mirror/ git
1
0
Fork 0
mirror of https://github.com/git/git.git synced 2024-05-04 00:46:12 +02:00
Code

Merge branch 'ab/compat-regex' 

* ab/compat-regex:
  Fix compat/regex ANSIfication on MinGW
  autoconf: regex library detection typofix
  autoconf: don't use platform regex if it lacks REG_STARTEND
  t/t7008-grep-binary.sh: un-TODO a test that needs REG_STARTEND
  compat/regex: get rid of old-style definition
  compat/regex: define out variables only used under RE_ENABLE_I18N
  Change regerror() declaration from K&R style to ANSI C (C89)
  compat/regex: get the gawk regex engine to compile within git
  compat/regex: use the regex engine from gawk for compat

Conflicts:
	compat/regex/regex.c
This commit is contained in:
Junio C Hamano 2010-09-03 09:43:45 -07:00
parent c6babe5762 ce518bbd6c
commit cd1547d204
10 changed files with 11171 additions and 5083 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

4
Makefile
View File

@ -1900,6 +1900,10 @@ ifdef NO_EXPAT
http-walker.s http-walker.o: EXTRA_CPPFLAGS = -DNO_EXPAT
endif
ifdef NO_REGEX
compat/regex/regex.o: EXTRA_CPPFLAGS = -DGAWK -DNO_MBSUPPORT
endif
git-%$X: %.o $(GITLIBS)
$(QUIET_LINK)$(CC) $(ALL_CFLAGS) -o $@ $(ALL_LDFLAGS) $(filter %.o,$^) $(LIBS)

3884
compat/regex/regcomp.c Normal file
View File

File diff suppressed because it is too large Load Diff

4975
compat/regex/regex.c
View File

File diff suppressed because it is too large Load Diff

444
compat/regex/regex.h
View File

@ -1,70 +1,90 @@
#include <stdio.h>
#include <stddef.h>
/* Definitions for data structures and routines for the regular
expression library, version 0.12.
expression library.
Copyright (C) 1985,1989-93,1995-98,2000,2001,2002,2003,2005,2006,2008
Free Software Foundation, Inc.
This file is part of the GNU C Library.
Copyright (C) 1985, 1989, 1990, 1991, 1992, 1993 Free Software Foundation, Inc.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; either version 2, or (at your option)
any later version.
This program is distributed in the hope that it will be useful,
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 675 Mass Ave, Cambridge, MA 02139, USA. */
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, write to the Free
Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
02110-1301 USA. */
#ifndef __REGEXP_LIBRARY_H__
#define __REGEXP_LIBRARY_H__
#ifndef _REGEX_H
#define _REGEX_H 1
/* POSIX says that <sys/types.h> must be included (by the caller) before
<regex.h>. */
#ifdef VMS
/* VMS doesn't have `size_t' in <sys/types.h>, even though POSIX says it
should be there. */
#ifdef HAVE_STDDEF_H
#include <stddef.h>
#endif
#ifdef HAVE_SYS_TYPES_H
#include <sys/types.h>
#endif
#ifndef _LIBC
#define __USE_GNU 1
#endif
/* Allow the use in C++ code. */
#ifdef __cplusplus
extern "C" {
#endif
/* The following two types have to be signed and unsigned integer type
wide enough to hold a value of a pointer. For most ANSI compilers
ptrdiff_t and size_t should be likely OK. Still size of these two
types is 2 for Microsoft C. Ugh... */
typedef long int s_reg_t;
typedef unsigned long int active_reg_t;
/* The following bits are used to determine the regexp syntax we
recognize. The set/not-set meanings are chosen so that Emacs syntax
remains the value 0. The bits are given in alphabetical order, and
the definitions shifted by one from the previous bit; thus, when we
add or remove a bit, only one other definition need change. */
typedef unsigned reg_syntax_t;
typedef unsigned long int reg_syntax_t;
#ifdef __USE_GNU
/* If this bit is not set, then \ inside a bracket expression is literal.
If set, then such a \ quotes the following character. */
#define RE_BACKSLASH_ESCAPE_IN_LISTS (1)
# define RE_BACKSLASH_ESCAPE_IN_LISTS ((unsigned long int) 1)
/* If this bit is not set, then + and ? are operators, and \+ and \? are
literals.
If set, then \+ and \? are operators and + and ? are literals. */
#define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
# define RE_BK_PLUS_QM (RE_BACKSLASH_ESCAPE_IN_LISTS << 1)
/* If this bit is set, then character classes are supported. They are:
[:alpha:], [:upper:], [:lower:], [:digit:], [:alnum:], [:xdigit:],
[:space:], [:print:], [:punct:], [:graph:], and [:cntrl:].
If not set, then character classes are not supported. */
#define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
# define RE_CHAR_CLASSES (RE_BK_PLUS_QM << 1)
/* If this bit is set, then ^ and $ are always anchors (outside bracket
expressions, of course).
If this bit is not set, then it depends:
^ is an anchor if it is at the beginning of a regular
expression or after an open-group or an alternation operator;
$ is an anchor if it is at the end of a regular expression, or
before a close-group or an alternation operator.
^ is an anchor if it is at the beginning of a regular
expression or after an open-group or an alternation operator;
$ is an anchor if it is at the end of a regular expression, or
before a close-group or an alternation operator.
This bit could be (re)combined with RE_CONTEXT_INDEP_OPS, because
POSIX draft 11.2 says that * etc. in leading positions is undefined.
We already implemented a previous draft which made those constructs
invalid, though, so we haven't changed the code back. */
#define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
# define RE_CONTEXT_INDEP_ANCHORS (RE_CHAR_CLASSES << 1)
/* If this bit is set, then special characters are always special
regardless of where they are in the pattern.
@ -72,63 +92,94 @@ typedef unsigned reg_syntax_t;
some contexts; otherwise they are ordinary. Specifically,
* + ? and intervals are only special when not after the beginning,
open-group, or alternation operator. */
#define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
# define RE_CONTEXT_INDEP_OPS (RE_CONTEXT_INDEP_ANCHORS << 1)
/* If this bit is set, then *, +, ?, and { cannot be first in an re or
immediately after an alternation or begin-group operator. */
#define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
# define RE_CONTEXT_INVALID_OPS (RE_CONTEXT_INDEP_OPS << 1)
/* If this bit is set, then . matches newline.
If not set, then it doesn't. */
#define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
# define RE_DOT_NEWLINE (RE_CONTEXT_INVALID_OPS << 1)
/* If this bit is set, then . doesn't match NUL.
If not set, then it does. */
#define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
# define RE_DOT_NOT_NULL (RE_DOT_NEWLINE << 1)
/* If this bit is set, nonmatching lists [^...] do not match newline.
If not set, they do. */
#define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
# define RE_HAT_LISTS_NOT_NEWLINE (RE_DOT_NOT_NULL << 1)
/* If this bit is set, either \{...\} or {...} defines an
interval, depending on RE_NO_BK_BRACES.
If not set, \{, \}, {, and } are literals. */
#define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
# define RE_INTERVALS (RE_HAT_LISTS_NOT_NEWLINE << 1)
/* If this bit is set, +, ? and | aren't recognized as operators.
If not set, they are. */
#define RE_LIMITED_OPS (RE_INTERVALS << 1)
# define RE_LIMITED_OPS (RE_INTERVALS << 1)
/* If this bit is set, newline is an alternation operator.
If not set, newline is literal. */
#define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
# define RE_NEWLINE_ALT (RE_LIMITED_OPS << 1)
/* If this bit is set, then `{...}' defines an interval, and \{ and \}
are literals.
If not set, then `\{...\}' defines an interval. */
#define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
# define RE_NO_BK_BRACES (RE_NEWLINE_ALT << 1)
/* If this bit is set, (...) defines a group, and \( and \) are literals.
If not set, \(...\) defines a group, and ( and ) are literals. */
#define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
# define RE_NO_BK_PARENS (RE_NO_BK_BRACES << 1)
/* If this bit is set, then \<digit> matches <digit>.
If not set, then \<digit> is a back-reference. */
#define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
# define RE_NO_BK_REFS (RE_NO_BK_PARENS << 1)
/* If this bit is set, then | is an alternation operator, and \| is literal.
If not set, then \| is an alternation operator, and | is literal. */
#define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
# define RE_NO_BK_VBAR (RE_NO_BK_REFS << 1)
/* If this bit is set, then an ending range point collating higher
than the starting range point, as in [z-a], is invalid.
If not set, then when ending range point collates higher than the
starting range point, the range is ignored. */
#define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
# define RE_NO_EMPTY_RANGES (RE_NO_BK_VBAR << 1)
/* If this bit is set, then an unmatched ) is ordinary.
If not set, then an unmatched ) is invalid. */
#define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
# define RE_UNMATCHED_RIGHT_PAREN_ORD (RE_NO_EMPTY_RANGES << 1)
/* If this bit is set, succeed as soon as we match the whole pattern,
without further backtracking. */
# define RE_NO_POSIX_BACKTRACKING (RE_UNMATCHED_RIGHT_PAREN_ORD << 1)
/* If this bit is set, do not process the GNU regex operators.
If not set, then the GNU regex operators are recognized. */
# define RE_NO_GNU_OPS (RE_NO_POSIX_BACKTRACKING << 1)
/* If this bit is set, a syntactically invalid interval is treated as
a string of ordinary characters. For example, the ERE 'a{1' is
treated as 'a\{1'. */
# define RE_INVALID_INTERVAL_ORD (RE_NO_GNU_OPS << 1)
/* If this bit is set, then ignore case when matching.
If not set, then case is significant. */
# define RE_ICASE (RE_INVALID_INTERVAL_ORD << 1)
/* This bit is used internally like RE_CONTEXT_INDEP_ANCHORS but only
for ^, because it is difficult to scan the regex backwards to find
whether ^ should be special. */
# define RE_CARET_ANCHORS_HERE (RE_ICASE << 1)
/* If this bit is set, then \{ cannot be first in an bre or
immediately after an alternation or begin-group operator. */
# define RE_CONTEXT_INVALID_DUP (RE_CARET_ANCHORS_HERE << 1)
/* If this bit is set, then no_sub will be set to 1 during
re_compile_pattern. */
#define RE_NO_SUB (RE_CONTEXT_INVALID_DUP << 1)
#endif
/* This global variable defines the particular regexp syntax to use (for
some interfaces). When a regexp is compiled, the syntax used is
@ -136,6 +187,7 @@ typedef unsigned reg_syntax_t;
already-compiled regexps. */
extern reg_syntax_t re_syntax_options;
#ifdef __USE_GNU
/* Define combinations of the above bits for the standard possibilities.
(The [[[ comments delimit what gets put into the Texinfo file, so
don't delete them!) */
@ -143,13 +195,22 @@ extern reg_syntax_t re_syntax_options;
#define RE_SYNTAX_EMACS 0
#define RE_SYNTAX_AWK \
(RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \
| RE_NO_BK_PARENS | RE_NO_BK_REFS \
| RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \
| RE_UNMATCHED_RIGHT_PAREN_ORD)
(RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DOT_NOT_NULL \
| RE_NO_BK_PARENS | RE_NO_BK_REFS \
| RE_NO_BK_VBAR | RE_NO_EMPTY_RANGES \
| RE_DOT_NEWLINE | RE_CONTEXT_INDEP_ANCHORS \
| RE_UNMATCHED_RIGHT_PAREN_ORD | RE_NO_GNU_OPS)
#define RE_SYNTAX_POSIX_AWK \
(RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS)
#define RE_SYNTAX_GNU_AWK \
((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \
| RE_INVALID_INTERVAL_ORD) \
& ~(RE_DOT_NOT_NULL | RE_CONTEXT_INDEP_OPS \
| RE_CONTEXT_INVALID_OPS ))
#define RE_SYNTAX_POSIX_AWK \
(RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \
| RE_INTERVALS | RE_NO_GNU_OPS \
| RE_INVALID_INTERVAL_ORD)
#define RE_SYNTAX_GREP \
(RE_BK_PLUS_QM | RE_CHAR_CLASSES \
@ -163,7 +224,8 @@ extern reg_syntax_t re_syntax_options;
| RE_NO_BK_VBAR)
#define RE_SYNTAX_POSIX_EGREP \
(RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES)
(RE_SYNTAX_EGREP | RE_INTERVALS | RE_NO_BK_BRACES \
| RE_INVALID_INTERVAL_ORD)
/* P1003.2/D11.2, section 4.20.7.1, lines 5078ff. */
#define RE_SYNTAX_ED RE_SYNTAX_POSIX_BASIC
@ -176,7 +238,7 @@ extern reg_syntax_t re_syntax_options;
| RE_INTERVALS | RE_NO_EMPTY_RANGES)
#define RE_SYNTAX_POSIX_BASIC \
(_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM)
(_RE_SYNTAX_POSIX_COMMON | RE_BK_PLUS_QM | RE_CONTEXT_INVALID_DUP)
/* Differs from ..._POSIX_BASIC only in that RE_BK_PLUS_QM becomes
RE_LIMITED_OPS, i.e., \? \+ \| are not recognized. Actually, this
@ -185,13 +247,13 @@ extern reg_syntax_t re_syntax_options;
(_RE_SYNTAX_POSIX_COMMON | RE_LIMITED_OPS)
#define RE_SYNTAX_POSIX_EXTENDED \
(_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \
| RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \
| RE_NO_BK_PARENS | RE_NO_BK_VBAR \
| RE_UNMATCHED_RIGHT_PAREN_ORD)
(_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \
| RE_CONTEXT_INDEP_OPS | RE_NO_BK_BRACES \
| RE_NO_BK_PARENS | RE_NO_BK_VBAR \
| RE_CONTEXT_INVALID_OPS | RE_UNMATCHED_RIGHT_PAREN_ORD)
/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INVALID_OPS
replaces RE_CONTEXT_INDEP_OPS and RE_NO_BK_REFS is added. */
/* Differs from ..._POSIX_EXTENDED in that RE_CONTEXT_INDEP_OPS is
removed and RE_NO_BK_REFS is added. */
#define RE_SYNTAX_POSIX_MINIMAL_EXTENDED \
(_RE_SYNTAX_POSIX_COMMON | RE_CONTEXT_INDEP_ANCHORS \
| RE_CONTEXT_INVALID_OPS | RE_NO_BK_BRACES \
@ -202,10 +264,12 @@ extern reg_syntax_t re_syntax_options;
/* Maximum number of duplicates an interval can allow. Some systems
(erroneously) define this in other header files, but we want our
value, so remove any previous define. */
#ifdef RE_DUP_MAX
#undef RE_DUP_MAX
# ifdef RE_DUP_MAX
# undef RE_DUP_MAX
# endif
/* If sizeof(int) == 2, then ((1 << 15) - 1) overflows. */
# define RE_DUP_MAX (0x7fff)
#endif
#define RE_DUP_MAX ((1 << 15) - 1)
/* POSIX `cflags' bits (i.e., information for `regcomp'). */
@ -240,18 +304,26 @@ extern reg_syntax_t re_syntax_options;
/* Like REG_NOTBOL, except for the end-of-line. */
#define REG_NOTEOL (1 << 1)
/* Use PMATCH[0] to delimit the start and end of the search in the
buffer. */
#define REG_STARTEND (1 << 2)
/* If any error codes are removed, changed, or added, update the
`re_error_msg' table in regex.c. */
typedef enum
{
#if defined _XOPEN_SOURCE || defined __USE_XOPEN2K
REG_ENOSYS = -1, /* This will never happen for this implementation. */
#endif
REG_NOERROR = 0, /* Success. */
REG_NOMATCH, /* Didn't find a match (for regexec). */
/* POSIX regcomp return error codes. (In the order listed in the
standard.) */
REG_BADPAT, /* Invalid pattern. */
REG_ECOLLATE, /* Not implemented. */
REG_ECOLLATE, /* Inalid collating element. */
REG_ECTYPE, /* Invalid character class name. */
REG_EESCAPE, /* Trailing backslash. */
REG_ESUBREG, /* Invalid back reference. */
@ -275,85 +347,92 @@ typedef enum
compiled, the `re_nsub' field is available. All other fields are
private to the regex routines. */
#ifndef RE_TRANSLATE_TYPE
# define __RE_TRANSLATE_TYPE unsigned char *
# ifdef __USE_GNU
# define RE_TRANSLATE_TYPE __RE_TRANSLATE_TYPE
# endif
#endif
#ifdef __USE_GNU
# define __REPB_PREFIX(name) name
#else
# define __REPB_PREFIX(name) __##name
#endif
struct re_pattern_buffer
{
/* [[[begin pattern_buffer]]] */
/* Space that holds the compiled pattern. It is declared as
`unsigned char *' because its elements are
sometimes used as array indexes. */
unsigned char *buffer;
/* Space that holds the compiled pattern. It is declared as
`unsigned char *' because its elements are sometimes used as
array indexes. */
unsigned char *__REPB_PREFIX(buffer);
/* Number of bytes to which `buffer' points. */
unsigned long allocated;
/* Number of bytes to which `buffer' points. */
unsigned long int __REPB_PREFIX(allocated);
/* Number of bytes actually used in `buffer'. */
unsigned long used;
/* Number of bytes actually used in `buffer'. */
unsigned long int __REPB_PREFIX(used);
/* Syntax setting with which the pattern was compiled. */
reg_syntax_t syntax;
/* Syntax setting with which the pattern was compiled. */
reg_syntax_t __REPB_PREFIX(syntax);
/* Pointer to a fastmap, if any, otherwise zero. re_search uses
the fastmap, if there is one, to skip over impossible
starting points for matches. */
char *fastmap;
/* Pointer to a fastmap, if any, otherwise zero. re_search uses the
fastmap, if there is one, to skip over impossible starting points
for matches. */
char *__REPB_PREFIX(fastmap);
/* Either a translate table to apply to all characters before
comparing them, or zero for no translation. The translation
is applied to a pattern when it is compiled and to a string
when it is matched. */
char *translate;
/* Either a translate table to apply to all characters before
comparing them, or zero for no translation. The translation is
applied to a pattern when it is compiled and to a string when it
is matched. */
__RE_TRANSLATE_TYPE __REPB_PREFIX(translate);
/* Number of subexpressions found by the compiler. */
/* Number of subexpressions found by the compiler. */
size_t re_nsub;
/* Zero if this pattern cannot match the empty string, one else.
Well, in truth it's used only in `re_search_2', to see
whether or not we should use the fastmap, so we don't set
this absolutely perfectly; see `re_compile_fastmap' (the
`duplicate' case). */
unsigned can_be_null : 1;
/* Zero if this pattern cannot match the empty string, one else.
Well, in truth it's used only in `re_search_2', to see whether or
not we should use the fastmap, so we don't set this absolutely
perfectly; see `re_compile_fastmap' (the `duplicate' case). */
unsigned __REPB_PREFIX(can_be_null) : 1;
/* If REGS_UNALLOCATED, allocate space in the `regs' structure
for `max (RE_NREGS, re_nsub + 1)' groups.
If REGS_REALLOCATE, reallocate space if necessary.
If REGS_FIXED, use what's there. */
#define REGS_UNALLOCATED 0
#define REGS_REALLOCATE 1
#define REGS_FIXED 2
unsigned regs_allocated : 2;
/* If REGS_UNALLOCATED, allocate space in the `regs' structure
for `max (RE_NREGS, re_nsub + 1)' groups.
If REGS_REALLOCATE, reallocate space if necessary.
If REGS_FIXED, use what's there. */
#ifdef __USE_GNU
# define REGS_UNALLOCATED 0
# define REGS_REALLOCATE 1
# define REGS_FIXED 2
#endif
unsigned __REPB_PREFIX(regs_allocated) : 2;
/* Set to zero when `regex_compile' compiles a pattern; set to one
by `re_compile_fastmap' if it updates the fastmap. */
unsigned fastmap_accurate : 1;
/* Set to zero when `regex_compile' compiles a pattern; set to one
by `re_compile_fastmap' if it updates the fastmap. */
unsigned __REPB_PREFIX(fastmap_accurate) : 1;
/* If set, `re_match_2' does not return information about
subexpressions. */
unsigned no_sub : 1;
/* If set, `re_match_2' does not return information about
subexpressions. */
unsigned __REPB_PREFIX(no_sub) : 1;
/* If set, a beginning-of-line anchor doesn't match at the
beginning of the string. */
unsigned not_bol : 1;
/* If set, a beginning-of-line anchor doesn't match at the beginning
of the string. */
unsigned __REPB_PREFIX(not_bol) : 1;
/* Similarly for an end-of-line anchor. */
unsigned not_eol : 1;
/* Similarly for an end-of-line anchor. */
unsigned __REPB_PREFIX(not_eol) : 1;
/* If true, an anchor at a newline matches. */
unsigned newline_anchor : 1;
/* [[[end pattern_buffer]]] */
/* If true, an anchor at a newline matches. */
unsigned __REPB_PREFIX(newline_anchor) : 1;
};
typedef struct re_pattern_buffer regex_t;
/* search.c (search_buffer) in Emacs needs this one opcode value. It is
defined both in `regex.c' and here. */
#define RE_EXACTN_VALUE 1
/* Type for byte offsets within the string. POSIX mandates this. */
typedef int regoff_t;
#ifdef __USE_GNU
/* This is the structure we store register match data in. See
regex.texinfo for a full description of what registers match. */
struct re_registers
@ -367,8 +446,9 @@ struct re_registers
/* If `regs_allocated' is REGS_UNALLOCATED in the pattern buffer,
`re_match_2' returns information about at least this many registers
the first time a `regs' structure is passed. */
#ifndef RE_NREGS
#define RE_NREGS 30
# ifndef RE_NREGS
# define RE_NREGS 30
# endif
#endif
@ -383,38 +463,22 @@ typedef struct
/* Declarations for routines. */
/* To avoid duplicating every routine declaration -- once with a
prototype (if we are ANSI), and once without (if we aren't) -- we
use the following macro to declare argument types. This
unfortunately clutters up the declarations a bit, but I think it's
worth it. */
#if __STDC__
#define _RE_ARGS(args) args
#else /* not __STDC__ */
#define _RE_ARGS(args) ()
#endif /* not __STDC__ */
#ifdef __USE_GNU
/* Sets the current default syntax to SYNTAX, and return the old syntax.
You can also simply assign to the `re_syntax_options' variable. */
extern reg_syntax_t re_set_syntax _RE_ARGS ((reg_syntax_t syntax));
extern reg_syntax_t re_set_syntax (reg_syntax_t __syntax);
/* Compile the regular expression PATTERN, with length LENGTH
and syntax given by the global `re_syntax_options', into the buffer
BUFFER. Return NULL if successful, and an error string if not. */
extern const char *re_compile_pattern
_RE_ARGS ((const char *pattern, int length,
struct re_pattern_buffer *buffer));
extern const char *re_compile_pattern (const char *__pattern, size_t __length,
struct re_pattern_buffer *__buffer);
/* Compile a fastmap for the compiled pattern in BUFFER; used to
accelerate searches. Return 0 if successful and -2 if was an
internal error. */
extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer));
extern int re_compile_fastmap (struct re_pattern_buffer *__buffer);
/* Search in the string STRING (with length LENGTH) for the pattern
@ -422,31 +486,30 @@ extern int re_compile_fastmap _RE_ARGS ((struct re_pattern_buffer *buffer));
characters. Return the starting position of the match, -1 for no
match, or -2 for an internal error. Also return register
information in REGS (if REGS and BUFFER->no_sub are nonzero). */
extern int re_search
_RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
int length, int start, int range, struct re_registers *regs));
extern int re_search (struct re_pattern_buffer *__buffer, const char *__cstring,
int __length, int __start, int __range,
struct re_registers *__regs);
/* Like `re_search', but search in the concatenation of STRING1 and
STRING2. Also, stop searching at index START + STOP. */
extern int re_search_2
_RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
int length1, const char *string2, int length2,
int start, int range, struct re_registers *regs, int stop));
extern int re_search_2 (struct re_pattern_buffer *__buffer,
const char *__string1, int __length1,
const char *__string2, int __length2, int __start,
int __range, struct re_registers *__regs, int __stop);
/* Like `re_search', but return how many characters in STRING the regexp
in BUFFER matched, starting at position START. */
extern int re_match
_RE_ARGS ((struct re_pattern_buffer *buffer, const char *string,
int length, int start, struct re_registers *regs));
extern int re_match (struct re_pattern_buffer *__buffer, const char *__cstring,
int __length, int __start, struct re_registers *__regs);
/* Relates to `re_match' as `re_search_2' relates to `re_search'. */
extern int re_match_2
_RE_ARGS ((struct re_pattern_buffer *buffer, const char *string1,
int length1, const char *string2, int length2,
int start, struct re_registers *regs, int stop));
extern int re_match_2 (struct re_pattern_buffer *__buffer,
const char *__string1, int __length1,
const char *__string2, int __length2, int __start,
struct re_registers *__regs, int __stop);
/* Set REGS to hold NUM_REGS registers, storing them in STARTS and
@ -461,30 +524,59 @@ extern int re_match_2
Unless this function is called, the first search or match using
PATTERN_BUFFER will allocate its own register data, without
freeing the old data. */
extern void re_set_registers
_RE_ARGS ((struct re_pattern_buffer *buffer, struct re_registers *regs,
unsigned num_regs, regoff_t *starts, regoff_t *ends));
extern void re_set_registers (struct re_pattern_buffer *__buffer,
struct re_registers *__regs,
unsigned int __num_regs,
regoff_t *__starts, regoff_t *__ends);
#endif /* Use GNU */
#if defined _REGEX_RE_COMP || (defined _LIBC && defined __USE_BSD)
# ifndef _CRAY
/* 4.2 bsd compatibility. */
extern char *re_comp _RE_ARGS ((const char *));
extern int re_exec _RE_ARGS ((const char *));
extern char *re_comp (const char *);
extern int re_exec (const char *);
# endif
#endif
/* GCC 2.95 and later have "__restrict"; C99 compilers have
"restrict", and "configure" may have defined "restrict". */
#ifndef __restrict
# if ! (2 < __GNUC__ || (2 == __GNUC__ && 95 <= __GNUC_MINOR__))
# if defined restrict || 199901L <= __STDC_VERSION__
# define __restrict restrict
# else
# define __restrict
# endif
# endif
#endif
/* gcc 3.1 and up support the [restrict] syntax. */
#ifndef __restrict_arr
# if (__GNUC__ > 3 || (__GNUC__ == 3 && __GNUC_MINOR__ >= 1)) \
&& !defined __GNUG__
# define __restrict_arr __restrict
# else
# define __restrict_arr
# endif
#endif
/* POSIX compatibility. */
extern int regcomp _RE_ARGS ((regex_t *preg, const char *pattern, int cflags));
extern int regexec
_RE_ARGS ((const regex_t *preg, const char *string, size_t nmatch,
regmatch_t pmatch[], int eflags));
extern size_t regerror
_RE_ARGS ((int errcode, const regex_t *preg, char *errbuf,
size_t errbuf_size));
extern void regfree _RE_ARGS ((regex_t *preg));
extern int regcomp (regex_t *__restrict __preg,
const char *__restrict __pattern,
int __cflags);
#endif /* not __REGEXP_LIBRARY_H__ */
/*
Local variables:
make-backup-files: t
version-control: t
trim-versions-without-asking: nil
End:
*/
extern int regexec (const regex_t *__restrict __preg,
const char *__restrict __cstring, size_t __nmatch,
regmatch_t __pmatch[__restrict_arr],
int __eflags);
extern size_t regerror (int __errcode, const regex_t *__restrict __preg,
char *__restrict __errbuf, size_t __errbuf_size);
extern void regfree (regex_t *__preg);
#ifdef __cplusplus
}
#endif /* C++ */
#endif /* regex.h */

1744
compat/regex/regex_internal.c Normal file
View File

@ -0,0 +1,1744 @@
/* Extended regular expression matching and search library.
Copyright (C) 2002-2006, 2010 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, write to the Free
Software Foundation, Inc., 51 Franklin Street, Fifth Floor, Boston, MA
02110-1301 USA. */
static void re_string_construct_common (const char *str, int len,
re_string_t *pstr,
RE_TRANSLATE_TYPE trans, int icase,
const re_dfa_t *dfa) internal_function;
static re_dfastate_t *create_ci_newstate (const re_dfa_t *dfa,
const re_node_set *nodes,
unsigned int hash) internal_function;
static re_dfastate_t *create_cd_newstate (const re_dfa_t *dfa,
const re_node_set *nodes,
unsigned int context,
unsigned int hash) internal_function;
#ifdef GAWK
#undef MAX /* safety */
static int
MAX(size_t a, size_t b)
{
return (a > b ? a : b);
}
#endif
/* Functions for string operation. */
/* This function allocate the buffers. It is necessary to call
re_string_reconstruct before using the object. */
static reg_errcode_t
internal_function
re_string_allocate (re_string_t *pstr, const char *str, int len, int init_len,
RE_TRANSLATE_TYPE trans, int icase, const re_dfa_t *dfa)
{
reg_errcode_t ret;
int init_buf_len;
/* Ensure at least one character fits into the buffers. */
if (init_len < dfa->mb_cur_max)
init_len = dfa->mb_cur_max;
init_buf_len = (len + 1 < init_len) ? len + 1: init_len;
re_string_construct_common (str, len, pstr, trans, icase, dfa);
ret = re_string_realloc_buffers (pstr, init_buf_len);
if (BE (ret != REG_NOERROR, 0))
return ret;
pstr->word_char = dfa->word_char;
pstr->word_ops_used = dfa->word_ops_used;
pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str;
pstr->valid_len = (pstr->mbs_allocated || dfa->mb_cur_max > 1) ? 0 : len;
pstr->valid_raw_len = pstr->valid_len;
return REG_NOERROR;
}
/* This function allocate the buffers, and initialize them. */
static reg_errcode_t
internal_function
re_string_construct (re_string_t *pstr, const char *str, int len,
RE_TRANSLATE_TYPE trans, int icase, const re_dfa_t *dfa)
{
reg_errcode_t ret;
memset (pstr, '\0', sizeof (re_string_t));
re_string_construct_common (str, len, pstr, trans, icase, dfa);
if (len > 0)
{
ret = re_string_realloc_buffers (pstr, len + 1);
if (BE (ret != REG_NOERROR, 0))
return ret;
}
pstr->mbs = pstr->mbs_allocated ? pstr->mbs : (unsigned char *) str;
if (icase)
{
#ifdef RE_ENABLE_I18N
if (dfa->mb_cur_max > 1)
{
while (1)
{
ret = build_wcs_upper_buffer (pstr);
if (BE (ret != REG_NOERROR, 0))
return ret;
if (pstr->valid_raw_len >= len)
break;
if (pstr->bufs_len > pstr->valid_len + dfa->mb_cur_max)
break;
ret = re_string_realloc_buffers (pstr, pstr->bufs_len * 2);
if (BE (ret != REG_NOERROR, 0))
return ret;
}
}
else
#endif /* RE_ENABLE_I18N */
build_upper_buffer (pstr);
}
else
{
#ifdef RE_ENABLE_I18N
if (dfa->mb_cur_max > 1)
build_wcs_buffer (pstr);
else
#endif /* RE_ENABLE_I18N */
{
if (trans != NULL)
re_string_translate_buffer (pstr);
else
{
pstr->valid_len = pstr->bufs_len;
pstr->valid_raw_len = pstr->bufs_len;
}
}
}
return REG_NOERROR;
}
/* Helper functions for re_string_allocate, and re_string_construct. */
static reg_errcode_t
internal_function
re_string_realloc_buffers (re_string_t *pstr, int new_buf_len)
{
#ifdef RE_ENABLE_I18N
if (pstr->mb_cur_max > 1)
{
wint_t *new_wcs;
/* Avoid overflow in realloc. */
const size_t max_object_size = MAX (sizeof (wint_t), sizeof (int));
if (BE (SIZE_MAX / max_object_size < new_buf_len, 0))
return REG_ESPACE;
new_wcs = re_realloc (pstr->wcs, wint_t, new_buf_len);
if (BE (new_wcs == NULL, 0))
return REG_ESPACE;
pstr->wcs = new_wcs;
if (pstr->offsets != NULL)
{
int *new_offsets = re_realloc (pstr->offsets, int, new_buf_len);
if (BE (new_offsets == NULL, 0))
return REG_ESPACE;
pstr->offsets = new_offsets;
}
}
#endif /* RE_ENABLE_I18N */
if (pstr->mbs_allocated)
{
unsigned char *new_mbs = re_realloc (pstr->mbs, unsigned char,
new_buf_len);
if (BE (new_mbs == NULL, 0))
return REG_ESPACE;
pstr->mbs = new_mbs;
}
pstr->bufs_len = new_buf_len;
return REG_NOERROR;
}
static void
internal_function
re_string_construct_common (const char *str, int len, re_string_t *pstr,
RE_TRANSLATE_TYPE trans, int icase,
const re_dfa_t *dfa)
{
pstr->raw_mbs = (const unsigned char *) str;
pstr->len = len;
pstr->raw_len = len;
pstr->trans = trans;
pstr->icase = icase ? 1 : 0;
pstr->mbs_allocated = (trans != NULL || icase);
pstr->mb_cur_max = dfa->mb_cur_max;
pstr->is_utf8 = dfa->is_utf8;
pstr->map_notascii = dfa->map_notascii;
pstr->stop = pstr->len;
pstr->raw_stop = pstr->stop;
}
#ifdef RE_ENABLE_I18N
/* Build wide character buffer PSTR->WCS.
If the byte sequence of the string are:
<mb1>(0), <mb1>(1), <mb2>(0), <mb2>(1), <sb3>
Then wide character buffer will be:
<wc1> , WEOF , <wc2> , WEOF , <wc3>
We use WEOF for padding, they indicate that the position isn't
a first byte of a multibyte character.
Note that this function assumes PSTR->VALID_LEN elements are already
built and starts from PSTR->VALID_LEN. */
static void
internal_function
build_wcs_buffer (re_string_t *pstr)
{
#ifdef _LIBC
unsigned char buf[MB_LEN_MAX];
assert (MB_LEN_MAX >= pstr->mb_cur_max);
#else
unsigned char buf[64];
#endif
mbstate_t prev_st;
int byte_idx, end_idx, remain_len;
size_t mbclen;
/* Build the buffers from pstr->valid_len to either pstr->len or
pstr->bufs_len. */
end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
for (byte_idx = pstr->valid_len; byte_idx < end_idx;)
{
wchar_t wc;
const char *p;
remain_len = end_idx - byte_idx;
prev_st = pstr->cur_state;
/* Apply the translation if we need. */
if (BE (pstr->trans != NULL, 0))
{
int i, ch;
for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i)
{
ch = pstr->raw_mbs [pstr->raw_mbs_idx + byte_idx + i];
buf[i] = pstr->mbs[byte_idx + i] = pstr->trans[ch];
}
p = (const char *) buf;
}
else
p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx;
mbclen = __mbrtowc (&wc, p, remain_len, &pstr->cur_state);
if (BE (mbclen == (size_t) -2, 0))
{
/* The buffer doesn't have enough space, finish to build. */
pstr->cur_state = prev_st;
break;
}
else if (BE (mbclen == (size_t) -1 || mbclen == 0, 0))
{
/* We treat these cases as a singlebyte character. */
mbclen = 1;
wc = (wchar_t) pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
if (BE (pstr->trans != NULL, 0))
wc = pstr->trans[wc];
pstr->cur_state = prev_st;
}
/* Write wide character and padding. */
pstr->wcs[byte_idx++] = wc;
/* Write paddings. */
for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
pstr->wcs[byte_idx++] = WEOF;
}
pstr->valid_len = byte_idx;
pstr->valid_raw_len = byte_idx;
}
/* Build wide character buffer PSTR->WCS like build_wcs_buffer,
but for REG_ICASE. */
static reg_errcode_t
internal_function
build_wcs_upper_buffer (re_string_t *pstr)
{
mbstate_t prev_st;
int src_idx, byte_idx, end_idx, remain_len;
size_t mbclen;
#ifdef _LIBC
char buf[MB_LEN_MAX];
assert (MB_LEN_MAX >= pstr->mb_cur_max);
#else
char buf[64];
#endif
byte_idx = pstr->valid_len;
end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
/* The following optimization assumes that ASCII characters can be
mapped to wide characters with a simple cast. */
if (! pstr->map_notascii && pstr->trans == NULL && !pstr->offsets_needed)
{
while (byte_idx < end_idx)
{
wchar_t wc;
if (isascii (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx])
&& mbsinit (&pstr->cur_state))
{
/* In case of a singlebyte character. */
pstr->mbs[byte_idx]
= toupper (pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx]);
/* The next step uses the assumption that wchar_t is encoded
ASCII-safe: all ASCII values can be converted like this. */
pstr->wcs[byte_idx] = (wchar_t) pstr->mbs[byte_idx];
++byte_idx;
continue;
}
remain_len = end_idx - byte_idx;
prev_st = pstr->cur_state;
mbclen = __mbrtowc (&wc,
((const char *) pstr->raw_mbs + pstr->raw_mbs_idx
+ byte_idx), remain_len, &pstr->cur_state);
if (BE (mbclen + 2 > 2, 1))
{
wchar_t wcu = wc;
if (iswlower (wc))
{
size_t mbcdlen;
wcu = towupper (wc);
mbcdlen = wcrtomb (buf, wcu, &prev_st);
if (BE (mbclen == mbcdlen, 1))
memcpy (pstr->mbs + byte_idx, buf, mbclen);
else
{
src_idx = byte_idx;
goto offsets_needed;
}
}
else
memcpy (pstr->mbs + byte_idx,
pstr->raw_mbs + pstr->raw_mbs_idx + byte_idx, mbclen);
pstr->wcs[byte_idx++] = wcu;
/* Write paddings. */
for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
pstr->wcs[byte_idx++] = WEOF;
}
else if (mbclen == (size_t) -1 || mbclen == 0)
{
/* It is an invalid character or '\0'. Just use the byte. */
int ch = pstr->raw_mbs[pstr->raw_mbs_idx + byte_idx];
pstr->mbs[byte_idx] = ch;
/* And also cast it to wide char. */
pstr->wcs[byte_idx++] = (wchar_t) ch;
if (BE (mbclen == (size_t) -1, 0))
pstr->cur_state = prev_st;
}
else
{
/* The buffer doesn't have enough space, finish to build. */
pstr->cur_state = prev_st;
break;
}
}
pstr->valid_len = byte_idx;
pstr->valid_raw_len = byte_idx;
return REG_NOERROR;
}
else
for (src_idx = pstr->valid_raw_len; byte_idx < end_idx;)
{
wchar_t wc;
const char *p;
offsets_needed:
remain_len = end_idx - byte_idx;
prev_st = pstr->cur_state;
if (BE (pstr->trans != NULL, 0))
{
int i, ch;
for (i = 0; i < pstr->mb_cur_max && i < remain_len; ++i)
{
ch = pstr->raw_mbs [pstr->raw_mbs_idx + src_idx + i];
buf[i] = pstr->trans[ch];
}
p = (const char *) buf;
}
else
p = (const char *) pstr->raw_mbs + pstr->raw_mbs_idx + src_idx;
mbclen = __mbrtowc (&wc, p, remain_len, &pstr->cur_state);
if (BE (mbclen + 2 > 2, 1))
{
wchar_t wcu = wc;
if (iswlower (wc))
{
size_t mbcdlen;
wcu = towupper (wc);
mbcdlen = wcrtomb ((char *) buf, wcu, &prev_st);
if (BE (mbclen == mbcdlen, 1))
memcpy (pstr->mbs + byte_idx, buf, mbclen);
else if (mbcdlen != (size_t) -1)
{
size_t i;
if (byte_idx + mbcdlen > pstr->bufs_len)
{
pstr->cur_state = prev_st;
break;
}
if (pstr->offsets == NULL)
{
pstr->offsets = re_malloc (int, pstr->bufs_len);
if (pstr->offsets == NULL)
return REG_ESPACE;
}
if (!pstr->offsets_needed)
{
for (i = 0; i < (size_t) byte_idx; ++i)
pstr->offsets[i] = i;
pstr->offsets_needed = 1;
}
memcpy (pstr->mbs + byte_idx, buf, mbcdlen);
pstr->wcs[byte_idx] = wcu;
pstr->offsets[byte_idx] = src_idx;
for (i = 1; i < mbcdlen; ++i)
{
pstr->offsets[byte_idx + i]
= src_idx + (i < mbclen ? i : mbclen - 1);
pstr->wcs[byte_idx + i] = WEOF;
}
pstr->len += mbcdlen - mbclen;
if (pstr->raw_stop > src_idx)
pstr->stop += mbcdlen - mbclen;
end_idx = (pstr->bufs_len > pstr->len)
? pstr->len : pstr->bufs_len;
byte_idx += mbcdlen;
src_idx += mbclen;
continue;
}
else
memcpy (pstr->mbs + byte_idx, p, mbclen);
}
else
memcpy (pstr->mbs + byte_idx, p, mbclen);
if (BE (pstr->offsets_needed != 0, 0))
{
size_t i;
for (i = 0; i < mbclen; ++i)
pstr->offsets[byte_idx + i] = src_idx + i;
}
src_idx += mbclen;
pstr->wcs[byte_idx++] = wcu;
/* Write paddings. */
for (remain_len = byte_idx + mbclen - 1; byte_idx < remain_len ;)
pstr->wcs[byte_idx++] = WEOF;
}
else if (mbclen == (size_t) -1 || mbclen == 0)
{
/* It is an invalid character or '\0'. Just use the byte. */
int ch = pstr->raw_mbs[pstr->raw_mbs_idx + src_idx];
if (BE (pstr->trans != NULL, 0))
ch = pstr->trans [ch];
pstr->mbs[byte_idx] = ch;
if (BE (pstr->offsets_needed != 0, 0))
pstr->offsets[byte_idx] = src_idx;
++src_idx;
/* And also cast it to wide char. */
pstr->wcs[byte_idx++] = (wchar_t) ch;
if (BE (mbclen == (size_t) -1, 0))
pstr->cur_state = prev_st;
}
else
{
/* The buffer doesn't have enough space, finish to build. */
pstr->cur_state = prev_st;
break;
}
}
pstr->valid_len = byte_idx;
pstr->valid_raw_len = src_idx;
return REG_NOERROR;
}
/* Skip characters until the index becomes greater than NEW_RAW_IDX.
Return the index. */
static int
internal_function
re_string_skip_chars (re_string_t *pstr, int new_raw_idx, wint_t *last_wc)
{
mbstate_t prev_st;
int rawbuf_idx;
size_t mbclen;
wint_t wc = WEOF;
/* Skip the characters which are not necessary to check. */
for (rawbuf_idx = pstr->raw_mbs_idx + pstr->valid_raw_len;
rawbuf_idx < new_raw_idx;)
{
wchar_t wc2;
int remain_len = pstr->len - rawbuf_idx;
prev_st = pstr->cur_state;
mbclen = __mbrtowc (&wc2, (const char *) pstr->raw_mbs + rawbuf_idx,
remain_len, &pstr->cur_state);
if (BE (mbclen == (size_t) -2 || mbclen == (size_t) -1 || mbclen == 0, 0))
{
/* We treat these cases as a single byte character. */
if (mbclen == 0 || remain_len == 0)
wc = L'\0';
else
wc = *(unsigned char *) (pstr->raw_mbs + rawbuf_idx);
mbclen = 1;
pstr->cur_state = prev_st;
}
else
wc = (wint_t) wc2;
/* Then proceed the next character. */
rawbuf_idx += mbclen;
}
*last_wc = (wint_t) wc;
return rawbuf_idx;
}
#endif /* RE_ENABLE_I18N */
/* Build the buffer PSTR->MBS, and apply the translation if we need.
This function is used in case of REG_ICASE. */
static void
internal_function
build_upper_buffer (re_string_t *pstr)
{
int char_idx, end_idx;
end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
for (char_idx = pstr->valid_len; char_idx < end_idx; ++char_idx)
{
int ch = pstr->raw_mbs[pstr->raw_mbs_idx + char_idx];
if (BE (pstr->trans != NULL, 0))
ch = pstr->trans[ch];
if (islower (ch))
pstr->mbs[char_idx] = toupper (ch);
else
pstr->mbs[char_idx] = ch;
}
pstr->valid_len = char_idx;
pstr->valid_raw_len = char_idx;
}
/* Apply TRANS to the buffer in PSTR. */
static void
internal_function
re_string_translate_buffer (re_string_t *pstr)
{
int buf_idx, end_idx;
end_idx = (pstr->bufs_len > pstr->len) ? pstr->len : pstr->bufs_len;
for (buf_idx = pstr->valid_len; buf_idx < end_idx; ++buf_idx)
{
int ch = pstr->raw_mbs[pstr->raw_mbs_idx + buf_idx];
pstr->mbs[buf_idx] = pstr->trans[ch];
}
pstr->valid_len = buf_idx;
pstr->valid_raw_len = buf_idx;
}
/* This function re-construct the buffers.
Concretely, convert to wide character in case of pstr->mb_cur_max > 1,
convert to upper case in case of REG_ICASE, apply translation. */
static reg_errcode_t
internal_function
re_string_reconstruct (re_string_t *pstr, int idx, int eflags)
{
int offset = idx - pstr->raw_mbs_idx;
if (BE (offset < 0, 0))
{
/* Reset buffer. */
#ifdef RE_ENABLE_I18N
if (pstr->mb_cur_max > 1)
memset (&pstr->cur_state, '\0', sizeof (mbstate_t));
#endif /* RE_ENABLE_I18N */
pstr->len = pstr->raw_len;
pstr->stop = pstr->raw_stop;
pstr->valid_len = 0;
pstr->raw_mbs_idx = 0;
pstr->valid_raw_len = 0;
pstr->offsets_needed = 0;
pstr->tip_context = ((eflags & REG_NOTBOL) ? CONTEXT_BEGBUF
: CONTEXT_NEWLINE | CONTEXT_BEGBUF);
if (!pstr->mbs_allocated)
pstr->mbs = (unsigned char *) pstr->raw_mbs;
offset = idx;
}
if (BE (offset != 0, 1))
{
/* Should the already checked characters be kept? */
if (BE (offset < pstr->valid_raw_len, 1))
{
/* Yes, move them to the front of the buffer. */
#ifdef RE_ENABLE_I18N
if (BE (pstr->offsets_needed, 0))
{
int low = 0, high = pstr->valid_len, mid;
do
{
mid = (high + low) / 2;
if (pstr->offsets[mid] > offset)
high = mid;
else if (pstr->offsets[mid] < offset)
low = mid + 1;
else
break;
}
while (low < high);
if (pstr->offsets[mid] < offset)
++mid;
pstr->tip_context = re_string_context_at (pstr, mid - 1,
eflags);
/* This can be quite complicated, so handle specially
only the common and easy case where the character with
different length representation of lower and upper
case is present at or after offset. */
if (pstr->valid_len > offset
&& mid == offset && pstr->offsets[mid] == offset)
{
memmove (pstr->wcs, pstr->wcs + offset,
(pstr->valid_len - offset) * sizeof (wint_t));
memmove (pstr->mbs, pstr->mbs + offset, pstr->valid_len - offset);
pstr->valid_len -= offset;
pstr->valid_raw_len -= offset;
for (low = 0; low < pstr->valid_len; low++)
pstr->offsets[low] = pstr->offsets[low + offset] - offset;
}
else
{
/* Otherwise, just find out how long the partial multibyte
character at offset is and fill it with WEOF/255. */
pstr->len = pstr->raw_len - idx + offset;
pstr->stop = pstr->raw_stop - idx + offset;
pstr->offsets_needed = 0;
while (mid > 0 && pstr->offsets[mid - 1] == offset)
--mid;
while (mid < pstr->valid_len)
if (pstr->wcs[mid] != WEOF)
break;
else
++mid;
if (mid == pstr->valid_len)
pstr->valid_len = 0;
else
{
pstr->valid_len = pstr->offsets[mid] - offset;
if (pstr->valid_len)
{
for (low = 0; low < pstr->valid_len; ++low)
pstr->wcs[low] = WEOF;
memset (pstr->mbs, 255, pstr->valid_len);
}
}
pstr->valid_raw_len = pstr->valid_len;
}
}
else
#endif
{
pstr->tip_context = re_string_context_at (pstr, offset - 1,
eflags);
#ifdef RE_ENABLE_I18N
if (pstr->mb_cur_max > 1)
memmove (pstr->wcs, pstr->wcs + offset,
(pstr->valid_len - offset) * sizeof (wint_t));
#endif /* RE_ENABLE_I18N */
if (BE (pstr->mbs_allocated, 0))
memmove (pstr->mbs, pstr->mbs + offset,
pstr->valid_len - offset);
pstr->valid_len -= offset;
pstr->valid_raw_len -= offset;
#if DEBUG
assert (pstr->valid_len > 0);
#endif
}
}
else
{
#ifdef RE_ENABLE_I18N
/* No, skip all characters until IDX. */
int prev_valid_len = pstr->valid_len;
if (BE (pstr->offsets_needed, 0))
{
pstr->len = pstr->raw_len - idx + offset;
pstr->stop = pstr->raw_stop - idx + offset;
pstr->offsets_needed = 0;
}
#endif
pstr->valid_len = 0;
#ifdef RE_ENABLE_I18N
if (pstr->mb_cur_max > 1)
{
int wcs_idx;
wint_t wc = WEOF;
if (pstr->is_utf8)
{
const unsigned char *raw, *p, *end;
/* Special case UTF-8. Multi-byte chars start with any
byte other than 0x80 - 0xbf. */
raw = pstr->raw_mbs + pstr->raw_mbs_idx;
end = raw + (offset - pstr->mb_cur_max);
if (end < pstr->raw_mbs)
end = pstr->raw_mbs;
p = raw + offset - 1;
#ifdef _LIBC
/* We know the wchar_t encoding is UCS4, so for the simple
case, ASCII characters, skip the conversion step. */
if (isascii (*p) && BE (pstr->trans == NULL, 1))
{
memset (&pstr->cur_state, '\0', sizeof (mbstate_t));
/* pstr->valid_len = 0; */
wc = (wchar_t) *p;
}
else
#endif
for (; p >= end; --p)
if ((*p & 0xc0) != 0x80)
{
mbstate_t cur_state;
wchar_t wc2;
int mlen = raw + pstr->len - p;
unsigned char buf[6];
size_t mbclen;
if (BE (pstr->trans != NULL, 0))
{
int i = mlen < 6 ? mlen : 6;
while (--i >= 0)
buf[i] = pstr->trans[p[i]];
}
/* XXX Don't use mbrtowc, we know which conversion
to use (UTF-8 -> UCS4). */
memset (&cur_state, 0, sizeof (cur_state));
mbclen = __mbrtowc (&wc2, (const char *) p, mlen,
&cur_state);
if (raw + offset - p <= mbclen
&& mbclen < (size_t) -2)
{
memset (&pstr->cur_state, '\0',
sizeof (mbstate_t));
pstr->valid_len = mbclen - (raw + offset - p);
wc = wc2;
}
break;
}
}
if (wc == WEOF)
pstr->valid_len = re_string_skip_chars (pstr, idx, &wc) - idx;
if (wc == WEOF)
pstr->tip_context
= re_string_context_at (pstr, prev_valid_len - 1, eflags);
else
pstr->tip_context = ((BE (pstr->word_ops_used != 0, 0)
&& IS_WIDE_WORD_CHAR (wc))
? CONTEXT_WORD
: ((IS_WIDE_NEWLINE (wc)
&& pstr->newline_anchor)
? CONTEXT_NEWLINE : 0));
if (BE (pstr->valid_len, 0))
{
for (wcs_idx = 0; wcs_idx < pstr->valid_len; ++wcs_idx)
pstr->wcs[wcs_idx] = WEOF;
if (pstr->mbs_allocated)
memset (pstr->mbs, 255, pstr->valid_len);
}
pstr->valid_raw_len = pstr->valid_len;
}
else
#endif /* RE_ENABLE_I18N */
{
int c = pstr->raw_mbs[pstr->raw_mbs_idx + offset - 1];
pstr->valid_raw_len = 0;
if (pstr->trans)
c = pstr->trans[c];
pstr->tip_context = (bitset_contain (pstr->word_char, c)
? CONTEXT_WORD
: ((IS_NEWLINE (c) && pstr->newline_anchor)
? CONTEXT_NEWLINE : 0));
}
}
if (!BE (pstr->mbs_allocated, 0))
pstr->mbs += offset;
}
pstr->raw_mbs_idx = idx;
pstr->len -= offset;
pstr->stop -= offset;
/* Then build the buffers. */
#ifdef RE_ENABLE_I18N
if (pstr->mb_cur_max > 1)
{
if (pstr->icase)
{
reg_errcode_t ret = build_wcs_upper_buffer (pstr);
if (BE (ret != REG_NOERROR, 0))
return ret;
}
else
build_wcs_buffer (pstr);
}
else
#endif /* RE_ENABLE_I18N */
if (BE (pstr->mbs_allocated, 0))
{
if (pstr->icase)
build_upper_buffer (pstr);
else if (pstr->trans != NULL)
re_string_translate_buffer (pstr);
}
else
pstr->valid_len = pstr->len;
pstr->cur_idx = 0;
return REG_NOERROR;
}
static unsigned char
internal_function __attribute ((pure))
re_string_peek_byte_case (const re_string_t *pstr, int idx)
{
int ch, off;
/* Handle the common (easiest) cases first. */
if (BE (!pstr->mbs_allocated, 1))
return re_string_peek_byte (pstr, idx);
#ifdef RE_ENABLE_I18N
if (pstr->mb_cur_max > 1
&& ! re_string_is_single_byte_char (pstr, pstr->cur_idx + idx))
return re_string_peek_byte (pstr, idx);
#endif
off = pstr->cur_idx + idx;
#ifdef RE_ENABLE_I18N
if (pstr->offsets_needed)
off = pstr->offsets[off];
#endif
ch = pstr->raw_mbs[pstr->raw_mbs_idx + off];
#ifdef RE_ENABLE_I18N
/* Ensure that e.g. for tr_TR.UTF-8 BACKSLASH DOTLESS SMALL LETTER I
this function returns CAPITAL LETTER I instead of first byte of
DOTLESS SMALL LETTER I. The latter would confuse the parser,
since peek_byte_case doesn't advance cur_idx in any way. */
if (pstr->offsets_needed && !isascii (ch))
return re_string_peek_byte (pstr, idx);
#endif
return ch;
}
static unsigned char
internal_function __attribute ((pure))
re_string_fetch_byte_case (re_string_t *pstr)
{
if (BE (!pstr->mbs_allocated, 1))
return re_string_fetch_byte (pstr);
#ifdef RE_ENABLE_I18N
if (pstr->offsets_needed)
{
int off, ch;
/* For tr_TR.UTF-8 [[:islower:]] there is
[[: CAPITAL LETTER I WITH DOT lower:]] in mbs. Skip
in that case the whole multi-byte character and return
the original letter. On the other side, with
[[: DOTLESS SMALL LETTER I return [[:I, as doing
anything else would complicate things too much. */
if (!re_string_first_byte (pstr, pstr->cur_idx))
return re_string_fetch_byte (pstr);
off = pstr->offsets[pstr->cur_idx];
ch = pstr->raw_mbs[pstr->raw_mbs_idx + off];
if (! isascii (ch))
return re_string_fetch_byte (pstr);
re_string_skip_bytes (pstr,
re_string_char_size_at (pstr, pstr->cur_idx));
return ch;
}
#endif
return pstr->raw_mbs[pstr->raw_mbs_idx + pstr->cur_idx++];
}
static void
internal_function
re_string_destruct (re_string_t *pstr)
{
#ifdef RE_ENABLE_I18N
re_free (pstr->wcs);
re_free (pstr->offsets);
#endif /* RE_ENABLE_I18N */
if (pstr->mbs_allocated)
re_free (pstr->mbs);
}
/* Return the context at IDX in INPUT. */
static unsigned int
internal_function
re_string_context_at (const re_string_t *input, int idx, int eflags)
{
int c;
if (BE (idx < 0, 0))
/* In this case, we use the value stored in input->tip_context,
since we can't know the character in input->mbs[-1] here. */
return input->tip_context;
if (BE (idx == input->len, 0))
return ((eflags & REG_NOTEOL) ? CONTEXT_ENDBUF
: CONTEXT_NEWLINE | CONTEXT_ENDBUF);
#ifdef RE_ENABLE_I18N
if (input->mb_cur_max > 1)
{
wint_t wc;
int wc_idx = idx;
while(input->wcs[wc_idx] == WEOF)
{
#ifdef DEBUG
/* It must not happen. */
assert (wc_idx >= 0);
#endif
--wc_idx;
if (wc_idx < 0)
return input->tip_context;
}
wc = input->wcs[wc_idx];
if (BE (input->word_ops_used != 0, 0) && IS_WIDE_WORD_CHAR (wc))
return CONTEXT_WORD;
return (IS_WIDE_NEWLINE (wc) && input->newline_anchor
? CONTEXT_NEWLINE : 0);
}
else
#endif
{
c = re_string_byte_at (input, idx);
if (bitset_contain (input->word_char, c))
return CONTEXT_WORD;
return IS_NEWLINE (c) && input->newline_anchor ? CONTEXT_NEWLINE : 0;
}
}
/* Functions for set operation. */
static reg_errcode_t
internal_function
re_node_set_alloc (re_node_set *set, int size)
{
/*
* ADR: valgrind says size can be 0, which then doesn't
* free the block of size 0. Harumph. This seems
* to work ok, though.
*/
if (size == 0)
{
memset(set, 0, sizeof(*set));
return REG_NOERROR;
}
set->alloc = size;
set->nelem = 0;
set->elems = re_malloc (int, size);
if (BE (set->elems == NULL, 0))
return REG_ESPACE;
return REG_NOERROR;
}
static reg_errcode_t
internal_function
re_node_set_init_1 (re_node_set *set, int elem)
{
set->alloc = 1;
set->nelem = 1;
set->elems = re_malloc (int, 1);
if (BE (set->elems == NULL, 0))
{
set->alloc = set->nelem = 0;
return REG_ESPACE;
}
set->elems[0] = elem;
return REG_NOERROR;
}
static reg_errcode_t
internal_function
re_node_set_init_2 (re_node_set *set, int elem1, int elem2)
{
set->alloc = 2;
set->elems = re_malloc (int, 2);
if (BE (set->elems == NULL, 0))
return REG_ESPACE;
if (elem1 == elem2)
{
set->nelem = 1;
set->elems[0] = elem1;
}
else
{
set->nelem = 2;
if (elem1 < elem2)
{
set->elems[0] = elem1;
set->elems[1] = elem2;
}
else
{
set->elems[0] = elem2;
set->elems[1] = elem1;
}
}
return REG_NOERROR;
}
static reg_errcode_t
internal_function
re_node_set_init_copy (re_node_set *dest, const re_node_set *src)
{
dest->nelem = src->nelem;
if (src->nelem > 0)
{
dest->alloc = dest->nelem;
dest->elems = re_malloc (int, dest->alloc);
if (BE (dest->elems == NULL, 0))
{
dest->alloc = dest->nelem = 0;
return REG_ESPACE;
}
memcpy (dest->elems, src->elems, src->nelem * sizeof (int));
}
else
re_node_set_init_empty (dest);
return REG_NOERROR;
}
/* Calculate the intersection of the sets SRC1 and SRC2. And merge it to
DEST. Return value indicate the error code or REG_NOERROR if succeeded.
Note: We assume dest->elems is NULL, when dest->alloc is 0. */
static reg_errcode_t
internal_function
re_node_set_add_intersect (re_node_set *dest, const re_node_set *src1,
const re_node_set *src2)
{
int i1, i2, is, id, delta, sbase;
if (src1->nelem == 0 || src2->nelem == 0)
return REG_NOERROR;
/* We need dest->nelem + 2 * elems_in_intersection; this is a
conservative estimate. */
if (src1->nelem + src2->nelem + dest->nelem > dest->alloc)
{
int new_alloc = src1->nelem + src2->nelem + dest->alloc;
int *new_elems = re_realloc (dest->elems, int, new_alloc);
if (BE (new_elems == NULL, 0))
return REG_ESPACE;
dest->elems = new_elems;
dest->alloc = new_alloc;
}
/* Find the items in the intersection of SRC1 and SRC2, and copy
into the top of DEST those that are not already in DEST itself. */
sbase = dest->nelem + src1->nelem + src2->nelem;
i1 = src1->nelem - 1;
i2 = src2->nelem - 1;
id = dest->nelem - 1;
for (;;)
{
if (src1->elems[i1] == src2->elems[i2])
{
/* Try to find the item in DEST. Maybe we could binary search? */
while (id >= 0 && dest->elems[id] > src1->elems[i1])
--id;
if (id < 0 || dest->elems[id] != src1->elems[i1])
dest->elems[--sbase] = src1->elems[i1];
if (--i1 < 0 || --i2 < 0)
break;
}
/* Lower the highest of the two items. */
else if (src1->elems[i1] < src2->elems[i2])
{
if (--i2 < 0)
break;
}
else
{
if (--i1 < 0)
break;
}
}
id = dest->nelem - 1;
is = dest->nelem + src1->nelem + src2->nelem - 1;
delta = is - sbase + 1;
/* Now copy. When DELTA becomes zero, the remaining
DEST elements are already in place; this is more or
less the same loop that is in re_node_set_merge. */
dest->nelem += delta;
if (delta > 0 && id >= 0)
for (;;)
{
if (dest->elems[is] > dest->elems[id])
{
/* Copy from the top. */
dest->elems[id + delta--] = dest->elems[is--];
if (delta == 0)
break;
}
else
{
/* Slide from the bottom. */
dest->elems[id + delta] = dest->elems[id];
if (--id < 0)
break;
}
}
/* Copy remaining SRC elements. */
memcpy (dest->elems, dest->elems + sbase, delta * sizeof (int));
return REG_NOERROR;
}
/* Calculate the union set of the sets SRC1 and SRC2. And store it to
DEST. Return value indicate the error code or REG_NOERROR if succeeded. */
static reg_errcode_t
internal_function
re_node_set_init_union (re_node_set *dest, const re_node_set *src1,
const re_node_set *src2)
{
int i1, i2, id;
if (src1 != NULL && src1->nelem > 0 && src2 != NULL && src2->nelem > 0)
{
dest->alloc = src1->nelem + src2->nelem;
dest->elems = re_malloc (int, dest->alloc);
if (BE (dest->elems == NULL, 0))
return REG_ESPACE;
}
else
{
if (src1 != NULL && src1->nelem > 0)
return re_node_set_init_copy (dest, src1);
else if (src2 != NULL && src2->nelem > 0)
return re_node_set_init_copy (dest, src2);
else
re_node_set_init_empty (dest);
return REG_NOERROR;
}
for (i1 = i2 = id = 0 ; i1 < src1->nelem && i2 < src2->nelem ;)
{
if (src1->elems[i1] > src2->elems[i2])
{
dest->elems[id++] = src2->elems[i2++];
continue;
}
if (src1->elems[i1] == src2->elems[i2])
++i2;
dest->elems[id++] = src1->elems[i1++];
}
if (i1 < src1->nelem)
{
memcpy (dest->elems + id, src1->elems + i1,
(src1->nelem - i1) * sizeof (int));
id += src1->nelem - i1;
}
else if (i2 < src2->nelem)
{
memcpy (dest->elems + id, src2->elems + i2,
(src2->nelem - i2) * sizeof (int));
id += src2->nelem - i2;
}
dest->nelem = id;
return REG_NOERROR;
}
/* Calculate the union set of the sets DEST and SRC. And store it to
DEST. Return value indicate the error code or REG_NOERROR if succeeded. */
static reg_errcode_t
internal_function
re_node_set_merge (re_node_set *dest, const re_node_set *src)
{
int is, id, sbase, delta;
if (src == NULL || src->nelem == 0)
return REG_NOERROR;
if (dest->alloc < 2 * src->nelem + dest->nelem)
{
int new_alloc = 2 * (src->nelem + dest->alloc);
int *new_buffer = re_realloc (dest->elems, int, new_alloc);
if (BE (new_buffer == NULL, 0))
return REG_ESPACE;
dest->elems = new_buffer;
dest->alloc = new_alloc;
}
if (BE (dest->nelem == 0, 0))
{
dest->nelem = src->nelem;
memcpy (dest->elems, src->elems, src->nelem * sizeof (int));
return REG_NOERROR;
}
/* Copy into the top of DEST the items of SRC that are not
found in DEST. Maybe we could binary search in DEST? */
for (sbase = dest->nelem + 2 * src->nelem,
is = src->nelem - 1, id = dest->nelem - 1; is >= 0 && id >= 0; )
{
if (dest->elems[id] == src->elems[is])
is--, id--;
else if (dest->elems[id] < src->elems[is])
dest->elems[--sbase] = src->elems[is--];
else /* if (dest->elems[id] > src->elems[is]) */
--id;
}
if (is >= 0)
{
/* If DEST is exhausted, the remaining items of SRC must be unique. */
sbase -= is + 1;
memcpy (dest->elems + sbase, src->elems, (is + 1) * sizeof (int));
}
id = dest->nelem - 1;
is = dest->nelem + 2 * src->nelem - 1;
delta = is - sbase + 1;
if (delta == 0)
return REG_NOERROR;
/* Now copy. When DELTA becomes zero, the remaining
DEST elements are already in place. */
dest->nelem += delta;
for (;;)
{
if (dest->elems[is] > dest->elems[id])
{
/* Copy from the top. */
dest->elems[id + delta--] = dest->elems[is--];
if (delta == 0)
break;
}
else
{
/* Slide from the bottom. */
dest->elems[id + delta] = dest->elems[id];
if (--id < 0)
{
/* Copy remaining SRC elements. */
memcpy (dest->elems, dest->elems + sbase,
delta * sizeof (int));
break;
}
}
}
return REG_NOERROR;
}
/* Insert the new element ELEM to the re_node_set* SET.
SET should not already have ELEM.
return -1 if an error is occured, return 1 otherwise. */
static int
internal_function
re_node_set_insert (re_node_set *set, int elem)
{
int idx;
/* In case the set is empty. */
if (set->alloc == 0)
{
if (BE (re_node_set_init_1 (set, elem) == REG_NOERROR, 1))
return 1;
else
return -1;
}
if (BE (set->nelem, 0) == 0)
{
/* We already guaranteed above that set->alloc != 0. */
set->elems[0] = elem;
++set->nelem;
return 1;
}
/* Realloc if we need. */
if (set->alloc == set->nelem)
{
int *new_elems;
set->alloc = set->alloc * 2;
new_elems = re_realloc (set->elems, int, set->alloc);
if (BE (new_elems == NULL, 0))
return -1;
set->elems = new_elems;
}
/* Move the elements which follows the new element. Test the
first element separately to skip a check in the inner loop. */
if (elem < set->elems[0])
{
idx = 0;
for (idx = set->nelem; idx > 0; idx--)
set->elems[idx] = set->elems[idx - 1];
}
else
{
for (idx = set->nelem; set->elems[idx - 1] > elem; idx--)
set->elems[idx] = set->elems[idx - 1];
}
/* Insert the new element. */
set->elems[idx] = elem;
++set->nelem;
return 1;
}
/* Insert the new element ELEM to the re_node_set* SET.
SET should not already have any element greater than or equal to ELEM.
Return -1 if an error is occured, return 1 otherwise. */
static int
internal_function
re_node_set_insert_last (re_node_set *set, int elem)
{
/* Realloc if we need. */
if (set->alloc == set->nelem)
{
int *new_elems;
set->alloc = (set->alloc + 1) * 2;
new_elems = re_realloc (set->elems, int, set->alloc);
if (BE (new_elems == NULL, 0))
return -1;
set->elems = new_elems;
}
/* Insert the new element. */
set->elems[set->nelem++] = elem;
return 1;
}
/* Compare two node sets SET1 and SET2.
return 1 if SET1 and SET2 are equivalent, return 0 otherwise. */
static int
internal_function __attribute ((pure))
re_node_set_compare (const re_node_set *set1, const re_node_set *set2)
{
int i;
if (set1 == NULL || set2 == NULL || set1->nelem != set2->nelem)
return 0;
for (i = set1->nelem ; --i >= 0 ; )
if (set1->elems[i] != set2->elems[i])
return 0;
return 1;
}
/* Return (idx + 1) if SET contains the element ELEM, return 0 otherwise. */
static int
internal_function __attribute ((pure))
re_node_set_contains (const re_node_set *set, int elem)
{
unsigned int idx, right, mid;
if (set->nelem <= 0)
return 0;
/* Binary search the element. */
idx = 0;
right = set->nelem - 1;
while (idx < right)
{
mid = (idx + right) / 2;
if (set->elems[mid] < elem)
idx = mid + 1;
else
right = mid;
}
return set->elems[idx] == elem ? idx + 1 : 0;
}
static void
internal_function
re_node_set_remove_at (re_node_set *set, int idx)
{
if (idx < 0 || idx >= set->nelem)
return;
--set->nelem;
for (; idx < set->nelem; idx++)
set->elems[idx] = set->elems[idx + 1];
}
/* Add the token TOKEN to dfa->nodes, and return the index of the token.
Or return -1, if an error will be occured. */
static int
internal_function
re_dfa_add_node (re_dfa_t *dfa, re_token_t token)
{
if (BE (dfa->nodes_len >= dfa->nodes_alloc, 0))
{
size_t new_nodes_alloc = dfa->nodes_alloc * 2;
int *new_nexts, *new_indices;
re_node_set *new_edests, *new_eclosures;
re_token_t *new_nodes;
/* Avoid overflows in realloc. */
const size_t max_object_size = MAX (sizeof (re_token_t),
MAX (sizeof (re_node_set),
sizeof (int)));
if (BE (SIZE_MAX / max_object_size < new_nodes_alloc, 0))
return -1;
new_nodes = re_realloc (dfa->nodes, re_token_t, new_nodes_alloc);
if (BE (new_nodes == NULL, 0))
return -1;
dfa->nodes = new_nodes;
new_nexts = re_realloc (dfa->nexts, int, new_nodes_alloc);
new_indices = re_realloc (dfa->org_indices, int, new_nodes_alloc);
new_edests = re_realloc (dfa->edests, re_node_set, new_nodes_alloc);
new_eclosures = re_realloc (dfa->eclosures, re_node_set, new_nodes_alloc);
if (BE (new_nexts == NULL || new_indices == NULL
|| new_edests == NULL || new_eclosures == NULL, 0))
return -1;
dfa->nexts = new_nexts;
dfa->org_indices = new_indices;
dfa->edests = new_edests;
dfa->eclosures = new_eclosures;
dfa->nodes_alloc = new_nodes_alloc;
}
dfa->nodes[dfa->nodes_len] = token;
dfa->nodes[dfa->nodes_len].constraint = 0;
#ifdef RE_ENABLE_I18N
dfa->nodes[dfa->nodes_len].accept_mb =
(token.type == OP_PERIOD && dfa->mb_cur_max > 1) || token.type == COMPLEX_BRACKET;
#endif
dfa->nexts[dfa->nodes_len] = -1;
re_node_set_init_empty (dfa->edests + dfa->nodes_len);
re_node_set_init_empty (dfa->eclosures + dfa->nodes_len);
return dfa->nodes_len++;
}
static inline unsigned int
internal_function
calc_state_hash (const re_node_set *nodes, unsigned int context)
{
unsigned int hash = nodes->nelem + context;
int i;
for (i = 0 ; i < nodes->nelem ; i++)
hash += nodes->elems[i];
return hash;
}
/* Search for the state whose node_set is equivalent to NODES.
Return the pointer to the state, if we found it in the DFA.
Otherwise create the new one and return it. In case of an error
return NULL and set the error code in ERR.
Note: - We assume NULL as the invalid state, then it is possible that
return value is NULL and ERR is REG_NOERROR.
- We never return non-NULL value in case of any errors, it is for
optimization. */
static re_dfastate_t *
internal_function
re_acquire_state (reg_errcode_t *err, const re_dfa_t *dfa,
const re_node_set *nodes)
{
unsigned int hash;
re_dfastate_t *new_state;
struct re_state_table_entry *spot;
int i;
if (BE (nodes->nelem == 0, 0))
{
*err = REG_NOERROR;
return NULL;
}
hash = calc_state_hash (nodes, 0);
spot = dfa->state_table + (hash & dfa->state_hash_mask);
for (i = 0 ; i < spot->num ; i++)
{
re_dfastate_t *state = spot->array[i];
if (hash != state->hash)
continue;
if (re_node_set_compare (&state->nodes, nodes))
return state;
}
/* There are no appropriate state in the dfa, create the new one. */
new_state = create_ci_newstate (dfa, nodes, hash);
if (BE (new_state == NULL, 0))
*err = REG_ESPACE;
return new_state;
}
/* Search for the state whose node_set is equivalent to NODES and
whose context is equivalent to CONTEXT.
Return the pointer to the state, if we found it in the DFA.
Otherwise create the new one and return it. In case of an error
return NULL and set the error code in ERR.
Note: - We assume NULL as the invalid state, then it is possible that
return value is NULL and ERR is REG_NOERROR.
- We never return non-NULL value in case of any errors, it is for
optimization. */
static re_dfastate_t *
internal_function
re_acquire_state_context (reg_errcode_t *err, const re_dfa_t *dfa,
const re_node_set *nodes, unsigned int context)
{
unsigned int hash;
re_dfastate_t *new_state;
struct re_state_table_entry *spot;
int i;
if (nodes->nelem == 0)
{
*err = REG_NOERROR;
return NULL;
}
hash = calc_state_hash (nodes, context);
spot = dfa->state_table + (hash & dfa->state_hash_mask);
for (i = 0 ; i < spot->num ; i++)
{
re_dfastate_t *state = spot->array[i];
if (state->hash == hash
&& state->context == context
&& re_node_set_compare (state->entrance_nodes, nodes))
return state;
}
/* There are no appropriate state in `dfa', create the new one. */
new_state = create_cd_newstate (dfa, nodes, context, hash);
if (BE (new_state == NULL, 0))
*err = REG_ESPACE;
return new_state;
}
/* Finish initialization of the new state NEWSTATE, and using its hash value
HASH put in the appropriate bucket of DFA's state table. Return value
indicates the error code if failed. */
static reg_errcode_t
register_state (const re_dfa_t *dfa, re_dfastate_t *newstate,
unsigned int hash)
{
struct re_state_table_entry *spot;
reg_errcode_t err;
int i;
newstate->hash = hash;
err = re_node_set_alloc (&newstate->non_eps_nodes, newstate->nodes.nelem);
if (BE (err != REG_NOERROR, 0))
return REG_ESPACE;
for (i = 0; i < newstate->nodes.nelem; i++)
{
int elem = newstate->nodes.elems[i];
if (!IS_EPSILON_NODE (dfa->nodes[elem].type))
if (re_node_set_insert_last (&newstate->non_eps_nodes, elem) < 0)
return REG_ESPACE;
}
spot = dfa->state_table + (hash & dfa->state_hash_mask);
if (BE (spot->alloc <= spot->num, 0))
{
int new_alloc = 2 * spot->num + 2;
re_dfastate_t **new_array = re_realloc (spot->array, re_dfastate_t *,
new_alloc);
if (BE (new_array == NULL, 0))
return REG_ESPACE;
spot->array = new_array;
spot->alloc = new_alloc;
}
spot->array[spot->num++] = newstate;
return REG_NOERROR;
}
static void
free_state (re_dfastate_t *state)
{
re_node_set_free (&state->non_eps_nodes);
re_node_set_free (&state->inveclosure);
if (state->entrance_nodes != &state->nodes)
{
re_node_set_free (state->entrance_nodes);
re_free (state->entrance_nodes);
}
re_node_set_free (&state->nodes);
re_free (state->word_trtable);
re_free (state->trtable);
re_free (state);
}
/* Create the new state which is independ of contexts.
Return the new state if succeeded, otherwise return NULL. */
static re_dfastate_t *
internal_function
create_ci_newstate (const re_dfa_t *dfa, const re_node_set *nodes,
unsigned int hash)
{
int i;
reg_errcode_t err;
re_dfastate_t *newstate;
newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1);
if (BE (newstate == NULL, 0))
return NULL;
err = re_node_set_init_copy (&newstate->nodes, nodes);
if (BE (err != REG_NOERROR, 0))
{
re_free (newstate);
return NULL;
}
newstate->entrance_nodes = &newstate->nodes;
for (i = 0 ; i < nodes->nelem ; i++)
{
re_token_t *node = dfa->nodes + nodes->elems[i];
re_token_type_t type = node->type;
if (type == CHARACTER && !node->constraint)
continue;
#ifdef RE_ENABLE_I18N
newstate->accept_mb |= node->accept_mb;
#endif /* RE_ENABLE_I18N */
/* If the state has the halt node, the state is a halt state. */
if (type == END_OF_RE)
newstate->halt = 1;
else if (type == OP_BACK_REF)
newstate->has_backref = 1;
else if (type == ANCHOR || node->constraint)
newstate->has_constraint = 1;
}
err = register_state (dfa, newstate, hash);
if (BE (err != REG_NOERROR, 0))
{
free_state (newstate);
newstate = NULL;
}
return newstate;
}
/* Create the new state which is depend on the context CONTEXT.
Return the new state if succeeded, otherwise return NULL. */
static re_dfastate_t *
internal_function
create_cd_newstate (const re_dfa_t *dfa, const re_node_set *nodes,
unsigned int context, unsigned int hash)
{
int i, nctx_nodes = 0;
reg_errcode_t err;
re_dfastate_t *newstate;
newstate = (re_dfastate_t *) calloc (sizeof (re_dfastate_t), 1);
if (BE (newstate == NULL, 0))
return NULL;
err = re_node_set_init_copy (&newstate->nodes, nodes);
if (BE (err != REG_NOERROR, 0))
{
re_free (newstate);
return NULL;
}
newstate->context = context;
newstate->entrance_nodes = &newstate->nodes;
for (i = 0 ; i < nodes->nelem ; i++)
{
re_token_t *node = dfa->nodes + nodes->elems[i];
re_token_type_t type = node->type;
unsigned int constraint = node->constraint;
if (type == CHARACTER && !constraint)
continue;
#ifdef RE_ENABLE_I18N
newstate->accept_mb |= node->accept_mb;
#endif /* RE_ENABLE_I18N */
/* If the state has the halt node, the state is a halt state. */
if (type == END_OF_RE)
newstate->halt = 1;
else if (type == OP_BACK_REF)
newstate->has_backref = 1;
if (constraint)
{
if (newstate->entrance_nodes == &newstate->nodes)
{
newstate->entrance_nodes = re_malloc (re_node_set, 1);
if (BE (newstate->entrance_nodes == NULL, 0))
{
free_state (newstate);
return NULL;
}
if (re_node_set_init_copy (newstate->entrance_nodes, nodes)
!= REG_NOERROR)
return NULL;
nctx_nodes = 0;
newstate->has_constraint = 1;
}
if (NOT_SATISFY_PREV_CONSTRAINT (constraint,context))
{
re_node_set_remove_at (&newstate->nodes, i - nctx_nodes);
++nctx_nodes;
}
}
}
err = register_state (dfa, newstate, hash);
if (BE (err != REG_NOERROR, 0))
{
free_state (newstate);
newstate = NULL;
}
return newstate;
}

810
compat/regex/regex_internal.h Normal file
View File

@ -0,0 +1,810 @@
/* Extended regular expression matching and search library.
Copyright (C) 2002-2005, 2007, 2008, 2010 Free Software Foundation, Inc.
This file is part of the GNU C Library.
Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
The GNU C Library is free software; you can redistribute it and/or
modify it under the terms of the GNU Lesser General Public
License as published by the Free Software Foundation; either
version 2.1 of the License, or (at your option) any later version.
The GNU C Library is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
Lesser General Public License for more details.
You should have received a copy of the GNU Lesser General Public
License along with the GNU C Library; if not, write to the Free
Software Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA
02111-1307 USA. */
#ifndef _REGEX_INTERNAL_H
#define _REGEX_INTERNAL_H 1
#include <assert.h>
#include <ctype.h>
#include <stdio.h>
#include <stdlib.h>
#include <string.h>
#if defined HAVE_LANGINFO_H || defined HAVE_LANGINFO_CODESET || defined _LIBC
# include <langinfo.h>
#endif
#if defined HAVE_LOCALE_H || defined _LIBC
# include <locale.h>
#endif
#if defined HAVE_WCHAR_H || defined _LIBC
# include <wchar.h>
#endif /* HAVE_WCHAR_H || _LIBC */
#if defined HAVE_WCTYPE_H || defined _LIBC
# include <wctype.h>
#endif /* HAVE_WCTYPE_H || _LIBC */
#if defined HAVE_STDBOOL_H || defined _LIBC
# include <stdbool.h>
#endif /* HAVE_STDBOOL_H || _LIBC */
#if !defined(ZOS_USS)
#if defined HAVE_STDINT_H || defined _LIBC
# include <stdint.h>
#endif /* HAVE_STDINT_H || _LIBC */
#endif /* !ZOS_USS */
#if defined _LIBC
# include <bits/libc-lock.h>
#else
# define __libc_lock_define(CLASS,NAME)
# define __libc_lock_init(NAME) do { } while (0)
# define __libc_lock_lock(NAME) do { } while (0)
# define __libc_lock_unlock(NAME) do { } while (0)
#endif
#ifndef GAWK
/* In case that the system doesn't have isblank(). */
#if !defined _LIBC && !defined HAVE_ISBLANK && !defined isblank
# define isblank(ch) ((ch) == ' ' || (ch) == '\t')
#endif
#else /* GAWK */
/*
* This is a freaking mess. On glibc systems you have to define
* a magic constant to get isblank() out of <ctype.h>, since it's
* a C99 function. To heck with all that and borrow a page from
* dfa.c's book.
*/
static int
is_blank (int c)
{
return (c == ' ' || c == '\t');
}
#endif /* GAWK */
#ifdef _LIBC
# ifndef _RE_DEFINE_LOCALE_FUNCTIONS
# define _RE_DEFINE_LOCALE_FUNCTIONS 1
# include <locale/localeinfo.h>
# include <locale/elem-hash.h>
# include <locale/coll-lookup.h>
# endif
#endif
/* This is for other GNU distributions with internationalized messages. */
#if (HAVE_LIBINTL_H && ENABLE_NLS) || defined _LIBC
# include <libintl.h>
# ifdef _LIBC
# undef gettext
# define gettext(msgid) \
INTUSE(__dcgettext) (_libc_intl_domainname, msgid, LC_MESSAGES)
# endif
#else
# define gettext(msgid) (msgid)
#endif
#ifndef gettext_noop
/* This define is so xgettext can find the internationalizable
strings. */
# define gettext_noop(String) String
#endif
/* For loser systems without the definition. */
#ifndef SIZE_MAX
# define SIZE_MAX ((size_t) -1)
#endif
#ifndef NO_MBSUPPORT
#include "mbsupport.h" /* gawk */
#endif
#ifndef MB_CUR_MAX
#define MB_CUR_MAX 1
#endif
#if (defined MBS_SUPPORT) || _LIBC
# define RE_ENABLE_I18N
#endif
#if __GNUC__ >= 3
# define BE(expr, val) __builtin_expect (expr, val)
#else
# define BE(expr, val) (expr)
# ifdef inline
# undef inline
# endif
# define inline
#endif
/* Number of single byte character. */
#define SBC_MAX 256
#define COLL_ELEM_LEN_MAX 8
/* The character which represents newline. */
#define NEWLINE_CHAR '\n'
#define WIDE_NEWLINE_CHAR L'\n'
/* Rename to standard API for using out of glibc. */
#ifndef _LIBC
# ifdef __wctype
# undef __wctype
# endif
# define __wctype wctype
# ifdef __iswctype
# undef __iswctype
# endif
# define __iswctype iswctype
# define __btowc btowc
# define __mbrtowc mbrtowc
#undef __mempcpy /* GAWK */
# define __mempcpy mempcpy
# define __wcrtomb wcrtomb
# define __regfree regfree
# define attribute_hidden
#endif /* not _LIBC */
#ifdef __GNUC__
# define __attribute(arg) __attribute__ (arg)
#else
# define __attribute(arg)
#endif
extern const char __re_error_msgid[] attribute_hidden;
extern const size_t __re_error_msgid_idx[] attribute_hidden;
/* An integer used to represent a set of bits. It must be unsigned,
and must be at least as wide as unsigned int. */
typedef unsigned long int bitset_word_t;
/* All bits set in a bitset_word_t. */
#define BITSET_WORD_MAX ULONG_MAX
/* Number of bits in a bitset_word_t. */
#define BITSET_WORD_BITS (sizeof (bitset_word_t) * CHAR_BIT)
/* Number of bitset_word_t in a bit_set. */
#define BITSET_WORDS (SBC_MAX / BITSET_WORD_BITS)
typedef bitset_word_t bitset_t[BITSET_WORDS];
typedef bitset_word_t *re_bitset_ptr_t;
typedef const bitset_word_t *re_const_bitset_ptr_t;
#define bitset_set(set,i) \
(set[i / BITSET_WORD_BITS] |= (bitset_word_t) 1 << i % BITSET_WORD_BITS)
#define bitset_clear(set,i) \
(set[i / BITSET_WORD_BITS] &= ~((bitset_word_t) 1 << i % BITSET_WORD_BITS))
#define bitset_contain(set,i) \
(set[i / BITSET_WORD_BITS] & ((bitset_word_t) 1 << i % BITSET_WORD_BITS))
#define bitset_empty(set) memset (set, '\0', sizeof (bitset_t))
#define bitset_set_all(set) memset (set, '\xff', sizeof (bitset_t))
#define bitset_copy(dest,src) memcpy (dest, src, sizeof (bitset_t))
#define PREV_WORD_CONSTRAINT 0x0001
#define PREV_NOTWORD_CONSTRAINT 0x0002
#define NEXT_WORD_CONSTRAINT 0x0004
#define NEXT_NOTWORD_CONSTRAINT 0x0008
#define PREV_NEWLINE_CONSTRAINT 0x0010
#define NEXT_NEWLINE_CONSTRAINT 0x0020
#define PREV_BEGBUF_CONSTRAINT 0x0040
#define NEXT_ENDBUF_CONSTRAINT 0x0080
#define WORD_DELIM_CONSTRAINT 0x0100
#define NOT_WORD_DELIM_CONSTRAINT 0x0200
typedef enum
{
INSIDE_WORD = PREV_WORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
WORD_FIRST = PREV_NOTWORD_CONSTRAINT | NEXT_WORD_CONSTRAINT,
WORD_LAST = PREV_WORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
INSIDE_NOTWORD = PREV_NOTWORD_CONSTRAINT | NEXT_NOTWORD_CONSTRAINT,
LINE_FIRST = PREV_NEWLINE_CONSTRAINT,
LINE_LAST = NEXT_NEWLINE_CONSTRAINT,
BUF_FIRST = PREV_BEGBUF_CONSTRAINT,
BUF_LAST = NEXT_ENDBUF_CONSTRAINT,
WORD_DELIM = WORD_DELIM_CONSTRAINT,
NOT_WORD_DELIM = NOT_WORD_DELIM_CONSTRAINT
} re_context_type;
typedef struct
{
int alloc;
int nelem;
int *elems;
} re_node_set;
typedef enum
{
NON_TYPE = 0,
/* Node type, These are used by token, node, tree. */
CHARACTER = 1,
END_OF_RE = 2,
SIMPLE_BRACKET = 3,
OP_BACK_REF = 4,
OP_PERIOD = 5,
#ifdef RE_ENABLE_I18N
COMPLEX_BRACKET = 6,
OP_UTF8_PERIOD = 7,
#endif /* RE_ENABLE_I18N */
/* We define EPSILON_BIT as a macro so that OP_OPEN_SUBEXP is used
when the debugger shows values of this enum type. */
#define EPSILON_BIT 8
OP_OPEN_SUBEXP = EPSILON_BIT | 0,
OP_CLOSE_SUBEXP = EPSILON_BIT | 1,
OP_ALT = EPSILON_BIT | 2,
OP_DUP_ASTERISK = EPSILON_BIT | 3,
ANCHOR = EPSILON_BIT | 4,
/* Tree type, these are used only by tree. */
CONCAT = 16,
SUBEXP = 17,
/* Token type, these are used only by token. */
OP_DUP_PLUS = 18,
OP_DUP_QUESTION,
OP_OPEN_BRACKET,
OP_CLOSE_BRACKET,
OP_CHARSET_RANGE,
OP_OPEN_DUP_NUM,
OP_CLOSE_DUP_NUM,
OP_NON_MATCH_LIST,
OP_OPEN_COLL_ELEM,
OP_CLOSE_COLL_ELEM,
OP_OPEN_EQUIV_CLASS,
OP_CLOSE_EQUIV_CLASS,
OP_OPEN_CHAR_CLASS,
OP_CLOSE_CHAR_CLASS,
OP_WORD,
OP_NOTWORD,
OP_SPACE,
OP_NOTSPACE,
BACK_SLASH
} re_token_type_t;
#ifdef RE_ENABLE_I18N
typedef struct
{
/* Multibyte characters. */
wchar_t *mbchars;
/* Collating symbols. */
# ifdef _LIBC
int32_t *coll_syms;
# endif
/* Equivalence classes. */
# ifdef _LIBC
int32_t *equiv_classes;
# endif
/* Range expressions. */
# ifdef _LIBC
uint32_t *range_starts;
uint32_t *range_ends;
# else /* not _LIBC */
wchar_t *range_starts;
wchar_t *range_ends;
# endif /* not _LIBC */
/* Character classes. */
wctype_t *char_classes;
/* If this character set is the non-matching list. */
unsigned int non_match : 1;
/* # of multibyte characters. */
int nmbchars;
/* # of collating symbols. */
int ncoll_syms;
/* # of equivalence classes. */
int nequiv_classes;
/* # of range expressions. */
int nranges;
/* # of character classes. */
int nchar_classes;
} re_charset_t;
#endif /* RE_ENABLE_I18N */
typedef struct
{
union
{
unsigned char c; /* for CHARACTER */
re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */
#ifdef RE_ENABLE_I18N
re_charset_t *mbcset; /* for COMPLEX_BRACKET */
#endif /* RE_ENABLE_I18N */
int idx; /* for BACK_REF */
re_context_type ctx_type; /* for ANCHOR */
} opr;
#if __GNUC__ >= 2
re_token_type_t type : 8;
#else
re_token_type_t type;
#endif
unsigned int constraint : 10; /* context constraint */
unsigned int duplicated : 1;
unsigned int opt_subexp : 1;
#ifdef RE_ENABLE_I18N
unsigned int accept_mb : 1;
/* These 2 bits can be moved into the union if needed (e.g. if running out
of bits; move opr.c to opr.c.c and move the flags to opr.c.flags). */
unsigned int mb_partial : 1;
#endif
unsigned int word_char : 1;
} re_token_t;
#define IS_EPSILON_NODE(type) ((type) & EPSILON_BIT)
struct re_string_t
{
/* Indicate the raw buffer which is the original string passed as an
argument of regexec(), re_search(), etc.. */
const unsigned char *raw_mbs;
/* Store the multibyte string. In case of "case insensitive mode" like
REG_ICASE, upper cases of the string are stored, otherwise MBS points
the same address that RAW_MBS points. */
unsigned char *mbs;
#ifdef RE_ENABLE_I18N
/* Store the wide character string which is corresponding to MBS. */
wint_t *wcs;
int *offsets;
mbstate_t cur_state;
#endif
/* Index in RAW_MBS. Each character mbs[i] corresponds to
raw_mbs[raw_mbs_idx + i]. */
int raw_mbs_idx;
/* The length of the valid characters in the buffers. */
int valid_len;
/* The corresponding number of bytes in raw_mbs array. */
int valid_raw_len;
/* The length of the buffers MBS and WCS. */
int bufs_len;
/* The index in MBS, which is updated by re_string_fetch_byte. */
int cur_idx;
/* length of RAW_MBS array. */
int raw_len;
/* This is RAW_LEN - RAW_MBS_IDX + VALID_LEN - VALID_RAW_LEN. */
int len;
/* End of the buffer may be shorter than its length in the cases such
as re_match_2, re_search_2. Then, we use STOP for end of the buffer
instead of LEN. */
int raw_stop;
/* This is RAW_STOP - RAW_MBS_IDX adjusted through OFFSETS. */
int stop;
/* The context of mbs[0]. We store the context independently, since
the context of mbs[0] may be different from raw_mbs[0], which is
the beginning of the input string. */
unsigned int tip_context;
/* The translation passed as a part of an argument of re_compile_pattern. */
RE_TRANSLATE_TYPE trans;
/* Copy of re_dfa_t's word_char. */
re_const_bitset_ptr_t word_char;
/* 1 if REG_ICASE. */
unsigned char icase;
unsigned char is_utf8;
unsigned char map_notascii;
unsigned char mbs_allocated;
unsigned char offsets_needed;
unsigned char newline_anchor;
unsigned char word_ops_used;
int mb_cur_max;
};
typedef struct re_string_t re_string_t;
struct re_dfa_t;
typedef struct re_dfa_t re_dfa_t;
#ifndef _LIBC
# ifdef __i386__
# define internal_function __attribute ((regparm (3), stdcall))
# else
# define internal_function
# endif
#endif
#ifndef NOT_IN_libc
static reg_errcode_t re_string_realloc_buffers (re_string_t *pstr,
int new_buf_len)
internal_function;
# ifdef RE_ENABLE_I18N
static void build_wcs_buffer (re_string_t *pstr) internal_function;
static reg_errcode_t build_wcs_upper_buffer (re_string_t *pstr)
internal_function;
# endif /* RE_ENABLE_I18N */
static void build_upper_buffer (re_string_t *pstr) internal_function;
static void re_string_translate_buffer (re_string_t *pstr) internal_function;
static unsigned int re_string_context_at (const re_string_t *input, int idx,
int eflags)
internal_function __attribute ((pure));
#endif
#define re_string_peek_byte(pstr, offset) \
((pstr)->mbs[(pstr)->cur_idx + offset])
#define re_string_fetch_byte(pstr) \
((pstr)->mbs[(pstr)->cur_idx++])
#define re_string_first_byte(pstr, idx) \
((idx) == (pstr)->valid_len || (pstr)->wcs[idx] != WEOF)
#define re_string_is_single_byte_char(pstr, idx) \
((pstr)->wcs[idx] != WEOF && ((pstr)->valid_len == (idx) + 1 \
|| (pstr)->wcs[(idx) + 1] != WEOF))
#define re_string_eoi(pstr) ((pstr)->stop <= (pstr)->cur_idx)
#define re_string_cur_idx(pstr) ((pstr)->cur_idx)
#define re_string_get_buffer(pstr) ((pstr)->mbs)
#define re_string_length(pstr) ((pstr)->len)
#define re_string_byte_at(pstr,idx) ((pstr)->mbs[idx])
#define re_string_skip_bytes(pstr,idx) ((pstr)->cur_idx += (idx))
#define re_string_set_index(pstr,idx) ((pstr)->cur_idx = (idx))
#ifndef _LIBC
# if HAVE_ALLOCA
# if (_MSC_VER)
# include <malloc.h>
# define __libc_use_alloca(n) 0
# else
# include <alloca.h>
/* The OS usually guarantees only one guard page at the bottom of the stack,
and a page size can be as small as 4096 bytes. So we cannot safely
allocate anything larger than 4096 bytes. Also care for the possibility
of a few compiler-allocated temporary stack slots. */
# define __libc_use_alloca(n) ((n) < 4032)
# endif
# else
/* alloca is implemented with malloc, so just use malloc. */
# define __libc_use_alloca(n) 0
# endif
#endif
#define re_malloc(t,n) ((t *) malloc ((n) * sizeof (t)))
/* SunOS 4.1.x realloc doesn't accept null pointers: pre-Standard C. Sigh. */
#define re_realloc(p,t,n) ((p != NULL) ? (t *) realloc (p,(n)*sizeof(t)) : (t *) calloc(n,sizeof(t)))
#define re_free(p) free (p)
struct bin_tree_t
{
struct bin_tree_t *parent;
struct bin_tree_t *left;
struct bin_tree_t *right;
struct bin_tree_t *first;
struct bin_tree_t *next;
re_token_t token;
/* `node_idx' is the index in dfa->nodes, if `type' == 0.
Otherwise `type' indicate the type of this node. */
int node_idx;
};
typedef struct bin_tree_t bin_tree_t;
#define BIN_TREE_STORAGE_SIZE \
((1024 - sizeof (void *)) / sizeof (bin_tree_t))
struct bin_tree_storage_t
{
struct bin_tree_storage_t *next;
bin_tree_t data[BIN_TREE_STORAGE_SIZE];
};
typedef struct bin_tree_storage_t bin_tree_storage_t;
#define CONTEXT_WORD 1
#define CONTEXT_NEWLINE (CONTEXT_WORD << 1)
#define CONTEXT_BEGBUF (CONTEXT_NEWLINE << 1)
#define CONTEXT_ENDBUF (CONTEXT_BEGBUF << 1)
#define IS_WORD_CONTEXT(c) ((c) & CONTEXT_WORD)
#define IS_NEWLINE_CONTEXT(c) ((c) & CONTEXT_NEWLINE)
#define IS_BEGBUF_CONTEXT(c) ((c) & CONTEXT_BEGBUF)
#define IS_ENDBUF_CONTEXT(c) ((c) & CONTEXT_ENDBUF)
#define IS_ORDINARY_CONTEXT(c) ((c) == 0)
#define IS_WORD_CHAR(ch) (isalnum (ch) || (ch) == '_')
#define IS_NEWLINE(ch) ((ch) == NEWLINE_CHAR)
#define IS_WIDE_WORD_CHAR(ch) (iswalnum (ch) || (ch) == L'_')
#define IS_WIDE_NEWLINE(ch) ((ch) == WIDE_NEWLINE_CHAR)
#define NOT_SATISFY_PREV_CONSTRAINT(constraint,context) \
((((constraint) & PREV_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
|| ((constraint & PREV_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
|| ((constraint & PREV_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context))\
|| ((constraint & PREV_BEGBUF_CONSTRAINT) && !IS_BEGBUF_CONTEXT (context)))
#define NOT_SATISFY_NEXT_CONSTRAINT(constraint,context) \
((((constraint) & NEXT_WORD_CONSTRAINT) && !IS_WORD_CONTEXT (context)) \
|| (((constraint) & NEXT_NOTWORD_CONSTRAINT) && IS_WORD_CONTEXT (context)) \
|| (((constraint) & NEXT_NEWLINE_CONSTRAINT) && !IS_NEWLINE_CONTEXT (context)) \
|| (((constraint) & NEXT_ENDBUF_CONSTRAINT) && !IS_ENDBUF_CONTEXT (context)))
struct re_dfastate_t
{
unsigned int hash;
re_node_set nodes;
re_node_set non_eps_nodes;
re_node_set inveclosure;
re_node_set *entrance_nodes;
struct re_dfastate_t **trtable, **word_trtable;
unsigned int context : 4;
unsigned int halt : 1;
/* If this state can accept `multi byte'.
Note that we refer to multibyte characters, and multi character
collating elements as `multi byte'. */
unsigned int accept_mb : 1;
/* If this state has backreference node(s). */
unsigned int has_backref : 1;
unsigned int has_constraint : 1;
};
typedef struct re_dfastate_t re_dfastate_t;
struct re_state_table_entry
{
int num;
int alloc;
re_dfastate_t **array;
};
/* Array type used in re_sub_match_last_t and re_sub_match_top_t. */
typedef struct
{
int next_idx;
int alloc;
re_dfastate_t **array;
} state_array_t;
/* Store information about the node NODE whose type is OP_CLOSE_SUBEXP. */
typedef struct
{
int node;
int str_idx; /* The position NODE match at. */
state_array_t path;
} re_sub_match_last_t;
/* Store information about the node NODE whose type is OP_OPEN_SUBEXP.
And information about the node, whose type is OP_CLOSE_SUBEXP,
corresponding to NODE is stored in LASTS. */
typedef struct
{
int str_idx;
int node;
state_array_t *path;
int alasts; /* Allocation size of LASTS. */
int nlasts; /* The number of LASTS. */
re_sub_match_last_t **lasts;
} re_sub_match_top_t;
struct re_backref_cache_entry
{
int node;
int str_idx;
int subexp_from;
int subexp_to;
char more;
char unused;
unsigned short int eps_reachable_subexps_map;
};
typedef struct
{
/* The string object corresponding to the input string. */
re_string_t input;
#if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)
const re_dfa_t *const dfa;
#else
const re_dfa_t *dfa;
#endif
/* EFLAGS of the argument of regexec. */
int eflags;
/* Where the matching ends. */
int match_last;
int last_node;
/* The state log used by the matcher. */
re_dfastate_t **state_log;
int state_log_top;
/* Back reference cache. */
int nbkref_ents;
int abkref_ents;
struct re_backref_cache_entry *bkref_ents;
int max_mb_elem_len;
int nsub_tops;
int asub_tops;
re_sub_match_top_t **sub_tops;
} re_match_context_t;
typedef struct
{
re_dfastate_t **sifted_states;
re_dfastate_t **limited_states;
int last_node;
int last_str_idx;
re_node_set limits;
} re_sift_context_t;
struct re_fail_stack_ent_t
{
int idx;
int node;
regmatch_t *regs;
re_node_set eps_via_nodes;
};
struct re_fail_stack_t
{
int num;
int alloc;
struct re_fail_stack_ent_t *stack;
};
struct re_dfa_t
{
re_token_t *nodes;
size_t nodes_alloc;
size_t nodes_len;
int *nexts;
int *org_indices;
re_node_set *edests;
re_node_set *eclosures;
re_node_set *inveclosures;
struct re_state_table_entry *state_table;
re_dfastate_t *init_state;
re_dfastate_t *init_state_word;
re_dfastate_t *init_state_nl;
re_dfastate_t *init_state_begbuf;
bin_tree_t *str_tree;
bin_tree_storage_t *str_tree_storage;
re_bitset_ptr_t sb_char;
int str_tree_storage_idx;
/* number of subexpressions `re_nsub' is in regex_t. */
unsigned int state_hash_mask;
int init_node;
int nbackref; /* The number of backreference in this dfa. */
/* Bitmap expressing which backreference is used. */
bitset_word_t used_bkref_map;
bitset_word_t completed_bkref_map;
unsigned int has_plural_match : 1;
/* If this dfa has "multibyte node", which is a backreference or
a node which can accept multibyte character or multi character
collating element. */
unsigned int has_mb_node : 1;
unsigned int is_utf8 : 1;
unsigned int map_notascii : 1;
unsigned int word_ops_used : 1;
int mb_cur_max;
bitset_t word_char;
reg_syntax_t syntax;
int *subexp_map;
#ifdef DEBUG
char* re_str;
#endif
#if defined _LIBC
__libc_lock_define (, lock)
#endif
};
#define re_node_set_init_empty(set) memset (set, '\0', sizeof (re_node_set))
#define re_node_set_remove(set,id) \
(re_node_set_remove_at (set, re_node_set_contains (set, id) - 1))
#define re_node_set_empty(p) ((p)->nelem = 0)
#define re_node_set_free(set) re_free ((set)->elems)
typedef enum
{
SB_CHAR,
MB_CHAR,
EQUIV_CLASS,
COLL_SYM,
CHAR_CLASS
} bracket_elem_type;
typedef struct
{
bracket_elem_type type;
union
{
unsigned char ch;
unsigned char *name;
wchar_t wch;
} opr;
} bracket_elem_t;
/* Inline functions for bitset operation. */
static inline void
bitset_not (bitset_t set)
{
int bitset_i;
for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
set[bitset_i] = ~set[bitset_i];
}
static inline void
bitset_merge (bitset_t dest, const bitset_t src)
{
int bitset_i;
for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
dest[bitset_i] |= src[bitset_i];
}
static inline void
bitset_mask (bitset_t dest, const bitset_t src)
{
int bitset_i;
for (bitset_i = 0; bitset_i < BITSET_WORDS; ++bitset_i)
dest[bitset_i] &= src[bitset_i];
}
#ifdef RE_ENABLE_I18N
/* Inline functions for re_string. */
static inline int
internal_function __attribute ((pure))
re_string_char_size_at (const re_string_t *pstr, int idx)
{
int byte_idx;
if (pstr->mb_cur_max == 1)
return 1;
for (byte_idx = 1; idx + byte_idx < pstr->valid_len; ++byte_idx)
if (pstr->wcs[idx + byte_idx] != WEOF)
break;
return byte_idx;
}
static inline wint_t
internal_function __attribute ((pure))
re_string_wchar_at (const re_string_t *pstr, int idx)
{
if (pstr->mb_cur_max == 1)
return (wint_t) pstr->mbs[idx];
return (wint_t) pstr->wcs[idx];
}
# ifndef NOT_IN_libc
static int
internal_function __attribute ((pure))
re_string_elem_size_at (const re_string_t *pstr, int idx)
{
# ifdef _LIBC
const unsigned char *p, *extra;
const int32_t *table, *indirect;
int32_t tmp;
# include <locale/weight.h>
uint_fast32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
if (nrules != 0)
{
table = (const int32_t *) _NL_CURRENT (LC_COLLATE, _NL_COLLATE_TABLEMB);
extra = (const unsigned char *)
_NL_CURRENT (LC_COLLATE, _NL_COLLATE_EXTRAMB);
indirect = (const int32_t *) _NL_CURRENT (LC_COLLATE,
_NL_COLLATE_INDIRECTMB);
p = pstr->mbs + idx;
tmp = findidx (&p);
return p - pstr->mbs - idx;
}
else
# endif /* _LIBC */
return 1;
}
# endif
#endif /* RE_ENABLE_I18N */
#endif /* _REGEX_INTERNAL_H */

4369
compat/regex/regexec.c Normal file
View File

File diff suppressed because it is too large Load Diff

1
config.mak.in
View File

@ -59,6 +59,7 @@ NO_INET_NTOP=@NO_INET_NTOP@
NO_INET_PTON=@NO_INET_PTON@
NO_ICONV=@NO_ICONV@
OLD_ICONV=@OLD_ICONV@
NO_REGEX=@NO_REGEX@
NO_DEFLATE_BOUND=@NO_DEFLATE_BOUND@
INLINE=@INLINE@
SOCKLEN_T=@SOCKLEN_T@

21
configure.ac
View File

@ -706,6 +706,27 @@ else
fi
AC_SUBST(NO_C99_FORMAT)
#
# Define NO_REGEX if you have no or inferior regex support in your C library.
AC_CACHE_CHECK([whether the platform regex can handle null bytes],
[ac_cv_c_excellent_regex], [
AC_EGREP_CPP(yippeeyeswehaveit,
AC_LANG_PROGRAM([AC_INCLUDES_DEFAULT
#include <regex.h>
],
[#ifdef REG_STARTEND
yippeeyeswehaveit
#endif
]),
[ac_cv_c_excellent_regex=yes],
[ac_cv_c_excellent_regex=no])
])
if test $ac_cv_c_excellent_regex = yes; then
NO_REGEX=
else
NO_REGEX=YesPlease
fi
AC_SUBST(NO_REGEX)
#
# Define FREAD_READS_DIRECTORIES if your are on a system which succeeds
# when attempting to read from an fopen'ed directory.
AC_CACHE_CHECK([whether system succeeds to read fopen'ed directory],

2
t/t7008-grep-binary.sh
View File

@ -61,7 +61,7 @@ test_expect_success 'git grep -Fi iLE a' '
# This test actually passes on platforms where regexec() supports the
# flag REG_STARTEND.
test_expect_failure 'git grep ile a' '
test_expect_success 'git grep ile a' '
git grep ile a
'