diff options
Diffstat (limited to 'usr.bin/indent')
| -rw-r--r-- | usr.bin/indent/Makefile | 6 | ||||
| -rw-r--r-- | usr.bin/indent/README | 97 | ||||
| -rw-r--r-- | usr.bin/indent/args.c | 300 | ||||
| -rw-r--r-- | usr.bin/indent/indent.1 | 452 | ||||
| -rw-r--r-- | usr.bin/indent/indent.c | 1181 | ||||
| -rw-r--r-- | usr.bin/indent/indent_codes.h | 69 | ||||
| -rw-r--r-- | usr.bin/indent/indent_globs.h | 310 | ||||
| -rw-r--r-- | usr.bin/indent/io.c | 625 | ||||
| -rw-r--r-- | usr.bin/indent/lexi.c | 559 | ||||
| -rw-r--r-- | usr.bin/indent/parse.c | 324 | ||||
| -rw-r--r-- | usr.bin/indent/pr_comment.c | 418 |
11 files changed, 4341 insertions, 0 deletions
diff --git a/usr.bin/indent/Makefile b/usr.bin/indent/Makefile new file mode 100644 index 0000000..8d9ff94 --- /dev/null +++ b/usr.bin/indent/Makefile @@ -0,0 +1,6 @@ +# @(#)Makefile 8.1 (Berkeley) 6/6/93 + +PROG= indent +SRCS= indent.c io.c lexi.c parse.c pr_comment.c args.c + +.include <bsd.prog.mk> diff --git a/usr.bin/indent/README b/usr.bin/indent/README new file mode 100644 index 0000000..618f1de --- /dev/null +++ b/usr.bin/indent/README @@ -0,0 +1,97 @@ +This is the C indenter, it originally came from the University of Illinois +via some distribution tape for PDP-11 Unix. It has subsequently been +hacked upon by James Gosling @ CMU. It isn't very pretty, and really needs +to be completely redone, but it is probably the nicest C pretty printer +around. + +Further additions to provide "Kernel Normal Form" were contributed +by the folks at Sun Microsystems. + +++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++++ +> From mnetor!yunexus!oz@uunet.UU.NET Wed Mar 9 15:30:55 1988 +> Date: Tue, 8 Mar 88 18:36:25 EST +> From: yunexus!oz@uunet.UU.NET (Ozan Yigit) +> To: bostic@okeeffe.berkeley.edu +> Cc: ccvaxa!willcox@uunet.UU.NET, jag@sun.com, rsalz@uunet.UU.NET +> In-Reply-To: Keith Bostic's message of Tue, 16 Feb 88 16:09:06 PST +> Subject: Re: Indent... + +Thank you for your response about indent. I was wrong in my original +observation (or mis-observation :-). UCB did keep the Illinois +copyright intact. + +The issue still is whether we can distribute indent, and if we can, which +version. David Willcox (the author) states that: + +| Several people have asked me on what basis I claim that indent is in +| the public domain. I knew I would be sorry I made that posting. +| +| Some history. Way back in 1976, the project I worked on at the +| University of Illinois Center for Advanced Computation had a huge +| battle about how to format C code. After about a week of fighting, I +| got disgusted and wrote a program, which I called indent, to reformat C +| code. It had a bunch of different options that would let you format +| the output the way you liked. In particular, all of the different +| formats being championed were supported. +| +| It was my first big C program. It was ugly. It wasn't designed, it +| just sort of grew. But it pretty much worked, and it stopped most of +| the fighting. +| +| As a matter of form, I included a University of Illinois Copyright +| notice. However, my understanding was that, since the work was done +| on an ARPA contract, it was in the public domain. +| +| Time passed. Some years later, indent showed up on one of the early +| emacs distributions. +| +| Later still, someone from UC Berlekey called the UofI and asked if +| indent was in the public domain. They wanted to include it in their +| UNIX distributions, along with the emacs stuff. I was no longer at the +| UofI, but Rob Kolstad, who was, asked me about it. I told him I didn't +| care if they used it, and since then it has been on the BSD distributions. +| +| Somewhere along the way, several other unnamed people have had their +| hands in it. It was converted to understand version 7 C. (The +| original was version 6.) It was converted from its original filter +| interface to its current "blow away the user's file" interface. +| The $HOME/.indent.pro file parsing was added. Some more formatting +| options were added. +| +| The source I have right now has two copyright notices. One is the +| original from the UofI. One is from Berkeley. +| +| I am not a lawyer, and I certainly do not understand copyright law. As +| far as I am concerned, the bulk of this program, everything covered by +| the UofI copyright, is in the public domain, and worth every penny. +| Berkeley's copyright probably should only cover their changes, and I +| don't know their feelings about sending it out. + +In any case, there appears to be noone at UofI to clarify/and change +that copyright, but I am confident (based on the statements of its +author) that the code, as it stands with its copyright, is +distributable, and will not cause any legal problems. + +Hence, the issue reduces to *which* one to distribute through +comp.sources.unix. I would suggest that with the permission of you +folks (given that you have parts copyrighted), we distribute the 4.3 +version of indent, which appears to be the most up-to-date version. I +happen to have just about every known version of indent, including the +very original submission from the author to a unix tape, later the +G-Emacs version, any 4.n version, sun version and the Unipress +version. I still think we should not have to "go-back-in-time" and +re-do all the work you people have done. + +I hope to hear from you as to what you think about this. You may of +course send 4.3 version to the moderator directly, or you can let me +know of your permission, and I will send the sources, or you can let +me know that 4.3 version is off-limits, in which case we would probably +have to revert to an older version. One way or another, I hope to get +a version of indent to comp.sources.unix. + +regards.. oz + +cc: ccvaxa!willcox + sun.com!jar + uunet!rsalz + diff --git a/usr.bin/indent/args.c b/usr.bin/indent/args.c new file mode 100644 index 0000000..e62f038 --- /dev/null +++ b/usr.bin/indent/args.c @@ -0,0 +1,300 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifndef lint +static char sccsid[] = "@(#)args.c 8.1 (Berkeley) 6/6/93"; +#endif /* not lint */ + +/* + * Argument scanning and profile reading code. Default parameters are set + * here as well. + */ + +#include <stdio.h> +#include <ctype.h> +#include <stdlib.h> +#include <string.h> +#include "indent_globs.h" + +/* profile types */ +#define PRO_SPECIAL 1 /* special case */ +#define PRO_BOOL 2 /* boolean */ +#define PRO_INT 3 /* integer */ +#define PRO_FONT 4 /* troff font */ + +/* profile specials for booleans */ +#define ON 1 /* turn it on */ +#define OFF 0 /* turn it off */ + +/* profile specials for specials */ +#define IGN 1 /* ignore it */ +#define CLI 2 /* case label indent (float) */ +#define STDIN 3 /* use stdin */ +#define KEY 4 /* type (keyword) */ + +char *option_source = "?"; + +/* + * N.B.: because of the way the table here is scanned, options whose names are + * substrings of other options must occur later; that is, with -lp vs -l, -lp + * must be first. Also, while (most) booleans occur more than once, the last + * default value is the one actually assigned. + */ +struct pro { + char *p_name; /* name, eg -bl, -cli */ + int p_type; /* type (int, bool, special) */ + int p_default; /* the default value (if int) */ + int p_special; /* depends on type */ + int *p_obj; /* the associated variable */ +} pro[] = { + + "T", PRO_SPECIAL, 0, KEY, 0, + "bacc", PRO_BOOL, false, ON, &blanklines_around_conditional_compilation, + "badp", PRO_BOOL, false, ON, &blanklines_after_declarations_at_proctop, + "bad", PRO_BOOL, false, ON, &blanklines_after_declarations, + "bap", PRO_BOOL, false, ON, &blanklines_after_procs, + "bbb", PRO_BOOL, false, ON, &blanklines_before_blockcomments, + "bc", PRO_BOOL, true, OFF, &ps.leave_comma, + "bl", PRO_BOOL, true, OFF, &btype_2, + "br", PRO_BOOL, true, ON, &btype_2, + "bs", PRO_BOOL, false, ON, &Bill_Shannon, + "cdb", PRO_BOOL, true, ON, &comment_delimiter_on_blankline, + "cd", PRO_INT, 0, 0, &ps.decl_com_ind, + "ce", PRO_BOOL, true, ON, &cuddle_else, + "ci", PRO_INT, 0, 0, &continuation_indent, + "cli", PRO_SPECIAL, 0, CLI, 0, + "c", PRO_INT, 33, 0, &ps.com_ind, + "di", PRO_INT, 16, 0, &ps.decl_indent, + "dj", PRO_BOOL, false, ON, &ps.ljust_decl, + "d", PRO_INT, 0, 0, &ps.unindent_displace, + "eei", PRO_BOOL, false, ON, &extra_expression_indent, + "ei", PRO_BOOL, true, ON, &ps.else_if, + "fbc", PRO_FONT, 0, 0, (int *) &blkcomf, + "fbx", PRO_FONT, 0, 0, (int *) &boxcomf, + "fb", PRO_FONT, 0, 0, (int *) &bodyf, + "fc1", PRO_BOOL, true, ON, &format_col1_comments, + "fc", PRO_FONT, 0, 0, (int *) &scomf, + "fk", PRO_FONT, 0, 0, (int *) &keywordf, + "fs", PRO_FONT, 0, 0, (int *) &stringf, + "ip", PRO_BOOL, true, ON, &ps.indent_parameters, + "i", PRO_INT, 8, 0, &ps.ind_size, + "lc", PRO_INT, 0, 0, &block_comment_max_col, + "lp", PRO_BOOL, true, ON, &lineup_to_parens, + "l", PRO_INT, 78, 0, &max_col, + "nbacc", PRO_BOOL, false, OFF, &blanklines_around_conditional_compilation, + "nbadp", PRO_BOOL, false, OFF, &blanklines_after_declarations_at_proctop, + "nbad", PRO_BOOL, false, OFF, &blanklines_after_declarations, + "nbap", PRO_BOOL, false, OFF, &blanklines_after_procs, + "nbbb", PRO_BOOL, false, OFF, &blanklines_before_blockcomments, + "nbc", PRO_BOOL, true, ON, &ps.leave_comma, + "nbs", PRO_BOOL, false, OFF, &Bill_Shannon, + "ncdb", PRO_BOOL, true, OFF, &comment_delimiter_on_blankline, + "nce", PRO_BOOL, true, OFF, &cuddle_else, + "ndj", PRO_BOOL, false, OFF, &ps.ljust_decl, + "neei", PRO_BOOL, false, OFF, &extra_expression_indent, + "nei", PRO_BOOL, true, OFF, &ps.else_if, + "nfc1", PRO_BOOL, true, OFF, &format_col1_comments, + "nip", PRO_BOOL, true, OFF, &ps.indent_parameters, + "nlp", PRO_BOOL, true, OFF, &lineup_to_parens, + "npcs", PRO_BOOL, false, OFF, &proc_calls_space, + "npro", PRO_SPECIAL, 0, IGN, 0, + "npsl", PRO_BOOL, true, OFF, &procnames_start_line, + "nps", PRO_BOOL, false, OFF, &pointer_as_binop, + "nsc", PRO_BOOL, true, OFF, &star_comment_cont, + "nsob", PRO_BOOL, false, OFF, &swallow_optional_blanklines, + "nv", PRO_BOOL, false, OFF, &verbose, + "pcs", PRO_BOOL, false, ON, &proc_calls_space, + "psl", PRO_BOOL, true, ON, &procnames_start_line, + "ps", PRO_BOOL, false, ON, &pointer_as_binop, + "sc", PRO_BOOL, true, ON, &star_comment_cont, + "sob", PRO_BOOL, false, ON, &swallow_optional_blanklines, + "st", PRO_SPECIAL, 0, STDIN, 0, + "troff", PRO_BOOL, false, ON, &troff, + "v", PRO_BOOL, false, ON, &verbose, + /* whew! */ + 0, 0, 0, 0, 0 +}; + +/* + * set_profile reads $HOME/.indent.pro and ./.indent.pro and handles arguments + * given in these files. + */ +set_profile() +{ + register FILE *f; + char fname[BUFSIZ]; + static char prof[] = ".indent.pro"; + + sprintf(fname, "%s/%s", getenv("HOME"), prof); + if ((f = fopen(option_source = fname, "r")) != NULL) { + scan_profile(f); + (void) fclose(f); + } + if ((f = fopen(option_source = prof, "r")) != NULL) { + scan_profile(f); + (void) fclose(f); + } + option_source = "Command line"; +} + +scan_profile(f) + register FILE *f; +{ + register int i; + register char *p; + char buf[BUFSIZ]; + + while (1) { + for (p = buf; (i = getc(f)) != EOF && (*p = i) > ' '; ++p); + if (p != buf) { + *p++ = 0; + if (verbose) + printf("profile: %s\n", buf); + set_option(buf); + } + else if (i == EOF) + return; + } +} + +char *param_start; + +eqin(s1, s2) + register char *s1; + register char *s2; +{ + while (*s1) { + if (*s1++ != *s2++) + return (false); + } + param_start = s2; + return (true); +} + +/* + * Set the defaults. + */ +set_defaults() +{ + register struct pro *p; + + /* + * Because ps.case_indent is a float, we can't initialize it from the + * table: + */ + ps.case_indent = 0.0; /* -cli0.0 */ + for (p = pro; p->p_name; p++) + if (p->p_type != PRO_SPECIAL && p->p_type != PRO_FONT) + *p->p_obj = p->p_default; +} + +set_option(arg) + register char *arg; +{ + register struct pro *p; + extern double atof(); + + arg++; /* ignore leading "-" */ + for (p = pro; p->p_name; p++) + if (*p->p_name == *arg && eqin(p->p_name, arg)) + goto found; + fprintf(stderr, "indent: %s: unknown parameter \"%s\"\n", option_source, arg - 1); + exit(1); +found: + switch (p->p_type) { + + case PRO_SPECIAL: + switch (p->p_special) { + + case IGN: + break; + + case CLI: + if (*param_start == 0) + goto need_param; + ps.case_indent = atof(param_start); + break; + + case STDIN: + if (input == 0) + input = stdin; + if (output == 0) + output = stdout; + break; + + case KEY: + if (*param_start == 0) + goto need_param; + { + register char *str = (char *) malloc(strlen(param_start) + 1); + strcpy(str, param_start); + addkey(str, 4); + } + break; + + default: + fprintf(stderr, "\ +indent: set_option: internal error: p_special %d\n", p->p_special); + exit(1); + } + break; + + case PRO_BOOL: + if (p->p_special == OFF) + *p->p_obj = false; + else + *p->p_obj = true; + break; + + case PRO_INT: + if (!isdigit(*param_start)) { + need_param: + fprintf(stderr, "indent: %s: ``%s'' requires a parameter\n", + option_source, arg - 1); + exit(1); + } + *p->p_obj = atoi(param_start); + break; + + case PRO_FONT: + parsefont((struct fstate *) p->p_obj, param_start); + break; + + default: + fprintf(stderr, "indent: set_option: internal error: p_type %d\n", + p->p_type); + exit(1); + } +} diff --git a/usr.bin/indent/indent.1 b/usr.bin/indent/indent.1 new file mode 100644 index 0000000..c44964a --- /dev/null +++ b/usr.bin/indent/indent.1 @@ -0,0 +1,452 @@ +.\" Copyright (c) 1980, 1990, 1993 +.\" The Regents of the University of California. All rights reserved. +.\" Copyright (c) 1976 Board of Trustees of the University of Illinois. +.\" All rights reserved. +.\" +.\" Redistribution and use in source and binary forms, with or without +.\" modification, are permitted provided that the following conditions +.\" are met: +.\" 1. Redistributions of source code must retain the above copyright +.\" notice, this list of conditions and the following disclaimer. +.\" 2. Redistributions in binary form must reproduce the above copyright +.\" notice, this list of conditions and the following disclaimer in the +.\" documentation and/or other materials provided with the distribution. +.\" 3. All advertising materials mentioning features or use of this software +.\" must display the following acknowledgement: +.\" This product includes software developed by the University of +.\" California, Berkeley and its contributors. +.\" 4. Neither the name of the University nor the names of its contributors +.\" may be used to endorse or promote products derived from this software +.\" without specific prior written permission. +.\" +.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND +.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE +.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE +.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE +.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL +.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS +.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) +.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT +.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY +.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF +.\" SUCH DAMAGE. +.\" +.\" @(#)indent.1 8.1 (Berkeley) 7/1/93 +.\" +.Dd July 1, 1993 +.Dt INDENT 1 +.Os BSD 4.2 +.Sh NAME +.Nm indent +.Nd indent and format C program source +.Sh SYNOPSIS +.Nm indent +.Op Ar input-file Op Ar output-file +.Op Fl bad | Fl nbad +.Op Fl bap | Fl nbap +.Bk -words +.Op Fl bbb | Fl nbbb +.Ek +.Op Fl \&bc | Fl nbc +.Op Fl \&bl +.Op Fl \&br +.Op Fl c Ns Ar n +.Op Fl \&cd Ns Ar n +.Bk -words +.Op Fl cdb | Fl ncdb +.Ek +.Op Fl \&ce | Fl nce +.Op Fl \&ci Ns Ar n +.Op Fl cli Ns Ar n +.Op Fl d Ns Ar n +.Op Fl \&di Ns Ar n +.Bk -words +.Op Fl fc1 | Fl nfc1 +.Ek +.Op Fl i Ns Ar n +.Op Fl \&ip | Fl nip +.Op Fl l Ns Ar n +.Op Fl \&lc Ns Ar n +.Op Fl \&lp | Fl nlp +.Op Fl npro +.Op Fl pcs | Fl npcs +.Op Fl psl | Fl npsl +.Op Fl \&sc | Fl nsc +.Bk -words +.Op Fl sob | Fl nsob +.Ek +.Op Fl \&st +.Op Fl troff +.Op Fl v | Fl \&nv +.Sh DESCRIPTION +.Nm Indent +is a +.Ar C +program formatter. It reformats the +.Ar C +program in the +.Ar input-file +according to the switches. The switches which can be +specified are described below. They may appear before or after the file +names. +.Pp +.Sy NOTE : +If you only specify an +.Ar input-file , +the formatting is +done `in-place', that is, the formatted file is written back into +.Ar input-file +and a backup copy of +.Ar input-file +is written in the current directory. If +.Ar input-file +is named +.Sq Pa /blah/blah/file , +the backup file is named +.Pa file.BAK . +.Pp +If +.Ar output-file +is specified, +.Nm indent +checks to make sure it is different from +.Ar input-file . +.Pp +The options listed below control the formatting style imposed by +.Nm indent . +.Bl -tag -width Op +.It Fl bad , nbad +If +.Fl bad +is specified, a blank line is forced after every block of +declarations. Default: +.Fl nbad . +.It Fl bap , nbap +If +.Fl bap +is specified, a blank line is forced after every procedure body. Default: +.Fl nbap . +.It Fl bbb , nbbb +If +.Fl bbb +is specified, a blank line is forced before every block comment. Default: +.Fl nbbb . +.It Fl \&bc , nbc +If +.Fl \&bc +is specified, then a newline is forced after each comma in a declaration. +.Fl nbc +turns off this option. The default is +.Fl \&bc . +.It Fl \&br , \&bl +Specifying +.Fl \&bl +lines up compound statements like this: +.ne 4 +.Bd -literal -offset indent +if (...) +{ + code +} +.Ed +.Pp +Specifying +.Fl \&br +(the default) makes them look like this: +.ne 3 +.Bd -literal -offset indent +if (...) { + code +} +.Ed +.Pp +.It Fl c n +The column in which comments on code start. The default is 33. +.It Fl cd n +The column in which comments on declarations start. The default +is for these comments to start in the same column as those on code. +.It Fl cdb , ncdb +Enables (disables) the placement of comment delimiters on blank lines. With +this option enabled, comments look like this: +.Bd -literal -offset indent +.ne 3 + /* + * this is a comment + */ +.Ed +.Pp +Rather than like this: +.Bd -literal -offset indent + /* this is a comment */ +.Ed +.Pp +This only affects block comments, not comments to the right of +code. The default is +.Fl cdb . +.It Fl ce , nce +Enables (disables) forcing `else's to cuddle up to the immediately preceding +`}'. The default is +.Fl \&ce . +.It Fl \&ci Ns Ar n +Sets the continuation indent to be +.Ar n . +Continuation +lines will be indented that far from the beginning of the first line of the +statement. Parenthesized expressions have extra indentation added to +indicate the nesting, unless +.Fl \&lp +is in effect. +.Fl \&ci +defaults to the same value as +.Fl i . +.It Fl cli Ns Ar n +Causes case labels to be indented +.Ar n +tab stops to the right of the containing +.Ic switch +statement. +.Fl cli0 .5 +causes case labels to be indented half a tab stop. The +default is +.Fl cli0 . +.It Fl d Ns Ar n +Controls the placement of comments which are not to the +right of code. The default +.Fl \&d\&1 +means that such comments are placed one indentation level to the +left of code. Specifying +.Fl \&d\&0 +lines up these comments with the code. See the section on comment +indentation below. +.It Fl \&di Ns Ar n +Specifies the indentation, in character positions, from a declaration keyword +to the following identifier. The default is +.Fl di16 . +.It Fl dj , ndj +.Fl \&dj +left justifies declarations. +.Fl ndj +indents declarations the same as code. The default is +.Fl ndj . +.It Fl \&ei , nei +Enables (disables) special +.Ic else-if +processing. If it's enabled, an +.Ic if +following an +.Ic else +will have the same indentation as the preceding +.Ic \&if +statement. +.It Fl fc1 , nfc1 +Enables (disables) the formatting of comments that start in column 1. +Often, comments whose leading `/' is in column 1 have been carefully +hand formatted by the programmer. In such cases, +.Fl nfc1 +should be +used. The default is +.Fl fc1 . +.It Fl i Ns Ar n +The number of spaces for one indentation level. The default is 4. +.It Fl \&ip , nip +Enables (disables) the indentation of parameter declarations from the left +margin. The default is +.Fl \&ip . +.It Fl l Ns Ar n +Maximum length of an output line. The default is 75. +.It Fl \&lp , nlp +Lines up code surrounded by parenthesis in continuation lines. If a line +has a left paren which is not closed on that line, then continuation lines +will be lined up to start at the character position just after the left +paren. For example, here is how a piece of continued code looks with +.Fl nlp +in effect: +.ne 2 +.Bd -literal -offset indent +p1 = first_procedure(second_procedure(p2, p3), +\ \ third_procedure(p4,p5)); +.Ed +.Pp +.ne 5 +With +.Fl lp +in effect (the default) the code looks somewhat clearer: +.Bd -literal -offset indent +p1\ =\ first_procedure(second_procedure(p2,\ p3), +\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ third_procedure(p4,p5)); +.Ed +.Pp +.ne 5 +Inserting two more newlines we get: +.Bd -literal -offset indent +p1\ =\ first_procedure(second_procedure(p2, +\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ p3), +\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ third_procedure(p4 +\ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ \ p5)); +.Ed +.It Fl npro +Causes the profile files, +.Sq Pa ./.indent.pro +and +.Sq Pa ~/.indent.pro , +to be ignored. +.It Fl pcs , npcs +If true +.Pq Fl pcs +all procedure calls will have a space inserted between +the name and the `('. The default is +.Fl npcs . +.It Fl psl , npsl +If true +.Pq Fl psl +the names of procedures being defined are placed in +column 1 \- their types, if any, will be left on the previous lines. The +default is +.Fl psl . +.It Fl \&sc , nsc +Enables (disables) the placement of asterisks (`*'s) at the left edge of all +comments. +.It Fl sob , nsob +If +.Fl sob +is specified, indent will swallow optional blank lines. You can use this to +get rid of blank lines after declarations. Default: +.Fl nsob . +.It Fl \&st +Causes +.Nm indent +to take its input from stdin, and put its output to stdout. +.It Fl T Ns Ar typename +Adds +.Ar typename +to the list of type keywords. Names accumulate: +.Fl T +can be specified more than once. You need to specify all the typenames that +appear in your program that are defined by +.Ic typedef +\- nothing will be +harmed if you miss a few, but the program won't be formatted as nicely as +it should. This sounds like a painful thing to have to do, but it's really +a symptom of a problem in C: +.Ic typedef +causes a syntactic change in the +language and +.Nm indent +can't find all +instances of +.Ic typedef . +.It Fl troff +Causes +.Nm indent +to format the program for processing by +.Xr troff 1 . +It will produce a fancy +listing in much the same spirit as +.Xr vgrind 1 . +If the output file is not specified, the default is standard output, +rather than formatting in place. +.It Fl v , \&nv +.Fl v +turns on `verbose' mode; +.Fl \&nv +turns it off. When in verbose mode, +.Nm indent +reports when it splits one line of input into two or more lines of output, +and gives some size statistics at completion. The default is +.Fl \&nv . +.El +.Pp +You may set up your own `profile' of defaults to +.Nm indent +by creating a file called +.Pa .indent.pro +in your login directory and/or the current directory and including +whatever switches you like. A `.indent.pro' in the current directory takes +precedence over the one in your login directory. If +.Nm indent +is run and a profile file exists, then it is read to set up the program's +defaults. Switches on the command line, though, always override profile +switches. The switches should be separated by spaces, tabs or newlines. +.Pp +.Ss Comments +.Sq Em Box +.Em comments . +.Nm Indent +assumes that any comment with a dash or star immediately after the start of +comment (that is, `/*\-' or `/**') is a comment surrounded by a box of stars. +Each line of such a comment is left unchanged, except that its indentation +may be adjusted to account for the change in indentation of the first line +of the comment. +.Pp +.Em Straight text . +All other comments are treated as straight text. +.Nm Indent +fits as many words (separated by blanks, tabs, or newlines) on a +line as possible. Blank lines break paragraphs. +.Pp +.Ss Comment indentation +If a comment is on a line with code it is started in the `comment column', +which is set by the +.Fl c Ns Ns Ar n +command line parameter. Otherwise, the comment is started at +.Ar n +indentation levels less than where code is currently being placed, where +.Ar n +is specified by the +.Fl d Ns Ns Ar n +command line parameter. If the code on a line extends past the comment +column, the comment starts further to the right, and the right margin may be +automatically extended in extreme cases. +.Pp +.Ss Preprocessor lines +In general, +.Nm indent +leaves preprocessor lines alone. The only +reformatting that it will do is to straighten up trailing comments. It +leaves embedded comments alone. Conditional compilation +.Pq Ic #ifdef...#endif +is recognized and +.Nm indent +attempts to correctly +compensate for the syntactic peculiarities introduced. +.Pp +.Ss C syntax +.Nm Indent +understands a substantial amount about the syntax of C, but it +has a `forgiving' parser. It attempts to cope with the usual sorts of +incomplete and misformed syntax. In particular, the use of macros like: +.Pp +.Dl #define forever for(;;) +.Pp +is handled properly. +.Sh ENVIRONMENT +.Nm Indent +uses the +.Ev HOME +environment variable. +.Sh FILES +.Bl -tag -width "./.indent.pro" -compact +.It Pa ./.indent.pro +profile file +.It Pa ~/.indent.pro +profile file +.El +.Sh HISTORY +The +.Nm indent +command appeared in +.Bx 4.2 . +.Sh BUGS +.Nm Indent +has even more switches than +.Xr ls 1 . +.Pp +.ne 5 +A common mistake that often causes grief is typing: +.Pp +.Dl indent *.c +.Pp +to the shell in an attempt to indent all the +.Nm C +programs in a directory. +This is probably a bug, not a feature. diff --git a/usr.bin/indent/indent.c b/usr.bin/indent/indent.c new file mode 100644 index 0000000..132d21f --- /dev/null +++ b/usr.bin/indent/indent.c @@ -0,0 +1,1181 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1976 Board of Trustees of the University of Illinois. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifndef lint +char copyright[] = +"@(#) Copyright (c) 1985 Sun Microsystems, Inc.\n\ +@(#) Copyright (c) 1976 Board of Trustees of the University of Illinois.\n\ +@(#) Copyright (c) 1980, 1993\n\ + The Regents of the University of California. All rights reserved.\n"; +#endif /* not lint */ + +#ifndef lint +static char sccsid[] = "@(#)indent.c 5.17 (Berkeley) 6/7/93"; +#endif /* not lint */ + +#include <sys/param.h> +#include <fcntl.h> +#include <unistd.h> +#include <stdio.h> +#include <stdlib.h> +#include <string.h> +#include "indent_globs.h" +#include "indent_codes.h" +#include <ctype.h> + +char *in_name = "Standard Input"; /* will always point to name of input + * file */ +char *out_name = "Standard Output"; /* will always point to name + * of output file */ +char bakfile[MAXPATHLEN] = ""; + +main(argc, argv) + int argc; + char **argv; +{ + + extern int found_err; /* flag set in diag() on error */ + int dec_ind; /* current indentation for declarations */ + int di_stack[20]; /* a stack of structure indentation levels */ + int flushed_nl; /* used when buffering up comments to remember + * that a newline was passed over */ + int force_nl; /* when true, code must be broken */ + int hd_type; /* used to store type of stmt for if (...), + * for (...), etc */ + register int i; /* local loop counter */ + int scase; /* set to true when we see a case, so we will + * know what to do with the following colon */ + int sp_sw; /* when true, we are in the expressin of + * if(...), while(...), etc. */ + int squest; /* when this is positive, we have seen a ? + * without the matching : in a <c>?<s>:<s> + * construct */ + register char *t_ptr; /* used for copying tokens */ + int type_code; /* the type of token, returned by lexi */ + + int last_else = 0; /* true iff last keyword was an else */ + + + /*-----------------------------------------------*\ + | INITIALIZATION | + \*-----------------------------------------------*/ + + + ps.p_stack[0] = stmt; /* this is the parser's stack */ + ps.last_nl = true; /* this is true if the last thing scanned was + * a newline */ + ps.last_token = semicolon; + combuf = (char *) malloc(bufsize); + labbuf = (char *) malloc(bufsize); + codebuf = (char *) malloc(bufsize); + tokenbuf = (char *) malloc(bufsize); + l_com = combuf + bufsize - 5; + l_lab = labbuf + bufsize - 5; + l_code = codebuf + bufsize - 5; + l_token = tokenbuf + bufsize - 5; + combuf[0] = codebuf[0] = labbuf[0] = ' '; /* set up code, label, and + * comment buffers */ + combuf[1] = codebuf[1] = labbuf[1] = '\0'; + ps.else_if = 1; /* Default else-if special processing to on */ + s_lab = e_lab = labbuf + 1; + s_code = e_code = codebuf + 1; + s_com = e_com = combuf + 1; + s_token = e_token = tokenbuf + 1; + + in_buffer = (char *) malloc(10); + in_buffer_limit = in_buffer + 8; + buf_ptr = buf_end = in_buffer; + line_no = 1; + had_eof = ps.in_decl = ps.decl_on_line = break_comma = false; + sp_sw = force_nl = false; + ps.in_or_st = false; + ps.bl_line = true; + dec_ind = 0; + di_stack[ps.dec_nest = 0] = 0; + ps.want_blank = ps.in_stmt = ps.ind_stmt = false; + + + scase = ps.pcase = false; + squest = 0; + sc_end = 0; + bp_save = 0; + be_save = 0; + + output = 0; + + + + /*--------------------------------------------------*\ + | COMMAND LINE SCAN | + \*--------------------------------------------------*/ + +#ifdef undef + max_col = 78; /* -l78 */ + lineup_to_parens = 1; /* -lp */ + ps.ljust_decl = 0; /* -ndj */ + ps.com_ind = 33; /* -c33 */ + star_comment_cont = 1; /* -sc */ + ps.ind_size = 8; /* -i8 */ + verbose = 0; + ps.decl_indent = 16; /* -di16 */ + ps.indent_parameters = 1; /* -ip */ + ps.decl_com_ind = 0; /* if this is not set to some positive value + * by an arg, we will set this equal to + * ps.com_ind */ + btype_2 = 1; /* -br */ + cuddle_else = 1; /* -ce */ + ps.unindent_displace = 0; /* -d0 */ + ps.case_indent = 0; /* -cli0 */ + format_col1_comments = 1; /* -fc1 */ + procnames_start_line = 1; /* -psl */ + proc_calls_space = 0; /* -npcs */ + comment_delimiter_on_blankline = 1; /* -cdb */ + ps.leave_comma = 1; /* -nbc */ +#endif + + for (i = 1; i < argc; ++i) + if (strcmp(argv[i], "-npro") == 0) + break; + set_defaults(); + if (i >= argc) + set_profile(); + + for (i = 1; i < argc; ++i) { + + /* + * look thru args (if any) for changes to defaults + */ + if (argv[i][0] != '-') {/* no flag on parameter */ + if (input == 0) { /* we must have the input file */ + in_name = argv[i]; /* remember name of input file */ + input = fopen(in_name, "r"); + if (input == 0) /* check for open error */ + err(in_name); + continue; + } + else if (output == 0) { /* we have the output file */ + out_name = argv[i]; /* remember name of output file */ + if (strcmp(in_name, out_name) == 0) { /* attempt to overwrite + * the file */ + fprintf(stderr, "indent: input and output files must be different\n"); + exit(1); + } + output = fopen(out_name, "w"); + if (output == 0) /* check for create error */ + err(out_name); + continue; + } + fprintf(stderr, "indent: unknown parameter: %s\n", argv[i]); + exit(1); + } + else + set_option(argv[i]); + } /* end of for */ + if (input == 0) { + fprintf(stderr, "indent: usage: indent file [ outfile ] [ options ]\n"); + exit(1); + } + if (output == 0) + if (troff) + output = stdout; + else { + out_name = in_name; + bakcopy(); + } + if (ps.com_ind <= 1) + ps.com_ind = 2; /* dont put normal comments before column 2 */ + if (troff) { + if (bodyf.font[0] == 0) + parsefont(&bodyf, "R"); + if (scomf.font[0] == 0) + parsefont(&scomf, "I"); + if (blkcomf.font[0] == 0) + blkcomf = scomf, blkcomf.size += 2; + if (boxcomf.font[0] == 0) + boxcomf = blkcomf; + if (stringf.font[0] == 0) + parsefont(&stringf, "L"); + if (keywordf.font[0] == 0) + parsefont(&keywordf, "B"); + writefdef(&bodyf, 'B'); + writefdef(&scomf, 'C'); + writefdef(&blkcomf, 'L'); + writefdef(&boxcomf, 'X'); + writefdef(&stringf, 'S'); + writefdef(&keywordf, 'K'); + } + if (block_comment_max_col <= 0) + block_comment_max_col = max_col; + if (ps.decl_com_ind <= 0) /* if not specified by user, set this */ + ps.decl_com_ind = ps.ljust_decl ? (ps.com_ind <= 10 ? 2 : ps.com_ind - 8) : ps.com_ind; + if (continuation_indent == 0) + continuation_indent = ps.ind_size; + fill_buffer(); /* get first batch of stuff into input buffer */ + + parse(semicolon); + { + register char *p = buf_ptr; + register col = 1; + + while (1) { + if (*p == ' ') + col++; + else if (*p == '\t') + col = ((col - 1) & ~7) + 9; + else + break; + p++; + } + if (col > ps.ind_size) + ps.ind_level = ps.i_l_follow = col / ps.ind_size; + } + if (troff) { + register char *p = in_name, + *beg = in_name; + + while (*p) + if (*p++ == '/') + beg = p; + fprintf(output, ".Fn \"%s\"\n", beg); + } + /* + * START OF MAIN LOOP + */ + + while (1) { /* this is the main loop. it will go until we + * reach eof */ + int is_procname; + + type_code = lexi(); /* lexi reads one token. The actual + * characters read are stored in "token". lexi + * returns a code indicating the type of token */ + is_procname = ps.procname[0]; + + /* + * The following code moves everything following an if (), while (), + * else, etc. up to the start of the following stmt to a buffer. This + * allows proper handling of both kinds of brace placement. + */ + + flushed_nl = false; + while (ps.search_brace) { /* if we scanned an if(), while(), + * etc., we might need to copy stuff + * into a buffer we must loop, copying + * stuff into save_com, until we find + * the start of the stmt which follows + * the if, or whatever */ + switch (type_code) { + case newline: + ++line_no; + flushed_nl = true; + case form_feed: + break; /* form feeds and newlines found here will be + * ignored */ + + case lbrace: /* this is a brace that starts the compound + * stmt */ + if (sc_end == 0) { /* ignore buffering if a comment wasnt + * stored up */ + ps.search_brace = false; + goto check_type; + } + if (btype_2) { + save_com[0] = '{'; /* we either want to put the brace + * right after the if */ + goto sw_buffer; /* go to common code to get out of + * this loop */ + } + case comment: /* we have a comment, so we must copy it into + * the buffer */ + if (!flushed_nl || sc_end != 0) { + if (sc_end == 0) { /* if this is the first comment, we + * must set up the buffer */ + save_com[0] = save_com[1] = ' '; + sc_end = &(save_com[2]); + } + else { + *sc_end++ = '\n'; /* add newline between + * comments */ + *sc_end++ = ' '; + --line_no; + } + *sc_end++ = '/'; /* copy in start of comment */ + *sc_end++ = '*'; + + for (;;) { /* loop until we get to the end of the comment */ + *sc_end = *buf_ptr++; + if (buf_ptr >= buf_end) + fill_buffer(); + + if (*sc_end++ == '*' && *buf_ptr == '/') + break; /* we are at end of comment */ + + if (sc_end >= &(save_com[sc_size])) { /* check for temp buffer + * overflow */ + diag(1, "Internal buffer overflow - Move big comment from right after if, while, or whatever."); + fflush(output); + exit(1); + } + } + *sc_end++ = '/'; /* add ending slash */ + if (++buf_ptr >= buf_end) /* get past / in buffer */ + fill_buffer(); + break; + } + default: /* it is the start of a normal statment */ + if (flushed_nl) /* if we flushed a newline, make sure it is + * put back */ + force_nl = true; + if (type_code == sp_paren && *token == 'i' + && last_else && ps.else_if + || type_code == sp_nparen && *token == 'e' + && e_code != s_code && e_code[-1] == '}') + force_nl = false; + + if (sc_end == 0) { /* ignore buffering if comment wasnt + * saved up */ + ps.search_brace = false; + goto check_type; + } + if (force_nl) { /* if we should insert a nl here, put it into + * the buffer */ + force_nl = false; + --line_no; /* this will be re-increased when the nl is + * read from the buffer */ + *sc_end++ = '\n'; + *sc_end++ = ' '; + if (verbose && !flushed_nl) /* print error msg if the line + * was not already broken */ + diag(0, "Line broken"); + flushed_nl = false; + } + for (t_ptr = token; *t_ptr; ++t_ptr) + *sc_end++ = *t_ptr; /* copy token into temp buffer */ + ps.procname[0] = 0; + + sw_buffer: + ps.search_brace = false; /* stop looking for start of + * stmt */ + bp_save = buf_ptr; /* save current input buffer */ + be_save = buf_end; + buf_ptr = save_com; /* fix so that subsequent calls to + * lexi will take tokens out of + * save_com */ + *sc_end++ = ' ';/* add trailing blank, just in case */ + buf_end = sc_end; + sc_end = 0; + break; + } /* end of switch */ + if (type_code != 0) /* we must make this check, just in case there + * was an unexpected EOF */ + type_code = lexi(); /* read another token */ + /* if (ps.search_brace) ps.procname[0] = 0; */ + if ((is_procname = ps.procname[0]) && flushed_nl + && !procnames_start_line && ps.in_decl + && type_code == ident) + flushed_nl = 0; + } /* end of while (search_brace) */ + last_else = 0; +check_type: + if (type_code == 0) { /* we got eof */ + if (s_lab != e_lab || s_code != e_code + || s_com != e_com) /* must dump end of line */ + dump_line(); + if (ps.tos > 1) /* check for balanced braces */ + diag(1, "Stuff missing from end of file."); + + if (verbose) { + printf("There were %d output lines and %d comments\n", + ps.out_lines, ps.out_coms); + printf("(Lines with comments)/(Lines with code): %6.3f\n", + (1.0 * ps.com_lines) / code_lines); + } + fflush(output); + exit(found_err); + } + if ( + (type_code != comment) && + (type_code != newline) && + (type_code != preesc) && + (type_code != form_feed)) { + if (force_nl && + (type_code != semicolon) && + (type_code != lbrace || !btype_2)) { + /* we should force a broken line here */ + if (verbose && !flushed_nl) + diag(0, "Line broken"); + flushed_nl = false; + dump_line(); + ps.want_blank = false; /* dont insert blank at line start */ + force_nl = false; + } + ps.in_stmt = true; /* turn on flag which causes an extra level of + * indentation. this is turned off by a ; or + * '}' */ + if (s_com != e_com) { /* the turkey has embedded a comment + * in a line. fix it */ + *e_code++ = ' '; + for (t_ptr = s_com; *t_ptr; ++t_ptr) { + CHECK_SIZE_CODE; + *e_code++ = *t_ptr; + } + *e_code++ = ' '; + *e_code = '\0'; /* null terminate code sect */ + ps.want_blank = false; + e_com = s_com; + } + } + else if (type_code != comment) /* preserve force_nl thru a comment */ + force_nl = false; /* cancel forced newline after newline, form + * feed, etc */ + + + + /*-----------------------------------------------------*\ + | do switch on type of token scanned | + \*-----------------------------------------------------*/ + CHECK_SIZE_CODE; + switch (type_code) { /* now, decide what to do with the token */ + + case form_feed: /* found a form feed in line */ + ps.use_ff = true; /* a form feed is treated much like a newline */ + dump_line(); + ps.want_blank = false; + break; + + case newline: + if (ps.last_token != comma || ps.p_l_follow > 0 + || !ps.leave_comma || ps.block_init || !break_comma || s_com != e_com) { + dump_line(); + ps.want_blank = false; + } + ++line_no; /* keep track of input line number */ + break; + + case lparen: /* got a '(' or '[' */ + ++ps.p_l_follow; /* count parens to make Healy happy */ + if (ps.want_blank && *token != '[' && + (ps.last_token != ident || proc_calls_space + || (ps.its_a_keyword && (!ps.sizeof_keyword || Bill_Shannon)))) + *e_code++ = ' '; + if (ps.in_decl && !ps.block_init) + if (troff && !ps.dumped_decl_indent && !is_procname && ps.last_token == decl) { + ps.dumped_decl_indent = 1; + sprintf(e_code, "\n.Du %dp+\200p \"%s\"\n", dec_ind * 7, token); + e_code += strlen(e_code); + } + else { + while ((e_code - s_code) < dec_ind) { + CHECK_SIZE_CODE; + *e_code++ = ' '; + } + *e_code++ = token[0]; + } + else + *e_code++ = token[0]; + ps.paren_indents[ps.p_l_follow - 1] = e_code - s_code; + if (sp_sw && ps.p_l_follow == 1 && extra_expression_indent + && ps.paren_indents[0] < 2 * ps.ind_size) + ps.paren_indents[0] = 2 * ps.ind_size; + ps.want_blank = false; + if (ps.in_or_st && *token == '(' && ps.tos <= 2) { + /* + * this is a kluge to make sure that declarations will be + * aligned right if proc decl has an explicit type on it, i.e. + * "int a(x) {..." + */ + parse(semicolon); /* I said this was a kluge... */ + ps.in_or_st = false; /* turn off flag for structure decl or + * initialization */ + } + if (ps.sizeof_keyword) + ps.sizeof_mask |= 1 << ps.p_l_follow; + break; + + case rparen: /* got a ')' or ']' */ + rparen_count--; + if (ps.cast_mask & (1 << ps.p_l_follow) & ~ps.sizeof_mask) { + ps.last_u_d = true; + ps.cast_mask &= (1 << ps.p_l_follow) - 1; + } + ps.sizeof_mask &= (1 << ps.p_l_follow) - 1; + if (--ps.p_l_follow < 0) { + ps.p_l_follow = 0; + diag(0, "Extra %c", *token); + } + if (e_code == s_code) /* if the paren starts the line */ + ps.paren_level = ps.p_l_follow; /* then indent it */ + + *e_code++ = token[0]; + ps.want_blank = true; + + if (sp_sw && (ps.p_l_follow == 0)) { /* check for end of if + * (...), or some such */ + sp_sw = false; + force_nl = true;/* must force newline after if */ + ps.last_u_d = true; /* inform lexi that a following + * operator is unary */ + ps.in_stmt = false; /* dont use stmt continuation + * indentation */ + + parse(hd_type); /* let parser worry about if, or whatever */ + } + ps.search_brace = btype_2; /* this should insure that constructs + * such as main(){...} and int[]{...} + * have their braces put in the right + * place */ + break; + + case unary_op: /* this could be any unary operation */ + if (ps.want_blank) + *e_code++ = ' '; + + if (troff && !ps.dumped_decl_indent && ps.in_decl && !is_procname) { + sprintf(e_code, "\n.Du %dp+\200p \"%s\"\n", dec_ind * 7, token); + ps.dumped_decl_indent = 1; + e_code += strlen(e_code); + } + else { + char *res = token; + + if (ps.in_decl && !ps.block_init) { /* if this is a unary op + * in a declaration, we + * should indent this + * token */ + for (i = 0; token[i]; ++i); /* find length of token */ + while ((e_code - s_code) < (dec_ind - i)) { + CHECK_SIZE_CODE; + *e_code++ = ' '; /* pad it */ + } + } + if (troff && token[0] == '-' && token[1] == '>') + res = "\\(->"; + for (t_ptr = res; *t_ptr; ++t_ptr) { + CHECK_SIZE_CODE; + *e_code++ = *t_ptr; + } + } + ps.want_blank = false; + break; + + case binary_op: /* any binary operation */ + if (ps.want_blank) + *e_code++ = ' '; + { + char *res = token; + + if (troff) + switch (token[0]) { + case '<': + if (token[1] == '=') + res = "\\(<="; + break; + case '>': + if (token[1] == '=') + res = "\\(>="; + break; + case '!': + if (token[1] == '=') + res = "\\(!="; + break; + case '|': + if (token[1] == '|') + res = "\\(br\\(br"; + else if (token[1] == 0) + res = "\\(br"; + break; + } + for (t_ptr = res; *t_ptr; ++t_ptr) { + CHECK_SIZE_CODE; + *e_code++ = *t_ptr; /* move the operator */ + } + } + ps.want_blank = true; + break; + + case postop: /* got a trailing ++ or -- */ + *e_code++ = token[0]; + *e_code++ = token[1]; + ps.want_blank = true; + break; + + case question: /* got a ? */ + squest++; /* this will be used when a later colon + * appears so we can distinguish the + * <c>?<n>:<n> construct */ + if (ps.want_blank) + *e_code++ = ' '; + *e_code++ = '?'; + ps.want_blank = true; + break; + + case casestmt: /* got word 'case' or 'default' */ + scase = true; /* so we can process the later colon properly */ + goto copy_id; + + case colon: /* got a ':' */ + if (squest > 0) { /* it is part of the <c>?<n>: <n> construct */ + --squest; + if (ps.want_blank) + *e_code++ = ' '; + *e_code++ = ':'; + ps.want_blank = true; + break; + } + if (ps.in_decl) { + *e_code++ = ':'; + ps.want_blank = false; + break; + } + ps.in_stmt = false; /* seeing a label does not imply we are in a + * stmt */ + for (t_ptr = s_code; *t_ptr; ++t_ptr) + *e_lab++ = *t_ptr; /* turn everything so far into a label */ + e_code = s_code; + *e_lab++ = ':'; + *e_lab++ = ' '; + *e_lab = '\0'; + + force_nl = ps.pcase = scase; /* ps.pcase will be used by + * dump_line to decide how to + * indent the label. force_nl + * will force a case n: to be + * on a line by itself */ + scase = false; + ps.want_blank = false; + break; + + case semicolon: /* got a ';' */ + ps.in_or_st = false;/* we are not in an initialization or + * structure declaration */ + scase = false; /* these will only need resetting in a error */ + squest = 0; + if (ps.last_token == rparen && rparen_count == 0) + ps.in_parameter_declaration = 0; + ps.cast_mask = 0; + ps.sizeof_mask = 0; + ps.block_init = 0; + ps.block_init_level = 0; + ps.just_saw_decl--; + + if (ps.in_decl && s_code == e_code && !ps.block_init) + while ((e_code - s_code) < (dec_ind - 1)) { + CHECK_SIZE_CODE; + *e_code++ = ' '; + } + + ps.in_decl = (ps.dec_nest > 0); /* if we were in a first level + * structure declaration, we + * arent any more */ + + if ((!sp_sw || hd_type != forstmt) && ps.p_l_follow > 0) { + + /* + * This should be true iff there were unbalanced parens in the + * stmt. It is a bit complicated, because the semicolon might + * be in a for stmt + */ + diag(1, "Unbalanced parens"); + ps.p_l_follow = 0; + if (sp_sw) { /* this is a check for a if, while, etc. with + * unbalanced parens */ + sp_sw = false; + parse(hd_type); /* dont lose the if, or whatever */ + } + } + *e_code++ = ';'; + ps.want_blank = true; + ps.in_stmt = (ps.p_l_follow > 0); /* we are no longer in the + * middle of a stmt */ + + if (!sp_sw) { /* if not if for (;;) */ + parse(semicolon); /* let parser know about end of stmt */ + force_nl = true;/* force newline after a end of stmt */ + } + break; + + case lbrace: /* got a '{' */ + ps.in_stmt = false; /* dont indent the {} */ + if (!ps.block_init) + force_nl = true;/* force other stuff on same line as '{' onto + * new line */ + else if (ps.block_init_level <= 0) + ps.block_init_level = 1; + else + ps.block_init_level++; + + if (s_code != e_code && !ps.block_init) { + if (!btype_2) { + dump_line(); + ps.want_blank = false; + } + else if (ps.in_parameter_declaration && !ps.in_or_st) { + ps.i_l_follow = 0; + dump_line(); + ps.want_blank = false; + } + } + if (ps.in_parameter_declaration) + prefix_blankline_requested = 0; + + if (ps.p_l_follow > 0) { /* check for preceeding unbalanced + * parens */ + diag(1, "Unbalanced parens"); + ps.p_l_follow = 0; + if (sp_sw) { /* check for unclosed if, for, etc. */ + sp_sw = false; + parse(hd_type); + ps.ind_level = ps.i_l_follow; + } + } + if (s_code == e_code) + ps.ind_stmt = false; /* dont put extra indentation on line + * with '{' */ + if (ps.in_decl && ps.in_or_st) { /* this is either a structure + * declaration or an init */ + di_stack[ps.dec_nest++] = dec_ind; + /* ? dec_ind = 0; */ + } + else { + ps.decl_on_line = false; /* we cant be in the middle of + * a declaration, so dont do + * special indentation of + * comments */ + if (blanklines_after_declarations_at_proctop + && ps.in_parameter_declaration) + postfix_blankline_requested = 1; + ps.in_parameter_declaration = 0; + } + dec_ind = 0; + parse(lbrace); /* let parser know about this */ + if (ps.want_blank) /* put a blank before '{' if '{' is not at + * start of line */ + *e_code++ = ' '; + ps.want_blank = false; + *e_code++ = '{'; + ps.just_saw_decl = 0; + break; + + case rbrace: /* got a '}' */ + if (ps.p_stack[ps.tos] == decl && !ps.block_init) /* semicolons can be + * omitted in + * declarations */ + parse(semicolon); + if (ps.p_l_follow) {/* check for unclosed if, for, else. */ + diag(1, "Unbalanced parens"); + ps.p_l_follow = 0; + sp_sw = false; + } + ps.just_saw_decl = 0; + ps.block_init_level--; + if (s_code != e_code && !ps.block_init) { /* '}' must be first on + * line */ + if (verbose) + diag(0, "Line broken"); + dump_line(); + } + *e_code++ = '}'; + ps.want_blank = true; + ps.in_stmt = ps.ind_stmt = false; + if (ps.dec_nest > 0) { /* we are in multi-level structure + * declaration */ + dec_ind = di_stack[--ps.dec_nest]; + if (ps.dec_nest == 0 && !ps.in_parameter_declaration) + ps.just_saw_decl = 2; + ps.in_decl = true; + } + prefix_blankline_requested = 0; + parse(rbrace); /* let parser know about this */ + ps.search_brace = cuddle_else && ps.p_stack[ps.tos] == ifhead + && ps.il[ps.tos] >= ps.ind_level; + if (ps.tos <= 1 && blanklines_after_procs && ps.dec_nest <= 0) + postfix_blankline_requested = 1; + break; + + case swstmt: /* got keyword "switch" */ + sp_sw = true; + hd_type = swstmt; /* keep this for when we have seen the + * expression */ + goto copy_id; /* go move the token into buffer */ + + case sp_paren: /* token is if, while, for */ + sp_sw = true; /* the interesting stuff is done after the + * expression is scanned */ + hd_type = (*token == 'i' ? ifstmt : + (*token == 'w' ? whilestmt : forstmt)); + + /* + * remember the type of header for later use by parser + */ + goto copy_id; /* copy the token into line */ + + case sp_nparen: /* got else, do */ + ps.in_stmt = false; + if (*token == 'e') { + if (e_code != s_code && (!cuddle_else || e_code[-1] != '}')) { + if (verbose) + diag(0, "Line broken"); + dump_line();/* make sure this starts a line */ + ps.want_blank = false; + } + force_nl = true;/* also, following stuff must go onto new line */ + last_else = 1; + parse(elselit); + } + else { + if (e_code != s_code) { /* make sure this starts a line */ + if (verbose) + diag(0, "Line broken"); + dump_line(); + ps.want_blank = false; + } + force_nl = true;/* also, following stuff must go onto new line */ + last_else = 0; + parse(dolit); + } + goto copy_id; /* move the token into line */ + + case decl: /* we have a declaration type (int, register, + * etc.) */ + parse(decl); /* let parser worry about indentation */ + if (ps.last_token == rparen && ps.tos <= 1) { + ps.in_parameter_declaration = 1; + if (s_code != e_code) { + dump_line(); + ps.want_blank = 0; + } + } + if (ps.in_parameter_declaration && ps.indent_parameters && ps.dec_nest == 0) { + ps.ind_level = ps.i_l_follow = 1; + ps.ind_stmt = 0; + } + ps.in_or_st = true; /* this might be a structure or initialization + * declaration */ + ps.in_decl = ps.decl_on_line = true; + if ( /* !ps.in_or_st && */ ps.dec_nest <= 0) + ps.just_saw_decl = 2; + prefix_blankline_requested = 0; + for (i = 0; token[i++];); /* get length of token */ + + /* + * dec_ind = e_code - s_code + (ps.decl_indent>i ? ps.decl_indent + * : i); + */ + dec_ind = ps.decl_indent > 0 ? ps.decl_indent : i; + goto copy_id; + + case ident: /* got an identifier or constant */ + if (ps.in_decl) { /* if we are in a declaration, we must indent + * identifier */ + if (ps.want_blank) + *e_code++ = ' '; + ps.want_blank = false; + if (is_procname == 0 || !procnames_start_line) { + if (!ps.block_init) + if (troff && !ps.dumped_decl_indent) { + sprintf(e_code, "\n.De %dp+\200p\n", dec_ind * 7); + ps.dumped_decl_indent = 1; + e_code += strlen(e_code); + } + else + while ((e_code - s_code) < dec_ind) { + CHECK_SIZE_CODE; + *e_code++ = ' '; + } + } + else { + if (dec_ind && s_code != e_code) + dump_line(); + dec_ind = 0; + ps.want_blank = false; + } + } + else if (sp_sw && ps.p_l_follow == 0) { + sp_sw = false; + force_nl = true; + ps.last_u_d = true; + ps.in_stmt = false; + parse(hd_type); + } + copy_id: + if (ps.want_blank) + *e_code++ = ' '; + if (troff && ps.its_a_keyword) { + e_code = chfont(&bodyf, &keywordf, e_code); + for (t_ptr = token; *t_ptr; ++t_ptr) { + CHECK_SIZE_CODE; + *e_code++ = keywordf.allcaps && islower(*t_ptr) + ? toupper(*t_ptr) : *t_ptr; + } + e_code = chfont(&keywordf, &bodyf, e_code); + } + else + for (t_ptr = token; *t_ptr; ++t_ptr) { + CHECK_SIZE_CODE; + *e_code++ = *t_ptr; + } + ps.want_blank = true; + break; + + case period: /* treat a period kind of like a binary + * operation */ + *e_code++ = '.'; /* move the period into line */ + ps.want_blank = false; /* dont put a blank after a period */ + break; + + case comma: + ps.want_blank = (s_code != e_code); /* only put blank after comma + * if comma does not start the + * line */ + if (ps.in_decl && is_procname == 0 && !ps.block_init) + while ((e_code - s_code) < (dec_ind - 1)) { + CHECK_SIZE_CODE; + *e_code++ = ' '; + } + + *e_code++ = ','; + if (ps.p_l_follow == 0) { + if (ps.block_init_level <= 0) + ps.block_init = 0; + if (break_comma && (!ps.leave_comma || compute_code_target() + (e_code - s_code) > max_col - 8)) + force_nl = true; + } + break; + + case preesc: /* got the character '#' */ + if ((s_com != e_com) || + (s_lab != e_lab) || + (s_code != e_code)) + dump_line(); + *e_lab++ = '#'; /* move whole line to 'label' buffer */ + { + int in_comment = 0; + int com_start = 0; + char quote = 0; + int com_end = 0; + + while (*buf_ptr == ' ' || *buf_ptr == '\t') { + buf_ptr++; + if (buf_ptr >= buf_end) + fill_buffer(); + } + while (*buf_ptr != '\n' || in_comment) { + CHECK_SIZE_LAB; + *e_lab = *buf_ptr++; + if (buf_ptr >= buf_end) + fill_buffer(); + switch (*e_lab++) { + case BACKSLASH: + if (troff) + *e_lab++ = BACKSLASH; + if (!in_comment) { + *e_lab++ = *buf_ptr++; + if (buf_ptr >= buf_end) + fill_buffer(); + } + break; + case '/': + if (*buf_ptr == '*' && !in_comment && !quote) { + in_comment = 1; + *e_lab++ = *buf_ptr++; + com_start = e_lab - s_lab - 2; + } + break; + case '"': + if (quote == '"') + quote = 0; + break; + case '\'': + if (quote == '\'') + quote = 0; + break; + case '*': + if (*buf_ptr == '/' && in_comment) { + in_comment = 0; + *e_lab++ = *buf_ptr++; + com_end = e_lab - s_lab; + } + break; + } + } + + while (e_lab > s_lab && (e_lab[-1] == ' ' || e_lab[-1] == '\t')) + e_lab--; + if (e_lab - s_lab == com_end && bp_save == 0) { /* comment on + * preprocessor line */ + if (sc_end == 0) /* if this is the first comment, we + * must set up the buffer */ + sc_end = &(save_com[0]); + else { + *sc_end++ = '\n'; /* add newline between + * comments */ + *sc_end++ = ' '; + --line_no; + } + bcopy(s_lab + com_start, sc_end, com_end - com_start); + sc_end += com_end - com_start; + if (sc_end >= &save_com[sc_size]) + abort(); + e_lab = s_lab + com_start; + while (e_lab > s_lab && (e_lab[-1] == ' ' || e_lab[-1] == '\t')) + e_lab--; + bp_save = buf_ptr; /* save current input buffer */ + be_save = buf_end; + buf_ptr = save_com; /* fix so that subsequent calls to + * lexi will take tokens out of + * save_com */ + *sc_end++ = ' '; /* add trailing blank, just in case */ + buf_end = sc_end; + sc_end = 0; + } + *e_lab = '\0'; /* null terminate line */ + ps.pcase = false; + } + + if (strncmp(s_lab, "#if", 3) == 0) { + if (blanklines_around_conditional_compilation) { + register c; + prefix_blankline_requested++; + while ((c = getc(input)) == '\n'); + ungetc(c, input); + } + if (ifdef_level < sizeof state_stack / sizeof state_stack[0]) { + match_state[ifdef_level].tos = -1; + state_stack[ifdef_level++] = ps; + } + else + diag(1, "#if stack overflow"); + } + else if (strncmp(s_lab, "#else", 5) == 0) + if (ifdef_level <= 0) + diag(1, "Unmatched #else"); + else { + match_state[ifdef_level - 1] = ps; + ps = state_stack[ifdef_level - 1]; + } + else if (strncmp(s_lab, "#endif", 6) == 0) { + if (ifdef_level <= 0) + diag(1, "Unmatched #endif"); + else { + ifdef_level--; + +#ifdef undef + /* + * This match needs to be more intelligent before the + * message is useful + */ + if (match_state[ifdef_level].tos >= 0 + && bcmp(&ps, &match_state[ifdef_level], sizeof ps)) + diag(0, "Syntactically inconsistant #ifdef alternatives."); +#endif + } + if (blanklines_around_conditional_compilation) { + postfix_blankline_requested++; + n_real_blanklines = 0; + } + } + break; /* subsequent processing of the newline + * character will cause the line to be printed */ + + case comment: /* we have gotten a /* this is a biggie */ + if (flushed_nl) { /* we should force a broken line here */ + flushed_nl = false; + dump_line(); + ps.want_blank = false; /* dont insert blank at line start */ + force_nl = false; + } + pr_comment(); + break; + } /* end of big switch stmt */ + + *e_code = '\0'; /* make sure code section is null terminated */ + if (type_code != comment && type_code != newline && type_code != preesc) + ps.last_token = type_code; + } /* end of main while (1) loop */ +} + +/* + * copy input file to backup file if in_name is /blah/blah/blah/file, then + * backup file will be ".Bfile" then make the backup file the input and + * original input file the output + */ +bakcopy() +{ + int n, + bakchn; + char buff[8 * 1024]; + register char *p; + + /* construct file name .Bfile */ + for (p = in_name; *p; p++); /* skip to end of string */ + while (p > in_name && *p != '/') /* find last '/' */ + p--; + if (*p == '/') + p++; + sprintf(bakfile, "%s.BAK", p); + + /* copy in_name to backup file */ + bakchn = creat(bakfile, 0600); + if (bakchn < 0) + err(bakfile); + while (n = read(fileno(input), buff, sizeof buff)) + if (write(bakchn, buff, n) != n) + err(bakfile); + if (n < 0) + err(in_name); + close(bakchn); + fclose(input); + + /* re-open backup file as the input file */ + input = fopen(bakfile, "r"); + if (input == 0) + err(bakfile); + /* now the original input file will be the output */ + output = fopen(in_name, "w"); + if (output == 0) { + unlink(bakfile); + err(in_name); + } +} + +err(msg) + char *msg; +{ + extern int errno; + char *strerror(); + + (void)fprintf(stderr, "indent: %s: %s\n", msg, strerror(errno)); + exit(1); +} diff --git a/usr.bin/indent/indent_codes.h b/usr.bin/indent/indent_codes.h new file mode 100644 index 0000000..8373d4e --- /dev/null +++ b/usr.bin/indent/indent_codes.h @@ -0,0 +1,69 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * @(#)indent_codes.h 8.1 (Berkeley) 6/6/93 + */ + +#define newline 1 +#define lparen 2 +#define rparen 3 +#define unary_op 4 +#define binary_op 5 +#define postop 6 +#define question 7 +#define casestmt 8 +#define colon 9 +#define semicolon 10 +#define lbrace 11 +#define rbrace 12 +#define ident 13 +#define comma 14 +#define comment 15 +#define swstmt 16 +#define preesc 17 +#define form_feed 18 +#define decl 19 +#define sp_paren 20 +#define sp_nparen 21 +#define ifstmt 22 +#define whilestmt 23 +#define forstmt 24 +#define stmt 25 +#define stmtl 26 +#define elselit 27 +#define dolit 28 +#define dohead 29 +#define ifhead 30 +#define elsehead 31 +#define period 32 diff --git a/usr.bin/indent/indent_globs.h b/usr.bin/indent/indent_globs.h new file mode 100644 index 0000000..8f911ec --- /dev/null +++ b/usr.bin/indent/indent_globs.h @@ -0,0 +1,310 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + * + * @(#)indent_globs.h 8.1 (Berkeley) 6/6/93 + */ + +#define BACKSLASH '\\' +#define bufsize 200 /* size of internal buffers */ +#define sc_size 5000 /* size of save_com buffer */ +#define label_offset 2 /* number of levels a label is placed to left + * of code */ + +#define tabsize 8 /* the size of a tab */ +#define tabmask 0177770 /* mask used when figuring length of lines + * with tabs */ + + +#define false 0 +#define true 1 + + +FILE *input; /* the fid for the input file */ +FILE *output; /* the output file */ + +#define CHECK_SIZE_CODE \ + if (e_code >= l_code) { \ + register nsize = l_code-s_code+400; \ + codebuf = (char *) realloc(codebuf, nsize); \ + e_code = codebuf + (e_code-s_code) + 1; \ + l_code = codebuf + nsize - 5; \ + s_code = codebuf + 1; \ + } +#define CHECK_SIZE_COM \ + if (e_com >= l_com) { \ + register nsize = l_com-s_com+400; \ + combuf = (char *) realloc(combuf, nsize); \ + e_com = combuf + (e_com-s_com) + 1; \ + l_com = combuf + nsize - 5; \ + s_com = combuf + 1; \ + } +#define CHECK_SIZE_LAB \ + if (e_lab >= l_lab) { \ + register nsize = l_lab-s_lab+400; \ + labbuf = (char *) realloc(labbuf, nsize); \ + e_lab = labbuf + (e_lab-s_lab) + 1; \ + l_lab = labbuf + nsize - 5; \ + s_lab = labbuf + 1; \ + } +#define CHECK_SIZE_TOKEN \ + if (e_token >= l_token) { \ + register nsize = l_token-s_token+400; \ + tokenbuf = (char *) realloc(tokenbuf, nsize); \ + e_token = tokenbuf + (e_token-s_token) + 1; \ + l_token = tokenbuf + nsize - 5; \ + s_token = tokenbuf + 1; \ + } + +char *labbuf; /* buffer for label */ +char *s_lab; /* start ... */ +char *e_lab; /* .. and end of stored label */ +char *l_lab; /* limit of label buffer */ + +char *codebuf; /* buffer for code section */ +char *s_code; /* start ... */ +char *e_code; /* .. and end of stored code */ +char *l_code; /* limit of code section */ + +char *combuf; /* buffer for comments */ +char *s_com; /* start ... */ +char *e_com; /* ... and end of stored comments */ +char *l_com; /* limit of comment buffer */ + +#define token s_token +char *tokenbuf; /* the last token scanned */ +char *s_token; +char *e_token; +char *l_token; + +char *in_buffer; /* input buffer */ +char *in_buffer_limit; /* the end of the input buffer */ +char *buf_ptr; /* ptr to next character to be taken from + * in_buffer */ +char *buf_end; /* ptr to first after last char in in_buffer */ + +char save_com[sc_size]; /* input text is saved here when looking for + * the brace after an if, while, etc */ +char *sc_end; /* pointer into save_com buffer */ + +char *bp_save; /* saved value of buf_ptr when taking input + * from save_com */ +char *be_save; /* similarly saved value of buf_end */ + + +int pointer_as_binop; +int blanklines_after_declarations; +int blanklines_before_blockcomments; +int blanklines_after_procs; +int blanklines_around_conditional_compilation; +int swallow_optional_blanklines; +int n_real_blanklines; +int prefix_blankline_requested; +int postfix_blankline_requested; +int break_comma; /* when true and not in parens, break after a + * comma */ +int btype_2; /* when true, brace should be on same line as + * if, while, etc */ +float case_ind; /* indentation level to be used for a "case + * n:" */ +int code_lines; /* count of lines with code */ +int had_eof; /* set to true when input is exhausted */ +int line_no; /* the current line number. */ +int max_col; /* the maximum allowable line length */ +int verbose; /* when true, non-essential error messages are + * printed */ +int cuddle_else; /* true if else should cuddle up to '}' */ +int star_comment_cont; /* true iff comment continuation lines should + * have stars at the beginning of each line. */ +int comment_delimiter_on_blankline; +int troff; /* true iff were generating troff input */ +int procnames_start_line; /* if true, the names of procedures + * being defined get placed in column + * 1 (ie. a newline is placed between + * the type of the procedure and its + * name) */ +int proc_calls_space; /* If true, procedure calls look like: + * foo(bar) rather than foo (bar) */ +int format_col1_comments; /* If comments which start in column 1 + * are to be magically reformatted + * (just like comments that begin in + * later columns) */ +int inhibit_formatting; /* true if INDENT OFF is in effect */ +int suppress_blanklines;/* set iff following blanklines should be + * suppressed */ +int continuation_indent;/* set to the indentation between the edge of + * code and continuation lines */ +int lineup_to_parens; /* if true, continued code within parens will + * be lined up to the open paren */ +int Bill_Shannon; /* true iff a blank should always be inserted + * after sizeof */ +int blanklines_after_declarations_at_proctop; /* This is vaguely + * similar to + * blanklines_after_decla + * rations except that + * it only applies to + * the first set of + * declarations in a + * procedure (just after + * the first '{') and it + * causes a blank line + * to be generated even + * if there are no + * declarations */ +int block_comment_max_col; +int extra_expression_indent; /* True if continuation lines from the + * expression part of "if(e)", + * "while(e)", "for(e;e;e)" should be + * indented an extra tab stop so that + * they don't conflict with the code + * that follows */ + +/* -troff font state information */ + +struct fstate { + char font[4]; + char size; + int allcaps:1; +}; +char *chfont(); + +struct fstate + keywordf, /* keyword font */ + stringf, /* string font */ + boxcomf, /* Box comment font */ + blkcomf, /* Block comment font */ + scomf, /* Same line comment font */ + bodyf; /* major body font */ + + +#define STACKSIZE 150 + +struct parser_state { + int last_token; + struct fstate cfont; /* Current font */ + int p_stack[STACKSIZE]; /* this is the parsers stack */ + int il[STACKSIZE]; /* this stack stores indentation levels */ + float cstk[STACKSIZE];/* used to store case stmt indentation levels */ + int box_com; /* set to true when we are in a "boxed" + * comment. In that case, the first non-blank + * char should be lined up with the / in /* */ + int comment_delta, + n_comment_delta; + int cast_mask; /* indicates which close parens close off + * casts */ + int sizeof_mask; /* indicates which close parens close off + * sizeof''s */ + int block_init; /* true iff inside a block initialization */ + int block_init_level; /* The level of brace nesting in an + * initialization */ + int last_nl; /* this is true if the last thing scanned was + * a newline */ + int in_or_st; /* Will be true iff there has been a + * declarator (e.g. int or char) and no left + * paren since the last semicolon. When true, + * a '{' is starting a structure definition or + * an initialization list */ + int bl_line; /* set to 1 by dump_line if the line is blank */ + int col_1; /* set to true if the last token started in + * column 1 */ + int com_col; /* this is the column in which the current + * coment should start */ + int com_ind; /* the column in which comments to the right + * of code should start */ + int com_lines; /* the number of lines with comments, set by + * dump_line */ + int dec_nest; /* current nesting level for structure or init */ + int decl_com_ind; /* the column in which comments after + * declarations should be put */ + int decl_on_line; /* set to true if this line of code has part + * of a declaration on it */ + int i_l_follow; /* the level to which ind_level should be set + * after the current line is printed */ + int in_decl; /* set to true when we are in a declaration + * stmt. The processing of braces is then + * slightly different */ + int in_stmt; /* set to 1 while in a stmt */ + int ind_level; /* the current indentation level */ + int ind_size; /* the size of one indentation level */ + int ind_stmt; /* set to 1 if next line should have an extra + * indentation level because we are in the + * middle of a stmt */ + int last_u_d; /* set to true after scanning a token which + * forces a following operator to be unary */ + int leave_comma; /* if true, never break declarations after + * commas */ + int ljust_decl; /* true if declarations should be left + * justified */ + int out_coms; /* the number of comments processed, set by + * pr_comment */ + int out_lines; /* the number of lines written, set by + * dump_line */ + int p_l_follow; /* used to remember how to indent following + * statement */ + int paren_level; /* parenthesization level. used to indent + * within stmts */ + short paren_indents[20]; /* column positions of each paren */ + int pcase; /* set to 1 if the current line label is a + * case. It is printed differently from a + * regular label */ + int search_brace; /* set to true by parse when it is necessary + * to buffer up all info up to the start of a + * stmt after an if, while, etc */ + int unindent_displace; /* comments not to the right of code + * will be placed this many + * indentation levels to the left of + * code */ + int use_ff; /* set to one if the current line should be + * terminated with a form feed */ + int want_blank; /* set to true when the following token should + * be prefixed by a blank. (Said prefixing is + * ignored in some cases.) */ + int else_if; /* True iff else if pairs should be handled + * specially */ + int decl_indent; /* column to indent declared identifiers to */ + int its_a_keyword; + int sizeof_keyword; + int dumped_decl_indent; + float case_indent; /* The distance to indent case labels from the + * switch statement */ + int in_parameter_declaration; + int indent_parameters; + int tos; /* pointer to top of stack */ + char procname[100]; /* The name of the current procedure */ + int just_saw_decl; +} ps; + +int ifdef_level; +int rparen_count; +struct parser_state state_stack[5]; +struct parser_state match_state[5]; diff --git a/usr.bin/indent/io.c b/usr.bin/indent/io.c new file mode 100644 index 0000000..4615db2 --- /dev/null +++ b/usr.bin/indent/io.c @@ -0,0 +1,625 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifndef lint +static char sccsid[] = "@(#)io.c 8.1 (Berkeley) 6/6/93"; +#endif /* not lint */ + +#include <stdio.h> +#include <ctype.h> +#include <stdlib.h> +#include <string.h> +#include "indent_globs.h" + + +int comment_open; +static paren_target; + +dump_line() +{ /* dump_line is the routine that actually + * effects the printing of the new source. It + * prints the label section, followed by the + * code section with the appropriate nesting + * level, followed by any comments */ + register int cur_col, + target_col; + static not_first_line; + + if (ps.procname[0]) { + if (troff) { + if (comment_open) { + comment_open = 0; + fprintf(output, ".*/\n"); + } + fprintf(output, ".Pr \"%s\"\n", ps.procname); + } + ps.ind_level = 0; + ps.procname[0] = 0; + } + if (s_code == e_code && s_lab == e_lab && s_com == e_com) { + if (suppress_blanklines > 0) + suppress_blanklines--; + else { + ps.bl_line = true; + n_real_blanklines++; + } + } + else if (!inhibit_formatting) { + suppress_blanklines = 0; + ps.bl_line = false; + if (prefix_blankline_requested && not_first_line) + if (swallow_optional_blanklines) { + if (n_real_blanklines == 1) + n_real_blanklines = 0; + } + else { + if (n_real_blanklines == 0) + n_real_blanklines = 1; + } + while (--n_real_blanklines >= 0) + putc('\n', output); + n_real_blanklines = 0; + if (ps.ind_level == 0) + ps.ind_stmt = 0; /* this is a class A kludge. dont do + * additional statement indentation if we are + * at bracket level 0 */ + + if (e_lab != s_lab || e_code != s_code) + ++code_lines; /* keep count of lines with code */ + + + if (e_lab != s_lab) { /* print lab, if any */ + if (comment_open) { + comment_open = 0; + fprintf(output, ".*/\n"); + } + while (e_lab > s_lab && (e_lab[-1] == ' ' || e_lab[-1] == '\t')) + e_lab--; + cur_col = pad_output(1, compute_label_target()); + if (s_lab[0] == '#' && (strncmp(s_lab, "#else", 5) == 0 + || strncmp(s_lab, "#endif", 6) == 0)) { + register char *s = s_lab; + if (e_lab[-1] == '\n') e_lab--; + do putc(*s++, output); + while (s < e_lab && 'a' <= *s && *s<='z'); + while ((*s == ' ' || *s == '\t') && s < e_lab) + s++; + if (s < e_lab) + fprintf(output, s[0]=='/' && s[1]=='*' ? "\t%.*s" : "\t/* %.*s */", + e_lab - s, s); + } + else fprintf(output, "%.*s", e_lab - s_lab, s_lab); + cur_col = count_spaces(cur_col, s_lab); + } + else + cur_col = 1; /* there is no label section */ + + ps.pcase = false; + + if (s_code != e_code) { /* print code section, if any */ + register char *p; + + if (comment_open) { + comment_open = 0; + fprintf(output, ".*/\n"); + } + target_col = compute_code_target(); + { + register i; + + for (i = 0; i < ps.p_l_follow; i++) + if (ps.paren_indents[i] >= 0) + ps.paren_indents[i] = -(ps.paren_indents[i] + target_col); + } + cur_col = pad_output(cur_col, target_col); + for (p = s_code; p < e_code; p++) + if (*p == (char) 0200) + fprintf(output, "%d", target_col * 7); + else + putc(*p, output); + cur_col = count_spaces(cur_col, s_code); + } + if (s_com != e_com) + if (troff) { + int all_here = 0; + register char *p; + + if (e_com[-1] == '/' && e_com[-2] == '*') + e_com -= 2, all_here++; + while (e_com > s_com && e_com[-1] == ' ') + e_com--; + *e_com = 0; + p = s_com; + while (*p == ' ') + p++; + if (p[0] == '/' && p[1] == '*') + p += 2, all_here++; + else if (p[0] == '*') + p += p[1] == '/' ? 2 : 1; + while (*p == ' ') + p++; + if (*p == 0) + goto inhibit_newline; + if (comment_open < 2 && ps.box_com) { + comment_open = 0; + fprintf(output, ".*/\n"); + } + if (comment_open == 0) { + if ('a' <= *p && *p <= 'z') + *p = *p + 'A' - 'a'; + if (e_com - p < 50 && all_here == 2) { + register char *follow = p; + fprintf(output, "\n.nr C! \\w\1"); + while (follow < e_com) { + switch (*follow) { + case '\n': + putc(' ', output); + case 1: + break; + case '\\': + putc('\\', output); + default: + putc(*follow, output); + } + follow++; + } + putc(1, output); + } + fprintf(output, "\n./* %dp %d %dp\n", + ps.com_col * 7, + (s_code != e_code || s_lab != e_lab) - ps.box_com, + target_col * 7); + } + comment_open = 1 + ps.box_com; + while (*p) { + if (*p == BACKSLASH) + putc(BACKSLASH, output); + putc(*p++, output); + } + } + else { /* print comment, if any */ + register target = ps.com_col; + register char *com_st = s_com; + + target += ps.comment_delta; + while (*com_st == '\t') + com_st++, target += 8; /* ? */ + while (target <= 0) + if (*com_st == ' ') + target++, com_st++; + else if (*com_st == '\t') + target = ((target - 1) & ~7) + 9, com_st++; + else + target = 1; + if (cur_col > target) { /* if comment cant fit on this line, + * put it on next line */ + putc('\n', output); + cur_col = 1; + ++ps.out_lines; + } + while (e_com > com_st && isspace(e_com[-1])) + e_com--; + cur_col = pad_output(cur_col, target); + if (!ps.box_com) { + if (star_comment_cont && (com_st[1] != '*' || e_com <= com_st + 1)) + if (com_st[1] == ' ' && com_st[0] == ' ' && e_com > com_st + 1) + com_st[1] = '*'; + else + fwrite(" * ", com_st[0] == '\t' ? 2 : com_st[0] == '*' ? 1 : 3, 1, output); + } + fwrite(com_st, e_com - com_st, 1, output); + ps.comment_delta = ps.n_comment_delta; + cur_col = count_spaces(cur_col, com_st); + ++ps.com_lines; /* count lines with comments */ + } + if (ps.use_ff) + putc('\014', output); + else + putc('\n', output); +inhibit_newline: + ++ps.out_lines; + if (ps.just_saw_decl == 1 && blanklines_after_declarations) { + prefix_blankline_requested = 1; + ps.just_saw_decl = 0; + } + else + prefix_blankline_requested = postfix_blankline_requested; + postfix_blankline_requested = 0; + } + ps.decl_on_line = ps.in_decl; /* if we are in the middle of a + * declaration, remember that fact for + * proper comment indentation */ + ps.ind_stmt = ps.in_stmt & ~ps.in_decl; /* next line should be + * indented if we have not + * completed this stmt and if + * we are not in the middle of + * a declaration */ + ps.use_ff = false; + ps.dumped_decl_indent = 0; + *(e_lab = s_lab) = '\0'; /* reset buffers */ + *(e_code = s_code) = '\0'; + *(e_com = s_com) = '\0'; + ps.ind_level = ps.i_l_follow; + ps.paren_level = ps.p_l_follow; + paren_target = -ps.paren_indents[ps.paren_level - 1]; + not_first_line = 1; + return; +} + +compute_code_target() +{ + register target_col = ps.ind_size * ps.ind_level + 1; + + if (ps.paren_level) + if (!lineup_to_parens) + target_col += continuation_indent * ps.paren_level; + else { + register w; + register t = paren_target; + + if ((w = count_spaces(t, s_code) - max_col) > 0 + && count_spaces(target_col, s_code) <= max_col) { + t -= w + 1; + if (t > target_col) + target_col = t; + } + else + target_col = t; + } + else if (ps.ind_stmt) + target_col += continuation_indent; + return target_col; +} + +compute_label_target() +{ + return + ps.pcase ? (int) (case_ind * ps.ind_size) + 1 + : *s_lab == '#' ? 1 + : ps.ind_size * (ps.ind_level - label_offset) + 1; +} + + +/* + * Copyright (C) 1976 by the Board of Trustees of the University of Illinois + * + * All rights reserved + * + * + * NAME: fill_buffer + * + * FUNCTION: Reads one block of input into input_buffer + * + * HISTORY: initial coding November 1976 D A Willcox of CAC 1/7/77 A + * Willcox of CAC Added check for switch back to partly full input + * buffer from temporary buffer + * + */ +int +fill_buffer() +{ /* this routine reads stuff from the input */ + register char *p; + register int i; + register FILE *f = input; + + if (bp_save != 0) { /* there is a partly filled input buffer left */ + buf_ptr = bp_save; /* dont read anything, just switch buffers */ + buf_end = be_save; + bp_save = be_save = 0; + if (buf_ptr < buf_end) + return; /* only return if there is really something in + * this buffer */ + } + for (p = in_buffer;;) { + if (p >= in_buffer_limit) { + register size = (in_buffer_limit - in_buffer) * 2 + 10; + register offset = p - in_buffer; + in_buffer = (char *) realloc(in_buffer, size); + if (in_buffer == 0) + err("input line too long"); + p = in_buffer + offset; + in_buffer_limit = in_buffer + size - 2; + } + if ((i = getc(f)) == EOF) { + *p++ = ' '; + *p++ = '\n'; + had_eof = true; + break; + } + *p++ = i; + if (i == '\n') + break; + } + buf_ptr = in_buffer; + buf_end = p; + if (p[-2] == '/' && p[-3] == '*') { + if (in_buffer[3] == 'I' && strncmp(in_buffer, "/**INDENT**", 11) == 0) + fill_buffer(); /* flush indent error message */ + else { + int com = 0; + + p = in_buffer; + while (*p == ' ' || *p == '\t') + p++; + if (*p == '/' && p[1] == '*') { + p += 2; + while (*p == ' ' || *p == '\t') + p++; + if (p[0] == 'I' && p[1] == 'N' && p[2] == 'D' && p[3] == 'E' + && p[4] == 'N' && p[5] == 'T') { + p += 6; + while (*p == ' ' || *p == '\t') + p++; + if (*p == '*') + com = 1; + else if (*p == 'O') + if (*++p == 'N') + p++, com = 1; + else if (*p == 'F' && *++p == 'F') + p++, com = 2; + while (*p == ' ' || *p == '\t') + p++; + if (p[0] == '*' && p[1] == '/' && p[2] == '\n' && com) { + if (s_com != e_com || s_lab != e_lab || s_code != e_code) + dump_line(); + if (!(inhibit_formatting = com - 1)) { + n_real_blanklines = 0; + postfix_blankline_requested = 0; + prefix_blankline_requested = 0; + suppress_blanklines = 1; + } + } + } + } + } + } + if (inhibit_formatting) { + p = in_buffer; + do + putc(*p, output); + while (*p++ != '\n'); + } + return; +} + +/* + * Copyright (C) 1976 by the Board of Trustees of the University of Illinois + * + * All rights reserved + * + * + * NAME: pad_output + * + * FUNCTION: Writes tabs and spaces to move the current column up to the desired + * position. + * + * ALGORITHM: Put tabs and/or blanks into pobuf, then write pobuf. + * + * PARAMETERS: current integer The current column target + * nteger The desired column + * + * RETURNS: Integer value of the new column. (If current >= target, no action is + * taken, and current is returned. + * + * GLOBALS: None + * + * CALLS: write (sys) + * + * CALLED BY: dump_line + * + * HISTORY: initial coding November 1976 D A Willcox of CAC + * + */ +pad_output(current, target) /* writes tabs and blanks (if necessary) to + * get the current output position up to the + * target column */ + int current; /* the current column value */ + int target; /* position we want it at */ +{ + register int curr; /* internal column pointer */ + register int tcur; + + if (troff) + fprintf(output, "\\h'|%dp'", (target - 1) * 7); + else { + if (current >= target) + return (current); /* line is already long enough */ + curr = current; + while ((tcur = ((curr - 1) & tabmask) + tabsize + 1) <= target) { + putc('\t', output); + curr = tcur; + } + while (curr++ < target) + putc(' ', output); /* pad with final blanks */ + } + return (target); +} + +/* + * Copyright (C) 1976 by the Board of Trustees of the University of Illinois + * + * All rights reserved + * + * + * NAME: count_spaces + * + * FUNCTION: Find out where printing of a given string will leave the current + * character position on output. + * + * ALGORITHM: Run thru input string and add appropriate values to current + * position. + * + * RETURNS: Integer value of position after printing "buffer" starting in column + * "current". + * + * HISTORY: initial coding November 1976 D A Willcox of CAC + * + */ +int +count_spaces(current, buffer) +/* + * this routine figures out where the character position will be after + * printing the text in buffer starting at column "current" + */ + int current; + char *buffer; +{ + register char *buf; /* used to look thru buffer */ + register int cur; /* current character counter */ + + cur = current; + + for (buf = buffer; *buf != '\0'; ++buf) { + switch (*buf) { + + case '\n': + case 014: /* form feed */ + cur = 1; + break; + + case '\t': + cur = ((cur - 1) & tabmask) + tabsize + 1; + break; + + case 010: /* backspace */ + --cur; + break; + + default: + ++cur; + break; + } /* end of switch */ + } /* end of for loop */ + return (cur); +} + +int found_err; +/* VARARGS2 */ +diag(level, msg, a, b) + char *msg; +{ + if (level) + found_err = 1; + if (output == stdout) { + fprintf(stdout, "/**INDENT** %s@%d: ", level == 0 ? "Warning" : "Error", line_no); + fprintf(stdout, msg, a, b); + fprintf(stdout, " */\n"); + } + else { + fprintf(stderr, "%s@%d: ", level == 0 ? "Warning" : "Error", line_no); + fprintf(stderr, msg, a, b); + fprintf(stderr, "\n"); + } +} + +writefdef(f, nm) + register struct fstate *f; +{ + fprintf(output, ".ds f%c %s\n.nr s%c %d\n", + nm, f->font, nm, f->size); +} + +char * +chfont(of, nf, s) + register struct fstate *of, + *nf; + char *s; +{ + if (of->font[0] != nf->font[0] + || of->font[1] != nf->font[1]) { + *s++ = '\\'; + *s++ = 'f'; + if (nf->font[1]) { + *s++ = '('; + *s++ = nf->font[0]; + *s++ = nf->font[1]; + } + else + *s++ = nf->font[0]; + } + if (nf->size != of->size) { + *s++ = '\\'; + *s++ = 's'; + if (nf->size < of->size) { + *s++ = '-'; + *s++ = '0' + of->size - nf->size; + } + else { + *s++ = '+'; + *s++ = '0' + nf->size - of->size; + } + } + return s; +} + + +parsefont(f, s0) + register struct fstate *f; + char *s0; +{ + register char *s = s0; + int sizedelta = 0; + bzero(f, sizeof *f); + while (*s) { + if (isdigit(*s)) + f->size = f->size * 10 + *s - '0'; + else if (isupper(*s)) + if (f->font[0]) + f->font[1] = *s; + else + f->font[0] = *s; + else if (*s == 'c') + f->allcaps = 1; + else if (*s == '+') + sizedelta++; + else if (*s == '-') + sizedelta--; + else { + fprintf(stderr, "indent: bad font specification: %s\n", s0); + exit(1); + } + s++; + } + if (f->font[0] == 0) + f->font[0] = 'R'; + if (bodyf.size == 0) + bodyf.size = 11; + if (f->size == 0) + f->size = bodyf.size + sizedelta; + else if (sizedelta > 0) + f->size += bodyf.size; + else + f->size = bodyf.size - f->size; +} diff --git a/usr.bin/indent/lexi.c b/usr.bin/indent/lexi.c new file mode 100644 index 0000000..8da9d2c --- /dev/null +++ b/usr.bin/indent/lexi.c @@ -0,0 +1,559 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifndef lint +static char sccsid[] = "@(#)lexi.c 8.1 (Berkeley) 6/6/93"; +#endif /* not lint */ + +/* + * Here we have the token scanner for indent. It scans off one token and puts + * it in the global variable "token". It returns a code, indicating the type + * of token scanned. + */ + +#include <stdio.h> +#include <ctype.h> +#include <stdlib.h> +#include <string.h> +#include "indent_globs.h" +#include "indent_codes.h" + +#define alphanum 1 +#define opchar 3 + +struct templ { + char *rwd; + int rwcode; +}; + +struct templ specials[100] = +{ + "switch", 1, + "case", 2, + "break", 0, + "struct", 3, + "union", 3, + "enum", 3, + "default", 2, + "int", 4, + "char", 4, + "float", 4, + "double", 4, + "long", 4, + "short", 4, + "typdef", 4, + "unsigned", 4, + "register", 4, + "static", 4, + "global", 4, + "extern", 4, + "void", 4, + "goto", 0, + "return", 0, + "if", 5, + "while", 5, + "for", 5, + "else", 6, + "do", 6, + "sizeof", 7, + 0, 0 +}; + +char chartype[128] = +{ /* this is used to facilitate the decision of + * what type (alphanumeric, operator) each + * character is */ + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 0, 0, 0, 0, 0, 0, 0, + 0, 3, 0, 0, 1, 3, 3, 0, + 0, 0, 3, 3, 0, 3, 0, 3, + 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 0, 0, 3, 3, 3, 3, + 0, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 0, 0, 0, 3, 1, + 0, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 1, 1, 1, 1, 1, + 1, 1, 1, 0, 3, 0, 3, 0 +}; + + + + +int +lexi() +{ + int unary_delim; /* this is set to 1 if the current token + * + * forces a following operator to be unary */ + static int last_code; /* the last token type returned */ + static int l_struct; /* set to 1 if the last token was 'struct' */ + int code; /* internal code to be returned */ + char qchar; /* the delimiter character for a string */ + + e_token = s_token; /* point to start of place to save token */ + unary_delim = false; + ps.col_1 = ps.last_nl; /* tell world that this token started in + * column 1 iff the last thing scanned was nl */ + ps.last_nl = false; + + while (*buf_ptr == ' ' || *buf_ptr == '\t') { /* get rid of blanks */ + ps.col_1 = false; /* leading blanks imply token is not in column + * 1 */ + if (++buf_ptr >= buf_end) + fill_buffer(); + } + + /* Scan an alphanumeric token */ + if (chartype[*buf_ptr] == alphanum || buf_ptr[0] == '.' && isdigit(buf_ptr[1])) { + /* + * we have a character or number + */ + register char *j; /* used for searching thru list of + * + * reserved words */ + register struct templ *p; + + if (isdigit(*buf_ptr) || buf_ptr[0] == '.' && isdigit(buf_ptr[1])) { + int seendot = 0, + seenexp = 0; + if (*buf_ptr == '0' && + (buf_ptr[1] == 'x' || buf_ptr[1] == 'X')) { + *e_token++ = *buf_ptr++; + *e_token++ = *buf_ptr++; + while (isxdigit(*buf_ptr)) { + CHECK_SIZE_TOKEN; + *e_token++ = *buf_ptr++; + } + } + else + while (1) { + if (*buf_ptr == '.') + if (seendot) + break; + else + seendot++; + CHECK_SIZE_TOKEN; + *e_token++ = *buf_ptr++; + if (!isdigit(*buf_ptr) && *buf_ptr != '.') + if ((*buf_ptr != 'E' && *buf_ptr != 'e') || seenexp) + break; + else { + seenexp++; + seendot++; + CHECK_SIZE_TOKEN; + *e_token++ = *buf_ptr++; + if (*buf_ptr == '+' || *buf_ptr == '-') + *e_token++ = *buf_ptr++; + } + } + if (*buf_ptr == 'L' || *buf_ptr == 'l') + *e_token++ = *buf_ptr++; + } + else + while (chartype[*buf_ptr] == alphanum) { /* copy it over */ + CHECK_SIZE_TOKEN; + *e_token++ = *buf_ptr++; + if (buf_ptr >= buf_end) + fill_buffer(); + } + *e_token++ = '\0'; + while (*buf_ptr == ' ' || *buf_ptr == '\t') { /* get rid of blanks */ + if (++buf_ptr >= buf_end) + fill_buffer(); + } + ps.its_a_keyword = false; + ps.sizeof_keyword = false; + if (l_struct) { /* if last token was 'struct', then this token + * should be treated as a declaration */ + l_struct = false; + last_code = ident; + ps.last_u_d = true; + return (decl); + } + ps.last_u_d = false; /* Operator after indentifier is binary */ + last_code = ident; /* Remember that this is the code we will + * return */ + + /* + * This loop will check if the token is a keyword. + */ + for (p = specials; (j = p->rwd) != 0; p++) { + register char *p = s_token; /* point at scanned token */ + if (*j++ != *p++ || *j++ != *p++) + continue; /* This test depends on the fact that + * identifiers are always at least 1 character + * long (ie. the first two bytes of the + * identifier are always meaningful) */ + if (p[-1] == 0) + break; /* If its a one-character identifier */ + while (*p++ == *j) + if (*j++ == 0) + goto found_keyword; /* I wish that C had a multi-level + * break... */ + } + if (p->rwd) { /* we have a keyword */ + found_keyword: + ps.its_a_keyword = true; + ps.last_u_d = true; + switch (p->rwcode) { + case 1: /* it is a switch */ + return (swstmt); + case 2: /* a case or default */ + return (casestmt); + + case 3: /* a "struct" */ + if (ps.p_l_follow) + break; /* inside parens: cast */ + l_struct = true; + + /* + * Next time around, we will want to know that we have had a + * 'struct' + */ + case 4: /* one of the declaration keywords */ + if (ps.p_l_follow) { + ps.cast_mask |= 1 << ps.p_l_follow; + break; /* inside parens: cast */ + } + last_code = decl; + return (decl); + + case 5: /* if, while, for */ + return (sp_paren); + + case 6: /* do, else */ + return (sp_nparen); + + case 7: + ps.sizeof_keyword = true; + default: /* all others are treated like any other + * identifier */ + return (ident); + } /* end of switch */ + } /* end of if (found_it) */ + if (*buf_ptr == '(' && ps.tos <= 1 && ps.ind_level == 0) { + register char *tp = buf_ptr; + while (tp < buf_end) + if (*tp++ == ')' && (*tp == ';' || *tp == ',')) + goto not_proc; + strncpy(ps.procname, token, sizeof ps.procname - 1); + ps.in_parameter_declaration = 1; + rparen_count = 1; + not_proc:; + } + /* + * The following hack attempts to guess whether or not the current + * token is in fact a declaration keyword -- one that has been + * typedefd + */ + if (((*buf_ptr == '*' && buf_ptr[1] != '=') || isalpha(*buf_ptr) || *buf_ptr == '_') + && !ps.p_l_follow + && !ps.block_init + && (ps.last_token == rparen || ps.last_token == semicolon || + ps.last_token == decl || + ps.last_token == lbrace || ps.last_token == rbrace)) { + ps.its_a_keyword = true; + ps.last_u_d = true; + last_code = decl; + return decl; + } + if (last_code == decl) /* if this is a declared variable, then + * following sign is unary */ + ps.last_u_d = true; /* will make "int a -1" work */ + last_code = ident; + return (ident); /* the ident is not in the list */ + } /* end of procesing for alpanum character */ + + /* Scan a non-alphanumeric token */ + + *e_token++ = *buf_ptr; /* if it is only a one-character token, it is + * moved here */ + *e_token = '\0'; + if (++buf_ptr >= buf_end) + fill_buffer(); + + switch (*token) { + case '\n': + unary_delim = ps.last_u_d; + ps.last_nl = true; /* remember that we just had a newline */ + code = (had_eof ? 0 : newline); + + /* + * if data has been exausted, the newline is a dummy, and we should + * return code to stop + */ + break; + + case '\'': /* start of quoted character */ + case '"': /* start of string */ + qchar = *token; + if (troff) { + e_token[-1] = '`'; + if (qchar == '"') + *e_token++ = '`'; + e_token = chfont(&bodyf, &stringf, e_token); + } + do { /* copy the string */ + while (1) { /* move one character or [/<char>]<char> */ + if (*buf_ptr == '\n') { + printf("%d: Unterminated literal\n", line_no); + goto stop_lit; + } + CHECK_SIZE_TOKEN; /* Only have to do this once in this loop, + * since CHECK_SIZE guarantees that there + * are at least 5 entries left */ + *e_token = *buf_ptr++; + if (buf_ptr >= buf_end) + fill_buffer(); + if (*e_token == BACKSLASH) { /* if escape, copy extra char */ + if (*buf_ptr == '\n') /* check for escaped newline */ + ++line_no; + if (troff) { + *++e_token = BACKSLASH; + if (*buf_ptr == BACKSLASH) + *++e_token = BACKSLASH; + } + *++e_token = *buf_ptr++; + ++e_token; /* we must increment this again because we + * copied two chars */ + if (buf_ptr >= buf_end) + fill_buffer(); + } + else + break; /* we copied one character */ + } /* end of while (1) */ + } while (*e_token++ != qchar); + if (troff) { + e_token = chfont(&stringf, &bodyf, e_token - 1); + if (qchar == '"') + *e_token++ = '\''; + } +stop_lit: + code = ident; + break; + + case ('('): + case ('['): + unary_delim = true; + code = lparen; + break; + + case (')'): + case (']'): + code = rparen; + break; + + case '#': + unary_delim = ps.last_u_d; + code = preesc; + break; + + case '?': + unary_delim = true; + code = question; + break; + + case (':'): + code = colon; + unary_delim = true; + break; + + case (';'): + unary_delim = true; + code = semicolon; + break; + + case ('{'): + unary_delim = true; + + /* + * if (ps.in_or_st) ps.block_init = 1; + */ + /* ? code = ps.block_init ? lparen : lbrace; */ + code = lbrace; + break; + + case ('}'): + unary_delim = true; + /* ? code = ps.block_init ? rparen : rbrace; */ + code = rbrace; + break; + + case 014: /* a form feed */ + unary_delim = ps.last_u_d; + ps.last_nl = true; /* remember this so we can set 'ps.col_1' + * right */ + code = form_feed; + break; + + case (','): + unary_delim = true; + code = comma; + break; + + case '.': + unary_delim = false; + code = period; + break; + + case '-': + case '+': /* check for -, +, --, ++ */ + code = (ps.last_u_d ? unary_op : binary_op); + unary_delim = true; + + if (*buf_ptr == token[0]) { + /* check for doubled character */ + *e_token++ = *buf_ptr++; + /* buffer overflow will be checked at end of loop */ + if (last_code == ident || last_code == rparen) { + code = (ps.last_u_d ? unary_op : postop); + /* check for following ++ or -- */ + unary_delim = false; + } + } + else if (*buf_ptr == '=') + /* check for operator += */ + *e_token++ = *buf_ptr++; + else if (*buf_ptr == '>') { + /* check for operator -> */ + *e_token++ = *buf_ptr++; + if (!pointer_as_binop) { + unary_delim = false; + code = unary_op; + ps.want_blank = false; + } + } + break; /* buffer overflow will be checked at end of + * switch */ + + case '=': + if (ps.in_or_st) + ps.block_init = 1; +#ifdef undef + if (chartype[*buf_ptr] == opchar) { /* we have two char assignment */ + e_token[-1] = *buf_ptr++; + if ((e_token[-1] == '<' || e_token[-1] == '>') && e_token[-1] == *buf_ptr) + *e_token++ = *buf_ptr++; + *e_token++ = '='; /* Flip =+ to += */ + *e_token = 0; + } +#else + if (*buf_ptr == '=') {/* == */ + *e_token++ = '='; /* Flip =+ to += */ + buf_ptr++; + *e_token = 0; + } +#endif + code = binary_op; + unary_delim = true; + break; + /* can drop thru!!! */ + + case '>': + case '<': + case '!': /* ops like <, <<, <=, !=, etc */ + if (*buf_ptr == '>' || *buf_ptr == '<' || *buf_ptr == '=') { + *e_token++ = *buf_ptr; + if (++buf_ptr >= buf_end) + fill_buffer(); + } + if (*buf_ptr == '=') + *e_token++ = *buf_ptr++; + code = (ps.last_u_d ? unary_op : binary_op); + unary_delim = true; + break; + + default: + if (token[0] == '/' && *buf_ptr == '*') { + /* it is start of comment */ + *e_token++ = '*'; + + if (++buf_ptr >= buf_end) + fill_buffer(); + + code = comment; + unary_delim = ps.last_u_d; + break; + } + while (*(e_token - 1) == *buf_ptr || *buf_ptr == '=') { + /* + * handle ||, &&, etc, and also things as in int *****i + */ + *e_token++ = *buf_ptr; + if (++buf_ptr >= buf_end) + fill_buffer(); + } + code = (ps.last_u_d ? unary_op : binary_op); + unary_delim = true; + + + } /* end of switch */ + if (code != newline) { + l_struct = false; + last_code = code; + } + if (buf_ptr >= buf_end) /* check for input buffer empty */ + fill_buffer(); + ps.last_u_d = unary_delim; + *e_token = '\0'; /* null terminate the token */ + return (code); +} + +/* + * Add the given keyword to the keyword table, using val as the keyword type + */ +addkey(key, val) + char *key; +{ + register struct templ *p = specials; + while (p->rwd) + if (p->rwd[0] == key[0] && strcmp(p->rwd, key) == 0) + return; + else + p++; + if (p >= specials + sizeof specials / sizeof specials[0]) + return; /* For now, table overflows are silently + * ignored */ + p->rwd = key; + p->rwcode = val; + p[1].rwd = 0; + p[1].rwcode = 0; + return; +} diff --git a/usr.bin/indent/parse.c b/usr.bin/indent/parse.c new file mode 100644 index 0000000..fef22cf --- /dev/null +++ b/usr.bin/indent/parse.c @@ -0,0 +1,324 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifndef lint +static char sccsid[] = "@(#)parse.c 8.1 (Berkeley) 6/6/93"; +#endif /* not lint */ + +#include <stdio.h> +#include "indent_globs.h" +#include "indent_codes.h" + +parse(tk) + int tk; /* the code for the construct scanned */ +{ + int i; + +#ifdef debug + printf("%2d - %s\n", tk, token); +#endif + + while (ps.p_stack[ps.tos] == ifhead && tk != elselit) { + /* true if we have an if without an else */ + ps.p_stack[ps.tos] = stmt; /* apply the if(..) stmt ::= stmt + * reduction */ + reduce(); /* see if this allows any reduction */ + } + + + switch (tk) { /* go on and figure out what to do with the + * input */ + + case decl: /* scanned a declaration word */ + ps.search_brace = btype_2; + /* indicate that following brace should be on same line */ + if (ps.p_stack[ps.tos] != decl) { /* only put one declaration + * onto stack */ + break_comma = true; /* while in declaration, newline should be + * forced after comma */ + ps.p_stack[++ps.tos] = decl; + ps.il[ps.tos] = ps.i_l_follow; + + if (ps.ljust_decl) {/* only do if we want left justified + * declarations */ + ps.ind_level = 0; + for (i = ps.tos - 1; i > 0; --i) + if (ps.p_stack[i] == decl) + ++ps.ind_level; /* indentation is number of + * declaration levels deep we are */ + ps.i_l_follow = ps.ind_level; + } + } + break; + + case ifstmt: /* scanned if (...) */ + if (ps.p_stack[ps.tos] == elsehead && ps.else_if) /* "else if ..." */ + ps.i_l_follow = ps.il[ps.tos]; + case dolit: /* 'do' */ + case forstmt: /* for (...) */ + ps.p_stack[++ps.tos] = tk; + ps.il[ps.tos] = ps.ind_level = ps.i_l_follow; + ++ps.i_l_follow; /* subsequent statements should be indented 1 */ + ps.search_brace = btype_2; + break; + + case lbrace: /* scanned { */ + break_comma = false; /* don't break comma in an initial list */ + if (ps.p_stack[ps.tos] == stmt || ps.p_stack[ps.tos] == decl + || ps.p_stack[ps.tos] == stmtl) + ++ps.i_l_follow; /* it is a random, isolated stmt group or a + * declaration */ + else { + if (s_code == e_code) { + /* + * only do this if there is nothing on the line + */ + --ps.ind_level; + /* + * it is a group as part of a while, for, etc. + */ + if (ps.p_stack[ps.tos] == swstmt && ps.case_indent >= 1) + --ps.ind_level; + /* + * for a switch, brace should be two levels out from the code + */ + } + } + + ps.p_stack[++ps.tos] = lbrace; + ps.il[ps.tos] = ps.ind_level; + ps.p_stack[++ps.tos] = stmt; + /* allow null stmt between braces */ + ps.il[ps.tos] = ps.i_l_follow; + break; + + case whilestmt: /* scanned while (...) */ + if (ps.p_stack[ps.tos] == dohead) { + /* it is matched with do stmt */ + ps.ind_level = ps.i_l_follow = ps.il[ps.tos]; + ps.p_stack[++ps.tos] = whilestmt; + ps.il[ps.tos] = ps.ind_level = ps.i_l_follow; + } + else { /* it is a while loop */ + ps.p_stack[++ps.tos] = whilestmt; + ps.il[ps.tos] = ps.i_l_follow; + ++ps.i_l_follow; + ps.search_brace = btype_2; + } + + break; + + case elselit: /* scanned an else */ + + if (ps.p_stack[ps.tos] != ifhead) + diag(1, "Unmatched 'else'"); + else { + ps.ind_level = ps.il[ps.tos]; /* indentation for else should + * be same as for if */ + ps.i_l_follow = ps.ind_level + 1; /* everything following should + * be in 1 level */ + ps.p_stack[ps.tos] = elsehead; + /* remember if with else */ + ps.search_brace = btype_2 | ps.else_if; + } + break; + + case rbrace: /* scanned a } */ + /* stack should have <lbrace> <stmt> or <lbrace> <stmtl> */ + if (ps.p_stack[ps.tos - 1] == lbrace) { + ps.ind_level = ps.i_l_follow = ps.il[--ps.tos]; + ps.p_stack[ps.tos] = stmt; + } + else + diag(1, "Stmt nesting error."); + break; + + case swstmt: /* had switch (...) */ + ps.p_stack[++ps.tos] = swstmt; + ps.cstk[ps.tos] = case_ind; + /* save current case indent level */ + ps.il[ps.tos] = ps.i_l_follow; + case_ind = ps.i_l_follow + ps.case_indent; /* cases should be one + * level down from + * switch */ + ps.i_l_follow += ps.case_indent + 1; /* statements should be two + * levels in */ + ps.search_brace = btype_2; + break; + + case semicolon: /* this indicates a simple stmt */ + break_comma = false; /* turn off flag to break after commas in a + * declaration */ + ps.p_stack[++ps.tos] = stmt; + ps.il[ps.tos] = ps.ind_level; + break; + + default: /* this is an error */ + diag(1, "Unknown code to parser"); + return; + + + } /* end of switch */ + + reduce(); /* see if any reduction can be done */ + +#ifdef debug + for (i = 1; i <= ps.tos; ++i) + printf("(%d %d)", ps.p_stack[i], ps.il[i]); + printf("\n"); +#endif + + return; +} + +/* + * NAME: reduce + * + * FUNCTION: Implements the reduce part of the parsing algorithm + * + * ALGORITHM: The following reductions are done. Reductions are repeated + * until no more are possible. + * + * Old TOS New TOS + * <stmt> <stmt> <stmtl> + * <stmtl> <stmt> <stmtl> + * do <stmt> "dostmt" + * if <stmt> "ifstmt" + * switch <stmt> <stmt> + * decl <stmt> <stmt> + * "ifelse" <stmt> <stmt> + * for <stmt> <stmt> + * while <stmt> <stmt> + * "dostmt" while <stmt> + * + * On each reduction, ps.i_l_follow (the indentation for the following line) + * is set to the indentation level associated with the old TOS. + * + * PARAMETERS: None + * + * RETURNS: Nothing + * + * GLOBALS: ps.cstk ps.i_l_follow = ps.il ps.p_stack = ps.tos = + * + * CALLS: None + * + * CALLED BY: parse + * + * HISTORY: initial coding November 1976 D A Willcox of CAC + * + */ +/*----------------------------------------------*\ +| REDUCTION PHASE | +\*----------------------------------------------*/ +reduce() +{ + + register int i; + + for (;;) { /* keep looping until there is nothing left to + * reduce */ + + switch (ps.p_stack[ps.tos]) { + + case stmt: + switch (ps.p_stack[ps.tos - 1]) { + + case stmt: + case stmtl: + /* stmtl stmt or stmt stmt */ + ps.p_stack[--ps.tos] = stmtl; + break; + + case dolit: /* <do> <stmt> */ + ps.p_stack[--ps.tos] = dohead; + ps.i_l_follow = ps.il[ps.tos]; + break; + + case ifstmt: + /* <if> <stmt> */ + ps.p_stack[--ps.tos] = ifhead; + for (i = ps.tos - 1; + ( + ps.p_stack[i] != stmt + && + ps.p_stack[i] != stmtl + && + ps.p_stack[i] != lbrace + ); + --i); + ps.i_l_follow = ps.il[i]; + /* + * for the time being, we will assume that there is no else on + * this if, and set the indentation level accordingly. If an + * else is scanned, it will be fixed up later + */ + break; + + case swstmt: + /* <switch> <stmt> */ + case_ind = ps.cstk[ps.tos - 1]; + + case decl: /* finish of a declaration */ + case elsehead: + /* <<if> <stmt> else> <stmt> */ + case forstmt: + /* <for> <stmt> */ + case whilestmt: + /* <while> <stmt> */ + ps.p_stack[--ps.tos] = stmt; + ps.i_l_follow = ps.il[ps.tos]; + break; + + default: /* <anything else> <stmt> */ + return; + + } /* end of section for <stmt> on top of stack */ + break; + + case whilestmt: /* while (...) on top */ + if (ps.p_stack[ps.tos - 1] == dohead) { + /* it is termination of a do while */ + ps.p_stack[--ps.tos] = stmt; + break; + } + else + return; + + default: /* anything else on top */ + return; + + } + } +} diff --git a/usr.bin/indent/pr_comment.c b/usr.bin/indent/pr_comment.c new file mode 100644 index 0000000..2d4bc90 --- /dev/null +++ b/usr.bin/indent/pr_comment.c @@ -0,0 +1,418 @@ +/* + * Copyright (c) 1985 Sun Microsystems, Inc. + * Copyright (c) 1980, 1993 + * The Regents of the University of California. All rights reserved. + * All rights reserved. + * + * Redistribution and use in source and binary forms, with or without + * modification, are permitted provided that the following conditions + * are met: + * 1. Redistributions of source code must retain the above copyright + * notice, this list of conditions and the following disclaimer. + * 2. Redistributions in binary form must reproduce the above copyright + * notice, this list of conditions and the following disclaimer in the + * documentation and/or other materials provided with the distribution. + * 3. All advertising materials mentioning features or use of this software + * must display the following acknowledgement: + * This product includes software developed by the University of + * California, Berkeley and its contributors. + * 4. Neither the name of the University nor the names of its contributors + * may be used to endorse or promote products derived from this software + * without specific prior written permission. + * + * THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND + * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE + * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE + * ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE + * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL + * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS + * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) + * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT + * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY + * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF + * SUCH DAMAGE. + */ + +#ifndef lint +static char sccsid[] = "@(#)pr_comment.c 8.1 (Berkeley) 6/6/93"; +#endif /* not lint */ + +#include <stdio.h> +#include <stdlib.h> +#include "indent_globs.h" + +/* + * NAME: + * pr_comment + * + * FUNCTION: + * This routine takes care of scanning and printing comments. + * + * ALGORITHM: + * 1) Decide where the comment should be aligned, and if lines should + * be broken. + * 2) If lines should not be broken and filled, just copy up to end of + * comment. + * 3) If lines should be filled, then scan thru input_buffer copying + * characters to com_buf. Remember where the last blank, tab, or + * newline was. When line is filled, print up to last blank and + * continue copying. + * + * HISTORY: + * November 1976 D A Willcox of CAC Initial coding + * 12/6/76 D A Willcox of CAC Modification to handle + * UNIX-style comments + * + */ + +/* + * this routine processes comments. It makes an attempt to keep comments from + * going over the max line length. If a line is too long, it moves everything + * from the last blank to the next comment line. Blanks and tabs from the + * beginning of the input line are removed + */ + + +pr_comment() +{ + int now_col; /* column we are in now */ + int adj_max_col; /* Adjusted max_col for when we decide to + * spill comments over the right margin */ + char *last_bl; /* points to the last blank in the output + * buffer */ + char *t_ptr; /* used for moving string */ + int unix_comment; /* tri-state variable used to decide if it is + * a unix-style comment. 0 means only blanks + * since /*, 1 means regular style comment, 2 + * means unix style comment */ + int break_delim = comment_delimiter_on_blankline; + int l_just_saw_decl = ps.just_saw_decl; + /* + * int ps.last_nl = 0; /* true iff the last significant thing + * weve seen is a newline + */ + int one_liner = 1; /* true iff this comment is a one-liner */ + adj_max_col = max_col; + ps.just_saw_decl = 0; + last_bl = 0; /* no blanks found so far */ + ps.box_com = false; /* at first, assume that we are not in + * a boxed comment or some other + * comment that should not be touched */ + ++ps.out_coms; /* keep track of number of comments */ + unix_comment = 1; /* set flag to let us figure out if there is a + * unix-style comment ** DISABLED: use 0 to + * reenable this hack! */ + + /* Figure where to align and how to treat the comment */ + + if (ps.col_1 && !format_col1_comments) { /* if comment starts in column + * 1 it should not be touched */ + ps.box_com = true; + ps.com_col = 1; + } + else { + if (*buf_ptr == '-' || *buf_ptr == '*') { + ps.box_com = true; /* a comment with a '-' or '*' immediately + * after the /* is assumed to be a boxed + * comment */ + break_delim = 0; + } + if ( /* ps.bl_line && */ (s_lab == e_lab) && (s_code == e_code)) { + /* klg: check only if this line is blank */ + /* + * If this (*and previous lines are*) blank, dont put comment way + * out at left + */ + ps.com_col = (ps.ind_level - ps.unindent_displace) * ps.ind_size + 1; + adj_max_col = block_comment_max_col; + if (ps.com_col <= 1) + ps.com_col = 1 + !format_col1_comments; + } + else { + register target_col; + break_delim = 0; + if (s_code != e_code) + target_col = count_spaces(compute_code_target(), s_code); + else { + target_col = 1; + if (s_lab != e_lab) + target_col = count_spaces(compute_label_target(), s_lab); + } + ps.com_col = ps.decl_on_line || ps.ind_level == 0 ? ps.decl_com_ind : ps.com_ind; + if (ps.com_col < target_col) + ps.com_col = ((target_col + 7) & ~7) + 1; + if (ps.com_col + 24 > adj_max_col) + adj_max_col = ps.com_col + 24; + } + } + if (ps.box_com) { + buf_ptr[-2] = 0; + ps.n_comment_delta = 1 - count_spaces(1, in_buffer); + buf_ptr[-2] = '/'; + } + else { + ps.n_comment_delta = 0; + while (*buf_ptr == ' ' || *buf_ptr == '\t') + buf_ptr++; + } + ps.comment_delta = 0; + *e_com++ = '/'; /* put '/*' into buffer */ + *e_com++ = '*'; + if (*buf_ptr != ' ' && !ps.box_com) + *e_com++ = ' '; + + *e_com = '\0'; + if (troff) { + now_col = 1; + adj_max_col = 80; + } + else + now_col = count_spaces(ps.com_col, s_com); /* figure what column we + * would be in if we + * printed the comment + * now */ + + /* Start to copy the comment */ + + while (1) { /* this loop will go until the comment is + * copied */ + if (*buf_ptr > 040 && *buf_ptr != '*') + ps.last_nl = 0; + CHECK_SIZE_COM; + switch (*buf_ptr) { /* this checks for various spcl cases */ + case 014: /* check for a form feed */ + if (!ps.box_com) { /* in a text comment, break the line here */ + ps.use_ff = true; + /* fix so dump_line uses a form feed */ + dump_line(); + last_bl = 0; + *e_com++ = ' '; + *e_com++ = '*'; + *e_com++ = ' '; + while (*++buf_ptr == ' ' || *buf_ptr == '\t'); + } + else { + if (++buf_ptr >= buf_end) + fill_buffer(); + *e_com++ = 014; + } + break; + + case '\n': + if (had_eof) { /* check for unexpected eof */ + printf("Unterminated comment\n"); + *e_com = '\0'; + dump_line(); + return; + } + one_liner = 0; + if (ps.box_com || ps.last_nl) { /* if this is a boxed comment, + * we dont ignore the newline */ + if (s_com == e_com) { + *e_com++ = ' '; + *e_com++ = ' '; + } + *e_com = '\0'; + if (!ps.box_com && e_com - s_com > 3) { + if (break_delim == 1 && s_com[0] == '/' + && s_com[1] == '*' && s_com[2] == ' ') { + char *t = e_com; + break_delim = 2; + e_com = s_com + 2; + *e_com = 0; + if (blanklines_before_blockcomments) + prefix_blankline_requested = 1; + dump_line(); + e_com = t; + s_com[0] = s_com[1] = s_com[2] = ' '; + } + dump_line(); + CHECK_SIZE_COM; + *e_com++ = ' '; + *e_com++ = ' '; + } + dump_line(); + now_col = ps.com_col; + } + else { + ps.last_nl = 1; + if (unix_comment != 1) { /* we not are in unix_style + * comment */ + if (unix_comment == 0 && s_code == e_code) { + /* + * if it is a UNIX-style comment, ignore the + * requirement that previous line be blank for + * unindention + */ + ps.com_col = (ps.ind_level - ps.unindent_displace) * ps.ind_size + 1; + if (ps.com_col <= 1) + ps.com_col = 2; + } + unix_comment = 2; /* permanently remember that we are in + * this type of comment */ + dump_line(); + ++line_no; + now_col = ps.com_col; + *e_com++ = ' '; + /* + * fix so that the star at the start of the line will line + * up + */ + do /* flush leading white space */ + if (++buf_ptr >= buf_end) + fill_buffer(); + while (*buf_ptr == ' ' || *buf_ptr == '\t'); + break; + } + if (*(e_com - 1) == ' ' || *(e_com - 1) == '\t') + last_bl = e_com - 1; + /* + * if there was a space at the end of the last line, remember + * where it was + */ + else { /* otherwise, insert one */ + last_bl = e_com; + CHECK_SIZE_COM; + *e_com++ = ' '; + ++now_col; + } + } + ++line_no; /* keep track of input line number */ + if (!ps.box_com) { + int nstar = 1; + do { /* flush any blanks and/or tabs at start of + * next line */ + if (++buf_ptr >= buf_end) + fill_buffer(); + if (*buf_ptr == '*' && --nstar >= 0) { + if (++buf_ptr >= buf_end) + fill_buffer(); + if (*buf_ptr == '/') + goto end_of_comment; + } + } while (*buf_ptr == ' ' || *buf_ptr == '\t'); + } + else if (++buf_ptr >= buf_end) + fill_buffer(); + break; /* end of case for newline */ + + case '*': /* must check for possibility of being at end + * of comment */ + if (++buf_ptr >= buf_end) /* get to next char after * */ + fill_buffer(); + + if (unix_comment == 0) /* set flag to show we are not in + * unix-style comment */ + unix_comment = 1; + + if (*buf_ptr == '/') { /* it is the end!!! */ + end_of_comment: + if (++buf_ptr >= buf_end) + fill_buffer(); + + if (*(e_com - 1) != ' ' && !ps.box_com) { /* insure blank before + * end */ + *e_com++ = ' '; + ++now_col; + } + if (break_delim == 1 && !one_liner && s_com[0] == '/' + && s_com[1] == '*' && s_com[2] == ' ') { + char *t = e_com; + break_delim = 2; + e_com = s_com + 2; + *e_com = 0; + if (blanklines_before_blockcomments) + prefix_blankline_requested = 1; + dump_line(); + e_com = t; + s_com[0] = s_com[1] = s_com[2] = ' '; + } + if (break_delim == 2 && e_com > s_com + 3 + /* now_col > adj_max_col - 2 && !ps.box_com */ ) { + *e_com = '\0'; + dump_line(); + now_col = ps.com_col; + } + CHECK_SIZE_COM; + *e_com++ = '*'; + *e_com++ = '/'; + *e_com = '\0'; + ps.just_saw_decl = l_just_saw_decl; + return; + } + else { /* handle isolated '*' */ + *e_com++ = '*'; + ++now_col; + } + break; + default: /* we have a random char */ + if (unix_comment == 0 && *buf_ptr != ' ' && *buf_ptr != '\t') + unix_comment = 1; /* we are not in unix-style comment */ + + *e_com = *buf_ptr++; + if (buf_ptr >= buf_end) + fill_buffer(); + + if (*e_com == '\t') /* keep track of column */ + now_col = ((now_col - 1) & tabmask) + tabsize + 1; + else if (*e_com == '\b') /* this is a backspace */ + --now_col; + else + ++now_col; + + if (*e_com == ' ' || *e_com == '\t') + last_bl = e_com; + /* remember we saw a blank */ + + ++e_com; + if (now_col > adj_max_col && !ps.box_com && unix_comment == 1 && e_com[-1] > ' ') { + /* + * the comment is too long, it must be broken up + */ + if (break_delim == 1 && s_com[0] == '/' + && s_com[1] == '*' && s_com[2] == ' ') { + char *t = e_com; + break_delim = 2; + e_com = s_com + 2; + *e_com = 0; + if (blanklines_before_blockcomments) + prefix_blankline_requested = 1; + dump_line(); + e_com = t; + s_com[0] = s_com[1] = s_com[2] = ' '; + } + if (last_bl == 0) { /* we have seen no blanks */ + last_bl = e_com; /* fake it */ + *e_com++ = ' '; + } + *e_com = '\0'; /* print what we have */ + *last_bl = '\0'; + while (last_bl > s_com && last_bl[-1] < 040) + *--last_bl = 0; + e_com = last_bl; + dump_line(); + + *e_com++ = ' '; /* add blanks for continuation */ + *e_com++ = ' '; + *e_com++ = ' '; + + t_ptr = last_bl + 1; + last_bl = 0; + if (t_ptr >= e_com) { + while (*t_ptr == ' ' || *t_ptr == '\t') + t_ptr++; + while (*t_ptr != '\0') { /* move unprinted part of + * comment down in buffer */ + if (*t_ptr == ' ' || *t_ptr == '\t') + last_bl = e_com; + *e_com++ = *t_ptr++; + } + } + *e_com = '\0'; + now_col = count_spaces(ps.com_col, s_com); /* recompute current + * position */ + } + break; + } + } +} |
