Blame posix/regex.h

Packit 6c4009
/* Definitions for data structures and routines for the regular
Packit 6c4009
   expression library.
Packit 6c4009
   Copyright (C) 1985, 1989-2018 Free Software Foundation, Inc.
Packit 6c4009
   This file is part of the GNU C Library.
Packit 6c4009
Packit 6c4009
   The GNU C Library is free software; you can redistribute it and/or
Packit 6c4009
   modify it under the terms of the GNU Lesser General Public
Packit 6c4009
   License as published by the Free Software Foundation; either
Packit 6c4009
   version 2.1 of the License, or (at your option) any later version.
Packit 6c4009
Packit 6c4009
   The GNU C Library is distributed in the hope that it will be useful,
Packit 6c4009
   but WITHOUT ANY WARRANTY; without even the implied warranty of
Packit 6c4009
   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
Packit 6c4009
   Lesser General Public License for more details.
Packit 6c4009
Packit 6c4009
   You should have received a copy of the GNU Lesser General Public
Packit 6c4009
   License along with the GNU C Library; if not, see
Packit 6c4009
   <https://www.gnu.org/licenses/>.  */
Packit 6c4009
Packit 6c4009
#ifndef _REGEX_H
Packit 6c4009
#define _REGEX_H 1
Packit 6c4009
Packit 6c4009
#include <sys/types.h>
Packit 6c4009
Packit 6c4009
/* Allow the use in C++ code.  */
Packit 6c4009
#ifdef __cplusplus
Packit 6c4009
extern "C" {
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
/* Define __USE_GNU to declare GNU extensions that violate the
Packit 6c4009
   POSIX name space rules.  */
Packit 6c4009
#ifdef _GNU_SOURCE
Packit 6c4009
# define __USE_GNU 1
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
#ifdef _REGEX_LARGE_OFFSETS
Packit 6c4009
Packit 6c4009
/* Use types and values that are wide enough to represent signed and
Packit 6c4009
   unsigned byte offsets in memory.  This currently works only when
Packit 6c4009
   the regex code is used outside of the GNU C library; it is not yet
Packit 6c4009
   supported within glibc itself, and glibc users should not define
Packit 6c4009
   _REGEX_LARGE_OFFSETS.  */
Packit 6c4009
Packit 6c4009
/* The type of object sizes.  */
Packit 6c4009
typedef size_t __re_size_t;
Packit 6c4009
Packit 6c4009
/* The type of object sizes, in places where the traditional code
Packit 6c4009
   uses unsigned long int.  */
Packit 6c4009
typedef size_t __re_long_size_t;
Packit 6c4009
Packit 6c4009
#else
Packit 6c4009
Packit 6c4009
/* The traditional GNU regex implementation mishandles strings longer
Packit 6c4009
   than INT_MAX.  */
Packit 6c4009
typedef unsigned int __re_size_t;
Packit 6c4009
typedef unsigned long int __re_long_size_t;
Packit 6c4009
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
/* The following two types have to be signed and unsigned integer type
Packit 6c4009
   wide enough to hold a value of a pointer.  For most ANSI compilers
Packit 6c4009
   ptrdiff_t and size_t should be likely OK.  Still size of these two
Packit 6c4009
   types is 2 for Microsoft C.  Ugh... */
Packit 6c4009
typedef long int s_reg_t;
Packit 6c4009
typedef unsigned long int active_reg_t;
Packit 6c4009
Packit 6c4009
/* The following bits are used to determine the regexp syntax we
Packit 6c4009
   recognize.  The set/not-set meanings are chosen so that Emacs syntax
Packit 6c4009
   remains the value 0.  The bits are given in alphabetical order, and
Packit 6c4009
   the definitions shifted by one from the previous bit; thus, when we
Packit 6c4009
   add or remove a bit, only one other definition need change.  */
Packit 6c4009
typedef unsigned long int reg_syntax_t;
Packit 6c4009
Packit 6c4009
#ifdef __USE_GNU
Packit 6c4009
/* If this bit is not set, then \ inside a bracket expression is literal.
Packit 6c4009
   If set, then such a \ quotes the following character.  */
Packit 6c4009
# define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1)
Packit 6c4009
Packit 6c4009
/* If this bit is not set, then + and ? are operators, and \+ and \? are
Packit 6c4009
     literals.
Packit 6c4009
   If set, then \+ and \? are operators and + and ? are literals.  */
Packit 6c4009
# define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then character classes are supported.  They are:
Packit 6c4009
     [:alpha:], [:upper:], [:lower:],  [:digit:], [:alnum:], [:xdigit:],
Packit 6c4009
     [:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
Packit 6c4009
   If not set, then character classes are not supported.  */
Packit 6c4009
# define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then ^ and $ are always anchors (outside bracket
Packit 6c4009
     expressions, of course).
Packit 6c4009
   If this bit is not set, then it depends:
Packit 6c4009
	^  is an anchor if it is at the beginning of a regular
Packit 6c4009
	   expression or after an open-group or an alternation operator;
Packit 6c4009
	$  is an anchor if it is at the end of a regular expression, or
Packit 6c4009
	   before a close-group or an alternation operator.
Packit 6c4009
Packit 6c4009
   This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
Packit 6c4009
   POSIX draft 11.2 says that * etc. in leading positions is undefined.
Packit 6c4009
   We already implemented a previous draft which made those constructs
Packit 6c4009
   invalid, though, so we haven't changed the code back.  */
Packit 6c4009
# define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then special characters are always special
Packit 6c4009
     regardless of where they are in the pattern.
Packit 6c4009
   If this bit is not set, then special characters are special only in
Packit 6c4009
     some contexts; otherwise they are ordinary.  Specifically,
Packit 6c4009
     * + ? and intervals are only special when not after the beginning,
Packit 6c4009
     open-group, or alternation operator.  */
Packit 6c4009
# define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then *, +, ?, and { cannot be first in an re or
Packit 6c4009
     immediately after an alternation or begin-group operator.  */
Packit 6c4009
# define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then . matches newline.
Packit 6c4009
   If not set, then it doesn't.  */
Packit 6c4009
# define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then . doesn't match NUL.
Packit 6c4009
   If not set, then it does.  */
Packit 6c4009
# define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, nonmatching lists [^...] do not match newline.
Packit 6c4009
   If not set, they do.  */
Packit 6c4009
# define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, either \{...\} or {...} defines an
Packit 6c4009
     interval, depending on RE_NO_BK_BRACES.
Packit 6c4009
   If not set, \{, \}, {, and } are literals.  */
Packit 6c4009
# define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, +, ? and | aren't recognized as operators.
Packit 6c4009
   If not set, they are.  */
Packit 6c4009
# define RE_LIMITED_OPS (RE_INTERVALS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, newline is an alternation operator.
Packit 6c4009
   If not set, newline is literal.  */
Packit 6c4009
# define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then '{...}' defines an interval, and \{ and \}
Packit 6c4009
     are literals.
Packit 6c4009
  If not set, then '\{...\}' defines an interval.  */
Packit 6c4009
# define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, (...) defines a group, and \( and \) are literals.
Packit 6c4009
   If not set, \(...\) defines a group, and ( and ) are literals.  */
Packit 6c4009
# define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then \<digit> matches <digit>.
Packit 6c4009
   If not set, then \<digit> is a back-reference.  */
Packit 6c4009
# define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then | is an alternation operator, and \| is literal.
Packit 6c4009
   If not set, then \| is an alternation operator, and | is literal.  */
Packit 6c4009
# define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then an ending range point collating higher
Packit 6c4009
     than the starting range point, as in [z-a], is invalid.
Packit 6c4009
   If not set, then when ending range point collates higher than the
Packit 6c4009
     starting range point, the range is ignored.  */
Packit 6c4009
# define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then an unmatched ) is ordinary.
Packit 6c4009
   If not set, then an unmatched ) is invalid.  */
Packit 6c4009
# define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, succeed as soon as we match the whole pattern,
Packit 6c4009
   without further backtracking.  */
Packit 6c4009
# define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, do not process the GNU regex operators.
Packit 6c4009
   If not set, then the GNU regex operators are recognized. */
Packit 6c4009
# define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, turn on internal regex debugging.
Packit 6c4009
   If not set, and debugging was on, turn it off.
Packit 6c4009
   This only works if regex.c is compiled -DDEBUG.
Packit 6c4009
   We define this bit always, so that all that's needed to turn on
Packit 6c4009
   debugging is to recompile regex.c; the calling code can always have
Packit 6c4009
   this bit set, and it won't affect anything in the normal case. */
Packit 6c4009
# define RE_DEBUG (RE_NO_GNU_OPS << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, a syntactically invalid interval is treated as
Packit 6c4009
   a string of ordinary characters.  For example, the ERE 'a{1' is
Packit 6c4009
   treated as 'a\{1'.  */
Packit 6c4009
# define RE_INVALID_INTERVAL_ORD (RE_DEBUG << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then ignore case when matching.
Packit 6c4009
   If not set, then case is significant.  */
Packit 6c4009
# define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1)
Packit 6c4009
Packit 6c4009
/* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only
Packit 6c4009
   for ^, because it is difficult to scan the regex backwards to find
Packit 6c4009
   whether ^ should be special.  */
Packit 6c4009
# define RE_CARET_ANCHORS_HERE (RE_ICASE << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then \{ cannot be first in a regex or
Packit 6c4009
   immediately after an alternation, open-group or \} operator.  */
Packit 6c4009
# define RE_CONTEXT_INVALID_DUP (RE_CARET_ANCHORS_HERE << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then no_sub will be set to 1 during
Packit 6c4009
   re_compile_pattern.  */
Packit 6c4009
# define RE_NO_SUB (RE_CONTEXT_INVALID_DUP << 1)
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
/* This global variable defines the particular regexp syntax to use (for
Packit 6c4009
   some interfaces).  When a regexp is compiled, the syntax used is
Packit 6c4009
   stored in the pattern buffer, so changing this does not affect
Packit 6c4009
   already-compiled regexps.  */
Packit 6c4009
extern reg_syntax_t re_syntax_options;
Packit 6c4009

Packit 6c4009
#ifdef __USE_GNU
Packit 6c4009
/* Define combinations of the above bits for the standard possibilities.
Packit 6c4009
   (The [[[ comments delimit what gets put into the Texinfo file, so
Packit 6c4009
   don't delete them!)  */
Packit 6c4009
/* [[[begin syntaxes]]] */
Packit 6c4009
# define RE_SYNTAX_EMACS 0
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_AWK							\
Packit 6c4009
  (RE_BACKSLASH_ESCAPE_IN_LISTS   | RE_DOT_NOT_NULL			\
Packit 6c4009
   | RE_NO_BK_PARENS              | RE_NO_BK_REFS			\
Packit 6c4009
   | RE_NO_BK_VBAR                | RE_NO_EMPTY_RANGES			\
Packit 6c4009
   | RE_DOT_NEWLINE		  | RE_CONTEXT_INDEP_ANCHORS		\
Packit 6c4009
   | RE_CHAR_CLASSES							\
Packit 6c4009
   | RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS)
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_GNU_AWK						\
Packit 6c4009
  ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS		\
Packit 6c4009
    | RE_INVALID_INTERVAL_ORD)						\
Packit 6c4009
   & ~(RE_DOT_NOT_NULL | RE_CONTEXT_INDEP_OPS				\
Packit 6c4009
      | RE_CONTEXT_INVALID_OPS ))
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_POSIX_AWK						\
Packit 6c4009
  (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS		\
Packit 6c4009
   | RE_INTERVALS	    | RE_NO_GNU_OPS				\
Packit 6c4009
   | RE_INVALID_INTERVAL_ORD)
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_GREP							\
Packit 6c4009
  ((RE_SYNTAX_POSIX_BASIC | RE_NEWLINE_ALT)				\
Packit 6c4009
   & ~(RE_CONTEXT_INVALID_DUP | RE_DOT_NOT_NULL))
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_EGREP						\
Packit 6c4009
  ((RE_SYNTAX_POSIX_EXTENDED | RE_INVALID_INTERVAL_ORD | RE_NEWLINE_ALT) \
Packit 6c4009
   & ~(RE_CONTEXT_INVALID_OPS | RE_DOT_NOT_NULL))
Packit 6c4009
Packit 6c4009
/* POSIX grep -E behavior is no longer incompatible with GNU.  */
Packit 6c4009
# define RE_SYNTAX_POSIX_EGREP						\
Packit 6c4009
  RE_SYNTAX_EGREP
Packit 6c4009
Packit 6c4009
/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff.  */
Packit 6c4009
# define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_SED RE_SYNTAX_POSIX_BASIC
Packit 6c4009
Packit 6c4009
/* Syntax bits common to both basic and extended POSIX regex syntax.  */
Packit 6c4009
# define _RE_SYNTAX_POSIX_COMMON					\
Packit 6c4009
  (RE_CHAR_CLASSES | RE_DOT_NEWLINE      | RE_DOT_NOT_NULL		\
Packit 6c4009
   | RE_INTERVALS  | RE_NO_EMPTY_RANGES)
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_POSIX_BASIC						\
Packit 6c4009
  (_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM | RE_CONTEXT_INVALID_DUP)
Packit 6c4009
Packit 6c4009
/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
Packit 6c4009
   RE_LIMITED_OPS, i.e., \? \+ \| are not recognized.  Actually, this
Packit 6c4009
   isn't minimal, since other operators, such as \`, aren't disabled.  */
Packit 6c4009
# define RE_SYNTAX_POSIX_MINIMAL_BASIC					\
Packit 6c4009
  (_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
Packit 6c4009
Packit 6c4009
# define RE_SYNTAX_POSIX_EXTENDED					\
Packit 6c4009
  (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS			\
Packit 6c4009
   | RE_CONTEXT_INDEP_OPS   | RE_NO_BK_BRACES				\
Packit 6c4009
   | RE_NO_BK_PARENS        | RE_NO_BK_VBAR				\
Packit 6c4009
   | RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD)
Packit 6c4009
Packit 6c4009
/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is
Packit 6c4009
   removed and RE_NO_BK_REFS is added.  */
Packit 6c4009
# define RE_SYNTAX_POSIX_MINIMAL_EXTENDED				\
Packit 6c4009
  (_RE_SYNTAX_POSIX_COMMON  | RE_CONTEXT_INDEP_ANCHORS			\
Packit 6c4009
   | RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES				\
Packit 6c4009
   | RE_NO_BK_PARENS        | RE_NO_BK_REFS				\
Packit 6c4009
   | RE_NO_BK_VBAR	    | RE_UNMATCHED_RIGHT_PAREN_ORD)
Packit 6c4009
/* [[[end syntaxes]]] */
Packit 6c4009
Packit 6c4009
/* Maximum number of duplicates an interval can allow.  POSIX-conforming
Packit 6c4009
   systems might define this in <limits.h>, but we want our
Packit 6c4009
   value, so remove any previous define.  */
Packit 6c4009
# ifdef _REGEX_INCLUDE_LIMITS_H
Packit 6c4009
#  include <limits.h>
Packit 6c4009
# endif
Packit 6c4009
# ifdef RE_DUP_MAX
Packit 6c4009
#  undef RE_DUP_MAX
Packit 6c4009
# endif
Packit 6c4009
Packit 6c4009
/* RE_DUP_MAX is 2**15 - 1 because an earlier implementation stored
Packit 6c4009
   the counter as a 2-byte signed integer.  This is no longer true, so
Packit 6c4009
   RE_DUP_MAX could be increased to (INT_MAX / 10 - 1), or to
Packit 6c4009
   ((SIZE_MAX - 9) / 10) if _REGEX_LARGE_OFFSETS is defined.
Packit 6c4009
   However, there would be a huge performance problem if someone
Packit 6c4009
   actually used a pattern like a\{214748363\}, so RE_DUP_MAX retains
Packit 6c4009
   its historical value.  */
Packit 6c4009
# define RE_DUP_MAX (0x7fff)
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* POSIX 'cflags' bits (i.e., information for 'regcomp').  */
Packit 6c4009
Packit 6c4009
/* If this bit is set, then use extended regular expression syntax.
Packit 6c4009
   If not set, then use basic regular expression syntax.  */
Packit 6c4009
#define REG_EXTENDED 1
Packit 6c4009
Packit 6c4009
/* If this bit is set, then ignore case when matching.
Packit 6c4009
   If not set, then case is significant.  */
Packit 6c4009
#define REG_ICASE (1 << 1)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then anchors do not match at newline
Packit 6c4009
     characters in the string.
Packit 6c4009
   If not set, then anchors do match at newlines.  */
Packit 6c4009
#define REG_NEWLINE (1 << 2)
Packit 6c4009
Packit 6c4009
/* If this bit is set, then report only success or fail in regexec.
Packit 6c4009
   If not set, then returns differ between not matching and errors.  */
Packit 6c4009
#define REG_NOSUB (1 << 3)
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* POSIX 'eflags' bits (i.e., information for regexec).  */
Packit 6c4009
Packit 6c4009
/* If this bit is set, then the beginning-of-line operator doesn't match
Packit 6c4009
     the beginning of the string (presumably because it's not the
Packit 6c4009
     beginning of a line).
Packit 6c4009
   If not set, then the beginning-of-line operator does match the
Packit 6c4009
     beginning of the string.  */
Packit 6c4009
#define REG_NOTBOL 1
Packit 6c4009
Packit 6c4009
/* Like REG_NOTBOL, except for the end-of-line.  */
Packit 6c4009
#define REG_NOTEOL (1 << 1)
Packit 6c4009
Packit 6c4009
/* Use PMATCH[0] to delimit the start and end of the search in the
Packit 6c4009
   buffer.  */
Packit 6c4009
#define REG_STARTEND (1 << 2)
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* If any error codes are removed, changed, or added, update the
Packit 6c4009
   '__re_error_msgid' table in regcomp.c.  */
Packit 6c4009
Packit 6c4009
typedef enum
Packit 6c4009
{
Packit 6c4009
  _REG_ENOSYS = -1,	/* This will never happen for this implementation.  */
Packit 6c4009
  _REG_NOERROR = 0,	/* Success.  */
Packit 6c4009
  _REG_NOMATCH,		/* Didn't find a match (for regexec).  */
Packit 6c4009
Packit 6c4009
  /* POSIX regcomp return error codes.  (In the order listed in the
Packit 6c4009
     standard.)  */
Packit 6c4009
  _REG_BADPAT,		/* Invalid pattern.  */
Packit 6c4009
  _REG_ECOLLATE,	/* Invalid collating element.  */
Packit 6c4009
  _REG_ECTYPE,		/* Invalid character class name.  */
Packit 6c4009
  _REG_EESCAPE,		/* Trailing backslash.  */
Packit 6c4009
  _REG_ESUBREG,		/* Invalid back reference.  */
Packit 6c4009
  _REG_EBRACK,		/* Unmatched left bracket.  */
Packit 6c4009
  _REG_EPAREN,		/* Parenthesis imbalance.  */
Packit 6c4009
  _REG_EBRACE,		/* Unmatched \{.  */
Packit 6c4009
  _REG_BADBR,		/* Invalid contents of \{\}.  */
Packit 6c4009
  _REG_ERANGE,		/* Invalid range end.  */
Packit 6c4009
  _REG_ESPACE,		/* Ran out of memory.  */
Packit 6c4009
  _REG_BADRPT,		/* No preceding re for repetition op.  */
Packit 6c4009
Packit 6c4009
  /* Error codes we've added.  */
Packit 6c4009
  _REG_EEND,		/* Premature end.  */
Packit 6c4009
  _REG_ESIZE,		/* Too large (e.g., repeat count too large).  */
Packit 6c4009
  _REG_ERPAREN		/* Unmatched ) or \); not returned from regcomp.  */
Packit 6c4009
} reg_errcode_t;
Packit 6c4009
Packit 6c4009
#if defined _XOPEN_SOURCE || defined __USE_XOPEN2K
Packit 6c4009
# define REG_ENOSYS	_REG_ENOSYS
Packit 6c4009
#endif
Packit 6c4009
#define REG_NOERROR	_REG_NOERROR
Packit 6c4009
#define REG_NOMATCH	_REG_NOMATCH
Packit 6c4009
#define REG_BADPAT	_REG_BADPAT
Packit 6c4009
#define REG_ECOLLATE	_REG_ECOLLATE
Packit 6c4009
#define REG_ECTYPE	_REG_ECTYPE
Packit 6c4009
#define REG_EESCAPE	_REG_EESCAPE
Packit 6c4009
#define REG_ESUBREG	_REG_ESUBREG
Packit 6c4009
#define REG_EBRACK	_REG_EBRACK
Packit 6c4009
#define REG_EPAREN	_REG_EPAREN
Packit 6c4009
#define REG_EBRACE	_REG_EBRACE
Packit 6c4009
#define REG_BADBR	_REG_BADBR
Packit 6c4009
#define REG_ERANGE	_REG_ERANGE
Packit 6c4009
#define REG_ESPACE	_REG_ESPACE
Packit 6c4009
#define REG_BADRPT	_REG_BADRPT
Packit 6c4009
#define REG_EEND	_REG_EEND
Packit 6c4009
#define REG_ESIZE	_REG_ESIZE
Packit 6c4009
#define REG_ERPAREN	_REG_ERPAREN
Packit 6c4009

Packit 6c4009
/* This data structure represents a compiled pattern.  Before calling
Packit 6c4009
   the pattern compiler, the fields 'buffer', 'allocated', 'fastmap',
Packit 6c4009
   and 'translate' can be set.  After the pattern has been compiled,
Packit 6c4009
   the fields 're_nsub', 'not_bol' and 'not_eol' are available.  All
Packit 6c4009
   other fields are private to the regex routines.  */
Packit 6c4009
Packit 6c4009
#ifndef RE_TRANSLATE_TYPE
Packit 6c4009
# define __RE_TRANSLATE_TYPE unsigned char *
Packit 6c4009
# ifdef __USE_GNU
Packit 6c4009
#  define RE_TRANSLATE_TYPE __RE_TRANSLATE_TYPE
Packit 6c4009
# endif
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
#ifdef __USE_GNU
Packit 6c4009
# define __REPB_PREFIX(name) name
Packit 6c4009
#else
Packit 6c4009
# define __REPB_PREFIX(name) __##name
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
struct re_pattern_buffer
Packit 6c4009
{
Packit 6c4009
  /* Space that holds the compiled pattern.  The type
Packit 6c4009
     'struct re_dfa_t' is private and is not declared here.  */
Packit 6c4009
  struct re_dfa_t *__REPB_PREFIX(buffer);
Packit 6c4009
Packit 6c4009
  /* Number of bytes to which 'buffer' points.  */
Packit 6c4009
  __re_long_size_t __REPB_PREFIX(allocated);
Packit 6c4009
Packit 6c4009
  /* Number of bytes actually used in 'buffer'.  */
Packit 6c4009
  __re_long_size_t __REPB_PREFIX(used);
Packit 6c4009
Packit 6c4009
  /* Syntax setting with which the pattern was compiled.  */
Packit 6c4009
  reg_syntax_t __REPB_PREFIX(syntax);
Packit 6c4009
Packit 6c4009
  /* Pointer to a fastmap, if any, otherwise zero.  re_search uses the
Packit 6c4009
     fastmap, if there is one, to skip over impossible starting points
Packit 6c4009
     for matches.  */
Packit 6c4009
  char *__REPB_PREFIX(fastmap);
Packit 6c4009
Packit 6c4009
  /* Either a translate table to apply to all characters before
Packit 6c4009
     comparing them, or zero for no translation.  The translation is
Packit 6c4009
     applied to a pattern when it is compiled and to a string when it
Packit 6c4009
     is matched.  */
Packit 6c4009
  __RE_TRANSLATE_TYPE __REPB_PREFIX(translate);
Packit 6c4009
Packit 6c4009
  /* Number of subexpressions found by the compiler.  */
Packit 6c4009
  size_t re_nsub;
Packit 6c4009
Packit 6c4009
  /* Zero if this pattern cannot match the empty string, one else.
Packit 6c4009
     Well, in truth it's used only in 're_search_2', to see whether or
Packit 6c4009
     not we should use the fastmap, so we don't set this absolutely
Packit 6c4009
     perfectly; see 're_compile_fastmap' (the "duplicate" case).  */
Packit 6c4009
  unsigned __REPB_PREFIX(can_be_null) : 1;
Packit 6c4009
Packit 6c4009
  /* If REGS_UNALLOCATED, allocate space in the 'regs' structure
Packit 6c4009
     for 'max (RE_NREGS, re_nsub + 1)' groups.
Packit 6c4009
     If REGS_REALLOCATE, reallocate space if necessary.
Packit 6c4009
     If REGS_FIXED, use what's there.  */
Packit 6c4009
#ifdef __USE_GNU
Packit 6c4009
# define REGS_UNALLOCATED 0
Packit 6c4009
# define REGS_REALLOCATE 1
Packit 6c4009
# define REGS_FIXED 2
Packit 6c4009
#endif
Packit 6c4009
  unsigned __REPB_PREFIX(regs_allocated) : 2;
Packit 6c4009
Packit 6c4009
  /* Set to zero when 're_compile_pattern' compiles a pattern; set to
Packit 6c4009
     one by 're_compile_fastmap' if it updates the fastmap.  */
Packit 6c4009
  unsigned __REPB_PREFIX(fastmap_accurate) : 1;
Packit 6c4009
Packit 6c4009
  /* If set, 're_match_2' does not return information about
Packit 6c4009
     subexpressions.  */
Packit 6c4009
  unsigned __REPB_PREFIX(no_sub) : 1;
Packit 6c4009
Packit 6c4009
  /* If set, a beginning-of-line anchor doesn't match at the beginning
Packit 6c4009
     of the string.  */
Packit 6c4009
  unsigned __REPB_PREFIX(not_bol) : 1;
Packit 6c4009
Packit 6c4009
  /* Similarly for an end-of-line anchor.  */
Packit 6c4009
  unsigned __REPB_PREFIX(not_eol) : 1;
Packit 6c4009
Packit 6c4009
  /* If true, an anchor at a newline matches.  */
Packit 6c4009
  unsigned __REPB_PREFIX(newline_anchor) : 1;
Packit 6c4009
};
Packit 6c4009
Packit 6c4009
typedef struct re_pattern_buffer regex_t;
Packit 6c4009

Packit 6c4009
/* Type for byte offsets within the string.  POSIX mandates this.  */
Packit 6c4009
#ifdef _REGEX_LARGE_OFFSETS
Packit 6c4009
/* POSIX 1003.1-2008 requires that regoff_t be at least as wide as
Packit 6c4009
   ptrdiff_t and ssize_t.  We don't know of any hosts where ptrdiff_t
Packit 6c4009
   is wider than ssize_t, so ssize_t is safe.  ptrdiff_t is not
Packit 6c4009
   visible here, so use ssize_t.  */
Packit 6c4009
typedef ssize_t regoff_t;
Packit 6c4009
#else
Packit 6c4009
/* The traditional GNU regex implementation mishandles strings longer
Packit 6c4009
   than INT_MAX.  */
Packit 6c4009
typedef int regoff_t;
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
Packit 6c4009
#ifdef __USE_GNU
Packit 6c4009
/* This is the structure we store register match data in.  See
Packit 6c4009
   regex.texinfo for a full description of what registers match.  */
Packit 6c4009
struct re_registers
Packit 6c4009
{
Packit 6c4009
  __re_size_t num_regs;
Packit 6c4009
  regoff_t *start;
Packit 6c4009
  regoff_t *end;
Packit 6c4009
};
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* If 'regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
Packit 6c4009
   're_match_2' returns information about at least this many registers
Packit 6c4009
   the first time a 'regs' structure is passed.  */
Packit 6c4009
# ifndef RE_NREGS
Packit 6c4009
#  define RE_NREGS 30
Packit 6c4009
# endif
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* POSIX specification for registers.  Aside from the different names than
Packit 6c4009
   're_registers', POSIX uses an array of structures, instead of a
Packit 6c4009
   structure of arrays.  */
Packit 6c4009
typedef struct
Packit 6c4009
{
Packit 6c4009
  regoff_t rm_so;  /* Byte offset from string's start to substring's start.  */
Packit 6c4009
  regoff_t rm_eo;  /* Byte offset from string's start to substring's end.  */
Packit 6c4009
} regmatch_t;
Packit 6c4009

Packit 6c4009
/* Declarations for routines.  */
Packit 6c4009
Packit 6c4009
#ifdef __USE_GNU
Packit 6c4009
/* Sets the current default syntax to SYNTAX, and return the old syntax.
Packit 6c4009
   You can also simply assign to the 're_syntax_options' variable.  */
Packit 6c4009
extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax);
Packit 6c4009
Packit 6c4009
/* Compile the regular expression PATTERN, with length LENGTH
Packit 6c4009
   and syntax given by the global 're_syntax_options', into the buffer
Packit 6c4009
   BUFFER.  Return NULL if successful, and an error string if not.
Packit 6c4009
Packit 6c4009
   To free the allocated storage, you must call 'regfree' on BUFFER.
Packit 6c4009
   Note that the translate table must either have been initialized by
Packit 6c4009
   'regcomp', with a malloc'ed value, or set to NULL before calling
Packit 6c4009
   'regfree'.  */
Packit 6c4009
extern const char *re_compile_pattern (const char *__pattern, size_t __length,
Packit 6c4009
				       struct re_pattern_buffer *__buffer);
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* Compile a fastmap for the compiled pattern in BUFFER; used to
Packit 6c4009
   accelerate searches.  Return 0 if successful and -2 if was an
Packit 6c4009
   internal error.  */
Packit 6c4009
extern int re_compile_fastmap (struct re_pattern_buffer *__buffer);
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* Search in the string STRING (with length LENGTH) for the pattern
Packit 6c4009
   compiled into BUFFER.  Start searching at position START, for RANGE
Packit 6c4009
   characters.  Return the starting position of the match, -1 for no
Packit 6c4009
   match, or -2 for an internal error.  Also return register
Packit 6c4009
   information in REGS (if REGS and BUFFER->no_sub are nonzero).  */
Packit 6c4009
extern regoff_t re_search (struct re_pattern_buffer *__buffer,
Packit 6c4009
			   const char *__String, regoff_t __length,
Packit 6c4009
			   regoff_t __start, regoff_t __range,
Packit 6c4009
			   struct re_registers *__regs);
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* Like 're_search', but search in the concatenation of STRING1 and
Packit 6c4009
   STRING2.  Also, stop searching at index START + STOP.  */
Packit 6c4009
extern regoff_t re_search_2 (struct re_pattern_buffer *__buffer,
Packit 6c4009
			     const char *__string1, regoff_t __length1,
Packit 6c4009
			     const char *__string2, regoff_t __length2,
Packit 6c4009
			     regoff_t __start, regoff_t __range,
Packit 6c4009
			     struct re_registers *__regs,
Packit 6c4009
			     regoff_t __stop);
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* Like 're_search', but return how many characters in STRING the regexp
Packit 6c4009
   in BUFFER matched, starting at position START.  */
Packit 6c4009
extern regoff_t re_match (struct re_pattern_buffer *__buffer,
Packit 6c4009
			  const char *__String, regoff_t __length,
Packit 6c4009
			  regoff_t __start, struct re_registers *__regs);
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* Relates to 're_match' as 're_search_2' relates to 're_search'.  */
Packit 6c4009
extern regoff_t re_match_2 (struct re_pattern_buffer *__buffer,
Packit 6c4009
			    const char *__string1, regoff_t __length1,
Packit 6c4009
			    const char *__string2, regoff_t __length2,
Packit 6c4009
			    regoff_t __start, struct re_registers *__regs,
Packit 6c4009
			    regoff_t __stop);
Packit 6c4009
Packit 6c4009
Packit 6c4009
/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
Packit 6c4009
   ENDS.  Subsequent matches using BUFFER and REGS will use this memory
Packit 6c4009
   for recording register information.  STARTS and ENDS must be
Packit 6c4009
   allocated with malloc, and must each be at least 'NUM_REGS * sizeof
Packit 6c4009
   (regoff_t)' bytes long.
Packit 6c4009
Packit 6c4009
   If NUM_REGS == 0, then subsequent matches should allocate their own
Packit 6c4009
   register data.
Packit 6c4009
Packit 6c4009
   Unless this function is called, the first search or match using
Packit 6c4009
   BUFFER will allocate its own register data, without
Packit 6c4009
   freeing the old data.  */
Packit 6c4009
extern void re_set_registers (struct re_pattern_buffer *__buffer,
Packit 6c4009
			      struct re_registers *__regs,
Packit 6c4009
			      __re_size_t __num_regs,
Packit 6c4009
			      regoff_t *__starts, regoff_t *__ends);
Packit 6c4009
#endif	/* Use GNU */
Packit 6c4009
Packit 6c4009
#if defined _REGEX_RE_COMP || (defined _LIBC && defined __USE_MISC)
Packit 6c4009
# ifndef _CRAY
Packit 6c4009
/* 4.2 bsd compatibility.  */
Packit 6c4009
extern char *re_comp (const char *);
Packit 6c4009
extern int re_exec (const char *);
Packit 6c4009
# endif
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
/* For plain 'restrict', use glibc's __restrict if defined.
Packit 6c4009
   Otherwise, GCC 2.95 and later have "__restrict"; C99 compilers have
Packit 6c4009
   "restrict", and "configure" may have defined "restrict".
Packit 6c4009
   Other compilers use __restrict, __restrict__, and _Restrict, and
Packit 6c4009
   'configure' might #define 'restrict' to those words, so pick a
Packit 6c4009
   different name.  */
Packit 6c4009
#ifndef _Restrict_
Packit 6c4009
# if defined __restrict || 2 < __GNUC__ + (95 <= __GNUC_MINOR__)
Packit 6c4009
#  define _Restrict_ __restrict
Packit 6c4009
# elif 199901L <= __STDC_VERSION__ || defined restrict
Packit 6c4009
#  define _Restrict_ restrict
Packit 6c4009
# else
Packit 6c4009
#  define _Restrict_
Packit 6c4009
# endif
Packit 6c4009
#endif
Packit 6c4009
/* For [restrict], use glibc's __restrict_arr if available.
Packit 6c4009
   Otherwise, GCC 3.1 (not in C++ mode) and C99 support [restrict].  */
Packit 6c4009
#ifndef _Restrict_arr_
Packit 6c4009
# ifdef __restrict_arr
Packit 6c4009
#  define _Restrict_arr_ __restrict_arr
Packit 6c4009
# elif ((199901L <= __STDC_VERSION__ || 3 < __GNUC__ + (1 <= __GNUC_MINOR__)) \
Packit 6c4009
        && !defined __GNUG__)
Packit 6c4009
#  define _Restrict_arr_ _Restrict_
Packit 6c4009
# else
Packit 6c4009
#  define _Restrict_arr_
Packit 6c4009
# endif
Packit 6c4009
#endif
Packit 6c4009
Packit 6c4009
/* POSIX compatibility.  */
Packit 6c4009
extern int regcomp (regex_t *_Restrict_ __preg,
Packit 6c4009
		    const char *_Restrict_ __pattern,
Packit 6c4009
		    int __cflags);
Packit 6c4009
Packit 6c4009
extern int regexec (const regex_t *_Restrict_ __preg,
Packit 6c4009
		    const char *_Restrict_ __String, size_t __nmatch,
Packit 6c4009
		    regmatch_t __pmatch[_Restrict_arr_],
Packit 6c4009
		    int __eflags);
Packit 6c4009
Packit 6c4009
extern size_t regerror (int __errcode, const regex_t *_Restrict_ __preg,
Packit 6c4009
			char *_Restrict_ __errbuf, size_t __errbuf_size);
Packit 6c4009
Packit 6c4009
extern void regfree (regex_t *__preg);
Packit 6c4009
Packit 6c4009
Packit 6c4009
#ifdef __cplusplus
Packit 6c4009
}
Packit 6c4009
#endif	/* C++ */
Packit 6c4009
Packit 6c4009
#endif /* regex.h */