Blame gl/regex.h

Packit a4aae4
/* Definitions for data structures and routines for the regular
Packit a4aae4
   expression library.
Packit a4aae4
   Copyright (C) 1985, 1989-1993, 1995-1998, 2000-2003, 2005-2017 Free Software
Packit a4aae4
   Foundation, Inc.
Packit a4aae4
   This file is part of the GNU C Library.
Packit a4aae4
Packit a4aae4
   The GNU C Library is free software; you can redistribute it and/or
Packit a4aae4
   modify it under the terms of the GNU Lesser General Public
Packit a4aae4
   License as published by the Free Software Foundation; either
Packit a4aae4
   version 2.1 of the License, or (at your option) any later version.
Packit a4aae4
Packit a4aae4
   The GNU C Library is distributed in the hope that it will be useful,
Packit a4aae4
   but WITHOUT ANY WARRANTY; without even the implied warranty of
Packit a4aae4
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
Packit a4aae4
   Lesser General Public License for more details.
Packit a4aae4
Packit a4aae4
   You should have received a copy of the GNU Lesser General Public
Packit a4aae4
   License along with the GNU C Library; if not, see
Packit a4aae4
   <http://www.gnu.org/licenses/>.  */
Packit a4aae4
Packit a4aae4
#ifndef _REGEX_H
Packit a4aae4
#define _REGEX_H 1
Packit a4aae4
Packit a4aae4
#include <sys/types.h>
Packit a4aae4
Packit a4aae4
/* Allow the use in C++ code.  */
Packit a4aae4
#ifdef __cplusplus
Packit a4aae4
extern "C" {
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
/* Define __USE_GNU to declare GNU extensions that violate the
Packit a4aae4
   POSIX name space rules.  */
Packit a4aae4
#ifdef _GNU_SOURCE
Packit a4aae4
# define __USE_GNU 1
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
#ifdef _REGEX_LARGE_OFFSETS
Packit a4aae4
Packit a4aae4
/* Use types and values that are wide enough to represent signed and
Packit a4aae4
   unsigned byte offsets in memory.  This currently works only when
Packit a4aae4
   the regex code is used outside of the GNU C library; it is not yet
Packit a4aae4
   supported within glibc itself, and glibc users should not define
Packit a4aae4
   _REGEX_LARGE_OFFSETS.  */
Packit a4aae4
Packit a4aae4
/* The type of object sizes.  */
Packit a4aae4
typedef size_t __re_size_t;
Packit a4aae4
Packit a4aae4
/* The type of object sizes, in places where the traditional code
Packit a4aae4
   uses unsigned long int.  */
Packit a4aae4
typedef size_t __re_long_size_t;
Packit a4aae4
Packit a4aae4
#else
Packit a4aae4
Packit a4aae4
/* The traditional GNU regex implementation mishandles strings longer
Packit a4aae4
   than INT_MAX.  */
Packit a4aae4
typedef unsigned int __re_size_t;
Packit a4aae4
typedef unsigned long int __re_long_size_t;
Packit a4aae4
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
/* The following two types have to be signed and unsigned integer type
Packit a4aae4
   wide enough to hold a value of a pointer.  For most ANSI compilers
Packit a4aae4
   ptrdiff_t and size_t should be likely OK.  Still size of these two
Packit a4aae4
   types is 2 for Microsoft C.  Ugh... */
Packit a4aae4
typedef long int s_reg_t;
Packit a4aae4
typedef unsigned long int active_reg_t;
Packit a4aae4
Packit a4aae4
/* The following bits are used to determine the regexp syntax we
Packit a4aae4
   recognize.  The set/not-set meanings are chosen so that Emacs syntax
Packit a4aae4
   remains the value 0.  The bits are given in alphabetical order, and
Packit a4aae4
   the definitions shifted by one from the previous bit; thus, when we
Packit a4aae4
   add or remove a bit, only one other definition need change.  */
Packit a4aae4
typedef unsigned long int reg_syntax_t;
Packit a4aae4
Packit a4aae4
#ifdef __USE_GNU
Packit a4aae4
/* If this bit is not set, then \ inside a bracket expression is literal.
Packit a4aae4
   If set, then such a \ quotes the following character.  */
Packit a4aae4
# define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1)
Packit a4aae4
Packit a4aae4
/* If this bit is not set, then + and ? are operators, and \+ and \? are
Packit a4aae4
     literals.
Packit a4aae4
   If set, then \+ and \? are operators and + and ? are literals.  */
Packit a4aae4
# define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then character classes are supported.  They are:
Packit a4aae4
     [:alpha:], [:upper:], [:lower:],  [:digit:], [:alnum:], [:xdigit:],
Packit a4aae4
     [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
Packit a4aae4
   If not set, then character classes are not supported.  */
Packit a4aae4
# define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then ^ and $ are always anchors (outside bracket
Packit a4aae4
     expressions, of course).
Packit a4aae4
   If this bit is not set, then it depends:
Packit a4aae4
	^  is an anchor if it is at the beginning of a regular
Packit a4aae4
	   expression or after an open-group or an alternation operator;
Packit a4aae4
	$  is an anchor if it is at the end of a regular expression, or
Packit a4aae4
	   before a close-group or an alternation operator.
Packit a4aae4
Packit a4aae4
   This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
Packit a4aae4
   POSIX draft 11.2 says that * etc. in leading positions is undefined.
Packit a4aae4
   We already implemented a previous draft which made those constructs
Packit a4aae4
   invalid, though, so we haven't changed the code back.  */
Packit a4aae4
# define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then special characters are always special
Packit a4aae4
     regardless of where they are in the pattern.
Packit a4aae4
   If this bit is not set, then special characters are special only in
Packit a4aae4
     some contexts; otherwise they are ordinary.  Specifically,
Packit a4aae4
     * + ? and intervals are only special when not after the beginning,
Packit a4aae4
     open-group, or alternation operator.  */
Packit a4aae4
# define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then *, +, ?, and { cannot be first in an re or
Packit a4aae4
     immediately after an alternation or begin-group operator.  */
Packit a4aae4
# define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then . matches newline.
Packit a4aae4
   If not set, then it doesn't.  */
Packit a4aae4
# define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then . doesn't match NUL.
Packit a4aae4
   If not set, then it does.  */
Packit a4aae4
# define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, nonmatching lists [^...] do not match newline.
Packit a4aae4
   If not set, they do.  */
Packit a4aae4
# define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, either \{...\} or {...} defines an
Packit a4aae4
     interval, depending on RE_NO_BK_BRACES.
Packit a4aae4
   If not set, \{, \}, {, and } are literals.  */
Packit a4aae4
# define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, +, ? and | aren't recognized as operators.
Packit a4aae4
   If not set, they are.  */
Packit a4aae4
# define RE_LIMITED_OPS (RE_INTERVALS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, newline is an alternation operator.
Packit a4aae4
   If not set, newline is literal.  */
Packit a4aae4
# define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then '{...}' defines an interval, and \{ and \}
Packit a4aae4
     are literals.
Packit a4aae4
  If not set, then '\{...\}' defines an interval.  */
Packit a4aae4
# define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, (...) defines a group, and \( and \) are literals.
Packit a4aae4
   If not set, \(...\) defines a group, and ( and ) are literals.  */
Packit a4aae4
# define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then \<digit> matches <digit>.
Packit a4aae4
   If not set, then \<digit> is a back-reference.  */
Packit a4aae4
# define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then | is an alternation operator, and \| is literal.
Packit a4aae4
   If not set, then \| is an alternation operator, and | is literal.  */
Packit a4aae4
# define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then an ending range point collating higher
Packit a4aae4
     than the starting range point, as in [z-a], is invalid.
Packit a4aae4
   If not set, then when ending range point collates higher than the
Packit a4aae4
     starting range point, the range is ignored.  */
Packit a4aae4
# define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then an unmatched ) is ordinary.
Packit a4aae4
   If not set, then an unmatched ) is invalid.  */
Packit a4aae4
# define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, succeed as soon as we match the whole pattern,
Packit a4aae4
   without further backtracking.  */
Packit a4aae4
# define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, do not process the GNU regex operators.
Packit a4aae4
   If not set, then the GNU regex operators are recognized. */
Packit a4aae4
# define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, turn on internal regex debugging.
Packit a4aae4
   If not set, and debugging was on, turn it off.
Packit a4aae4
   This only works if regex.c is compiled -DDEBUG.
Packit a4aae4
   We define this bit always, so that all that's needed to turn on
Packit a4aae4
   debugging is to recompile regex.c; the calling code can always have
Packit a4aae4
   this bit set, and it won't affect anything in the normal case. */
Packit a4aae4
# define RE_DEBUG (RE_NO_GNU_OPS << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, a syntactically invalid interval is treated as
Packit a4aae4
   a string of ordinary characters.  For example, the ERE 'a{1' is
Packit a4aae4
   treated as 'a\{1'.  */
Packit a4aae4
# define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then ignore case when matching.
Packit a4aae4
   If not set, then case is significant.  */
Packit a4aae4
# define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1)
Packit a4aae4
Packit a4aae4
/* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only
Packit a4aae4
   for ^, because it is difficult to scan the regex backwards to find
Packit a4aae4
   whether ^ should be special.  */
Packit a4aae4
# define RE_CARET_ANCHORS_HERE (RE_ICASE << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then \{ cannot be first in a regex or
Packit a4aae4
   immediately after an alternation, open-group or \} operator.  */
Packit a4aae4
# define RE_CONTEXT_INVALID_DUP (RE_CARET_ANCHORS_HERE << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then no_sub will be set to 1 during
Packit a4aae4
   re_compile_pattern.  */
Packit a4aae4
# define RE_NO_SUB (RE_CONTEXT_INVALID_DUP << 1)
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
/* This global variable defines the particular regexp syntax to use (for
Packit a4aae4
   some interfaces).  When a regexp is compiled, the syntax used is
Packit a4aae4
   stored in the pattern buffer, so changing this does not affect
Packit a4aae4
   already-compiled regexps.  */
Packit a4aae4
extern reg_syntax_t re_syntax_options;
Packit a4aae4

Packit a4aae4
#ifdef __USE_GNU
Packit a4aae4
/* Define combinations of the above bits for the standard possibilities.
Packit a4aae4
   (The [[[ comments delimit what gets put into the Texinfo file, so
Packit a4aae4
   don't delete them!)  */
Packit a4aae4
/* [[[begin syntaxes]]] */
Packit a4aae4
# define RE_SYNTAX_EMACS 0
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_AWK							\
Packit a4aae4
  (RE_BACKSLASH_ESCAPE_IN_LISTS   | RE_DOT_NOT_NULL			\
Packit a4aae4
   | RE_NO_BK_PARENS              | RE_NO_BK_REFS			\
Packit a4aae4
   | RE_NO_BK_VBAR                | RE_NO_EMPTY_RANGES			\
Packit a4aae4
   | RE_DOT_NEWLINE		  | RE_CONTEXT_INDEP_ANCHORS		\
Packit a4aae4
   | RE_CHAR_CLASSES							\
Packit a4aae4
   | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS)
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_GNU_AWK						\
Packit a4aae4
  ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS		\
Packit a4aae4
    | RE_INVALID_INTERVAL_ORD)						\
Packit a4aae4
   & ~(RE_DOT_NOT_NULL | RE_CONTEXT_INDEP_OPS				\
Packit a4aae4
      | RE_CONTEXT_INVALID_OPS ))
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_POSIX_AWK						\
Packit a4aae4
  (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS		\
Packit a4aae4
   | RE_INTERVALS	    | RE_NO_GNU_OPS				\
Packit a4aae4
   | RE_INVALID_INTERVAL_ORD)
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_GREP							\
Packit a4aae4
  ((RE_SYNTAX_POSIX_BASIC | RE_NEWLINE_ALT)				\
Packit a4aae4
   & ~(RE_CONTEXT_INVALID_DUP | RE_DOT_NOT_NULL))
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_EGREP						\
Packit a4aae4
  ((RE_SYNTAX_POSIX_EXTENDED | RE_INVALID_INTERVAL_ORD | RE_NEWLINE_ALT) \
Packit a4aae4
   & ~(RE_CONTEXT_INVALID_OPS | RE_DOT_NOT_NULL))
Packit a4aae4
Packit a4aae4
/* POSIX grep -E behavior is no longer incompatible with GNU.  */
Packit a4aae4
# define RE_SYNTAX_POSIX_EGREP						\
Packit a4aae4
  RE_SYNTAX_EGREP
Packit a4aae4
Packit a4aae4
/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff.  */
Packit a4aae4
# define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC
Packit a4aae4
Packit a4aae4
/* Syntax bits common to both basic and extended POSIX regex syntax.  */
Packit a4aae4
# define _RE_SYNTAX_POSIX_COMMON					\
Packit a4aae4
  (RE_CHAR_CLASSES | RE_DOT_NEWLINE      | RE_DOT_NOT_NULL		\
Packit a4aae4
   | RE_INTERVALS  | RE_NO_EMPTY_RANGES)
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_POSIX_BASIC						\
Packit a4aae4
  (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM | RE_CONTEXT_INVALID_DUP)
Packit a4aae4
Packit a4aae4
/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
Packit a4aae4
   RE_LIMITED_OPS, i.e., \? \+ \| are not recognized.  Actually, this
Packit a4aae4
   isn't minimal, since other operators, such as \`, aren't disabled.  */
Packit a4aae4
# define RE_SYNTAX_POSIX_MINIMAL_BASIC					\
Packit a4aae4
  (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
Packit a4aae4
Packit a4aae4
# define RE_SYNTAX_POSIX_EXTENDED					\
Packit a4aae4
  (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS			\
Packit a4aae4
   | RE_CONTEXT_INDEP_OPS   | RE_NO_BK_BRACES				\
Packit a4aae4
   | RE_NO_BK_PARENS        | RE_NO_BK_VBAR				\
Packit a4aae4
   | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD)
Packit a4aae4
Packit a4aae4
/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is
Packit a4aae4
   removed and RE_NO_BK_REFS is added.  */
Packit a4aae4
# define RE_SYNTAX_POSIX_MINIMAL_EXTENDED				\
Packit a4aae4
  (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS			\
Packit a4aae4
   | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES				\
Packit a4aae4
   | RE_NO_BK_PARENS        | RE_NO_BK_REFS				\
Packit a4aae4
   | RE_NO_BK_VBAR	    | RE_UNMATCHED_RIGHT_PAREN_ORD)
Packit a4aae4
/* [[[end syntaxes]]] */
Packit a4aae4
Packit a4aae4
/* Maximum number of duplicates an interval can allow.  POSIX-conforming
Packit a4aae4
   systems might define this in <limits.h>, but we want our
Packit a4aae4
   value, so remove any previous define.  */
Packit a4aae4
# ifdef _REGEX_INCLUDE_LIMITS_H
Packit a4aae4
#  include <limits.h>
Packit a4aae4
# endif
Packit a4aae4
# ifdef RE_DUP_MAX
Packit a4aae4
#  undef RE_DUP_MAX
Packit a4aae4
# endif
Packit a4aae4
Packit a4aae4
/* RE_DUP_MAX is 2**15 - 1 because an earlier implementation stored
Packit a4aae4
   the counter as a 2-byte signed integer.  This is no longer true, so
Packit a4aae4
   RE_DUP_MAX could be increased to (INT_MAX / 10 - 1), or to
Packit a4aae4
   ((SIZE_MAX - 9) / 10) if _REGEX_LARGE_OFFSETS is defined.
Packit a4aae4
   However, there would be a huge performance problem if someone
Packit a4aae4
   actually used a pattern like a\{214748363\}, so RE_DUP_MAX retains
Packit a4aae4
   its historical value.  */
Packit a4aae4
# define RE_DUP_MAX (0x7fff)
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* POSIX 'cflags' bits (i.e., information for 'regcomp').  */
Packit a4aae4
Packit a4aae4
/* If this bit is set, then use extended regular expression syntax.
Packit a4aae4
   If not set, then use basic regular expression syntax.  */
Packit a4aae4
#define REG_EXTENDED 1
Packit a4aae4
Packit a4aae4
/* If this bit is set, then ignore case when matching.
Packit a4aae4
   If not set, then case is significant.  */
Packit a4aae4
#define REG_ICASE (1 << 1)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then anchors do not match at newline
Packit a4aae4
     characters in the string.
Packit a4aae4
   If not set, then anchors do match at newlines.  */
Packit a4aae4
#define REG_NEWLINE (1 << 2)
Packit a4aae4
Packit a4aae4
/* If this bit is set, then report only success or fail in regexec.
Packit a4aae4
   If not set, then returns differ between not matching and errors.  */
Packit a4aae4
#define REG_NOSUB (1 << 3)
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* POSIX 'eflags' bits (i.e., information for regexec).  */
Packit a4aae4
Packit a4aae4
/* If this bit is set, then the beginning-of-line operator doesn't match
Packit a4aae4
     the beginning of the string (presumably because it's not the
Packit a4aae4
     beginning of a line).
Packit a4aae4
   If not set, then the beginning-of-line operator does match the
Packit a4aae4
     beginning of the string.  */
Packit a4aae4
#define REG_NOTBOL 1
Packit a4aae4
Packit a4aae4
/* Like REG_NOTBOL, except for the end-of-line.  */
Packit a4aae4
#define REG_NOTEOL (1 << 1)
Packit a4aae4
Packit a4aae4
/* Use PMATCH[0] to delimit the start and end of the search in the
Packit a4aae4
   buffer.  */
Packit a4aae4
#define REG_STARTEND (1 << 2)
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* If any error codes are removed, changed, or added, update the
Packit a4aae4
   '__re_error_msgid' table in regcomp.c.  */
Packit a4aae4
Packit a4aae4
typedef enum
Packit a4aae4
{
Packit a4aae4
  _REG_ENOSYS = -1,	/* This will never happen for this implementation.  */
Packit a4aae4
  _REG_NOERROR = 0,	/* Success.  */
Packit a4aae4
  _REG_NOMATCH,		/* Didn't find a match (for regexec).  */
Packit a4aae4
Packit a4aae4
  /* POSIX regcomp return error codes.  (In the order listed in the
Packit a4aae4
     standard.)  */
Packit a4aae4
  _REG_BADPAT,		/* Invalid pattern.  */
Packit a4aae4
  _REG_ECOLLATE,	/* Invalid collating element.  */
Packit a4aae4
  _REG_ECTYPE,		/* Invalid character class name.  */
Packit a4aae4
  _REG_EESCAPE,		/* Trailing backslash.  */
Packit a4aae4
  _REG_ESUBREG,		/* Invalid back reference.  */
Packit a4aae4
  _REG_EBRACK,		/* Unmatched left bracket.  */
Packit a4aae4
  _REG_EPAREN,		/* Parenthesis imbalance.  */
Packit a4aae4
  _REG_EBRACE,		/* Unmatched \{.  */
Packit a4aae4
  _REG_BADBR,		/* Invalid contents of \{\}.  */
Packit a4aae4
  _REG_ERANGE,		/* Invalid range end.  */
Packit a4aae4
  _REG_ESPACE,		/* Ran out of memory.  */
Packit a4aae4
  _REG_BADRPT,		/* No preceding re for repetition op.  */
Packit a4aae4
Packit a4aae4
  /* Error codes we've added.  */
Packit a4aae4
  _REG_EEND,		/* Premature end.  */
Packit a4aae4
  _REG_ESIZE,		/* Too large (e.g., repeat count too large).  */
Packit a4aae4
  _REG_ERPAREN		/* Unmatched ) or \); not returned from regcomp.  */
Packit a4aae4
} reg_errcode_t;
Packit a4aae4
Packit a4aae4
#if defined _XOPEN_SOURCE || defined __USE_XOPEN2K
Packit a4aae4
# define REG_ENOSYS	_REG_ENOSYS
Packit a4aae4
#endif
Packit a4aae4
#define REG_NOERROR	_REG_NOERROR
Packit a4aae4
#define REG_NOMATCH	_REG_NOMATCH
Packit a4aae4
#define REG_BADPAT	_REG_BADPAT
Packit a4aae4
#define REG_ECOLLATE	_REG_ECOLLATE
Packit a4aae4
#define REG_ECTYPE	_REG_ECTYPE
Packit a4aae4
#define REG_EESCAPE	_REG_EESCAPE
Packit a4aae4
#define REG_ESUBREG	_REG_ESUBREG
Packit a4aae4
#define REG_EBRACK	_REG_EBRACK
Packit a4aae4
#define REG_EPAREN	_REG_EPAREN
Packit a4aae4
#define REG_EBRACE	_REG_EBRACE
Packit a4aae4
#define REG_BADBR	_REG_BADBR
Packit a4aae4
#define REG_ERANGE	_REG_ERANGE
Packit a4aae4
#define REG_ESPACE	_REG_ESPACE
Packit a4aae4
#define REG_BADRPT	_REG_BADRPT
Packit a4aae4
#define REG_EEND	_REG_EEND
Packit a4aae4
#define REG_ESIZE	_REG_ESIZE
Packit a4aae4
#define REG_ERPAREN	_REG_ERPAREN
Packit a4aae4

Packit a4aae4
/* This data structure represents a compiled pattern.  Before calling
Packit a4aae4
   the pattern compiler, the fields 'buffer', 'allocated', 'fastmap',
Packit a4aae4
   and 'translate' can be set.  After the pattern has been compiled,
Packit a4aae4
   the fields 're_nsub', 'not_bol' and 'not_eol' are available.  All
Packit a4aae4
   other fields are private to the regex routines.  */
Packit a4aae4
Packit a4aae4
#ifndef RE_TRANSLATE_TYPE
Packit a4aae4
# define __RE_TRANSLATE_TYPE unsigned char *
Packit a4aae4
# ifdef __USE_GNU
Packit a4aae4
#  define RE_TRANSLATE_TYPE __RE_TRANSLATE_TYPE
Packit a4aae4
# endif
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
#ifdef __USE_GNU
Packit a4aae4
# define __REPB_PREFIX(name) name
Packit a4aae4
#else
Packit a4aae4
# define __REPB_PREFIX(name) __##name
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
struct re_pattern_buffer
Packit a4aae4
{
Packit a4aae4
  /* Space that holds the compiled pattern.  The type
Packit a4aae4
     'struct re_dfa_t' is private and is not declared here.  */
Packit a4aae4
  struct re_dfa_t *__REPB_PREFIX(buffer);
Packit a4aae4
Packit a4aae4
  /* Number of bytes to which 'buffer' points.  */
Packit a4aae4
  __re_long_size_t __REPB_PREFIX(allocated);
Packit a4aae4
Packit a4aae4
  /* Number of bytes actually used in 'buffer'.  */
Packit a4aae4
  __re_long_size_t __REPB_PREFIX(used);
Packit a4aae4
Packit a4aae4
  /* Syntax setting with which the pattern was compiled.  */
Packit a4aae4
  reg_syntax_t __REPB_PREFIX(syntax);
Packit a4aae4
Packit a4aae4
  /* Pointer to a fastmap, if any, otherwise zero.  re_search uses the
Packit a4aae4
     fastmap, if there is one, to skip over impossible starting points
Packit a4aae4
     for matches.  */
Packit a4aae4
  char *__REPB_PREFIX(fastmap);
Packit a4aae4
Packit a4aae4
  /* Either a translate table to apply to all characters before
Packit a4aae4
     comparing them, or zero for no translation.  The translation is
Packit a4aae4
     applied to a pattern when it is compiled and to a string when it
Packit a4aae4
     is matched.  */
Packit a4aae4
  __RE_TRANSLATE_TYPE __REPB_PREFIX(translate);
Packit a4aae4
Packit a4aae4
  /* Number of subexpressions found by the compiler.  */
Packit a4aae4
  size_t re_nsub;
Packit a4aae4
Packit a4aae4
  /* Zero if this pattern cannot match the empty string, one else.
Packit a4aae4
     Well, in truth it's used only in 're_search_2', to see whether or
Packit a4aae4
     not we should use the fastmap, so we don't set this absolutely
Packit a4aae4
     perfectly; see 're_compile_fastmap' (the "duplicate" case).  */
Packit a4aae4
  unsigned __REPB_PREFIX(can_be_null) : 1;
Packit a4aae4
Packit a4aae4
  /* If REGS_UNALLOCATED, allocate space in the 'regs' structure
Packit a4aae4
     for 'max (RE_NREGS, re_nsub + 1)' groups.
Packit a4aae4
     If REGS_REALLOCATE, reallocate space if necessary.
Packit a4aae4
     If REGS_FIXED, use what's there.  */
Packit a4aae4
#ifdef __USE_GNU
Packit a4aae4
# define REGS_UNALLOCATED 0
Packit a4aae4
# define REGS_REALLOCATE 1
Packit a4aae4
# define REGS_FIXED 2
Packit a4aae4
#endif
Packit a4aae4
  unsigned __REPB_PREFIX(regs_allocated) : 2;
Packit a4aae4
Packit a4aae4
  /* Set to zero when 're_compile_pattern' compiles a pattern; set to
Packit a4aae4
     one by 're_compile_fastmap' if it updates the fastmap.  */
Packit a4aae4
  unsigned __REPB_PREFIX(fastmap_accurate) : 1;
Packit a4aae4
Packit a4aae4
  /* If set, 're_match_2' does not return information about
Packit a4aae4
     subexpressions.  */
Packit a4aae4
  unsigned __REPB_PREFIX(no_sub) : 1;
Packit a4aae4
Packit a4aae4
  /* If set, a beginning-of-line anchor doesn't match at the beginning
Packit a4aae4
     of the string.  */
Packit a4aae4
  unsigned __REPB_PREFIX(not_bol) : 1;
Packit a4aae4
Packit a4aae4
  /* Similarly for an end-of-line anchor.  */
Packit a4aae4
  unsigned __REPB_PREFIX(not_eol) : 1;
Packit a4aae4
Packit a4aae4
  /* If true, an anchor at a newline matches.  */
Packit a4aae4
  unsigned __REPB_PREFIX(newline_anchor) : 1;
Packit a4aae4
};
Packit a4aae4
Packit a4aae4
typedef struct re_pattern_buffer regex_t;
Packit a4aae4

Packit a4aae4
/* Type for byte offsets within the string.  POSIX mandates this.  */
Packit a4aae4
#ifdef _REGEX_LARGE_OFFSETS
Packit a4aae4
/* POSIX 1003.1-2008 requires that regoff_t be at least as wide as
Packit a4aae4
   ptrdiff_t and ssize_t.  We don't know of any hosts where ptrdiff_t
Packit a4aae4
   is wider than ssize_t, so ssize_t is safe.  ptrdiff_t is not
Packit a4aae4
   visible here, so use ssize_t.  */
Packit a4aae4
typedef ssize_t regoff_t;
Packit a4aae4
#else
Packit a4aae4
/* The traditional GNU regex implementation mishandles strings longer
Packit a4aae4
   than INT_MAX.  */
Packit a4aae4
typedef int regoff_t;
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
Packit a4aae4
#ifdef __USE_GNU
Packit a4aae4
/* This is the structure we store register match data in.  See
Packit a4aae4
   regex.texinfo for a full description of what registers match.  */
Packit a4aae4
struct re_registers
Packit a4aae4
{
Packit a4aae4
  __re_size_t num_regs;
Packit a4aae4
  regoff_t *start;
Packit a4aae4
  regoff_t *end;
Packit a4aae4
};
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* If 'regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
Packit a4aae4
   're_match_2' returns information about at least this many registers
Packit a4aae4
   the first time a 'regs' structure is passed.  */
Packit a4aae4
# ifndef RE_NREGS
Packit a4aae4
#  define RE_NREGS 30
Packit a4aae4
# endif
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* POSIX specification for registers.  Aside from the different names than
Packit a4aae4
   're_registers', POSIX uses an array of structures, instead of a
Packit a4aae4
   structure of arrays.  */
Packit a4aae4
typedef struct
Packit a4aae4
{
Packit a4aae4
  regoff_t rm_so;  /* Byte offset from string's start to substring's start.  */
Packit a4aae4
  regoff_t rm_eo;  /* Byte offset from string's start to substring's end.  */
Packit a4aae4
} regmatch_t;
Packit a4aae4

Packit a4aae4
/* Declarations for routines.  */
Packit a4aae4
Packit a4aae4
#ifdef __USE_GNU
Packit a4aae4
/* Sets the current default syntax to SYNTAX, and return the old syntax.
Packit a4aae4
   You can also simply assign to the 're_syntax_options' variable.  */
Packit a4aae4
extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax);
Packit a4aae4
Packit a4aae4
/* Compile the regular expression PATTERN, with length LENGTH
Packit a4aae4
   and syntax given by the global 're_syntax_options', into the buffer
Packit a4aae4
   BUFFER.  Return NULL if successful, and an error string if not.
Packit a4aae4
Packit a4aae4
   To free the allocated storage, you must call 'regfree' on BUFFER.
Packit a4aae4
   Note that the translate table must either have been initialized by
Packit a4aae4
   'regcomp', with a malloc'ed value, or set to NULL before calling
Packit a4aae4
   'regfree'.  */
Packit a4aae4
extern const char *re_compile_pattern (const char *__pattern, size_t __length,
Packit a4aae4
				       struct re_pattern_buffer *__buffer);
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* Compile a fastmap for the compiled pattern in BUFFER; used to
Packit a4aae4
   accelerate searches.  Return 0 if successful and -2 if was an
Packit a4aae4
   internal error.  */
Packit a4aae4
extern int re_compile_fastmap (struct re_pattern_buffer *__buffer);
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* Search in the string STRING (with length LENGTH) for the pattern
Packit a4aae4
   compiled into BUFFER.  Start searching at position START, for RANGE
Packit a4aae4
   characters.  Return the starting position of the match, -1 for no
Packit a4aae4
   match, or -2 for an internal error.  Also return register
Packit a4aae4
   information in REGS (if REGS and BUFFER->no_sub are nonzero).  */
Packit a4aae4
extern regoff_t re_search (struct re_pattern_buffer *__buffer,
Packit a4aae4
			   const char *__String, regoff_t __length,
Packit a4aae4
			   regoff_t __start, regoff_t __range,
Packit a4aae4
			   struct re_registers *__regs);
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* Like 're_search', but search in the concatenation of STRING1 and
Packit a4aae4
   STRING2.  Also, stop searching at index START + STOP.  */
Packit a4aae4
extern regoff_t re_search_2 (struct re_pattern_buffer *__buffer,
Packit a4aae4
			     const char *__string1, regoff_t __length1,
Packit a4aae4
			     const char *__string2, regoff_t __length2,
Packit a4aae4
			     regoff_t __start, regoff_t __range,
Packit a4aae4
			     struct re_registers *__regs,
Packit a4aae4
			     regoff_t __stop);
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* Like 're_search', but return how many characters in STRING the regexp
Packit a4aae4
   in BUFFER matched, starting at position START.  */
Packit a4aae4
extern regoff_t re_match (struct re_pattern_buffer *__buffer,
Packit a4aae4
			  const char *__String, regoff_t __length,
Packit a4aae4
			  regoff_t __start, struct re_registers *__regs);
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* Relates to 're_match' as 're_search_2' relates to 're_search'.  */
Packit a4aae4
extern regoff_t re_match_2 (struct re_pattern_buffer *__buffer,
Packit a4aae4
			    const char *__string1, regoff_t __length1,
Packit a4aae4
			    const char *__string2, regoff_t __length2,
Packit a4aae4
			    regoff_t __start, struct re_registers *__regs,
Packit a4aae4
			    regoff_t __stop);
Packit a4aae4
Packit a4aae4
Packit a4aae4
/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
Packit a4aae4
   ENDS.  Subsequent matches using BUFFER and REGS will use this memory
Packit a4aae4
   for recording register information.  STARTS and ENDS must be
Packit a4aae4
   allocated with malloc, and must each be at least 'NUM_REGS * sizeof
Packit a4aae4
   (regoff_t)' bytes long.
Packit a4aae4
Packit a4aae4
   If NUM_REGS == 0, then subsequent matches should allocate their own
Packit a4aae4
   register data.
Packit a4aae4
Packit a4aae4
   Unless this function is called, the first search or match using
Packit a4aae4
   BUFFER will allocate its own register data, without
Packit a4aae4
   freeing the old data.  */
Packit a4aae4
extern void re_set_registers (struct re_pattern_buffer *__buffer,
Packit a4aae4
			      struct re_registers *__regs,
Packit a4aae4
			      __re_size_t __num_regs,
Packit a4aae4
			      regoff_t *__starts, regoff_t *__ends);
Packit a4aae4
#endif	/* Use GNU */
Packit a4aae4
Packit a4aae4
#if defined _REGEX_RE_COMP || (defined _LIBC && defined __USE_MISC)
Packit a4aae4
# ifndef _CRAY
Packit a4aae4
/* 4.2 bsd compatibility.  */
Packit a4aae4
extern char *re_comp (const char *);
Packit a4aae4
extern int re_exec (const char *);
Packit a4aae4
# endif
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
/* GCC 2.95 and later have "__restrict"; C99 compilers have
Packit a4aae4
   "restrict", and "configure" may have defined "restrict".
Packit a4aae4
   Other compilers use __restrict, __restrict__, and _Restrict, and
Packit a4aae4
   'configure' might #define 'restrict' to those words, so pick a
Packit a4aae4
   different name.  */
Packit a4aae4
#ifndef _Restrict_
Packit a4aae4
# if 199901L <= __STDC_VERSION__
Packit a4aae4
#  define _Restrict_ restrict
Packit a4aae4
# elif 2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__)
Packit a4aae4
#  define _Restrict_ __restrict
Packit a4aae4
# else
Packit a4aae4
#  define _Restrict_
Packit a4aae4
# endif
Packit a4aae4
#endif
Packit a4aae4
/* gcc 3.1 and up support the [restrict] syntax.  Don't trust
Packit a4aae4
   sys/cdefs.h's definition of __restrict_arr, though, as it
Packit a4aae4
   mishandles gcc -ansi -pedantic.  */
Packit a4aae4
#ifndef _Restrict_arr_
Packit a4aae4
# if ((199901L <= __STDC_VERSION__					\
Packit a4aae4
       || ((3 < __GNUC__ || (3 == __GNUC__ && 1 <= __GNUC_MINOR__))	\
Packit a4aae4
	   && !defined __STRICT_ANSI__))					\
Packit a4aae4
      && !defined __GNUG__)
Packit a4aae4
#  define _Restrict_arr_ _Restrict_
Packit a4aae4
# else
Packit a4aae4
#  define _Restrict_arr_
Packit a4aae4
# endif
Packit a4aae4
#endif
Packit a4aae4
Packit a4aae4
/* POSIX compatibility.  */
Packit a4aae4
extern int regcomp (regex_t *_Restrict_ __preg,
Packit a4aae4
		    const char *_Restrict_ __pattern,
Packit a4aae4
		    int __cflags);
Packit a4aae4
Packit a4aae4
extern int regexec (const regex_t *_Restrict_ __preg,
Packit a4aae4
		    const char *_Restrict_ __String, size_t __nmatch,
Packit a4aae4
		    regmatch_t __pmatch[_Restrict_arr_],
Packit a4aae4
		    int __eflags);
Packit a4aae4
Packit a4aae4
extern size_t regerror (int __errcode, const regex_t *_Restrict_ __preg,
Packit a4aae4
			char *_Restrict_ __errbuf, size_t __errbuf_size);
Packit a4aae4
Packit a4aae4
extern void regfree (regex_t *__preg);
Packit a4aae4
Packit a4aae4
Packit a4aae4
#ifdef __cplusplus
Packit a4aae4
}
Packit a4aae4
#endif	/* C++ */
Packit a4aae4
Packit a4aae4
#endif /* regex.h */