checking for EOF and error. */ + +void +file_block_read (struct file_data *current, size_t size) +{ + if (size && ! current->eof) + { + size_t s = block_read (current->desc, + FILE_BUFFER (current) + current->buffered, size); + if (s == SIZE_MAX) + pfatal_with_name (current->name); + current->buffered += s; + current->eof = s < size; + } +} + +/* Check for binary files and compare them for exact identity. */ + +/* Return 1 if BUF contains a non text character. + SIZE is the number of characters in BUF. */ + +#define binary_file_p(buf, size) (memchr (buf, 0, size) != 0) + +/* Get ready to read the current file. + Return nonzero if SKIP_TEST is zero, + and if it appears to be a binary file. */ + +static bool +sip (struct file_data *current, bool skip_test) +{ + /* If we have a nonexistent file at this stage, treat it as empty. */ + if (current->desc < 0) + { + /* Leave room for a sentinel. */ + current->bufsize = sizeof (word); + current->buffer = xmalloc (current->bufsize); + } + else + { + current->bufsize = buffer_lcm (sizeof (word), + STAT_BLOCKSIZE (current->stat), + PTRDIFF_MAX - 2 * sizeof (word)); + current->buffer = xmalloc (current->bufsize); + + if (! skip_test) + { + /* Check first part of file to see if it's a binary file. */ + + bool was_binary = set_binary_mode (current->desc, true); + off_t buffered; + file_block_read (current, current->bufsize); + buffered = current->buffered; + + if (! was_binary) + { + /* Revert to text mode and seek back to the beginning to + reread the file. Use relative seek, since file + descriptors like stdin might not start at offset + zero. */ + + if (lseek (current->desc, - buffered, SEEK_CUR) == -1) + pfatal_with_name (current->name); + set_binary_mode (current->desc, false); + current->buffered = 0; + current->eof = false; + } + + return binary_file_p (current->buffer, buffered); + } + } + + current->buffered = 0; + current->eof = false; + return false; +} + +/* Slurp the rest of the current file completely into memory. */ + +static void +slurp (struct file_data *current) +{ + size_t cc; + + if (current->desc < 0) + { + /* The file is nonexistent. */ + return; + } + + if (S_ISREG (current->stat.st_mode)) + { + /* It's a regular file; slurp in the rest all at once. */ + + /* Get the size out of the stat block. + Allocate just enough room for appended newline plus word sentinel, + plus word-alignment since we want the buffer word-aligned. */ + size_t file_size = current->stat.st_size; + cc = file_size + 2 * sizeof (word) - file_size % sizeof (word); + if (file_size != current->stat.st_size || cc < file_size + || PTRDIFF_MAX <= cc) + xalloc_die (); + + if (current->bufsize < cc) + { + current->bufsize = cc; + current->buffer = xrealloc (current->buffer, cc); + } + + /* Try to read at least 1 more byte than the size indicates, to + detect whether the file is growing. This is a nicety for + users who run 'diff' on files while they are changing. */ + + if (current->buffered <= file_size) + { + file_block_read (current, file_size + 1 - current->buffered); + if (current->buffered <= file_size) + return; + } + } + + /* It's not a regular file, or it's a growing regular file; read it, + growing the buffer as needed. */ + + file_block_read (current, current->bufsize - current->buffered); + + if (current->buffered) + { + while (current->buffered == current->bufsize) + { + if (PTRDIFF_MAX / 2 - sizeof (word) < current->bufsize) + xalloc_die (); + current->bufsize *= 2; + current->buffer = xrealloc (current->buffer, current->bufsize); + file_block_read (current, current->bufsize - current->buffered); + } + + /* Allocate just enough room for appended newline plus word + sentinel, plus word-alignment. */ + cc = current->buffered + 2 * sizeof (word); + current->bufsize = cc - cc % sizeof (word); + current->buffer = xrealloc (current->buffer, current->bufsize); + } +} + +/* Split the file into lines, simultaneously computing the equivalence + class for each line. */ + +static void +find_and_hash_each_line (struct file_data *current) +{ + hash_value h; + char const *p = current->prefix_end; + unsigned char c; + lin i, *bucket; + size_t length; + + /* Cache often-used quantities in local variables to help the compiler. */ + char const **linbuf = current->linbuf; + lin alloc_lines = current->alloc_lines; + lin line = 0; + lin linbuf_base = current->linbuf_base; + lin *cureqs = xmalloc (alloc_lines * sizeof *cureqs); + struct equivclass *eqs = equivs; + lin eqs_index = equivs_index; + lin eqs_alloc = equivs_alloc; + char const *suffix_begin = current->suffix_begin; + char const *bufend = FILE_BUFFER (current) + current->buffered; + bool diff_length_compare_anyway = + ignore_white_space != IGNORE_NO_WHITE_SPACE; + bool same_length_diff_contents_compare_anyway = + diff_length_compare_anyway | ignore_case; + + while (p < suffix_begin) + { + char const *ip = p; + + h = 0; + + /* Hash this line until we find a newline. */ + if (ignore_case) + switch (ignore_white_space) + { + case IGNORE_ALL_SPACE: + while ((c = *p++) != '\n') + if (! isspace (c)) + h = HASH (h, tolower (c)); + break; + + case IGNORE_SPACE_CHANGE: + while ((c = *p++) != '\n') + { + if (isspace (c)) + { + do + if ((c = *p++) == '\n') + goto hashing_done; + while (isspace (c)); + + h = HASH (h, ' '); + } + + /* C is now the first non-space. */ + h = HASH (h, tolower (c)); + } + break; + + case IGNORE_TAB_EXPANSION: + { + size_t column = 0; + while ((c = *p++) != '\n') + { + size_t repetitions = 1; + + switch (c) + { + case '\b': + column -= 0 < column; + break; + + case '\t': + c = ' '; + repetitions = tabsize - column % tabsize; + column = (column + repetitions < column + ? 0 + : column + repetitions); + break; + + case '\r': + column = 0; + break; + + default: + c = tolower (c); + column++; + break; + } + + do + h = HASH (h, c); + while (--repetitions != 0); + } + } + break; + + default: + while ((c = *p++) != '\n') + h = HASH (h, tolower (c)); + break; + } + else + switch (ignore_white_space) + { + case IGNORE_ALL_SPACE: + while ((c = *p++) != '\n') + if (! isspace (c)) + h = HASH (h, c); + break; + + case IGNORE_SPACE_CHANGE: + while ((c = *p++) != '\n') + { + if (isspace (c)) + { + do + if ((c = *p++) == '\n') + goto hashing_done; + while (isspace (c)); + + h = HASH (h, ' '); + } + + /* C is now the first non-space. */ + h = HASH (h, c); + } + break; + + case IGNORE_TAB_EXPANSION: + { + size_t column = 0; + while ((c = *p++) != '\n') + { + size_t repetitions = 1; + + switch (c) + { + case '\b': + column -= 0 < column; + break; + + case '\t': + c = ' '; + repetitions = tabsize - column % tabsize; + column = (column + repetitions < column + ? 0 + : column + repetitions); + break; + + case '\r': + column = 0; + break; + + default: + column++; + break; + } + + do + h = HASH (h, c); + while (--repetitions != 0); + } + } + break; + + default: + while ((c = *p++) != '\n') + h = HASH (h, c); + break; + } + + hashing_done:; + + bucket = &buckets[h % nbuckets]; + length = p - ip - 1; + + if (p == bufend + && current->missing_newline + && ROBUST_OUTPUT_STYLE (output_style)) + { + /* This line is incomplete. If this is significant, + put the line into buckets[-1]. */ + if (ignore_white_space < IGNORE_SPACE_CHANGE) + bucket = &buckets[-1]; + + /* Omit the inserted newline when computing linbuf later. */ + p--; + bufend = suffix_begin = p; + } + + for (i = *bucket; ; i = eqs[i].next) + if (!i) + { + /* Create a new equivalence class in this bucket. */ + i = eqs_index++; + if (i == eqs_alloc) + { + if (PTRDIFF_MAX / (2 * sizeof *eqs) <= eqs_alloc) + xalloc_die (); + eqs_alloc *= 2; + eqs = xrealloc (eqs, eqs_alloc * sizeof *eqs); + } + eqs[i].next = *bucket; + eqs[i].hash = h; + eqs[i].line = ip; + eqs[i].length = length; + *bucket = i; + break; + } + else if (eqs[i].hash == h) + { + char const *eqline = eqs[i].line; + + /* Reuse existing class if lines_differ reports the lines + equal. */ + if (eqs[i].length == length) + { + /* Reuse existing equivalence class if the lines are identical. + This detects the common case of exact identity + faster than lines_differ would. */ + if (memcmp (eqline, ip, length) == 0) + break; + if (!same_length_diff_contents_compare_anyway) + continue; + } + else if (!diff_length_compare_anyway) + continue; + + if (! lines_differ (eqline, ip)) + break; + } + + /* Maybe increase the size of the line table. */ + if (line == alloc_lines) + { + /* Double (alloc_lines - linbuf_base) by adding to alloc_lines. */ + if (PTRDIFF_MAX / 3 <= alloc_lines + || PTRDIFF_MAX / sizeof *cureqs <= 2 * alloc_lines - linbuf_base + || PTRDIFF_MAX / sizeof *linbuf <= alloc_lines - linbuf_base) + xalloc_die (); + alloc_lines = 2 * alloc_lines - linbuf_base; + cureqs = xrealloc (cureqs, alloc_lines * sizeof *cureqs); + linbuf += linbuf_base; + linbuf = xrealloc (linbuf, + (alloc_lines - linbuf_base) * sizeof *linbuf); + linbuf -= linbuf_base; + } + linbuf[line] = ip; + cureqs[line] = i; + ++line; + } + + current->buffered_lines = line; + + for (i = 0; ; i++) + { + /* Record the line start for lines in the suffix that we care about. + Record one more line start than lines, + so that we can compute the length of any buffered line. */ + if (line == alloc_lines) + { + /* Double (alloc_lines - linbuf_base) by adding to alloc_lines. */ + if (PTRDIFF_MAX / 3 <= alloc_lines + || PTRDIFF_MAX / sizeof *cureqs <= 2 * alloc_lines - linbuf_base + || PTRDIFF_MAX / sizeof *linbuf <= alloc_lines - linbuf_base) + xalloc_die (); + alloc_lines = 2 * alloc_lines - linbuf_base; + linbuf += linbuf_base; + linbuf = xrealloc (linbuf, + (alloc_lines - linbuf_base) * sizeof *linbuf); + linbuf -= linbuf_base; + } + linbuf[line] = p; + + if (p == bufend) + break; + + if (context <= i && no_diff_means_no_output) + break; + + line++; + + while (*p++ != '\n') + continue; + } + + /* Done with cache in local variables. */ + current->linbuf = linbuf; + current->valid_lines = line; + current->alloc_lines = alloc_lines; + current->equivs = cureqs; + equivs = eqs; + equivs_alloc = eqs_alloc; + equivs_index = eqs_index; +} + +/* Prepare the text. Make sure the text end is initialized. + Make sure text ends in a newline, + but remember that we had to add one. + Strip trailing CRs, if that was requested. */ + +static void +prepare_text (struct file_data *current) +{ + size_t buffered = current->buffered; + char *p = FILE_BUFFER (current); + char *dst; + + if (buffered == 0 || p[buffered - 1] == '\n') + current->missing_newline = false; + else + { + p[buffered++] = '\n'; + current->missing_newline = true; + } + + if (!p) + return; + + /* Don't use uninitialized storage when planting or using sentinels. */ + memset (p + buffered, 0, sizeof (word)); + + if (strip_trailing_cr && (dst = memchr (p, '\r', buffered))) + { + char const *src = dst; + char const *srclim = p + buffered; + + do + dst += ! ((*dst = *src++) == '\r' && *src == '\n'); + while (src < srclim); + + buffered -= src - dst; + } + + current->buffered = buffered; +} + +/* We have found N lines in a buffer of size S; guess the + proportionate number of lines that will be found in a buffer of + size T. However, do not guess a number of lines so large that the + resulting line table might cause overflow in size calculations. */ +static lin +guess_lines (lin n, size_t s, size_t t) +{ + size_t guessed_bytes_per_line = n < 10 ? 32 : s / (n - 1); + lin guessed_lines = MAX (1, t / guessed_bytes_per_line); + return MIN (guessed_lines, PTRDIFF_MAX / (2 * sizeof (char *) + 1) - 5) + 5; +} + +/* Given a vector of two file_data objects, find the identical + prefixes and suffixes of each object. */ + +static void +find_identical_ends (struct file_data filevec[]) +{ + word *w0, *w1; + char *p0, *p1, *buffer0, *buffer1; + char const *end0, *beg0; + char const **linbuf0, **linbuf1; + lin i, lines; + size_t n0, n1; + lin alloc_lines0, alloc_lines1; + lin buffered_prefix, prefix_count, prefix_mask; + lin middle_guess, suffix_guess; + + slurp (&filevec[0]); + prepare_text (&filevec[0]); + if (filevec[0].desc != filevec[1].desc) + { + slurp (&filevec[1]); + prepare_text (&filevec[1]); + } + else + { + filevec[1].buffer = filevec[0].buffer; + filevec[1].bufsize = filevec[0].bufsize; + filevec[1].buffered = filevec[0].buffered; + filevec[1].missing_newline = filevec[0].missing_newline; + } + + /* Find identical prefix. */ + + w0 = filevec[0].buffer; + w1 = filevec[1].buffer; + p0 = buffer0 = (char *) w0; + p1 = buffer1 = (char *) w1; + n0 = filevec[0].buffered; + n1 = filevec[1].buffered; + + if (p0 == p1) + /* The buffers are the same; sentinels won't work. */ + p0 = p1 += n1; + else + { + /* Insert end sentinels, in this case characters that are guaranteed + to make the equality test false, and thus terminate the loop. */ + + if (n0 < n1) + p0[n0] = ~p1[n0]; + else + p1[n1] = ~p0[n1]; + + /* Loop until first mismatch, or to the sentinel characters. */ + + /* Compare a word at a time for speed. */ + while (*w0 == *w1) + w0++, w1++; + + /* Do the last few bytes of comparison a byte at a time. */ + p0 = (char *) w0; + p1 = (char *) w1; + while (*p0 == *p1) + p0++, p1++; + + /* Don't mistakenly count missing newline as part of prefix. */ + if (ROBUST_OUTPUT_STYLE (output_style) + && ((buffer0 + n0 - filevec[0].missing_newline < p0) + != + (buffer1 + n1 - filevec[1].missing_newline < p1))) + p0--, p1--; + } + + /* Now P0 and P1 point at the first nonmatching characters. */ + + /* Skip back to last line-beginning in the prefix, + and then discard up to HORIZON_LINES lines from the prefix. */ + i = horizon_lines; + while (p0 != buffer0 && (p0[-1] != '\n' || i--)) + p0--, p1--; + + /* Record the prefix. */ + filevec[0].prefix_end = p0; + filevec[1].prefix_end = p1; + + /* Find identical suffix. */ + + /* P0 and P1 point beyond the last chars not yet compared. */ + p0 = buffer0 + n0; + p1 = buffer1 + n1; + + if (! ROBUST_OUTPUT_STYLE (output_style) + || filevec[0].missing_newline == filevec[1].missing_newline) + { + end0 = p0; /* Addr of last char in file 0. */ + + /* Get value of P0 at which we should stop scanning backward: + this is when either P0 or P1 points just past the last char + of the identical prefix. */ + beg0 = filevec[0].prefix_end + (n0 < n1 ? 0 : n0 - n1); + + /* Scan back until chars don't match or we reach that point. */ + for (; p0 != beg0; p0--, p1--) + if (*p0 != *p1) + { + /* Point at the first char of the matching suffix. */ + beg0 = p0; + break; + } + + /* Are we at a line-beginning in both files? If not, add the rest of + this line to the main body. Discard up to HORIZON_LINES lines from + the identical suffix. Also, discard one extra line, + because shift_boundaries may need it. */ + i = horizon_lines + !((buffer0 == p0 || p0[-1] == '\n') + && + (buffer1 == p1 || p1[-1] == '\n')); + while (i-- && p0 != end0) + while (*p0++ != '\n') + continue; + + p1 += p0 - beg0; + } + + /* Record the suffix. */ + filevec[0].suffix_begin = p0; + filevec[1].suffix_begin = p1; + + /* Calculate number of lines of prefix to save. + + prefix_count == 0 means save the whole prefix; + we need this for options like -D that output the whole file, + or for enormous contexts (to avoid worrying about arithmetic overflow). + We also need it for options like -F that output some preceding line; + at least we will need to find the last few lines, + but since we don't know how many, it's easiest to find them all. + + Otherwise, prefix_count != 0. Save just prefix_count lines at start + of the line buffer; they'll be moved to the proper location later. + Handle 1 more line than the context says (because we count 1 too many), + rounded up to the next power of 2 to speed index computation. */ + + if (no_diff_means_no_output && ! function_regexp.fastmap + && context < LIN_MAX / 4 && context < n0) + { + middle_guess = guess_lines (0, 0, p0 - filevec[0].prefix_end); + suffix_guess = guess_lines (0, 0, buffer0 + n0 - p0); + for (prefix_count = 1; prefix_count <= context; prefix_count *= 2) + continue; + alloc_lines0 = (prefix_count + middle_guess + + MIN (context, suffix_guess)); + } + else + { + prefix_count = 0; + alloc_lines0 = guess_lines (0, 0, n0); + } + + prefix_mask = prefix_count - 1; + lines = 0; + linbuf0 = xmalloc (alloc_lines0 * sizeof *linbuf0); + p0 = buffer0; + + /* If the prefix is needed, find the prefix lines. */ + if (! (no_diff_means_no_output + && filevec[0].prefix_end == p0 + && filevec[1].prefix_end == p1)) + { + end0 = filevec[0].prefix_end; + while (p0 != end0) + { + lin l = lines++ & prefix_mask; + if (l == alloc_lines0) + { + if (PTRDIFF_MAX / (2 * sizeof *linbuf0) <= alloc_lines0) + xalloc_die (); + alloc_lines0 *= 2; + linbuf0 = xrealloc (linbuf0, alloc_lines0 * sizeof *linbuf0); + } + linbuf0[l] = p0; + while (*p0++ != '\n') + continue; + } + } + buffered_prefix = prefix_count && context < lines ? context : lines; + + /* Allocate line buffer 1. */ + + middle_guess = guess_lines (lines, p0 - buffer0, p1 - filevec[1].prefix_end); + suffix_guess = guess_lines (lines, p0 - buffer0, buffer1 + n1 - p1); + alloc_lines1 = buffered_prefix + middle_guess + MIN (context, suffix_guess); + if (alloc_lines1 < buffered_prefix + || PTRDIFF_MAX / sizeof *linbuf1 <= alloc_lines1) + xalloc_die (); + linbuf1 = xmalloc (alloc_lines1 * sizeof *linbuf1); + + if (buffered_prefix != lines) + { + /* Rotate prefix lines to proper location. */ + for (i = 0; i < buffered_prefix; i++) + linbuf1[i] = linbuf0[(lines - context + i) & prefix_mask]; + for (i = 0; i < buffered_prefix; i++) + linbuf0[i] = linbuf1[i]; + } + + /* Initialize line buffer 1 from line buffer 0. */ + for (i = 0; i < buffered_prefix; i++) + linbuf1[i] = linbuf0[i] - buffer0 + buffer1; + + /* Record the line buffer, adjusted so that + linbuf[0] points at the first differing line. */ + filevec[0].linbuf = linbuf0 + buffered_prefix; + filevec[1].linbuf = linbuf1 + buffered_prefix; + filevec[0].linbuf_base = filevec[1].linbuf_base = - buffered_prefix; + filevec[0].alloc_lines = alloc_lines0 - buffered_prefix; + filevec[1].alloc_lines = alloc_lines1 - buffered_prefix; + filevec[0].prefix_lines = filevec[1].prefix_lines = lines; +} + +/* If 1 < k, then (2**k - prime_offset[k]) is the largest prime less + than 2**k. This table is derived from Chris K. Caldwell's list + . */ + +static unsigned char const prime_offset[] = +{ + 0, 0, 1, 1, 3, 1, 3, 1, 5, 3, 3, 9, 3, 1, 3, 19, 15, 1, 5, 1, 3, 9, 3, + 15, 3, 39, 5, 39, 57, 3, 35, 1, 5, 9, 41, 31, 5, 25, 45, 7, 87, 21, + 11, 57, 17, 55, 21, 115, 59, 81, 27, 129, 47, 111, 33, 55, 5, 13, 27, + 55, 93, 1, 57, 25 +}; + +/* Verify that this host's size_t is not too wide for the above table. */ + +verify (enough_prime_offsets, + sizeof (size_t) * CHAR_BIT <= sizeof prime_offset); + +/* Given a vector of two file_data objects, read the file associated + with each one, and build the table of equivalence classes. + Return nonzero if either file appears to be a binary file. + If PRETEND_BINARY is nonzero, pretend they are binary regardless. */ + +bool +read_files (struct file_data filevec[], bool pretend_binary) +{ + int i; + bool skip_test = text | pretend_binary; + bool appears_binary = pretend_binary | sip (&filevec[0], skip_test); + + if (filevec[0].desc != filevec[1].desc) + appears_binary |= sip (&filevec[1], skip_test | appears_binary); + else + { + filevec[1].buffer = filevec[0].buffer; + filevec[1].bufsize = filevec[0].bufsize; + filevec[1].buffered = filevec[0].buffered; + } + if (appears_binary) + { + set_binary_mode (filevec[0].desc, true); + set_binary_mode (filevec[1].desc, true); + return true; + } + + find_identical_ends (filevec); + + equivs_alloc = filevec[0].alloc_lines + filevec[1].alloc_lines + 1; + if (PTRDIFF_MAX / sizeof *equivs <= equivs_alloc) + xalloc_die (); + equivs = xmalloc (equivs_alloc * sizeof *equivs); + /* Equivalence class 0 is permanently safe for lines that were not + hashed. Real equivalence classes start at 1. */ + equivs_index = 1; + + /* Allocate (one plus) a prime number of hash buckets. Use a prime + number between 1/3 and 2/3 of the value of equiv_allocs, + approximately. */ + for (i = 9; (size_t) 1 << i < equivs_alloc / 3; i++) + continue; + nbuckets = ((size_t) 1 << i) - prime_offset[i]; + if (PTRDIFF_MAX / sizeof *buckets <= nbuckets) + xalloc_die (); + buckets = zalloc ((nbuckets + 1) * sizeof *buckets); + buckets++; + + for (i = 0; i < 2; i++) + find_and_hash_each_line (&filevec[i]); + + filevec[0].equiv_max = filevec[1].equiv_max = equivs_index; + + free (equivs); + free (buckets - 1); + + return false; +} diff --git a/contrib/diff/src/normal.c b/contrib/diff/src/normal.c new file mode 100644 index 0000000..4a01596 --- /dev/null +++ b/contrib/diff/src/normal.c @@ -0,0 +1,71 @@ +/* Normal-format output routines for GNU DIFF. + + Copyright (C) 1988, 1989, 1993, 1995, 1998, 2001 Free Software + Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "diff.h" + +static void print_normal_hunk (struct change *); + +/* Print the edit-script SCRIPT as a normal diff. + INF points to an array of descriptions of the two files. */ + +void +print_normal_script (struct change *script) +{ + print_script (script, find_change, print_normal_hunk); +} + +/* Print a hunk of a normal diff. + This is a contiguous portion of a complete edit script, + describing changes in consecutive lines. */ + +static void +print_normal_hunk (struct change *hunk) +{ + lin first0, last0, first1, last1; + register lin i; + + /* Determine range of line numbers involved in each file. */ + enum changes changes = analyze_hunk (hunk, &first0, &last0, &first1, &last1); + if (!changes) + return; + + begin_output (); + + /* Print out the line number header for this hunk */ + print_number_range (',', &files[0], first0, last0); + fprintf (outfile, "%c", change_letter[changes]); + print_number_range (',', &files[1], first1, last1); + fprintf (outfile, "\n"); + + /* Print the lines that the first file has. */ + if (changes & OLD) + for (i = first0; i <= last0; i++) + print_1_line ("<", &files[0].linbuf[i]); + + if (changes == CHANGED) + fprintf (outfile, "---\n"); + + /* Print the lines that the second file has. */ + if (changes & NEW) + for (i = first1; i <= last1; i++) + print_1_line (">", &files[1].linbuf[i]); +} diff --git a/contrib/diff/src/sdiff.c b/contrib/diff/src/sdiff.c new file mode 100644 index 0000000..2b0a2ae --- /dev/null +++ b/contrib/diff/src/sdiff.c @@ -0,0 +1,1225 @@ +/* sdiff - side-by-side merge of file differences + + Copyright (C) 1992, 1993, 1994, 1995, 1996, 1998, 2001, 2002, 2004 + Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. + See the GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "system.h" +#include "paths.h" + +#include +#include + +#include +#include +#include +#include +#include +#include +#include +#include +#include +#include + +/* Size of chunks read from files which must be parsed into lines. */ +#define SDIFF_BUFSIZE ((size_t) 65536) + +char *program_name; + +static char const *editor_program = DEFAULT_EDITOR_PROGRAM; +static char const **diffargv; + +static char * volatile tmpname; +static FILE *tmp; + +#if HAVE_WORKING_FORK || HAVE_WORKING_VFORK +static pid_t volatile diffpid; +#endif + +struct line_filter; + +static void catchsig (int); +static bool edit (struct line_filter *, char const *, lin, lin, struct line_filter *, char const *, lin, lin, FILE *); +static bool interact (struct line_filter *, struct line_filter *, char const *, struct line_filter *, char const *, FILE *); +static void checksigs (void); +static void diffarg (char const *); +static void fatal (char const *) __attribute__((noreturn)); +static void perror_fatal (char const *) __attribute__((noreturn)); +static void trapsigs (void); +static void untrapsig (int); + +#define NUM_SIGS (sizeof sigs / sizeof *sigs) +static int const sigs[] = { +#ifdef SIGHUP + SIGHUP, +#endif +#ifdef SIGQUIT + SIGQUIT, +#endif +#ifdef SIGTERM + SIGTERM, +#endif +#ifdef SIGXCPU + SIGXCPU, +#endif +#ifdef SIGXFSZ + SIGXFSZ, +#endif + SIGINT, + SIGPIPE +}; +#define handler_index_of_SIGINT (NUM_SIGS - 2) +#define handler_index_of_SIGPIPE (NUM_SIGS - 1) + +#if HAVE_SIGACTION + /* Prefer `sigaction' if available, since `signal' can lose signals. */ + static struct sigaction initial_action[NUM_SIGS]; +# define initial_handler(i) (initial_action[i].sa_handler) + static void signal_handler (int, void (*) (int)); +#else + static void (*initial_action[NUM_SIGS]) (); +# define initial_handler(i) (initial_action[i]) +# define signal_handler(sig, handler) signal (sig, handler) +#endif + +#if ! HAVE_SIGPROCMASK +# define sigset_t int +# define sigemptyset(s) (*(s) = 0) +# ifndef sigmask +# define sigmask(sig) (1 << ((sig) - 1)) +# endif +# define sigaddset(s, sig) (*(s) |= sigmask (sig)) +# ifndef SIG_BLOCK +# define SIG_BLOCK 0 +# endif +# ifndef SIG_SETMASK +# define SIG_SETMASK (! SIG_BLOCK) +# endif +# define sigprocmask(how, n, o) \ + ((how) == SIG_BLOCK ? *(o) = sigblock (*(n)) : sigsetmask (*(n))) +#endif + +static bool diraccess (char const *); +static int temporary_file (void); + +/* Options: */ + +/* Name of output file if -o specified. */ +static char const *output; + +/* Do not print common lines. */ +static bool suppress_common_lines; + +/* Value for the long option that does not have single-letter equivalents. */ +enum +{ + DIFF_PROGRAM_OPTION = CHAR_MAX + 1, + HELP_OPTION, + STRIP_TRAILING_CR_OPTION, + TABSIZE_OPTION +}; + +static struct option const longopts[] = +{ + {"diff-program", 1, 0, DIFF_PROGRAM_OPTION}, + {"expand-tabs", 0, 0, 't'}, + {"help", 0, 0, HELP_OPTION}, + {"ignore-all-space", 0, 0, 'W'}, /* swap W and w for historical reasons */ + {"ignore-blank-lines", 0, 0, 'B'}, + {"ignore-case", 0, 0, 'i'}, + {"ignore-matching-lines", 1, 0, 'I'}, + {"ignore-space-change", 0, 0, 'b'}, + {"ignore-tab-expansion", 0, 0, 'E'}, + {"left-column", 0, 0, 'l'}, + {"minimal", 0, 0, 'd'}, + {"output", 1, 0, 'o'}, + {"speed-large-files", 0, 0, 'H'}, + {"strip-trailing-cr", 0, 0, STRIP_TRAILING_CR_OPTION}, + {"suppress-common-lines", 0, 0, 's'}, + {"tabsize", 1, 0, TABSIZE_OPTION}, + {"text", 0, 0, 'a'}, + {"version", 0, 0, 'v'}, + {"width", 1, 0, 'w'}, + {0, 0, 0, 0} +}; + +static void try_help (char const *, char const *) __attribute__((noreturn)); +static void +try_help (char const *reason_msgid, char const *operand) +{ + if (reason_msgid) + error (0, 0, _(reason_msgid), operand); + error (EXIT_TROUBLE, 0, _("Try `%s --help' for more information."), + program_name); + abort (); +} + +static void +check_stdout (void) +{ + if (ferror (stdout)) + fatal ("write failed"); + else if (fclose (stdout) != 0) + perror_fatal (_("standard output")); +} + +static char const * const option_help_msgid[] = { + N_("-o FILE --output=FILE Operate interactively, sending output to FILE."), + "", + N_("-i --ignore-case Consider upper- and lower-case to be the same."), + N_("-E --ignore-tab-expansion Ignore changes due to tab expansion."), + N_("-b --ignore-space-change Ignore changes in the amount of white space."), + N_("-W --ignore-all-space Ignore all white space."), + N_("-B --ignore-blank-lines Ignore changes whose lines are all blank."), + N_("-I RE --ignore-matching-lines=RE Ignore changes whose lines all match RE."), + N_("--strip-trailing-cr Strip trailing carriage return on input."), + N_("-a --text Treat all files as text."), + "", + N_("-w NUM --width=NUM Output at most NUM (default 130) print columns."), + N_("-l --left-column Output only the left column of common lines."), + N_("-s --suppress-common-lines Do not output common lines."), + "", + N_("-t --expand-tabs Expand tabs to spaces in output."), + N_("--tabsize=NUM Tab stops are every NUM (default 8) print columns."), + "", + N_("-d --minimal Try hard to find a smaller set of changes."), + N_("-H --speed-large-files Assume large files and many scattered small changes."), + N_("--diff-program=PROGRAM Use PROGRAM to compare files."), + "", + N_("-v --version Output version info."), + N_("--help Output this help."), + 0 +}; + +static void +usage (void) +{ + char const * const *p; + + printf (_("Usage: %s [OPTION]... FILE1 FILE2\n"), program_name); + printf ("%s\n\n", _("Side-by-side merge of file differences.")); + for (p = option_help_msgid; *p; p++) + if (**p) + printf (" %s\n", _(*p)); + else + putchar ('\n'); + printf ("\n%s\n%s\n\n%s\n", + _("If a FILE is `-', read standard input."), + _("Exit status is 0 if inputs are the same, 1 if different, 2 if trouble."), + _("Report bugs to .")); +} + +/* Clean up after a signal or other failure. This function is + async-signal-safe. */ +static void +cleanup (int signo __attribute__((unused))) +{ +#if HAVE_WORKING_FORK || HAVE_WORKING_VFORK + if (0 < diffpid) + kill (diffpid, SIGPIPE); +#endif + if (tmpname) + unlink (tmpname); +} + +static void exiterr (void) __attribute__((noreturn)); +static void +exiterr (void) +{ + cleanup (0); + untrapsig (0); + checksigs (); + exit (EXIT_TROUBLE); +} + +static void +fatal (char const *msgid) +{ + error (0, 0, "%s", _(msgid)); + exiterr (); +} + +static void +perror_fatal (char const *msg) +{ + int e = errno; + checksigs (); + error (0, e, "%s", msg); + exiterr (); +} + +static void +check_child_status (int werrno, int wstatus, int max_ok_status, + char const *subsidiary_program) +{ + int status = (! werrno && WIFEXITED (wstatus) + ? WEXITSTATUS (wstatus) + : INT_MAX); + + if (max_ok_status < status) + { + error (0, werrno, + _(status == 126 + ? "subsidiary program `%s' could not be invoked" + : status == 127 + ? "subsidiary program `%s' not found" + : status == INT_MAX + ? "subsidiary program `%s' failed" + : "subsidiary program `%s' failed (exit status %d)"), + subsidiary_program, status); + exiterr (); + } +} + +static FILE * +ck_fopen (char const *fname, char const *type) +{ + FILE *r = fopen (fname, type); + if (! r) + perror_fatal (fname); + return r; +} + +static void +ck_fclose (FILE *f) +{ + if (fclose (f)) + perror_fatal ("fclose"); +} + +static size_t +ck_fread (char *buf, size_t size, FILE *f) +{ + size_t r = fread (buf, sizeof (char), size, f); + if (r == 0 && ferror (f)) + perror_fatal (_("read failed")); + return r; +} + +static void +ck_fwrite (char const *buf, size_t size, FILE *f) +{ + if (fwrite (buf, sizeof (char), size, f) != size) + perror_fatal (_("write failed")); +} + +static void +ck_fflush (FILE *f) +{ + if (fflush (f) != 0) + perror_fatal (_("write failed")); +} + +static char const * +expand_name (char *name, bool is_dir, char const *other_name) +{ + if (strcmp (name, "-") == 0) + fatal ("cannot interactively merge standard input"); + if (! is_dir) + return name; + else + { + /* Yield NAME/BASE, where BASE is OTHER_NAME's basename. */ + char const *base = base_name (other_name); + size_t namelen = strlen (name), baselen = strlen (base); + bool insert_slash = *base_name (name) && name[namelen - 1] != '/'; + char *r = xmalloc (namelen + insert_slash + baselen + 1); + memcpy (r, name, namelen); + r[namelen] = '/'; + memcpy (r + namelen + insert_slash, base, baselen + 1); + return r; + } +} + +struct line_filter { + FILE *infile; + char *bufpos; + char *buffer; + char *buflim; +}; + +static void +lf_init (struct line_filter *lf, FILE *infile) +{ + lf->infile = infile; + lf->bufpos = lf->buffer = lf->buflim = xmalloc (SDIFF_BUFSIZE + 1); + lf->buflim[0] = '\n'; +} + +/* Fill an exhausted line_filter buffer from its INFILE */ +static size_t +lf_refill (struct line_filter *lf) +{ + size_t s = ck_fread (lf->buffer, SDIFF_BUFSIZE, lf->infile); + lf->bufpos = lf->buffer; + lf->buflim = lf->buffer + s; + lf->buflim[0] = '\n'; + checksigs (); + return s; +} + +/* Advance LINES on LF's infile, copying lines to OUTFILE */ +static void +lf_copy (struct line_filter *lf, lin lines, FILE *outfile) +{ + char *start = lf->bufpos; + + while (lines) + { + lf->bufpos = (char *) memchr (lf->bufpos, '\n', lf->buflim - lf->bufpos); + if (! lf->bufpos) + { + ck_fwrite (start, lf->buflim - start, outfile); + if (! lf_refill (lf)) + return; + start = lf->bufpos; + } + else + { + --lines; + ++lf->bufpos; + } + } + + ck_fwrite (start, lf->bufpos - start, outfile); +} + +/* Advance LINES on LF's infile without doing output */ +static void +lf_skip (struct line_filter *lf, lin lines) +{ + while (lines) + { + lf->bufpos = (char *) memchr (lf->bufpos, '\n', lf->buflim - lf->bufpos); + if (! lf->bufpos) + { + if (! lf_refill (lf)) + break; + } + else + { + --lines; + ++lf->bufpos; + } + } +} + +/* Snarf a line into a buffer. Return EOF if EOF, 0 if error, 1 if OK. */ +static int +lf_snarf (struct line_filter *lf, char *buffer, size_t bufsize) +{ + for (;;) + { + char *start = lf->bufpos; + char *next = (char *) memchr (start, '\n', lf->buflim + 1 - start); + size_t s = next - start; + if (bufsize <= s) + return 0; + memcpy (buffer, start, s); + if (next < lf->buflim) + { + buffer[s] = 0; + lf->bufpos = next + 1; + return 1; + } + if (! lf_refill (lf)) + return s ? 0 : EOF; + buffer += s; + bufsize -= s; + } +} + +int +main (int argc, char *argv[]) +{ + int opt; + char const *prog; + + exit_failure = EXIT_TROUBLE; + initialize_main (&argc, &argv); + program_name = argv[0]; + setlocale (LC_ALL, ""); + bindtextdomain (PACKAGE, LOCALEDIR); + textdomain (PACKAGE); + c_stack_action (cleanup); + + prog = getenv ("EDITOR"); + if (prog) + editor_program = prog; + + diffarg (DEFAULT_DIFF_PROGRAM); + + /* parse command line args */ + while ((opt = getopt_long (argc, argv, "abBdEHiI:lo:stvw:W", longopts, 0)) + != -1) + { + switch (opt) + { + case 'a': + diffarg ("-a"); + break; + + case 'b': + diffarg ("-b"); + break; + + case 'B': + diffarg ("-B"); + break; + + case 'd': + diffarg ("-d"); + break; + + case 'E': + diffarg ("-E"); + break; + + case 'H': + diffarg ("-H"); + break; + + case 'i': + diffarg ("-i"); + break; + + case 'I': + diffarg ("-I"); + diffarg (optarg); + break; + + case 'l': + diffarg ("--left-column"); + break; + + case 'o': + output = optarg; + break; + + case 's': + suppress_common_lines = true; + break; + + case 't': + diffarg ("-t"); + break; + + case 'v': + version_etc (stdout, "sdiff", PACKAGE_NAME, PACKAGE_VERSION, + "Thomas Lord", (char *) 0); + check_stdout (); + return EXIT_SUCCESS; + + case 'w': + diffarg ("-W"); + diffarg (optarg); + break; + + case 'W': + diffarg ("-w"); + break; + + case DIFF_PROGRAM_OPTION: + diffargv[0] = optarg; + break; + + case HELP_OPTION: + usage (); + check_stdout (); + return EXIT_SUCCESS; + + case STRIP_TRAILING_CR_OPTION: + diffarg ("--strip-trailing-cr"); + break; + + case TABSIZE_OPTION: + diffarg ("--tabsize"); + diffarg (optarg); + break; + + default: + try_help (0, 0); + } + } + + if (argc - optind != 2) + { + if (argc - optind < 2) + try_help ("missing operand after `%s'", argv[argc - 1]); + else + try_help ("extra operand `%s'", argv[optind + 2]); + } + + if (! output) + { + /* easy case: diff does everything for us */ + if (suppress_common_lines) + diffarg ("--suppress-common-lines"); + diffarg ("-y"); + diffarg ("--"); + diffarg (argv[optind]); + diffarg (argv[optind + 1]); + diffarg (0); + execvp (diffargv[0], (char **) diffargv); + perror_fatal (diffargv[0]); + } + else + { + char const *lname, *rname; + FILE *left, *right, *out, *diffout; + bool interact_ok; + struct line_filter lfilt; + struct line_filter rfilt; + struct line_filter diff_filt; + bool leftdir = diraccess (argv[optind]); + bool rightdir = diraccess (argv[optind + 1]); + + if (leftdir & rightdir) + fatal ("both files to be compared are directories"); + + lname = expand_name (argv[optind], leftdir, argv[optind + 1]); + left = ck_fopen (lname, "r"); + rname = expand_name (argv[optind + 1], rightdir, argv[optind]); + right = ck_fopen (rname, "r"); + out = ck_fopen (output, "w"); + + diffarg ("--sdiff-merge-assist"); + diffarg ("--"); + diffarg (argv[optind]); + diffarg (argv[optind + 1]); + diffarg (0); + + trapsigs (); + +#if ! (HAVE_WORKING_FORK || HAVE_WORKING_VFORK) + { + size_t cmdsize = 1; + char *p, *command; + int i; + + for (i = 0; diffargv[i]; i++) + cmdsize += quote_system_arg (0, diffargv[i]) + 1; + command = p = xmalloc (cmdsize); + for (i = 0; diffargv[i]; i++) + { + p += quote_system_arg (p, diffargv[i]); + *p++ = ' '; + } + p[-1] = 0; + errno = 0; + diffout = popen (command, "r"); + if (! diffout) + perror_fatal (command); + free (command); + } +#else + { + int diff_fds[2]; +# if HAVE_WORKING_VFORK + sigset_t procmask; + sigset_t blocked; +# endif + + if (pipe (diff_fds) != 0) + perror_fatal ("pipe"); + +# if HAVE_WORKING_VFORK + /* Block SIGINT and SIGPIPE. */ + sigemptyset (&blocked); + sigaddset (&blocked, SIGINT); + sigaddset (&blocked, SIGPIPE); + sigprocmask (SIG_BLOCK, &blocked, &procmask); +# endif + diffpid = vfork (); + if (diffpid < 0) + perror_fatal ("fork"); + if (! diffpid) + { + /* Alter the child's SIGINT and SIGPIPE handlers; + this may munge the parent. + The child ignores SIGINT in case the user interrupts the editor. + The child does not ignore SIGPIPE, even if the parent does. */ + if (initial_handler (handler_index_of_SIGINT) != SIG_IGN) + signal_handler (SIGINT, SIG_IGN); + signal_handler (SIGPIPE, SIG_DFL); +# if HAVE_WORKING_VFORK + /* Stop blocking SIGINT and SIGPIPE in the child. */ + sigprocmask (SIG_SETMASK, &procmask, 0); +# endif + close (diff_fds[0]); + if (diff_fds[1] != STDOUT_FILENO) + { + dup2 (diff_fds[1], STDOUT_FILENO); + close (diff_fds[1]); + } + + execvp (diffargv[0], (char **) diffargv); + _exit (errno == ENOENT ? 127 : 126); + } + +# if HAVE_WORKING_VFORK + /* Restore the parent's SIGINT and SIGPIPE behavior. */ + if (initial_handler (handler_index_of_SIGINT) != SIG_IGN) + signal_handler (SIGINT, catchsig); + if (initial_handler (handler_index_of_SIGPIPE) != SIG_IGN) + signal_handler (SIGPIPE, catchsig); + else + signal_handler (SIGPIPE, SIG_IGN); + + /* Stop blocking SIGINT and SIGPIPE in the parent. */ + sigprocmask (SIG_SETMASK, &procmask, 0); +# endif + + close (diff_fds[1]); + diffout = fdopen (diff_fds[0], "r"); + if (! diffout) + perror_fatal ("fdopen"); + } +#endif + + lf_init (&diff_filt, diffout); + lf_init (&lfilt, left); + lf_init (&rfilt, right); + + interact_ok = interact (&diff_filt, &lfilt, lname, &rfilt, rname, out); + + ck_fclose (left); + ck_fclose (right); + ck_fclose (out); + + { + int wstatus; + int werrno = 0; + +#if ! (HAVE_WORKING_FORK || HAVE_WORKING_VFORK) + wstatus = pclose (diffout); + if (wstatus == -1) + werrno = errno; +#else + ck_fclose (diffout); + while (waitpid (diffpid, &wstatus, 0) < 0) + if (errno == EINTR) + checksigs (); + else + perror_fatal ("waitpid"); + diffpid = 0; +#endif + + if (tmpname) + { + unlink (tmpname); + tmpname = 0; + } + + if (! interact_ok) + exiterr (); + + check_child_status (werrno, wstatus, EXIT_FAILURE, diffargv[0]); + untrapsig (0); + checksigs (); + exit (WEXITSTATUS (wstatus)); + } + } + return EXIT_SUCCESS; /* Fool `-Wall'. */ +} + +static void +diffarg (char const *a) +{ + static size_t diffargs, diffarglim; + + if (diffargs == diffarglim) + { + if (! diffarglim) + diffarglim = 16; + else if (PTRDIFF_MAX / (2 * sizeof *diffargv) <= diffarglim) + xalloc_die (); + else + diffarglim *= 2; + diffargv = xrealloc (diffargv, diffarglim * sizeof *diffargv); + } + diffargv[diffargs++] = a; +} + +/* Signal handling */ + +static bool volatile ignore_SIGINT; +static int volatile signal_received; +static bool sigs_trapped; + +static void +catchsig (int s) +{ +#if ! HAVE_SIGACTION + signal (s, SIG_IGN); +#endif + if (! (s == SIGINT && ignore_SIGINT)) + signal_received = s; +} + +#if HAVE_SIGACTION +static struct sigaction catchaction; + +static void +signal_handler (int sig, void (*handler) (int)) +{ + catchaction.sa_handler = handler; + sigaction (sig, &catchaction, 0); +} +#endif + +static void +trapsigs (void) +{ + int i; + +#if HAVE_SIGACTION + catchaction.sa_flags = SA_RESTART; + sigemptyset (&catchaction.sa_mask); + for (i = 0; i < NUM_SIGS; i++) + sigaddset (&catchaction.sa_mask, sigs[i]); +#endif + + for (i = 0; i < NUM_SIGS; i++) + { +#if HAVE_SIGACTION + sigaction (sigs[i], 0, &initial_action[i]); +#else + initial_action[i] = signal (sigs[i], SIG_IGN); +#endif + if (initial_handler (i) != SIG_IGN) + signal_handler (sigs[i], catchsig); + } + +#ifdef SIGCHLD + /* System V fork+wait does not work if SIGCHLD is ignored. */ + signal (SIGCHLD, SIG_DFL); +#endif + + sigs_trapped = true; +} + +/* Untrap signal S, or all trapped signals if S is zero. */ +static void +untrapsig (int s) +{ + int i; + + if (sigs_trapped) + for (i = 0; i < NUM_SIGS; i++) + if ((! s || sigs[i] == s) && initial_handler (i) != SIG_IGN) + { +#if HAVE_SIGACTION + sigaction (sigs[i], &initial_action[i], 0); +#else + signal (sigs[i], initial_action[i]); +#endif + } +} + +/* Exit if a signal has been received. */ +static void +checksigs (void) +{ + int s = signal_received; + if (s) + { + cleanup (0); + + /* Yield an exit status indicating that a signal was received. */ + untrapsig (s); + kill (getpid (), s); + + /* That didn't work, so exit with error status. */ + exit (EXIT_TROUBLE); + } +} + +static void +give_help (void) +{ + fprintf (stderr, "%s", _("\ +ed:\tEdit then use both versions, each decorated with a header.\n\ +eb:\tEdit then use both versions.\n\ +el:\tEdit then use the left version.\n\ +er:\tEdit then use the right version.\n\ +e:\tEdit a new version.\n\ +l:\tUse the left version.\n\ +r:\tUse the right version.\n\ +s:\tSilently include common lines.\n\ +v:\tVerbosely include common lines.\n\ +q:\tQuit.\n\ +")); +} + +static int +skip_white (void) +{ + int c; + for (;;) + { + c = getchar (); + if (! isspace (c) || c == '\n') + break; + checksigs (); + } + if (ferror (stdin)) + perror_fatal (_("read failed")); + return c; +} + +static void +flush_line (void) +{ + int c; + while ((c = getchar ()) != '\n' && c != EOF) + continue; + if (ferror (stdin)) + perror_fatal (_("read failed")); +} + + +/* interpret an edit command */ +static bool +edit (struct line_filter *left, char const *lname, lin lline, lin llen, + struct line_filter *right, char const *rname, lin rline, lin rlen, + FILE *outfile) +{ + for (;;) + { + int cmd0, cmd1; + bool gotcmd = false; + + cmd1 = 0; /* Pacify `gcc -W'. */ + + while (! gotcmd) + { + if (putchar ('%') != '%') + perror_fatal (_("write failed")); + ck_fflush (stdout); + + cmd0 = skip_white (); + switch (cmd0) + { + case 'l': case 'r': case 's': case 'v': case 'q': + if (skip_white () != '\n') + { + give_help (); + flush_line (); + continue; + } + gotcmd = true; + break; + + case 'e': + cmd1 = skip_white (); + switch (cmd1) + { + case 'b': case 'd': case 'l': case 'r': + if (skip_white () != '\n') + { + give_help (); + flush_line (); + continue; + } + gotcmd = true; + break; + case '\n': + gotcmd = true; + break; + default: + give_help (); + flush_line (); + continue; + } + break; + + case EOF: + if (feof (stdin)) + { + gotcmd = true; + cmd0 = 'q'; + break; + } + /* Fall through. */ + default: + flush_line (); + /* Fall through. */ + case '\n': + give_help (); + continue; + } + } + + switch (cmd0) + { + case 'l': + lf_copy (left, llen, outfile); + lf_skip (right, rlen); + return true; + case 'r': + lf_copy (right, rlen, outfile); + lf_skip (left, llen); + return true; + case 's': + suppress_common_lines = true; + break; + case 'v': + suppress_common_lines = false; + break; + case 'q': + return false; + case 'e': + { + int fd; + + if (tmpname) + tmp = fopen (tmpname, "w"); + else + { + if ((fd = temporary_file ()) < 0) + perror_fatal ("mkstemp"); + tmp = fdopen (fd, "w"); + } + + if (! tmp) + perror_fatal (tmpname); + + switch (cmd1) + { + case 'd': + if (llen) + { + if (llen == 1) + fprintf (tmp, "--- %s %ld\n", lname, (long int) lline); + else + fprintf (tmp, "--- %s %ld,%ld\n", lname, + (long int) lline, + (long int) (lline + llen - 1)); + } + /* Fall through. */ + case 'b': case 'l': + lf_copy (left, llen, tmp); + break; + + default: + lf_skip (left, llen); + break; + } + + switch (cmd1) + { + case 'd': + if (rlen) + { + if (rlen == 1) + fprintf (tmp, "+++ %s %ld\n", rname, (long int) rline); + else + fprintf (tmp, "+++ %s %ld,%ld\n", rname, + (long int) rline, + (long int) (rline + rlen - 1)); + } + /* Fall through. */ + case 'b': case 'r': + lf_copy (right, rlen, tmp); + break; + + default: + lf_skip (right, rlen); + break; + } + + ck_fclose (tmp); + + { + int wstatus; + int werrno = 0; + ignore_SIGINT = true; + checksigs (); + + { +#if ! (HAVE_WORKING_FORK || HAVE_WORKING_VFORK) + char *command = + xmalloc (quote_system_arg (0, editor_program) + + 1 + strlen (tmpname) + 1); + sprintf (command + quote_system_arg (command, editor_program), + " %s", tmpname); + wstatus = system (command); + if (wstatus == -1) + werrno = errno; + free (command); +#else + pid_t pid; + + pid = vfork (); + if (pid == 0) + { + char const *argv[3]; + int i = 0; + + argv[i++] = editor_program; + argv[i++] = tmpname; + argv[i] = 0; + + execvp (editor_program, (char **) argv); + _exit (errno == ENOENT ? 127 : 126); + } + + if (pid < 0) + perror_fatal ("fork"); + + while (waitpid (pid, &wstatus, 0) < 0) + if (errno == EINTR) + checksigs (); + else + perror_fatal ("waitpid"); +#endif + } + + ignore_SIGINT = false; + check_child_status (werrno, wstatus, EXIT_SUCCESS, + editor_program); + } + + { + char buf[SDIFF_BUFSIZE]; + size_t size; + tmp = ck_fopen (tmpname, "r"); + while ((size = ck_fread (buf, SDIFF_BUFSIZE, tmp)) != 0) + { + checksigs (); + ck_fwrite (buf, size, outfile); + } + ck_fclose (tmp); + } + return true; + } + default: + give_help (); + break; + } + } +} + +/* Alternately reveal bursts of diff output and handle user commands. */ +static bool +interact (struct line_filter *diff, + struct line_filter *left, char const *lname, + struct line_filter *right, char const *rname, + FILE *outfile) +{ + lin lline = 1, rline = 1; + + for (;;) + { + char diff_help[256]; + int snarfed = lf_snarf (diff, diff_help, sizeof diff_help); + + if (snarfed <= 0) + return snarfed != 0; + + checksigs (); + + if (diff_help[0] == ' ') + puts (diff_help + 1); + else + { + char *numend; + uintmax_t val; + lin llen, rlen, lenmax; + errno = 0; + llen = val = strtoumax (diff_help + 1, &numend, 10); + if (llen < 0 || llen != val || errno || *numend != ',') + fatal (diff_help); + rlen = val = strtoumax (numend + 1, &numend, 10); + if (rlen < 0 || rlen != val || errno || *numend) + fatal (diff_help); + + lenmax = MAX (llen, rlen); + + switch (diff_help[0]) + { + case 'i': + if (suppress_common_lines) + lf_skip (diff, lenmax); + else + lf_copy (diff, lenmax, stdout); + + lf_copy (left, llen, outfile); + lf_skip (right, rlen); + break; + + case 'c': + lf_copy (diff, lenmax, stdout); + if (! edit (left, lname, lline, llen, + right, rname, rline, rlen, + outfile)) + return false; + break; + + default: + fatal (diff_help); + } + + lline += llen; + rline += rlen; + } + } +} + +/* Return true if DIR is an existing directory. */ +static bool +diraccess (char const *dir) +{ + struct stat buf; + return stat (dir, &buf) == 0 && S_ISDIR (buf.st_mode); +} + +#ifndef P_tmpdir +# define P_tmpdir "/tmp" +#endif +#ifndef TMPDIR_ENV +# define TMPDIR_ENV "TMPDIR" +#endif + +/* Open a temporary file and return its file descriptor. Put into + tmpname the address of a newly allocated buffer that holds the + file's name. Use the prefix "sdiff". */ +static int +temporary_file (void) +{ + char const *tmpdir = getenv (TMPDIR_ENV); + char const *dir = tmpdir ? tmpdir : P_tmpdir; + char *buf = xmalloc (strlen (dir) + 1 + 5 + 6 + 1); + int fd; + int e; + sigset_t procmask; + sigset_t blocked; + sprintf (buf, "%s/sdiffXXXXXX", dir); + sigemptyset (&blocked); + sigaddset (&blocked, SIGINT); + sigprocmask (SIG_BLOCK, &blocked, &procmask); + fd = mkstemp (buf); + e = errno; + if (0 <= fd) + tmpname = buf; + sigprocmask (SIG_SETMASK, &procmask, 0); + errno = e; + return fd; +} diff --git a/contrib/diff/src/side.c b/contrib/diff/src/side.c new file mode 100644 index 0000000..3eaef4d --- /dev/null +++ b/contrib/diff/src/side.c @@ -0,0 +1,281 @@ +/* sdiff-format output routines for GNU DIFF. + + Copyright (C) 1991, 1992, 1993, 1998, 2001, 2002, 2004 Free + Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY. No author or distributor + accepts responsibility to anyone for the consequences of using it + or for whether it serves any particular purpose or works at all, + unless he says so in writing. Refer to the GNU DIFF General Public + License for full details. + + Everyone is granted permission to copy, modify and redistribute + GNU DIFF, but only under the conditions described in the + GNU DIFF General Public License. A copy of this license is + supposed to have been given to you along with GNU DIFF so you + can know your rights and responsibilities. It should be in a + file named COPYING. Among other things, the copyright notice + and this notice must be preserved on all copies. */ + +#include "diff.h" + +static void print_sdiff_common_lines (lin, lin); +static void print_sdiff_hunk (struct change *); + +/* Next line number to be printed in the two input files. */ +static lin next0, next1; + +/* Print the edit-script SCRIPT as a sdiff style output. */ + +void +print_sdiff_script (struct change *script) +{ + begin_output (); + + next0 = next1 = - files[0].prefix_lines; + print_script (script, find_change, print_sdiff_hunk); + + print_sdiff_common_lines (files[0].valid_lines, files[1].valid_lines); +} + +/* Tab from column FROM to column TO, where FROM <= TO. Yield TO. */ + +static size_t +tab_from_to (size_t from, size_t to) +{ + FILE *out = outfile; + size_t tab; + size_t tab_size = tabsize; + + if (!expand_tabs) + for (tab = from + tab_size - from % tab_size; tab <= to; tab += tab_size) + { + putc ('\t', out); + from = tab; + } + while (from++ < to) + putc (' ', out); + return to; +} + +/* Print the text for half an sdiff line. This means truncate to + width observing tabs, and trim a trailing newline. Return the + last column written (not the number of chars). */ + +static size_t +print_half_line (char const *const *line, size_t indent, size_t out_bound) +{ + FILE *out = outfile; + register size_t in_position = 0; + register size_t out_position = 0; + register char const *text_pointer = line[0]; + register char const *text_limit = line[1]; + + while (text_pointer < text_limit) + { + register unsigned char c = *text_pointer++; + + switch (c) + { + case '\t': + { + size_t spaces = tabsize - in_position % tabsize; + if (in_position == out_position) + { + size_t tabstop = out_position + spaces; + if (expand_tabs) + { + if (out_bound < tabstop) + tabstop = out_bound; + for (; out_position < tabstop; out_position++) + putc (' ', out); + } + else + if (tabstop < out_bound) + { + out_position = tabstop; + putc (c, out); + } + } + in_position += spaces; + } + break; + + case '\r': + { + putc (c, out); + tab_from_to (0, indent); + in_position = out_position = 0; + } + break; + + case '\b': + if (in_position != 0 && --in_position < out_bound) + { + if (out_position <= in_position) + /* Add spaces to make up for suppressed tab past out_bound. */ + for (; out_position < in_position; out_position++) + putc (' ', out); + else + { + out_position = in_position; + putc (c, out); + } + } + break; + + case '\f': + case '\v': + control_char: + if (in_position < out_bound) + putc (c, out); + break; + + default: + if (! isprint (c)) + goto control_char; + /* falls through */ + case ' ': + if (in_position++ < out_bound) + { + out_position = in_position; + putc (c, out); + } + break; + + case '\n': + return out_position; + } + } + + return out_position; +} + +/* Print side by side lines with a separator in the middle. + 0 parameters are taken to indicate white space text. + Blank lines that can easily be caught are reduced to a single newline. */ + +static void +print_1sdiff_line (char const *const *left, char sep, + char const *const *right) +{ + FILE *out = outfile; + size_t hw = sdiff_half_width; + size_t c2o = sdiff_column2_offset; + size_t col = 0; + bool put_newline = false; + + if (left) + { + put_newline |= left[1][-1] == '\n'; + col = print_half_line (left, 0, hw); + } + + if (sep != ' ') + { + col = tab_from_to (col, (hw + c2o - 1) / 2) + 1; + if (sep == '|' && put_newline != (right[1][-1] == '\n')) + sep = put_newline ? '/' : '\\'; + putc (sep, out); + } + + if (right) + { + put_newline |= right[1][-1] == '\n'; + if (**right != '\n') + { + col = tab_from_to (col, c2o); + print_half_line (right, col, hw); + } + } + + if (put_newline) + putc ('\n', out); +} + +/* Print lines common to both files in side-by-side format. */ +static void +print_sdiff_common_lines (lin limit0, lin limit1) +{ + lin i0 = next0, i1 = next1; + + if (!suppress_common_lines && (i0 != limit0 || i1 != limit1)) + { + if (sdiff_merge_assist) + { + long int len0 = limit0 - i0; + long int len1 = limit1 - i1; + fprintf (outfile, "i%ld,%ld\n", len0, len1); + } + + if (!left_column) + { + while (i0 != limit0 && i1 != limit1) + print_1sdiff_line (&files[0].linbuf[i0++], ' ', + &files[1].linbuf[i1++]); + while (i1 != limit1) + print_1sdiff_line (0, ')', &files[1].linbuf[i1++]); + } + while (i0 != limit0) + print_1sdiff_line (&files[0].linbuf[i0++], '(', 0); + } + + next0 = limit0; + next1 = limit1; +} + +/* Print a hunk of an sdiff diff. + This is a contiguous portion of a complete edit script, + describing changes in consecutive lines. */ + +static void +print_sdiff_hunk (struct change *hunk) +{ + lin first0, last0, first1, last1; + register lin i, j; + + /* Determine range of line numbers involved in each file. */ + enum changes changes = + analyze_hunk (hunk, &first0, &last0, &first1, &last1); + if (!changes) + return; + + /* Print out lines up to this change. */ + print_sdiff_common_lines (first0, first1); + + if (sdiff_merge_assist) + { + long int len0 = last0 - first0 + 1; + long int len1 = last1 - first1 + 1; + fprintf (outfile, "c%ld,%ld\n", len0, len1); + } + + /* Print ``xxx | xxx '' lines */ + if (changes == CHANGED) + { + for (i = first0, j = first1; i <= last0 && j <= last1; i++, j++) + print_1sdiff_line (&files[0].linbuf[i], '|', &files[1].linbuf[j]); + changes = (i <= last0 ? OLD : 0) + (j <= last1 ? NEW : 0); + next0 = first0 = i; + next1 = first1 = j; + } + + /* Print `` > xxx '' lines */ + if (changes & NEW) + { + for (j = first1; j <= last1; ++j) + print_1sdiff_line (0, '>', &files[1].linbuf[j]); + next1 = j; + } + + /* Print ``xxx < '' lines */ + if (changes & OLD) + { + for (i = first0; i <= last0; ++i) + print_1sdiff_line (&files[0].linbuf[i], '<', 0); + next0 = i; + } +} diff --git a/contrib/diff/src/system.h b/contrib/diff/src/system.h new file mode 100644 index 0000000..e480cc5 --- /dev/null +++ b/contrib/diff/src/system.h @@ -0,0 +1,326 @@ +/* System dependent declarations. + + Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1998, 2001, 2002, + 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include + +/* Don't bother to support K&R C compilers any more; it's not worth + the trouble. These macros prevent some library modules from being + compiled in K&R C mode. */ +#define PARAMS(Args) Args +#define PROTOTYPES 1 + +/* Define `__attribute__' and `volatile' first + so that they're used consistently in all system includes. */ +#if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 6) || __STRICT_ANSI__ +# define __attribute__(x) +#endif +#if defined const && !defined volatile +# define volatile +#endif + +/* Verify a requirement at compile-time (unlike assert, which is runtime). */ +#define verify(name, assertion) struct name { char a[(assertion) ? 1 : -1]; } + + +/* Determine whether an integer type is signed, and its bounds. + This code assumes two's (or one's!) complement with no holes. */ + +/* The extra casts work around common compiler bugs, + e.g. Cray C 5.0.3.0 when t == time_t. */ +#ifndef TYPE_SIGNED +# define TYPE_SIGNED(t) (! ((t) 0 < (t) -1)) +#endif +#ifndef TYPE_MINIMUM +# define TYPE_MINIMUM(t) ((t) (TYPE_SIGNED (t) \ + ? ~ (t) 0 << (sizeof (t) * CHAR_BIT - 1) \ + : (t) 0)) +#endif +#ifndef TYPE_MAXIMUM +# define TYPE_MAXIMUM(t) ((t) (~ (t) 0 - TYPE_MINIMUM (t))) +#endif + +#include +#include + +#ifndef S_IRWXU +# define S_IRWXU 0700 +#endif +#ifndef S_IRWXG +# define S_IRWXG 0070 +#endif +#ifndef S_IRWXO +# define S_IRWXO 0007 +#endif + +#if HAVE_UNISTD_H +# include +#endif + +#ifndef SEEK_SET +# define SEEK_SET 0 +#endif +#ifndef SEEK_CUR +# define SEEK_CUR 1 +#endif + +#ifndef STDIN_FILENO +# define STDIN_FILENO 0 +#endif +#ifndef STDOUT_FILENO +# define STDOUT_FILENO 1 +#endif +#ifndef STDERR_FILENO +# define STDERR_FILENO 2 +#endif + +#include + +#if HAVE_FCNTL_H +# include +#else +# if HAVE_SYS_FILE_H +# include +# endif +#endif + +#if !HAVE_DUP2 +# define dup2(f, t) (close (t), fcntl (f, F_DUPFD, t)) +#endif + +#ifndef O_RDONLY +# define O_RDONLY 0 +#endif +#ifndef O_RDWR +# define O_RDWR 2 +#endif +#ifndef S_IRUSR +# define S_IRUSR 0400 +#endif +#ifndef S_IWUSR +# define S_IWUSR 0200 +#endif + +#if HAVE_SYS_WAIT_H +# include +#endif +#ifndef WEXITSTATUS +# define WEXITSTATUS(stat_val) ((unsigned int) (stat_val) >> 8) +#endif +#ifndef WIFEXITED +# define WIFEXITED(stat_val) (((stat_val) & 255) == 0) +#endif + +#ifndef STAT_BLOCKSIZE +# if HAVE_STRUCT_STAT_ST_BLKSIZE +# define STAT_BLOCKSIZE(s) ((s).st_blksize) +# else +# define STAT_BLOCKSIZE(s) (8 * 1024) +# endif +#endif + +#if HAVE_DIRENT_H +# include +# define NAMLEN(dirent) strlen ((dirent)->d_name) +#else +# define dirent direct +# define NAMLEN(dirent) ((dirent)->d_namlen) +# if HAVE_SYS_NDIR_H +# include +# endif +# if HAVE_SYS_DIR_H +# include +# endif +# if HAVE_NDIR_H +# include +# endif +#endif + +#include +#define EXIT_TROUBLE 2 + +#include + +#if HAVE_INTTYPES_H +# include +#endif +#ifndef PTRDIFF_MAX +# define PTRDIFF_MAX TYPE_MAXIMUM (ptrdiff_t) +#endif +#ifndef SIZE_MAX +# define SIZE_MAX TYPE_MAXIMUM (size_t) +#endif +#ifndef UINTMAX_MAX +# define UINTMAX_MAX TYPE_MAXIMUM (uintmax_t) +#endif +#if ! HAVE_STRTOUMAX && ! defined strtoumax +uintmax_t strtoumax (char const *, char **, int); +#endif + +#include + +#include +#if ! HAVE_STRCASECOLL +# if HAVE_STRICOLL || defined stricoll +# define strcasecoll(a, b) stricoll (a, b) +# else +# define strcasecoll(a, b) strcasecmp (a, b) /* best we can do */ +# endif +#endif +#if ! (HAVE_STRCASECMP || defined strcasecmp) +int strcasecmp (char const *, char const *); +#endif + +#if HAVE_LOCALE_H +# include +#else +# define setlocale(category, locale) +#endif + +#include + +#define _(msgid) gettext (msgid) +#define N_(msgid) msgid + +#include + +/* ISDIGIT differs from isdigit, as follows: + - Its arg may be any int or unsigned int; it need not be an unsigned char. + - It's guaranteed to evaluate its argument exactly once. + - It's typically faster. + POSIX 1003.1-2001 says that only '0' through '9' are digits. + Prefer ISDIGIT to isdigit unless it's important to use the locale's + definition of `digit' even when the host does not conform to POSIX. */ +#define ISDIGIT(c) ((unsigned int) (c) - '0' <= 9) + +#include + +#include +#ifndef SA_RESTART +# ifdef SA_INTERRUPT /* e.g. SunOS 4.1.x */ +# define SA_RESTART SA_INTERRUPT +# else +# define SA_RESTART 0 +# endif +#endif +#if !defined SIGCHLD && defined SIGCLD +# define SIGCHLD SIGCLD +#endif + +#undef MIN +#undef MAX +#define MIN(a, b) ((a) <= (b) ? (a) : (b)) +#define MAX(a, b) ((a) >= (b) ? (a) : (b)) + +#include + +#if HAVE_VFORK_H +# include +#endif + +#if ! HAVE_WORKING_VFORK +# define vfork fork +#endif + +/* Type used for fast comparison of several bytes at a time. */ + +#ifndef word +# define word uintmax_t +#endif + +/* The integer type of a line number. Since files are read into main + memory, ptrdiff_t should be wide enough. */ + +typedef ptrdiff_t lin; +#define LIN_MAX PTRDIFF_MAX +verify (lin_is_signed, TYPE_SIGNED (lin)); +verify (lin_is_wide_enough, sizeof (ptrdiff_t) <= sizeof (lin)); +verify (lin_is_printable_as_long_int, sizeof (lin) <= sizeof (long int)); + +/* This section contains POSIX-compliant defaults for macros + that are meant to be overridden by hand in config.h as needed. */ + +#ifndef file_name_cmp +# define file_name_cmp strcmp +#endif + +#ifndef initialize_main +# define initialize_main(argcp, argvp) +#endif + +#ifndef NULL_DEVICE +# define NULL_DEVICE "/dev/null" +#endif + +/* Do struct stat *S, *T describe the same special file? */ +#ifndef same_special_file +# if HAVE_ST_RDEV && defined S_ISBLK && defined S_ISCHR +# define same_special_file(s, t) \ + (((S_ISBLK ((s)->st_mode) && S_ISBLK ((t)->st_mode)) \ + || (S_ISCHR ((s)->st_mode) && S_ISCHR ((t)->st_mode))) \ + && (s)->st_rdev == (t)->st_rdev) +# else +# define same_special_file(s, t) 0 +# endif +#endif + +/* Do struct stat *S, *T describe the same file? Answer -1 if unknown. */ +#ifndef same_file +# define same_file(s, t) \ + ((((s)->st_ino == (t)->st_ino) && ((s)->st_dev == (t)->st_dev)) \ + || same_special_file (s, t)) +#endif + +/* Do struct stat *S, *T have the same file attributes? + + POSIX says that two files are identical if st_ino and st_dev are + the same, but many filesystems incorrectly assign the same (device, + inode) pair to two distinct files, including: + + - GNU/Linux NFS servers that export all local filesystems as a + single NFS filesystem, if a local device number (st_dev) exceeds + 255, or if a local inode number (st_ino) exceeds 16777215. + + - Network Appliance NFS servers in snapshot directories; see + Network Appliance bug #195. + + - ClearCase MVFS; see bug id ATRia04618. + + Check whether two files that purport to be the same have the same + attributes, to work around instances of this common bug. Do not + inspect all attributes, only attributes useful in checking for this + bug. + + It's possible for two distinct files on a buggy filesystem to have + the same attributes, but it's not worth slowing down all + implementations (or complicating the configuration) to cater to + these rare cases in buggy implementations. */ + +#ifndef same_file_attributes +# define same_file_attributes(s, t) \ + ((s)->st_mode == (t)->st_mode \ + && (s)->st_nlink == (t)->st_nlink \ + && (s)->st_uid == (t)->st_uid \ + && (s)->st_gid == (t)->st_gid \ + && (s)->st_size == (t)->st_size \ + && (s)->st_mtime == (t)->st_mtime \ + && (s)->st_ctime == (t)->st_ctime) +#endif diff --git a/contrib/diff/src/util.c b/contrib/diff/src/util.c new file mode 100644 index 0000000..809f125 --- /dev/null +++ b/contrib/diff/src/util.c @@ -0,0 +1,775 @@ +/* Support routines for GNU DIFF. + + Copyright (C) 1988, 1989, 1992, 1993, 1994, 1995, 1998, 2001, 2002, + 2004 Free Software Foundation, Inc. + + This file is part of GNU DIFF. + + GNU DIFF is free software; you can redistribute it and/or modify + it under the terms of the GNU General Public License as published by + the Free Software Foundation; either version 2, or (at your option) + any later version. + + GNU DIFF is distributed in the hope that it will be useful, + but WITHOUT ANY WARRANTY; without even the implied warranty of + MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the + GNU General Public License for more details. + + You should have received a copy of the GNU General Public License + along with this program; see the file COPYING. + If not, write to the Free Software Foundation, + 59 Temple Place - Suite 330, Boston, MA 02111-1307, USA. */ + +#include "diff.h" +#include +#include +#include +#include + +char const pr_program[] = PR_PROGRAM; + +/* Queue up one-line messages to be printed at the end, + when -l is specified. Each message is recorded with a `struct msg'. */ + +struct msg +{ + struct msg *next; + char args[1]; /* Format + 4 args, each '\0' terminated, concatenated. */ +}; + +/* Head of the chain of queues messages. */ + +static struct msg *msg_chain; + +/* Tail of the chain of queues messages. */ + +static struct msg **msg_chain_end = &msg_chain; + +/* Use when a system call returns non-zero status. + NAME should normally be the file name. */ + +void +perror_with_name (char const *name) +{ + error (0, errno, "%s", name); +} + +/* Use when a system call returns non-zero status and that is fatal. */ + +void +pfatal_with_name (char const *name) +{ + int e = errno; + print_message_queue (); + error (EXIT_TROUBLE, e, "%s", name); + abort (); +} + +/* Print an error message containing MSGID, then exit. */ + +void +fatal (char const *msgid) +{ + print_message_queue (); + error (EXIT_TROUBLE, 0, "%s", _(msgid)); + abort (); +} + +/* Like printf, except if -l in effect then save the message and print later. + This is used for things like "Only in ...". */ + +void +message (char const *format_msgid, char const *arg1, char const *arg2) +{ + message5 (format_msgid, arg1, arg2, 0, 0); +} + +void +message5 (char const *format_msgid, char const *arg1, char const *arg2, + char const *arg3, char const *arg4) +{ + if (paginate) + { + char *p; + char const *arg[5]; + int i; + size_t size[5]; + size_t total_size = offsetof (struct msg, args); + struct msg *new; + + arg[0] = format_msgid; + arg[1] = arg1; + arg[2] = arg2; + arg[3] = arg3 ? arg3 : ""; + arg[4] = arg4 ? arg4 : ""; + + for (i = 0; i < 5; i++) + total_size += size[i] = strlen (arg[i]) + 1; + + new = xmalloc (total_size); + + for (i = 0, p = new->args; i < 5; p += size[i++]) + memcpy (p, arg[i], size[i]); + + *msg_chain_end = new; + new->next = 0; + msg_chain_end = &new->next; + } + else + { + if (sdiff_merge_assist) + putchar (' '); + printf (_(format_msgid), arg1, arg2, arg3, arg4); + } +} + +/* Output all the messages that were saved up by calls to `message'. */ + +void +print_message_queue (void) +{ + char const *arg[5]; + int i; + struct msg *m = msg_chain; + + while (m) + { + struct msg *next = m->next; + arg[0] = m->args; + for (i = 0; i < 4; i++) + arg[i + 1] = arg[i] + strlen (arg[i]) + 1; + printf (_(arg[0]), arg[1], arg[2], arg[3], arg[4]); + free (m); + m = next; + } +} + +/* Call before outputting the results of comparing files NAME0 and NAME1 + to set up OUTFILE, the stdio stream for the output to go to. + + Usually, OUTFILE is just stdout. But when -l was specified + we fork off a `pr' and make OUTFILE a pipe to it. + `pr' then outputs to our stdout. */ + +static char const *current_name0; +static char const *current_name1; +static bool currently_recursive; + +void +setup_output (char const *name0, char const *name1, bool recursive) +{ + current_name0 = name0; + current_name1 = name1; + currently_recursive = recursive; + outfile = 0; +} + +#if HAVE_WORKING_FORK || HAVE_WORKING_VFORK +static pid_t pr_pid; +#endif + +void +begin_output (void) +{ + char *name; + + if (outfile != 0) + return; + + /* Construct the header of this piece of diff. */ + name = xmalloc (strlen (current_name0) + strlen (current_name1) + + strlen (switch_string) + 7); + + /* POSIX 1003.1-2001 specifies this format. But there are some bugs in + the standard: it says that we must print only the last component + of the pathnames, and it requires two spaces after "diff" if + there are no options. These requirements are silly and do not + match historical practice. */ + sprintf (name, "diff%s %s %s", switch_string, current_name0, current_name1); + + if (paginate) + { + if (fflush (stdout) != 0) + pfatal_with_name (_("write failed")); + + /* Make OUTFILE a pipe to a subsidiary `pr'. */ + { +#if HAVE_WORKING_FORK || HAVE_WORKING_VFORK + int pipes[2]; + + if (pipe (pipes) != 0) + pfatal_with_name ("pipe"); + + pr_pid = vfork (); + if (pr_pid < 0) + pfatal_with_name ("fork"); + + if (pr_pid == 0) + { + close (pipes[1]); + if (pipes[0] != STDIN_FILENO) + { + if (dup2 (pipes[0], STDIN_FILENO) < 0) + pfatal_with_name ("dup2"); + close (pipes[0]); + } + + execl (pr_program, pr_program, "-h", name, (char *) 0); + _exit (errno == ENOENT ? 127 : 126); + } + else + { + close (pipes[0]); + outfile = fdopen (pipes[1], "w"); + if (!outfile) + pfatal_with_name ("fdopen"); + } +#else + char *command = xmalloc (sizeof pr_program - 1 + 7 + + quote_system_arg ((char *) 0, name) + 1); + char *p; + sprintf (command, "%s -f -h ", pr_program); + p = command + sizeof pr_program - 1 + 7; + p += quote_system_arg (p, name); + *p = 0; + errno = 0; + outfile = popen (command, "w"); + if (!outfile) + pfatal_with_name (command); + free (command); +#endif + } + } + else + { + + /* If -l was not specified, output the diff straight to `stdout'. */ + + outfile = stdout; + + /* If handling multiple files (because scanning a directory), + print which files the following output is about. */ + if (currently_recursive) + printf ("%s\n", name); + } + + free (name); + + /* A special header is needed at the beginning of context output. */ + switch (output_style) + { + case OUTPUT_CONTEXT: + print_context_header (files, false); + break; + + case OUTPUT_UNIFIED: + print_context_header (files, true); + break; + + default: + break; + } +} + +/* Call after the end of output of diffs for one file. + Close OUTFILE and get rid of the `pr' subfork. */ + +void +finish_output (void) +{ + if (outfile != 0 && outfile != stdout) + { + int status; + int wstatus; + int werrno = 0; + if (ferror (outfile)) + fatal ("write failed"); +#if ! (HAVE_WORKING_FORK || HAVE_WORKING_VFORK) + wstatus = pclose (outfile); + if (wstatus == -1) + werrno = errno; +#else + if (fclose (outfile) != 0) + pfatal_with_name (_("write failed")); + if (waitpid (pr_pid, &wstatus, 0) < 0) + pfatal_with_name ("waitpid"); +#endif + status = (! werrno && WIFEXITED (wstatus) + ? WEXITSTATUS (wstatus) + : INT_MAX); + if (status) + error (EXIT_TROUBLE, werrno, + _(status == 126 + ? "subsidiary program `%s' could not be invoked" + : status == 127 + ? "subsidiary program `%s' not found" + : status == INT_MAX + ? "subsidiary program `%s' failed" + : "subsidiary program `%s' failed (exit status %d)"), + pr_program, status); + } + + outfile = 0; +} + +/* Compare two lines (typically one from each input file) + according to the command line options. + For efficiency, this is invoked only when the lines do not match exactly + but an option like -i might cause us to ignore the difference. + Return nonzero if the lines differ. */ + +bool +lines_differ (char const *s1, char const *s2) +{ + register char const *t1 = s1; + register char const *t2 = s2; + size_t column = 0; + + while (1) + { + register unsigned char c1 = *t1++; + register unsigned char c2 = *t2++; + + /* Test for exact char equality first, since it's a common case. */ + if (c1 != c2) + { + switch (ignore_white_space) + { + case IGNORE_ALL_SPACE: + /* For -w, just skip past any white space. */ + while (isspace (c1) && c1 != '\n') c1 = *t1++; + while (isspace (c2) && c2 != '\n') c2 = *t2++; + break; + + case IGNORE_SPACE_CHANGE: + /* For -b, advance past any sequence of white space in + line 1 and consider it just one space, or nothing at + all if it is at the end of the line. */ + if (isspace (c1)) + { + while (c1 != '\n') + { + c1 = *t1++; + if (! isspace (c1)) + { + --t1; + c1 = ' '; + break; + } + } + } + + /* Likewise for line 2. */ + if (isspace (c2)) + { + while (c2 != '\n') + { + c2 = *t2++; + if (! isspace (c2)) + { + --t2; + c2 = ' '; + break; + } + } + } + + if (c1 != c2) + { + /* If we went too far when doing the simple test + for equality, go back to the first non-white-space + character in both sides and try again. */ + if (c2 == ' ' && c1 != '\n' + && s1 + 1 < t1 + && isspace ((unsigned char) t1[-2])) + { + --t1; + continue; + } + if (c1 == ' ' && c2 != '\n' + && s2 + 1 < t2 + && isspace ((unsigned char) t2[-2])) + { + --t2; + continue; + } + } + + break; + + case IGNORE_TAB_EXPANSION: + if ((c1 == ' ' && c2 == '\t') + || (c1 == '\t' && c2 == ' ')) + { + size_t column2 = column; + for (;; c1 = *t1++) + { + if (c1 == ' ') + column++; + else if (c1 == '\t') + column += tabsize - column % tabsize; + else + break; + } + for (;; c2 = *t2++) + { + if (c2 == ' ') + column2++; + else if (c2 == '\t') + column2 += tabsize - column2 % tabsize; + else + break; + } + if (column != column2) + return true; + } + break; + + case IGNORE_NO_WHITE_SPACE: + break; + } + + /* Lowercase all letters if -i is specified. */ + + if (ignore_case) + { + c1 = tolower (c1); + c2 = tolower (c2); + } + + if (c1 != c2) + break; + } + if (c1 == '\n') + return false; + + column += c1 == '\t' ? tabsize - column % tabsize : 1; + } + + return true; +} + +/* Find the consecutive changes at the start of the script START. + Return the last link before the first gap. */ + +struct change * +find_change (struct change *start) +{ + return start; +} + +struct change * +find_reverse_change (struct change *start) +{ + return start; +} + +/* Divide SCRIPT into pieces by calling HUNKFUN and + print each piece with PRINTFUN. + Both functions take one arg, an edit script. + + HUNKFUN is called with the tail of the script + and returns the last link that belongs together with the start + of the tail. + + PRINTFUN takes a subscript which belongs together (with a null + link at the end) and prints it. */ + +void +print_script (struct change *script, + struct change * (*hunkfun) (struct change *), + void (*printfun) (struct change *)) +{ + struct change *next = script; + + while (next) + { + struct change *this, *end; + + /* Find a set of changes that belong together. */ + this = next; + end = (*hunkfun) (next); + + /* Disconnect them from the rest of the changes, + making them a hunk, and remember the rest for next iteration. */ + next = end->link; + end->link = 0; +#ifdef DEBUG + debug_script (this); +#endif + + /* Print this hunk. */ + (*printfun) (this); + + /* Reconnect the script so it will all be freed properly. */ + end->link = next; + } +} + +/* Print the text of a single line LINE, + flagging it with the characters in LINE_FLAG (which say whether + the line is inserted, deleted, changed, etc.). */ + +void +print_1_line (char const *line_flag, char const *const *line) +{ + char const *base = line[0], *limit = line[1]; /* Help the compiler. */ + FILE *out = outfile; /* Help the compiler some more. */ + char const *flag_format = 0; + + /* If -T was specified, use a Tab between the line-flag and the text. + Otherwise use a Space (as Unix diff does). + Print neither space nor tab if line-flags are empty. */ + + if (line_flag && *line_flag) + { + flag_format = initial_tab ? "%s\t" : "%s "; + fprintf (out, flag_format, line_flag); + } + + output_1_line (base, limit, flag_format, line_flag); + + if ((!line_flag || line_flag[0]) && limit[-1] != '\n') + fprintf (out, "\n\\ %s\n", _("No newline at end of file")); +} + +/* Output a line from BASE up to LIMIT. + With -t, expand white space characters to spaces, and if FLAG_FORMAT + is nonzero, output it with argument LINE_FLAG after every + internal carriage return, so that tab stops continue to line up. */ + +void +output_1_line (char const *base, char const *limit, char const *flag_format, + char const *line_flag) +{ + if (!expand_tabs) + fwrite (base, sizeof (char), limit - base, outfile); + else + { + register FILE *out = outfile; + register unsigned char c; + register char const *t = base; + register size_t column = 0; + size_t tab_size = tabsize; + + while (t < limit) + switch ((c = *t++)) + { + case '\t': + { + size_t spaces = tab_size - column % tab_size; + column += spaces; + do + putc (' ', out); + while (--spaces); + } + break; + + case '\r': + putc (c, out); + if (flag_format && t < limit && *t != '\n') + fprintf (out, flag_format, line_flag); + column = 0; + break; + + case '\b': + if (column == 0) + continue; + column--; + putc (c, out); + break; + + default: + column += isprint (c) != 0; + putc (c, out); + break; + } + } +} + +char const change_letter[] = { 0, 'd', 'a', 'c' }; + +/* Translate an internal line number (an index into diff's table of lines) + into an actual line number in the input file. + The internal line number is I. FILE points to the data on the file. + + Internal line numbers count from 0 starting after the prefix. + Actual line numbers count from 1 within the entire file. */ + +lin +translate_line_number (struct file_data const *file, lin i) +{ + return i + file->prefix_lines + 1; +} + +/* Translate a line number range. This is always done for printing, + so for convenience translate to long int rather than lin, so that the + caller can use printf with "%ld" without casting. */ + +void +translate_range (struct file_data const *file, + lin a, lin b, + long int *aptr, long int *bptr) +{ + *aptr = translate_line_number (file, a - 1) + 1; + *bptr = translate_line_number (file, b + 1) - 1; +} + +/* Print a pair of line numbers with SEPCHAR, translated for file FILE. + If the two numbers are identical, print just one number. + + Args A and B are internal line numbers. + We print the translated (real) line numbers. */ + +void +print_number_range (char sepchar, struct file_data *file, lin a, lin b) +{ + long int trans_a, trans_b; + translate_range (file, a, b, &trans_a, &trans_b); + + /* Note: we can have B < A in the case of a range of no lines. + In this case, we should print the line number before the range, + which is B. */ + if (trans_b > trans_a) + fprintf (outfile, "%ld%c%ld", trans_a, sepchar, trans_b); + else + fprintf (outfile, "%ld", trans_b); +} + +/* Look at a hunk of edit script and report the range of lines in each file + that it applies to. HUNK is the start of the hunk, which is a chain + of `struct change'. The first and last line numbers of file 0 are stored in + *FIRST0 and *LAST0, and likewise for file 1 in *FIRST1 and *LAST1. + Note that these are internal line numbers that count from 0. + + If no lines from file 0 are deleted, then FIRST0 is LAST0+1. + + Return UNCHANGED if only ignorable lines are inserted or deleted, + OLD if lines of file 0 are deleted, + NEW if lines of file 1 are inserted, + and CHANGED if both kinds of changes are found. */ + +enum changes +analyze_hunk (struct change *hunk, + lin *first0, lin *last0, + lin *first1, lin *last1) +{ + struct change *next; + lin l0, l1; + lin show_from, show_to; + lin i; + bool trivial = ignore_blank_lines || ignore_regexp.fastmap; + size_t trivial_length = ignore_blank_lines - 1; + /* If 0, ignore zero-length lines; + if SIZE_MAX, do not ignore lines just because of their length. */ + bool skip_leading_white_space = + (ignore_blank_lines && IGNORE_SPACE_CHANGE <= ignore_white_space); + + char const * const *linbuf0 = files[0].linbuf; /* Help the compiler. */ + char const * const *linbuf1 = files[1].linbuf; + + show_from = show_to = 0; + + *first0 = hunk->line0; + *first1 = hunk->line1; + + next = hunk; + do + { + l0 = next->line0 + next->deleted - 1; + l1 = next->line1 + next->inserted - 1; + show_from += next->deleted; + show_to += next->inserted; + + for (i = next->line0; i <= l0 && trivial; i++) + { + char const *line = linbuf0[i]; + char const *newline = linbuf0[i + 1] - 1; + size_t len = newline - line; + char const *p = line; + if (skip_leading_white_space) + while (isspace ((unsigned char) *p) && *p != '\n') + p++; + if (newline - p != trivial_length + && (! ignore_regexp.fastmap + || re_search (&ignore_regexp, line, len, 0, len, 0) < 0)) + trivial = 0; + } + + for (i = next->line1; i <= l1 && trivial; i++) + { + char const *line = linbuf1[i]; + char const *newline = linbuf1[i + 1] - 1; + size_t len = newline - line; + char const *p = line; + if (skip_leading_white_space) + while (isspace ((unsigned char) *p) && *p != '\n') + p++; + if (newline - p != trivial_length + && (! ignore_regexp.fastmap + || re_search (&ignore_regexp, line, len, 0, len, 0) < 0)) + trivial = 0; + } + } + while ((next = next->link) != 0); + + *last0 = l0; + *last1 = l1; + + /* If all inserted or deleted lines are ignorable, + tell the caller to ignore this hunk. */ + + if (trivial) + return UNCHANGED; + + return (show_from ? OLD : UNCHANGED) | (show_to ? NEW : UNCHANGED); +} + +/* Concatenate three strings, returning a newly malloc'd string. */ + +char * +concat (char const *s1, char const *s2, char const *s3) +{ + char *new = xmalloc (strlen (s1) + strlen (s2) + strlen (s3) + 1); + sprintf (new, "%s%s%s", s1, s2, s3); + return new; +} + +/* Yield a new block of SIZE bytes, initialized to zero. */ + +void * +zalloc (size_t size) +{ + void *p = xmalloc (size); + memset (p, 0, size); + return p; +} + +/* Yield the newly malloc'd pathname + of the file in DIR whose filename is FILE. */ + +char * +dir_file_pathname (char const *dir, char const *file) +{ + char const *base = base_name (dir); + bool omit_slash = !*base || base[strlen (base) - 1] == '/'; + return concat (dir, "/" + omit_slash, file); +} + +void +debug_script (struct change *sp) +{ + fflush (stdout); + + for (; sp; sp = sp->link) + { + long int line0 = sp->line0; + long int line1 = sp->line1; + long int deleted = sp->deleted; + long int inserted = sp->inserted; + fprintf (stderr, "%3ld %3ld delete %ld insert %ld\n", + line0, line1, deleted, inserted); + } + + fflush (stderr); +} -- cgit v1.1