Update GNU grep to version 2.0

1 files changed, 109 insertions, 199 deletions

diff --git a/gnu/usr.bin/grep/dfa.h b/gnu/usr.bin/grep/dfa.h
index 33e4bf1..32e05fc 100644
--- a/gnu/usr.bin/grep/dfa.h
+++ b/gnu/usr.bin/grep/dfa.h
@@ -16,210 +16,115 @@
    Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA.  */
 
 /* Written June, 1988 by Mike Haertel */
-
-#ifdef STDC_HEADERS
-
-#include <stddef.h>
-#include <stdlib.h>
-
-#else /* !STDC_HEADERS */
-
-#define const
-#include <sys/types.h>		/* For size_t.  */
-extern char *calloc(), *malloc(), *realloc();
-extern void free();
-
-#ifndef NULL
-#define NULL 0
-#endif
-
-#endif /* ! STDC_HEADERS */
-
-#include <ctype.h>
-#ifndef isascii
-#define ISALNUM(c) isalnum(c)
-#define ISALPHA(c) isalpha(c)
-#define ISUPPER(c) isupper(c)
-#define ISLOWER(c) islower(c)
-#else
-#define ISALNUM(c) (isascii(c) && isalnum(c))
-#define ISALPHA(c) (isascii(c) && isalpha(c))
-#define ISUPPER(c) (isascii(c) && isupper(c))
-#define ISLOWER(c) (isascii(c) && islower(c))
-#endif
-
-/* 1 means plain parentheses serve as grouping, and backslash
-     parentheses are needed for literal searching.
-   0 means backslash-parentheses are grouping, and plain parentheses
-     are for literal searching.  */
-#define RE_NO_BK_PARENS 1
-
-/* 1 means plain | serves as the "or"-operator, and \| is a literal.
-   0 means \| serves as the "or"-operator, and | is a literal.  */
-#define RE_NO_BK_VBAR 2
-
-/* 0 means plain + or ? serves as an operator, and \+, \? are literals.
-   1 means \+, \? are operators and plain +, ? are literals.  */
-#define RE_BK_PLUS_QM 4
-
-/* 1 means | binds tighter than ^ or $.
-   0 means the contrary.  */
-#define RE_TIGHT_VBAR 8
-
-/* 1 means treat \n as an _OR operator
-   0 means treat it as a normal character */
-#define RE_NEWLINE_OR 16
-
-/* 0 means that a special characters (such as *, ^, and $) always have
-     their special meaning regardless of the surrounding context.
-   1 means that special characters may act as normal characters in some
-     contexts.  Specifically, this applies to:
-	^ - only special at the beginning, or after ( or |
-	$ - only special at the end, or before ) or |
-	*, +, ? - only special when not after the beginning, (, or | */
-#define RE_CONTEXT_INDEP_OPS 32
-
-/* Now define combinations of bits for the standard possibilities.  */
-#define RE_SYNTAX_AWK (RE_NO_BK_PARENS | RE_NO_BK_VBAR | RE_CONTEXT_INDEP_OPS)
-#define RE_SYNTAX_EGREP (RE_SYNTAX_AWK | RE_NEWLINE_OR)
-#define RE_SYNTAX_GREP (RE_BK_PLUS_QM | RE_NEWLINE_OR)
-#define RE_SYNTAX_EMACS 0
+
+/* FIXME:
+   2.  We should not export so much of the DFA internals.
+   In addition to clobbering modularity, we eat up valuable
+   name space. */
 
 /* Number of bits in an unsigned char. */
 #define CHARBITS 8
 
 /* First integer value that is greater than any character code. */
-#define _NOTCHAR (1 << CHARBITS)
+#define NOTCHAR (1 << CHARBITS)
 
 /* INTBITS need not be exact, just a lower bound. */
 #define INTBITS (CHARBITS * sizeof (int))
 
 /* Number of ints required to hold a bit for every character. */
-#define _CHARSET_INTS ((_NOTCHAR + INTBITS - 1) / INTBITS)
+#define CHARCLASS_INTS ((NOTCHAR + INTBITS - 1) / INTBITS)
 
 /* Sets of unsigned characters are stored as bit vectors in arrays of ints. */
-typedef int _charset[_CHARSET_INTS];
+typedef int charclass[CHARCLASS_INTS];
 
 /* The regexp is parsed into an array of tokens in postfix form.  Some tokens
    are operators and others are terminal symbols.  Most (but not all) of these
    codes are returned by the lexical analyzer. */
-#if __STDC__
 
 typedef enum
 {
-  _END = -1,			/* _END is a terminal symbol that matches the
-				   end of input; any value of _END or less in
+  END = -1,			/* END is a terminal symbol that matches the
+				   end of input; any value of END or less in
 				   the parse tree is such a symbol.  Accepting
 				   states of the DFA are those that would have
-				   a transition on _END. */
+				   a transition on END. */
 
   /* Ordinary character values are terminal symbols that match themselves. */
 
-  _EMPTY = _NOTCHAR,		/* _EMPTY is a terminal symbol that matches
+  EMPTY = NOTCHAR,		/* EMPTY is a terminal symbol that matches
 				   the empty string. */
 
-  _BACKREF,			/* _BACKREF is generated by \<digit>; it
+  BACKREF,			/* BACKREF is generated by \<digit>; it
 				   it not completely handled.  If the scanner
 				   detects a transition on backref, it returns
 				   a kind of "semi-success" indicating that
 				   the match will have to be verified with
 				   a backtracking matcher. */
 
-  _BEGLINE,			/* _BEGLINE is a terminal symbol that matches
+  BEGLINE,			/* BEGLINE is a terminal symbol that matches
 				   the empty string if it is at the beginning
 				   of a line. */
 
-  _ALLBEGLINE,			/* _ALLBEGLINE is a terminal symbol that
-				   matches the empty string if it is at the
-				   beginning of a line; _ALLBEGLINE applies
-				   to the entire regexp and can only occur
-				   as the first token thereof.  _ALLBEGLINE
-				   never appears in the parse tree; a _BEGLINE
-				   is prepended with _CAT to the entire
-				   regexp instead. */
-
-  _ENDLINE,			/* _ENDLINE is a terminal symbol that matches
+  ENDLINE,			/* ENDLINE is a terminal symbol that matches
 				   the empty string if it is at the end of
 				   a line. */
 
-  _ALLENDLINE,			/* _ALLENDLINE is to _ENDLINE as _ALLBEGLINE
-				   is to _BEGLINE. */
-
-  _BEGWORD,			/* _BEGWORD is a terminal symbol that matches
+  BEGWORD,			/* BEGWORD is a terminal symbol that matches
 				   the empty string if it is at the beginning
 				   of a word. */
 
-  _ENDWORD,			/* _ENDWORD is a terminal symbol that matches
+  ENDWORD,			/* ENDWORD is a terminal symbol that matches
 				   the empty string if it is at the end of
 				   a word. */
 
-  _LIMWORD,			/* _LIMWORD is a terminal symbol that matches
+  LIMWORD,			/* LIMWORD is a terminal symbol that matches
 				   the empty string if it is at the beginning
 				   or the end of a word. */
 
-  _NOTLIMWORD,			/* _NOTLIMWORD is a terminal symbol that
+  NOTLIMWORD,			/* NOTLIMWORD is a terminal symbol that
 				   matches the empty string if it is not at
 				   the beginning or end of a word. */
 
-  _QMARK,			/* _QMARK is an operator of one argument that
+  QMARK,			/* QMARK is an operator of one argument that
 				   matches zero or one occurences of its
 				   argument. */
 
-  _STAR,			/* _STAR is an operator of one argument that
+  STAR,				/* STAR is an operator of one argument that
 				   matches the Kleene closure (zero or more
 				   occurrences) of its argument. */
 
-  _PLUS,			/* _PLUS is an operator of one argument that
+  PLUS,				/* PLUS is an operator of one argument that
 				   matches the positive closure (one or more
 				   occurrences) of its argument. */
 
-  _CAT,				/* _CAT is an operator of two arguments that
+  REPMN,			/* REPMN is a lexical token corresponding
+				   to the {m,n} construct.  REPMN never
+				   appears in the compiled token vector. */
+
+  CAT,				/* CAT is an operator of two arguments that
 				   matches the concatenation of its
-				   arguments.  _CAT is never returned by the
+				   arguments.  CAT is never returned by the
 				   lexical analyzer. */
 
-  _OR,				/* _OR is an operator of two arguments that
+  OR,				/* OR is an operator of two arguments that
 				   matches either of its arguments. */
 
-  _LPAREN,			/* _LPAREN never appears in the parse tree,
+  ORTOP,			/* OR at the toplevel in the parse tree.
+				   This is used for a boyer-moore heuristic. */
+
+  LPAREN,			/* LPAREN never appears in the parse tree,
 				   it is only a lexeme. */
 
-  _RPAREN,			/* _RPAREN never appears in the parse tree. */
+  RPAREN,			/* RPAREN never appears in the parse tree. */
 
-  _SET				/* _SET and (and any value greater) is a
+  CSET				/* CSET and (and any value greater) is a
 				   terminal symbol that matches any of a
 				   class of characters. */
-} _token;
-
-#else /* ! __STDC__ */
-
-typedef short _token;
-
-#define _END -1
-#define _EMPTY _NOTCHAR
-#define _BACKREF (_EMPTY + 1)
-#define _BEGLINE (_EMPTY + 2)
-#define _ALLBEGLINE (_EMPTY + 3)
-#define _ENDLINE (_EMPTY + 4)
-#define _ALLENDLINE (_EMPTY + 5)
-#define _BEGWORD (_EMPTY + 6)
-#define _ENDWORD (_EMPTY + 7)
-#define _LIMWORD (_EMPTY + 8)
-#define _NOTLIMWORD (_EMPTY + 9)
-#define _QMARK (_EMPTY + 10)
-#define _STAR (_EMPTY + 11)
-#define _PLUS (_EMPTY + 12)
-#define _CAT (_EMPTY + 13)
-#define _OR (_EMPTY + 14)
-#define _LPAREN (_EMPTY + 15)
-#define _RPAREN (_EMPTY + 16)
-#define _SET (_EMPTY + 17)
-
-#endif /* ! __STDC__ */
+} token;
 
-/* Sets are stored in an array in the compiled regexp; the index of the
-   array corresponding to a given set token is given by _SET_INDEX(t). */
-#define _SET_INDEX(t) ((t) - _SET)
+/* Sets are stored in an array in the compiled dfa; the index of the
+   array corresponding to a given set token is given by SET_INDEX(t). */
+#define SET_INDEX(t) ((t) - CSET)
 
 /* Sometimes characters can only be matched depending on the surrounding
    context.  Such context decisions depend on what the previous character
@@ -239,36 +144,36 @@ typedef short _token;
 
    Word-constituent characters are those that satisfy isalnum().
 
-   The macro _SUCCEEDS_IN_CONTEXT determines whether a a given constraint
+   The macro SUCCEEDS_IN_CONTEXT determines whether a a given constraint
    succeeds in a particular context.  Prevn is true if the previous character
    was a newline, currn is true if the lookahead character is a newline.
    Prevl and currl similarly depend upon whether the previous and current
    characters are word-constituent letters. */
-#define _MATCHES_NEWLINE_CONTEXT(constraint, prevn, currn) \
-  ((constraint) & 1 << ((prevn) ? 2 : 0) + ((currn) ? 1 : 0) + 4)
-#define _MATCHES_LETTER_CONTEXT(constraint, prevl, currl) \
-  ((constraint) & 1 << ((prevl) ? 2 : 0) + ((currl) ? 1 : 0))
-#define _SUCCEEDS_IN_CONTEXT(constraint, prevn, currn, prevl, currl) \
-  (_MATCHES_NEWLINE_CONTEXT(constraint, prevn, currn)		     \
-   && _MATCHES_LETTER_CONTEXT(constraint, prevl, currl))
+#define MATCHES_NEWLINE_CONTEXT(constraint, prevn, currn) \
+  ((constraint) & 1 << (((prevn) ? 2 : 0) + ((currn) ? 1 : 0) + 4))
+#define MATCHES_LETTER_CONTEXT(constraint, prevl, currl) \
+  ((constraint) & 1 << (((prevl) ? 2 : 0) + ((currl) ? 1 : 0)))
+#define SUCCEEDS_IN_CONTEXT(constraint, prevn, currn, prevl, currl) \
+  (MATCHES_NEWLINE_CONTEXT(constraint, prevn, currn)		     \
+   && MATCHES_LETTER_CONTEXT(constraint, prevl, currl))
 
 /* The following macros give information about what a constraint depends on. */
-#define _PREV_NEWLINE_DEPENDENT(constraint) \
+#define PREV_NEWLINE_DEPENDENT(constraint) \
   (((constraint) & 0xc0) >> 2 != ((constraint) & 0x30))
-#define _PREV_LETTER_DEPENDENT(constraint) \
+#define PREV_LETTER_DEPENDENT(constraint) \
   (((constraint) & 0x0c) >> 2 != ((constraint) & 0x03))
 
 /* Tokens that match the empty string subject to some constraint actually
    work by applying that constraint to determine what may follow them,
    taking into account what has gone before.  The following values are
    the constraints corresponding to the special tokens previously defined. */
-#define _NO_CONSTRAINT 0xff
-#define _BEGLINE_CONSTRAINT 0xcf
-#define _ENDLINE_CONSTRAINT 0xaf
-#define _BEGWORD_CONSTRAINT 0xf2
-#define _ENDWORD_CONSTRAINT 0xf4
-#define _LIMWORD_CONSTRAINT 0xf6
-#define _NOTLIMWORD_CONSTRAINT 0xf9
+#define NO_CONSTRAINT 0xff
+#define BEGLINE_CONSTRAINT 0xcf
+#define ENDLINE_CONSTRAINT 0xaf
+#define BEGWORD_CONSTRAINT 0xf2
+#define ENDWORD_CONSTRAINT 0xf4
+#define LIMWORD_CONSTRAINT 0xf6
+#define NOTLIMWORD_CONSTRAINT 0xf9
 
 /* States of the recognizer correspond to sets of positions in the parse
    tree, together with the constraints under which they may be matched.
@@ -278,44 +183,48 @@ typedef struct
 {
   unsigned index;		/* Index into the parse array. */
   unsigned constraint;		/* Constraint for matching this position. */
-} _position;
+} position;
 
 /* Sets of positions are stored as arrays. */
 typedef struct
 {
-  _position *elems;		/* Elements of this position set. */
+  position *elems;		/* Elements of this position set. */
   int nelem;			/* Number of elements in this set. */
-} _position_set;
+} position_set;
 
-/* A state of the regexp consists of a set of positions, some flags,
+/* A state of the dfa consists of a set of positions, some flags,
    and the token value of the lowest-numbered position of the state that
-   contains an _END token. */
+   contains an END token. */
 typedef struct
 {
   int hash;			/* Hash of the positions of this state. */
-  _position_set elems;		/* Positions this state could match. */
+  position_set elems;		/* Positions this state could match. */
   char newline;			/* True if previous state matched newline. */
   char letter;			/* True if previous state matched a letter. */
   char backref;			/* True if this state matches a \<digit>. */
   unsigned char constraint;	/* Constraint for this state to accept. */
-  int first_end;		/* Token value of the first _END in elems. */
-} _dfa_state;
+  int first_end;		/* Token value of the first END in elems. */
+} dfa_state;
 
-/* If an r.e. is at most MUST_MAX characters long, we look for a string which
-   must appear in it; whatever's found is dropped into the struct reg. */
-
-#define MUST_MAX	50
+/* Element of a list of strings, at least one of which is known to
+   appear in any R.E. matching the DFA. */
+struct dfamust
+{
+  int exact;
+  char *must;
+  struct dfamust *next;
+};
 
 /* A compiled regular expression. */
-struct regexp
+struct dfa
 {
   /* Stuff built by the scanner. */
-  _charset *charsets;		/* Array of character sets for _SET tokens. */
-  int cindex;			/* Index for adding new charsets. */
-  int calloc;			/* Number of charsets currently allocated. */
+  charclass *charclasses;	/* Array of character sets for CSET tokens. */
+  int cindex;			/* Index for adding new charclasses. */
+  int calloc;			/* Number of charclasses currently allocated. */
 
   /* Stuff built by the parser. */
-  _token *tokens;		/* Postfix parse array. */
+  token *tokens;		/* Postfix parse array. */
   int tindex;			/* Index for adding new tokens. */
   int talloc;			/* Number of tokens currently allocated. */
   int depth;			/* Depth required of an evaluation stack
@@ -323,15 +232,15 @@ struct regexp
 				   parse tree. */
   int nleaves;			/* Number of leaves on the parse tree. */
   int nregexps;			/* Count of parallel regexps being built
-				   with regparse(). */
+				   with dfaparse(). */
 
   /* Stuff owned by the state builder. */
-  _dfa_state *states;		/* States of the regexp. */
+  dfa_state *states;		/* States of the dfa. */
   int sindex;			/* Index for adding new states. */
   int salloc;			/* Number of states currently allocated. */
 
   /* Stuff built by the structure analyzer. */
-  _position_set *follows;	/* Array of follow sets, indexed by position
+  position_set *follows;	/* Array of follow sets, indexed by position
 				   index.  The follow of a position is the set
 				   of positions containing characters that
 				   could conceivably follow a character
@@ -361,7 +270,7 @@ struct regexp
   int **fails;			/* Transition tables after failing to accept
 				   on a state that potentially could do so. */
   int *success;			/* Table of acceptance conditions used in
-				   regexecute and computed in build_state. */
+				   dfaexec and computed in build_state. */
   int *newlines;		/* Transitions on newlines.  The entry for a
 				   newline in any transition table is always
 				   -1 so we can count lines without wasting
@@ -369,40 +278,41 @@ struct regexp
 				   newline is stored separately and handled
 				   as a special case.  Newline is also used
 				   as a sentinel at the end of the buffer. */
-  char must[MUST_MAX];
-  int mustn;
+  struct dfamust *musts;	/* List of strings, at least one of which
+				   is known to appear in any r.e. matching
+				   the dfa. */
 };
 
-/* Some macros for user access to regexp internals. */
+/* Some macros for user access to dfa internals. */
 
 /* ACCEPTING returns true if s could possibly be an accepting state of r. */
 #define ACCEPTING(s, r) ((r).states[s].constraint)
 
 /* ACCEPTS_IN_CONTEXT returns true if the given state accepts in the
    specified context. */
-#define ACCEPTS_IN_CONTEXT(prevn, currn, prevl, currl, state, reg) \
-  _SUCCEEDS_IN_CONTEXT((reg).states[state].constraint,		   \
+#define ACCEPTS_IN_CONTEXT(prevn, currn, prevl, currl, state, dfa) \
+  SUCCEEDS_IN_CONTEXT((dfa).states[state].constraint,		   \
 		       prevn, currn, prevl, currl)
 
 /* FIRST_MATCHING_REGEXP returns the index number of the first of parallel
    regexps that a given state could accept.  Parallel regexps are numbered
    starting at 1. */
-#define FIRST_MATCHING_REGEXP(state, reg) (-(reg).states[state].first_end)
+#define FIRST_MATCHING_REGEXP(state, dfa) (-(dfa).states[state].first_end)
 
 /* Entry points. */
 
 #if __STDC__
 
-/* Regsyntax() takes two arguments; the first sets the syntax bits described
+/* dfasyntax() takes two arguments; the first sets the syntax bits described
    earlier in this file, and the second sets the case-folding flag. */
-extern void regsyntax(int, int);
+extern void dfasyntax(int, int);
 
-/* Compile the given string of the given length into the given struct regexp.
+/* Compile the given string of the given length into the given struct dfa.
    Final argument is a flag specifying whether to build a searching or an
    exact matcher. */
-extern void regcompile(const char *, size_t, struct regexp *, int);
+extern void dfacomp(char *, size_t, struct dfa *, int);
 
-/* Execute the given struct regexp on the buffer of characters.  The
+/* Execute the given struct dfa on the buffer of characters.  The
    first char * points to the beginning, and the second points to the
    first character after the end of the buffer, which must be a writable
    place so a sentinel end-of-buffer marker can be stored there.  The
@@ -414,37 +324,37 @@ extern void regcompile(const char *, size_t, struct regexp *, int);
    order to verify backreferencing; otherwise the flag will be cleared.
    Returns NULL if no match is found, or a pointer to the first
    character after the first & shortest matching string in the buffer. */
-extern char *regexecute(struct regexp *, char *, char *, int, int *, int *);
+extern char *dfaexec(struct dfa *, char *, char *, int, int *, int *);
 
-/* Free the storage held by the components of a struct regexp. */
-extern void regfree(struct regexp *);
+/* Free the storage held by the components of a struct dfa. */
+extern void dfafree(struct dfa *);
 
 /* Entry points for people who know what they're doing. */
 
-/* Initialize the components of a struct regexp. */
-extern void reginit(struct regexp *);
+/* Initialize the components of a struct dfa. */
+extern void dfainit(struct dfa *);
 
-/* Incrementally parse a string of given length into a struct regexp. */
-extern void regparse(const char *, size_t, struct regexp *);
+/* Incrementally parse a string of given length into a struct dfa. */
+extern void dfaparse(char *, size_t, struct dfa *);
 
 /* Analyze a parsed regexp; second argument tells whether to build a searching
    or an exact matcher. */
-extern void reganalyze(struct regexp *, int);
+extern void dfaanalyze(struct dfa *, int);
 
 /* Compute, for each possible character, the transitions out of a given
    state, storing them in an array of integers. */
-extern void regstate(int, struct regexp *, int []);
+extern void dfastate(int, struct dfa *, int []);
 
 /* Error handling. */
 
-/* Regerror() is called by the regexp routines whenever an error occurs.  It
+/* dfaerror() is called by the regexp routines whenever an error occurs.  It
    takes a single argument, a NUL-terminated string describing the error.
-   The default regerror() prints the error message to stderr and exits.
-   The user can provide a different regfree() if so desired. */
-extern void regerror(const char *);
+   The default dfaerror() prints the error message to stderr and exits.
+   The user can provide a different dfafree() if so desired. */
+extern void dfaerror(char *);
 
 #else /* ! __STDC__ */
-extern void regsyntax(), regcompile(), regfree(), reginit(), regparse();
-extern void reganalyze(), regstate(), regerror();
-extern char *regexecute();
+extern void dfasyntax(), dfacomp(), dfafree(), dfainit(), dfaparse();
+extern void dfaanalyze(), dfastate(), dfaerror();
+extern char *dfaexec();
 #endif /* ! __STDC__ */