lib/glibc
1
0
Fork 0
mirror of https://sourceware.org/git/glibc.git synced 2025-07-29 11:41:21 +03:00
Code Activity

Update.

Browse Source 
2002-05-21  Isamu Hasegawa  <isamu@yamato.ibm.com>

	* posix/regex.c: Define `inline' as a macro into nothing for the
	compilers which lack the keyword.
	* posix/regex.h: (RE_SYNTAX_GNU_AWK): Remove RE_CONTEXT_INVALID_OPS
	for the compatibility of gawk.
	* posix/regcomp.c: Add fake implementation of isblank() for the
	environments which lack the function.
	Don't use free_charset() in case of non-i18n envs.
	(build_range_exp): Don't use i18n related code in case of non-i18n
	envs.
	(build_collating_symbol): Likewise.
	(build_equiv_class): Likewise.
	(build_charclass): Likewise.
	(re_compile_fastmap_iter): Likewise.
	(parse_bracket_exp): Likewise.
	(build_word_op): Likewise.
	(regfree): Don't use free_charset() in case of non-i18n envs.
	* posix/regex_internal.h: Remove COMPLEX_BRACKET from
	re_token_type_t in case of non-i18n envs.
	Don't define re_charset_t in case of non-i18n envs.
	Change the type of wcs of re_string_t from wchar_t to wint_t,
	since we store also WEOF.
	* posix/regex_internal.c: (re_string_realloc_buffers): Change
	the type of wcs of re_string_t from wchar_t to wint_t.
	(re_string_reconstruct): Likewise.
	(create_ci_newstate): Don't use i18n related code in case of
	non-i18n envs.
	(create_cd_newstate): Likewise.

2002-05-24  Ulrich Drepper  <drepper@redhat.com>

	* iconv/loop.c: Fix typo.

2002-05-23  Jakub Jelinek  <jakub@redhat.com>

	* inet/ether_line.c (ether_line): Fix a typo causing only
	lower 4 bits of each ethernet address byte being assigned.
	Don't modify what line points to.
	* inet/tst-ether_aton.c (main): Add ether_line tests.

2002-05-23  Marcus Brinkmann  <marcus@gnu.org>

	* manual/filesys.texi: Don't make readlink example leak memory
	when readlink fails.
This commit is contained in:
Ulrich Drepper
2002-05-24 08:49:00 +00:00
parent 12b7f470d2
commit c0a0f9a32c
12 changed files with 322 additions and 88 deletions
Download Patch File Download Diff File Expand all files Collapse all files

46
ChangeLog
View File

@ -1,3 +1,49 @@
2002-05-21 Isamu Hasegawa <isamu@yamato.ibm.com>
* posix/regex.c: Define `inline' as a macro into nothing for the
compilers which lack the keyword.
* posix/regex.h: (RE_SYNTAX_GNU_AWK): Remove RE_CONTEXT_INVALID_OPS
for the compatibility of gawk.
* posix/regcomp.c: Add fake implementation of isblank() for the
environments which lack the function.
Don't use free_charset() in case of non-i18n envs.
(build_range_exp): Don't use i18n related code in case of non-i18n
envs.
(build_collating_symbol): Likewise.
(build_equiv_class): Likewise.
(build_charclass): Likewise.
(re_compile_fastmap_iter): Likewise.
(parse_bracket_exp): Likewise.
(build_word_op): Likewise.
(regfree): Don't use free_charset() in case of non-i18n envs.
* posix/regex_internal.h: Remove COMPLEX_BRACKET from
re_token_type_t in case of non-i18n envs.
Don't define re_charset_t in case of non-i18n envs.
Change the type of wcs of re_string_t from wchar_t to wint_t,
since we store also WEOF.
* posix/regex_internal.c: (re_string_realloc_buffers): Change
the type of wcs of re_string_t from wchar_t to wint_t.
(re_string_reconstruct): Likewise.
(create_ci_newstate): Don't use i18n related code in case of
non-i18n envs.
(create_cd_newstate): Likewise.
2002-05-24 Ulrich Drepper <drepper@redhat.com>
* iconv/loop.c: Fix typo.
2002-05-23 Jakub Jelinek <jakub@redhat.com>
* inet/ether_line.c (ether_line): Fix a typo causing only
lower 4 bits of each ethernet address byte being assigned.
Don't modify what line points to.
* inet/tst-ether_aton.c (main): Add ether_line tests.
2002-05-23 Marcus Brinkmann <marcus@gnu.org>
* manual/filesys.texi: Don't make readlink example leak memory
when readlink fails.
2002-05-20 Ulrich Drepper <drepper@redhat.com> 2002-05-20 Ulrich Drepper <drepper@redhat.com>
* iconv/loop.c: Fix condition for defining unaligned loop. Add * iconv/loop.c: Fix condition for defining unaligned loop. Add

4
iconv/loop.c
View File

@ -145,8 +145,8 @@
/* We produce at least one byte in the next round. */ /* We produce at least one byte in the next round. */
#ifndef MIN_NEEDED_OUTPUT #ifndef MIN_NEEDED_OUTPUT
# error "MIN_NEEDED_OUTPUT definition missing" # error "MIN_NEEDED_OUTPUT definition missing"
#elif MIN_NEEDED_INPUT < 1 #elif MIN_NEEDED_OUTPUT < 1
# error "MIN_NEEDED_INPUT must be >= 1" # error "MIN_NEEDED_OUTPUT must be >= 1"
#endif #endif
/* Let's see how many bytes we produce. */ /* Let's see how many bytes we produce. */

10
inet/ether_line.c
View File

@ -1,4 +1,4 @@
/* Copyright (C) 1996, 1999 Free Software Foundation, Inc. /* Copyright (C) 1996, 1999, 2002 Free Software Foundation, Inc.
This file is part of the GNU C Library. This file is part of the GNU C Library.
The GNU C Library is free software; you can redistribute it and/or The GNU C Library is free software; you can redistribute it and/or
@ -46,7 +46,7 @@ ether_line (const char *line, struct ether_addr *addr, char *hostname)
if ((ch < '0' || ch > '9') && (ch < 'a' || ch > 'f')) if ((ch < '0' || ch > '9') && (ch < 'a' || ch > 'f'))
return -1; return -1;
number <<= 4; number <<= 4;
number = isdigit (ch) ? (ch - '0') : (ch - 'a' + 10); number += isdigit (ch) ? (ch - '0') : (ch - 'a' + 10);
ch = *line; ch = *line;
if (cnt < 5 && ch != ':') if (cnt < 5 && ch != ':')
@ -65,15 +65,15 @@ ether_line (const char *line, struct ether_addr *addr, char *hostname)
cp = __strchrnul (line, '#'); cp = __strchrnul (line, '#');
while (cp > line && isspace (cp[-1])) while (cp > line && isspace (cp[-1]))
--cp; --cp;
*cp = '\0';
if (*line == '\0') if (cp == line)
/* No hostname. */ /* No hostname. */
return -1; return -1;
/* XXX This can cause trouble because the hostname might be too long /* XXX This can cause trouble because the hostname might be too long
but we have no possibility to check it here. */ but we have no possibility to check it here. */
strcpy (hostname, line); memcpy (hostname, line, cp - line);
hostname [cp - line] = '\0';
return 0; return 0;
} }

58
inet/tst-ether_aton.c
View File

@ -1,29 +1,57 @@
#include <stdio.h> #include <stdio.h>
#include <string.h>
#include <netinet/ether.h> #include <netinet/ether.h>
int int
main (int argc, char *argv[]) main (int argc, char *argv[])
{ {
struct ether_addr *val; struct ether_addr *valp, val;
int result; int result, r;
char hostname[32], buf[64], *p;
val = ether_aton ("12:34:56:78:9a:bc"); valp = ether_aton ("12:34:56:78:9a:bc");
printf ("ether_aton (\"12:34:56:78:9a:bc\") = %hhx:%hhx:%hhx:%hhx:%hhx:%hhx\n", printf ("ether_aton (\"12:34:56:78:9a:bc\") = %hhx:%hhx:%hhx:%hhx:%hhx:%hhx\n",
val->ether_addr_octet[0], valp->ether_addr_octet[0],
val->ether_addr_octet[1], valp->ether_addr_octet[1],
val->ether_addr_octet[2], valp->ether_addr_octet[2],
val->ether_addr_octet[3], valp->ether_addr_octet[3],
val->ether_addr_octet[4], valp->ether_addr_octet[4],
val->ether_addr_octet[5]); valp->ether_addr_octet[5]);
result = (valp->ether_addr_octet[0] != 0x12
|| valp->ether_addr_octet[1] != 0x34
|| valp->ether_addr_octet[2] != 0x56
|| valp->ether_addr_octet[3] != 0x78
|| valp->ether_addr_octet[4] != 0x9a
|| valp->ether_addr_octet[5] != 0xbc);
result = (val->ether_addr_octet[0] != 0x12 if ((r = ether_line ("0:c0:f0:46:5f:97 host.ether.com \t# comment",
|| val->ether_addr_octet[1] != 0x34 &val, hostname)) == 0)
|| val->ether_addr_octet[2] != 0x56 {
|| val->ether_addr_octet[3] != 0x78 ether_ntoa_r (&val, buf);
|| val->ether_addr_octet[4] != 0x9a p = strchr (buf, '\0');
|| val->ether_addr_octet[5] != 0xbc); *p++ = ' ';
strcpy (p, hostname);
printf ("ether_line (\"0:c0:f0:46:5f:97 host.ether.com\") = \"%s\"\n",
buf);
result |= strcmp ("0:c0:f0:46:5f:97 host.ether.com", buf) != 0;
}
else
{
printf ("ether_line (\"0:c0:f0:46:5f:97 host.ether.com\") = %d\n", r);
result |= 1;
}
r = ether_line ("0:c0:2:d0 foo.bar ", &val, hostname);
printf ("ether_line (\"0:c0:2:d0 foo.bar \") = %d\n", r);
result |= r != -1;
r = ether_line ("0:c0:2:d0:1a:2a ", &val, hostname);
printf ("ether_line (\"0:c0:2:d0:1a:2a \") = %d\n", r);
result |= r != -1;
return result; return result;
} }

4
linuxthreads/ChangeLog
View File

@ -1,3 +1,7 @@
2002-05-24 Ulrich Drepper <drepper@redhat.com>
* sysdeps/ia64/pt-machine.h (MEMORY_BARRIER): Use __sync_synchronize.
2002-05-21 Ulrich Drepper <drepper@redhat.com> 2002-05-21 Ulrich Drepper <drepper@redhat.com>
* sysdeps/pthread/pthread.h (pthread_create): Rename first * sysdeps/pthread/pthread.h (pthread_create): Rename first

4
linuxthreads/sysdeps/ia64/pt-machine.h
View File

@ -21,6 +21,8 @@
#ifndef _PT_MACHINE_H #ifndef _PT_MACHINE_H
#define _PT_MACHINE_H 1 #define _PT_MACHINE_H 1
#include <ia64intrin.h>
#ifndef PT_EI #ifndef PT_EI
# define PT_EI extern inline # define PT_EI extern inline
#endif #endif
@ -69,7 +71,7 @@ register struct _pthread_descr_struct *__thread_self __asm__("r13");
/* Memory barrier */ /* Memory barrier */
#define MEMORY_BARRIER() __asm__ __volatile__("mf" : : : "memory") #define MEMORY_BARRIER() __sync_syncronize ()
#define HAS_COMPARE_AND_SWAP_WITH_RELEASE_SEMANTICS #define HAS_COMPARE_AND_SWAP_WITH_RELEASE_SEMANTICS

9
manual/filesys.texi
View File

@ -1141,16 +1141,19 @@ char *
readlink_malloc (const char *filename) readlink_malloc (const char *filename)
@{ @{
int size = 100; int size = 100;
char *buffer = NULL;
while (1) while (1)
@{ @{
char *buffer = (char *) xmalloc (size); buffer = (char *) xrealloc (buffer, size);
int nchars = readlink (filename, buffer, size); int nchars = readlink (filename, buffer, size);
if (nchars < 0) if (nchars < 0)
return NULL; @{
free (buffer);
return NULL;
@}
if (nchars < size) if (nchars < size)
return buffer; return buffer;
free (buffer);
size *= 2; size *= 2;
@} @}
@} @}

239
posix/regcomp.c
View File

@ -28,6 +28,11 @@
#include <wchar.h> #include <wchar.h>
#include <wctype.h> #include <wctype.h>
/* In case that the system doesn't have isblank(). */
#if !defined _LIBC && !defined HAVE_ISBLANK && !defined isblank
# define isblank(ch) ((ch) == ' ' || (ch) == '\t')
#endif
#ifdef _LIBC #ifdef _LIBC
# ifndef _RE_DEFINE_LOCALE_FUNCTIONS # ifndef _RE_DEFINE_LOCALE_FUNCTIONS
# define _RE_DEFINE_LOCALE_FUNCTIONS 1 # define _RE_DEFINE_LOCALE_FUNCTIONS 1
@ -65,7 +70,9 @@ static void re_compile_fastmap_iter (regex_t *bufp,
char *fastmap); char *fastmap);
static reg_errcode_t init_dfa (re_dfa_t *dfa, int pat_len); static reg_errcode_t init_dfa (re_dfa_t *dfa, int pat_len);
static reg_errcode_t init_word_char (re_dfa_t *dfa); static reg_errcode_t init_word_char (re_dfa_t *dfa);
#ifdef RE_ENABLE_I18N
static void free_charset (re_charset_t *cset); static void free_charset (re_charset_t *cset);
#endif /* RE_ENABLE_I18N */
static void free_workarea_compile (regex_t *preg); static void free_workarea_compile (regex_t *preg);
static reg_errcode_t create_initial_state (re_dfa_t *dfa); static reg_errcode_t create_initial_state (re_dfa_t *dfa);
static reg_errcode_t analyze (re_dfa_t *dfa); static reg_errcode_t analyze (re_dfa_t *dfa);
@ -115,24 +122,40 @@ static reg_errcode_t parse_bracket_symbol (bracket_elem_t *elem,
re_string_t *regexp, re_string_t *regexp,
re_token_t *token); re_token_t *token);
#ifndef _LIBC #ifndef _LIBC
static reg_errcode_t build_range_exp (re_charset_t *mbcset, # ifdef RE_ENABLE_I18N
re_bitset_ptr_t sbcset, int *range_alloc, static reg_errcode_t build_range_exp (re_bitset_ptr_t sbcset,
re_charset_t *mbcset, int *range_alloc,
bracket_elem_t *start_elem, bracket_elem_t *start_elem,
bracket_elem_t *end_elem); bracket_elem_t *end_elem);
static reg_errcode_t build_collating_symbol (re_charset_t *mbcset, static reg_errcode_t build_collating_symbol (re_bitset_ptr_t sbcset,
re_bitset_ptr_t sbcset, re_charset_t *mbcset,
int *coll_sym_alloc, int *coll_sym_alloc,
unsigned char *name); unsigned char *name);
# else /* not RE_ENABLE_I18N */
static reg_errcode_t build_range_exp (re_bitset_ptr_t sbcset,
bracket_elem_t *start_elem,
bracket_elem_t *end_elem);
static reg_errcode_t build_collating_symbol (re_bitset_ptr_t sbcset,
unsigned char *name);
# endif /* not RE_ENABLE_I18N */
#endif /* not _LIBC */ #endif /* not _LIBC */
static reg_errcode_t build_equiv_class (re_charset_t *mbcset, #ifdef RE_ENABLE_I18N
re_bitset_ptr_t sbcset, static reg_errcode_t build_equiv_class (re_bitset_ptr_t sbcset,
re_charset_t *mbcset,
int *equiv_class_alloc, int *equiv_class_alloc,
const unsigned char *name); const unsigned char *name);
static reg_errcode_t build_charclass (re_charset_t *mbcset, static reg_errcode_t build_charclass (re_bitset_ptr_t sbcset,
re_bitset_ptr_t sbcset, re_charset_t *mbcset,
int *char_class_alloc, int *char_class_alloc,
const unsigned char *class_name, const unsigned char *class_name,
reg_syntax_t syntax); reg_syntax_t syntax);
#else /* not RE_ENABLE_I18N */
static reg_errcode_t build_equiv_class (re_bitset_ptr_t sbcset,
const unsigned char *name);
static reg_errcode_t build_charclass (re_bitset_ptr_t sbcset,
const unsigned char *class_name,
reg_syntax_t syntax);
#endif /* not RE_ENABLE_I18N */
static bin_tree_t *build_word_op (re_dfa_t *dfa, int not, reg_errcode_t *err); static bin_tree_t *build_word_op (re_dfa_t *dfa, int not, reg_errcode_t *err);
static void free_bin_tree (bin_tree_t *tree); static void free_bin_tree (bin_tree_t *tree);
static bin_tree_t *create_tree (bin_tree_t *left, bin_tree_t *right, static bin_tree_t *create_tree (bin_tree_t *left, bin_tree_t *right,
@ -340,6 +363,7 @@ re_compile_fastmap_iter (bufp, init_state, fastmap)
if (dfa->nodes[node].opr.sbcset[i] & (1 << j)) if (dfa->nodes[node].opr.sbcset[i] & (1 << j))
fastmap[ch] = 1; fastmap[ch] = 1;
} }
#ifdef RE_ENABLE_I18N
else if (type == COMPLEX_BRACKET) else if (type == COMPLEX_BRACKET)
{ {
int i; int i;
@ -347,7 +371,7 @@ re_compile_fastmap_iter (bufp, init_state, fastmap)
if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes if (cset->non_match || cset->ncoll_syms || cset->nequiv_classes
|| cset->nranges || cset->nchar_classes) || cset->nranges || cset->nchar_classes)
{ {
#ifdef _LIBC # ifdef _LIBC
if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0) if (_NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES) != 0)
{ {
/* In this case we want to catch the bytes which are /* In this case we want to catch the bytes which are
@ -364,14 +388,12 @@ re_compile_fastmap_iter (bufp, init_state, fastmap)
if (table[ch] < 0) if (table[ch] < 0)
fastmap[ch] = 1; fastmap[ch] = 1;
} }
#else # else
# ifdef RE_ENABLE_I18N
if (MB_CUR_MAX > 1) if (MB_CUR_MAX > 1)
for (i = 0; i < SBC_MAX; ++i) for (i = 0; i < SBC_MAX; ++i)
if (__btowc (i) == WEOF) if (__btowc (i) == WEOF)
fastmap[i] = 1; fastmap[i] = 1;
# endif /* RE_ENABLE_I18N */ # endif /* not _LIBC */
#endif /* not _LIBC */
} }
for (i = 0; i < cset->nmbchars; ++i) for (i = 0; i < cset->nmbchars; ++i)
{ {
@ -380,8 +402,12 @@ re_compile_fastmap_iter (bufp, init_state, fastmap)
fastmap[buf[0]] = 1; fastmap[buf[0]] = 1;
} }
} }
else if (type == END_OF_RE || type == COMPLEX_BRACKET #endif /* RE_ENABLE_I18N */
|| type == OP_PERIOD) else if (type == END_OF_RE || type == OP_PERIOD
#ifdef RE_ENABLE_I18N
|| type == COMPLEX_BRACKET
#endif /* RE_ENABLE_I18N */
)
{ {
memset (fastmap, '\1', sizeof (char) * SBC_MAX); memset (fastmap, '\1', sizeof (char) * SBC_MAX);
if (type == END_OF_RE) if (type == END_OF_RE)
@ -550,9 +576,12 @@ regfree (preg)
for (i = 0; i < dfa->nodes_len; ++i) for (i = 0; i < dfa->nodes_len; ++i)
{ {
re_token_t *node = dfa->nodes + i; re_token_t *node = dfa->nodes + i;
#ifdef RE_ENABLE_I18N
if (node->type == COMPLEX_BRACKET && node->duplicated == 0) if (node->type == COMPLEX_BRACKET && node->duplicated == 0)
free_charset (node->opr.mbcset); free_charset (node->opr.mbcset);
else if (node->type == SIMPLE_BRACKET && node->duplicated == 0) else
#endif /* RE_ENABLE_I18N */
if (node->type == SIMPLE_BRACKET && node->duplicated == 0)
re_free (node->opr.sbcset); re_free (node->opr.sbcset);
else if (node->type == OP_CONTEXT_NODE) else if (node->type == OP_CONTEXT_NODE)
{ {
@ -993,7 +1022,9 @@ calc_first (dfa, node)
case OP_PERIOD: case OP_PERIOD:
case OP_DUP_ASTERISK: case OP_DUP_ASTERISK:
case OP_DUP_QUESTION: case OP_DUP_QUESTION:
#ifdef RE_ENABLE_I18N
case COMPLEX_BRACKET: case COMPLEX_BRACKET:
#endif /* RE_ENABLE_I18N */
case SIMPLE_BRACKET: case SIMPLE_BRACKET:
case OP_BACK_REF: case OP_BACK_REF:
case ANCHOR: case ANCHOR:
@ -2233,10 +2264,14 @@ parse_dup_op (dup_elem, regexp, dfa, token, syntax, err)
update it. */ update it. */
static reg_errcode_t static reg_errcode_t
build_range_exp (mbcset, sbcset, range_alloc, start_elem, end_elem) # ifdef RE_ENABLE_I18N
build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem)
re_charset_t *mbcset; re_charset_t *mbcset;
re_bitset_ptr_t sbcset;
int *range_alloc; int *range_alloc;
# else /* not RE_ENABLE_I18N */
build_range_exp (sbcset, start_elem, end_elem)
# endif /* not RE_ENABLE_I18N */
re_bitset_ptr_t sbcset;
bracket_elem_t *start_elem, *end_elem; bracket_elem_t *start_elem, *end_elem;
{ {
unsigned int start_ch, end_ch; unsigned int start_ch, end_ch;
@ -2338,10 +2373,14 @@ build_range_exp (mbcset, sbcset, range_alloc, start_elem, end_elem)
pointer argument since we may update it. */ pointer argument since we may update it. */
static reg_errcode_t static reg_errcode_t
build_collating_symbol (mbcset, sbcset, coll_sym_alloc, name) # ifdef RE_ENABLE_I18N
build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name)
re_charset_t *mbcset; re_charset_t *mbcset;
re_bitset_ptr_t sbcset;
int *coll_sym_alloc; int *coll_sym_alloc;
# else /* not RE_ENABLE_I18N */
build_collating_symbol (sbcset, name)
# endif /* not RE_ENABLE_I18N */
re_bitset_ptr_t sbcset;
unsigned char *name; unsigned char *name;
{ {
if (BE (strlen (name) != 1, 0)) if (BE (strlen (name) != 1, 0))
@ -2476,16 +2515,21 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
update it. */ update it. */
static inline reg_errcode_t static inline reg_errcode_t
build_range_exp (mbcset, sbcset, range_alloc, start_elem, end_elem) # ifdef RE_ENABLE_I18N
build_range_exp (sbcset, mbcset, range_alloc, start_elem, end_elem)
re_charset_t *mbcset; re_charset_t *mbcset;
re_bitset_ptr_t sbcset;
int *range_alloc; int *range_alloc;
# else /* not RE_ENABLE_I18N */
build_range_exp (sbcset, start_elem, end_elem)
# endif /* not RE_ENABLE_I18N */
re_bitset_ptr_t sbcset;
bracket_elem_t *start_elem, *end_elem; bracket_elem_t *start_elem, *end_elem;
{ {
unsigned int ch; unsigned int ch;
uint32_t start_collseq; uint32_t start_collseq;
uint32_t end_collseq; uint32_t end_collseq;
# ifdef RE_ENABLE_I18N
/* Check the space of the arrays. */ /* Check the space of the arrays. */
if (*range_alloc == mbcset->nranges) if (*range_alloc == mbcset->nranges)
{ {
@ -2510,6 +2554,7 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
mbcset->range_ends = new_array_end; mbcset->range_ends = new_array_end;
*range_alloc = new_nranges; *range_alloc = new_nranges;
} }
# endif /* RE_ENABLE_I18N */
/* Equivalence Classes and Character Classes can't be a range /* Equivalence Classes and Character Classes can't be a range
start/end. */ start/end. */
@ -2526,9 +2571,11 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
if (BE ((syntax & RE_NO_EMPTY_RANGES) && start_collseq > end_collseq, 0)) if (BE ((syntax & RE_NO_EMPTY_RANGES) && start_collseq > end_collseq, 0))
return REG_ERANGE; return REG_ERANGE;
# ifdef RE_ENABLE_I18N
/* Got valid collation sequence values, add them as a new entry. */ /* Got valid collation sequence values, add them as a new entry. */
mbcset->range_starts[mbcset->nranges] = start_collseq; mbcset->range_starts[mbcset->nranges] = start_collseq;
mbcset->range_ends[mbcset->nranges++] = end_collseq; mbcset->range_ends[mbcset->nranges++] = end_collseq;
# endif /* RE_ENABLE_I18N */
/* Build the table for single byte characters. */ /* Build the table for single byte characters. */
for (ch = 0; ch <= SBC_MAX; ch++) for (ch = 0; ch <= SBC_MAX; ch++)
@ -2554,10 +2601,14 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
pointer argument sinse we may update it. */ pointer argument sinse we may update it. */
static inline reg_errcode_t static inline reg_errcode_t
build_collating_symbol (mbcset, sbcset, coll_sym_alloc, name) # ifdef RE_ENABLE_I18N
build_collating_symbol (sbcset, mbcset, coll_sym_alloc, name)
re_charset_t *mbcset; re_charset_t *mbcset;
re_bitset_ptr_t sbcset;
int *coll_sym_alloc; int *coll_sym_alloc;
# else /* not RE_ENABLE_I18N */
build_collating_symbol (sbcset, name)
# endif /* not RE_ENABLE_I18N */
re_bitset_ptr_t sbcset;
unsigned char *name; unsigned char *name;
{ {
int32_t elem, idx; int32_t elem, idx;
@ -2581,6 +2632,7 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
else else
return REG_ECOLLATE; return REG_ECOLLATE;
# ifdef RE_ENABLE_I18N
/* Got valid collation sequence, add it as a new entry. */ /* Got valid collation sequence, add it as a new entry. */
/* Check the space of the arrays. */ /* Check the space of the arrays. */
if (*coll_sym_alloc == mbcset->ncoll_syms) if (*coll_sym_alloc == mbcset->ncoll_syms)
@ -2596,6 +2648,7 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
return REG_ESPACE; return REG_ESPACE;
} }
mbcset->coll_syms[mbcset->ncoll_syms++] = idx; mbcset->coll_syms[mbcset->ncoll_syms++] = idx;
# endif /* RE_ENABLE_I18N */
return REG_NOERROR; return REG_NOERROR;
} }
else else
@ -2613,11 +2666,15 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
re_token_t br_token; re_token_t br_token;
re_bitset_ptr_t sbcset; re_bitset_ptr_t sbcset;
#ifdef RE_ENABLE_I18N
re_charset_t *mbcset; re_charset_t *mbcset;
bin_tree_t *work_tree;
int token_len, new_idx;
int coll_sym_alloc = 0, range_alloc = 0, mbchar_alloc = 0; int coll_sym_alloc = 0, range_alloc = 0, mbchar_alloc = 0;
int equiv_class_alloc = 0, char_class_alloc = 0; int equiv_class_alloc = 0, char_class_alloc = 0;
#else /* not RE_ENABLE_I18N */
int non_match = 0;
#endif /* not RE_ENABLE_I18N */
bin_tree_t *work_tree;
int token_len, new_idx;
#ifdef _LIBC #ifdef _LIBC
collseqmb = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB); collseqmb = _NL_CURRENT (LC_COLLATE, _NL_COLLATE_COLLSEQMB);
nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
@ -2635,8 +2692,14 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
} }
#endif #endif
sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS); sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS);
#ifdef RE_ENABLE_I18N
mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1); mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1);
#endif /* RE_ENABLE_I18N */
#ifdef RE_ENABLE_I18N
if (BE (sbcset == NULL || mbcset == NULL, 0)) if (BE (sbcset == NULL || mbcset == NULL, 0))
#else
if (BE (sbcset == NULL, 0))
#endif /* RE_ENABLE_I18N */
{ {
*err = REG_ESPACE; *err = REG_ESPACE;
return NULL; return NULL;
@ -2650,8 +2713,12 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
} }
if (token->type == OP_NON_MATCH_LIST) if (token->type == OP_NON_MATCH_LIST)
{ {
#ifdef RE_ENABLE_I18N
int i; int i;
mbcset->non_match = 1; mbcset->non_match = 1;
#else /* not RE_ENABLE_I18N */
non_match = 1;
#endif /* not RE_ENABLE_I18N */
if (syntax & RE_HAT_LISTS_NOT_NEWLINE) if (syntax & RE_HAT_LISTS_NOT_NEWLINE)
bitset_set (sbcset, '\0'); bitset_set (sbcset, '\0');
re_string_skip_bytes (regexp, token_len); /* Skip a token. */ re_string_skip_bytes (regexp, token_len); /* Skip a token. */
@ -2661,10 +2728,12 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
*err = REG_BADPAT; *err = REG_BADPAT;
goto parse_bracket_exp_free_return; goto parse_bracket_exp_free_return;
} }
#ifdef RE_ENABLE_I18N
if (MB_CUR_MAX > 1) if (MB_CUR_MAX > 1)
for (i = 0; i < SBC_MAX; ++i) for (i = 0; i < SBC_MAX; ++i)
if (__btowc (i) == WEOF) if (__btowc (i) == WEOF)
bitset_set (sbcset, i); bitset_set (sbcset, i);
#endif /* RE_ENABLE_I18N */
} }
/* We treat the first ']' as a normal character. */ /* We treat the first ']' as a normal character. */
@ -2731,8 +2800,11 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
*err = REG_BADPAT; *err = REG_BADPAT;
goto parse_bracket_exp_free_return; goto parse_bracket_exp_free_return;
} }
*err = build_range_exp (mbcset, sbcset, &range_alloc, &start_elem, *err = build_range_exp (sbcset,
&end_elem); #ifdef RE_ENABLE_I18N
mbcset, &range_alloc,
#endif /* RE_ENABLE_I18N */
&start_elem, &end_elem);
if (BE (*err != REG_NOERROR, 0)) if (BE (*err != REG_NOERROR, 0))
goto parse_bracket_exp_free_return; goto parse_bracket_exp_free_return;
} }
@ -2743,6 +2815,7 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
case SB_CHAR: case SB_CHAR:
bitset_set (sbcset, start_elem.opr.ch); bitset_set (sbcset, start_elem.opr.ch);
break; break;
#ifdef RE_ENABLE_I18N
case MB_CHAR: case MB_CHAR:
/* Check whether the array has enough space. */ /* Check whether the array has enough space. */
if (mbchar_alloc == mbcset->nmbchars) if (mbchar_alloc == mbcset->nmbchars)
@ -2758,20 +2831,30 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
} }
mbcset->mbchars[mbcset->nmbchars++] = start_elem.opr.wch; mbcset->mbchars[mbcset->nmbchars++] = start_elem.opr.wch;
break; break;
#endif /* RE_ENABLE_I18N */
case EQUIV_CLASS: case EQUIV_CLASS:
*err = build_equiv_class (mbcset, sbcset, &equiv_class_alloc, *err = build_equiv_class (sbcset,
#ifdef RE_ENABLE_I18N
mbcset, &equiv_class_alloc,
#endif /* RE_ENABLE_I18N */
start_elem.opr.name); start_elem.opr.name);
if (BE (*err != REG_NOERROR, 0)) if (BE (*err != REG_NOERROR, 0))
goto parse_bracket_exp_free_return; goto parse_bracket_exp_free_return;
break; break;
case COLL_SYM: case COLL_SYM:
*err = build_collating_symbol (mbcset, sbcset, &coll_sym_alloc, *err = build_collating_symbol (sbcset,
#ifdef RE_ENABLE_I18N
mbcset, &coll_sym_alloc,
#endif /* RE_ENABLE_I18N */
start_elem.opr.name); start_elem.opr.name);
if (BE (*err != REG_NOERROR, 0)) if (BE (*err != REG_NOERROR, 0))
goto parse_bracket_exp_free_return; goto parse_bracket_exp_free_return;
break; break;
case CHAR_CLASS: case CHAR_CLASS:
ret = build_charclass (mbcset, sbcset, &char_class_alloc, ret = build_charclass (sbcset,
#ifdef RE_ENABLE_I18N
mbcset, &char_class_alloc,
#endif /* RE_ENABLE_I18N */
start_elem.opr.name, syntax); start_elem.opr.name, syntax);
if (BE (ret != REG_NOERROR, 0)) if (BE (ret != REG_NOERROR, 0))
goto parse_bracket_exp_espace; goto parse_bracket_exp_espace;
@ -2788,7 +2871,11 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
re_string_skip_bytes (regexp, token_len); /* Skip a token. */ re_string_skip_bytes (regexp, token_len); /* Skip a token. */
/* If it is non-matching list. */ /* If it is non-matching list. */
#ifdef RE_ENABLE_I18N
if (mbcset->non_match) if (mbcset->non_match)
#else /* not RE_ENABLE_I18N */
if (non_match)
#endif /* not RE_ENABLE_I18N */
bitset_not (sbcset); bitset_not (sbcset);
/* Build a tree for simple bracket. */ /* Build a tree for simple bracket. */
@ -2799,6 +2886,7 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
if (BE (new_idx == -1 || work_tree == NULL, 0)) if (BE (new_idx == -1 || work_tree == NULL, 0))
goto parse_bracket_exp_espace; goto parse_bracket_exp_espace;
#ifdef RE_ENABLE_I18N
if (mbcset->nmbchars || mbcset->ncoll_syms || mbcset->nequiv_classes if (mbcset->nmbchars || mbcset->ncoll_syms || mbcset->nequiv_classes
|| mbcset->nranges || (MB_CUR_MAX > 1 && (mbcset->nchar_classes || mbcset->nranges || (MB_CUR_MAX > 1 && (mbcset->nchar_classes
|| mbcset->non_match))) || mbcset->non_match)))
@ -2825,12 +2913,17 @@ parse_bracket_exp (regexp, dfa, token, syntax, err)
free_charset (mbcset); free_charset (mbcset);
return work_tree; return work_tree;
} }
#else /* not RE_ENABLE_I18N */
return work_tree;
#endif /* not RE_ENABLE_I18N */
parse_bracket_exp_espace: parse_bracket_exp_espace:
*err = REG_ESPACE; *err = REG_ESPACE;
parse_bracket_exp_free_return: parse_bracket_exp_free_return:
re_free (sbcset); re_free (sbcset);
#ifdef RE_ENABLE_I18N
free_charset (mbcset); free_charset (mbcset);
#endif /* RE_ENABLE_I18N */
return NULL; return NULL;
} }
@ -2915,13 +3008,17 @@ parse_bracket_symbol (elem, regexp, token)
is a pointer argument sinse we may update it. */ is a pointer argument sinse we may update it. */
static reg_errcode_t static reg_errcode_t
build_equiv_class (mbcset, sbcset, equiv_class_alloc, name) #ifdef RE_ENABLE_I18N
build_equiv_class (sbcset, mbcset, equiv_class_alloc, name)
re_charset_t *mbcset; re_charset_t *mbcset;
re_bitset_ptr_t sbcset;
int *equiv_class_alloc; int *equiv_class_alloc;
#else /* not RE_ENABLE_I18N */
build_equiv_class (sbcset, name)
#endif /* not RE_ENABLE_I18N */
re_bitset_ptr_t sbcset;
const unsigned char *name; const unsigned char *name;
{ {
#ifdef _LIBC #if defined _LIBC && defined RE_ENABLE_I18N
uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES); uint32_t nrules = _NL_CURRENT_WORD (LC_COLLATE, _NL_COLLATE_NRULES);
if (nrules != 0) if (nrules != 0)
{ {
@ -2987,7 +3084,7 @@ build_equiv_class (mbcset, sbcset, equiv_class_alloc, name)
mbcset->equiv_classes[mbcset->nequiv_classes++] = idx1; mbcset->equiv_classes[mbcset->nequiv_classes++] = idx1;
} }
else else
#endif #endif /* _LIBC && RE_ENABLE_I18N */
{ {
if (BE (strlen (name) != 1, 0)) if (BE (strlen (name) != 1, 0))
return REG_ECOLLATE; return REG_ECOLLATE;
@ -3003,15 +3100,27 @@ build_equiv_class (mbcset, sbcset, equiv_class_alloc, name)
is a pointer argument sinse we may update it. */ is a pointer argument sinse we may update it. */
static reg_errcode_t static reg_errcode_t
build_charclass (mbcset, sbcset, char_class_alloc, class_name, syntax) #ifdef RE_ENABLE_I18N
build_charclass (sbcset, mbcset, char_class_alloc, class_name, syntax)
re_charset_t *mbcset; re_charset_t *mbcset;
re_bitset_ptr_t sbcset;
int *char_class_alloc; int *char_class_alloc;
#else /* not RE_ENABLE_I18N */
build_charclass (sbcset, class_name, syntax)
#endif /* not RE_ENABLE_I18N */
re_bitset_ptr_t sbcset;
const unsigned char *class_name; const unsigned char *class_name;
reg_syntax_t syntax; reg_syntax_t syntax;
{ {
int i; int i;
const unsigned char *name = class_name; const unsigned char *name = class_name;
/* In case of REG_ICASE "upper" and "lower" match the both of
upper and lower cases. */
if ((syntax & RE_ICASE)
&& (strcmp (name, "upper") == 0 || strcmp (name, "lower") == 0))
name = "alpha";
#ifdef RE_ENABLE_I18N
/* Check the space of the arrays. */ /* Check the space of the arrays. */
if (*char_class_alloc == mbcset->nchar_classes) if (*char_class_alloc == mbcset->nchar_classes)
{ {
@ -3024,14 +3133,8 @@ build_charclass (mbcset, sbcset, char_class_alloc, class_name, syntax)
if (BE (mbcset->char_classes == NULL, 0)) if (BE (mbcset->char_classes == NULL, 0))
return REG_ESPACE; return REG_ESPACE;
} }
/* In case of REG_ICASE "upper" and "lower" match the both of
upper and lower cases. */
if ((syntax & RE_ICASE)
&& (strcmp (name, "upper") == 0 || strcmp (name, "lower") == 0))
name = "alpha";
mbcset->char_classes[mbcset->nchar_classes++] = __wctype (name); mbcset->char_classes[mbcset->nchar_classes++] = __wctype (name);
#endif /* RE_ENABLE_I18N */
#define BUILD_CHARCLASS_LOOP(ctype_func)\ #define BUILD_CHARCLASS_LOOP(ctype_func)\
for (i = 0; i < SBC_MAX; ++i) \ for (i = 0; i < SBC_MAX; ++i) \
@ -3077,15 +3180,27 @@ build_word_op (dfa, not, err)
reg_errcode_t *err; reg_errcode_t *err;
{ {
re_bitset_ptr_t sbcset; re_bitset_ptr_t sbcset;
#ifdef RE_ENABLE_I18N
re_charset_t *mbcset; re_charset_t *mbcset;
int alloc = 0;
#else /* not RE_ENABLE_I18N */
int non_match = 0;
#endif /* not RE_ENABLE_I18N */
reg_errcode_t ret; reg_errcode_t ret;
re_token_t br_token; re_token_t br_token;
bin_tree_t *tree; bin_tree_t *tree;
int new_idx, alloc = 0; int new_idx;
sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS); sbcset = (re_bitset_ptr_t) calloc (sizeof (unsigned int), BITSET_UINTS);
#ifdef RE_ENABLE_I18N
mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1); mbcset = (re_charset_t *) calloc (sizeof (re_charset_t), 1);
#endif /* RE_ENABLE_I18N */
#ifdef RE_ENABLE_I18N
if (BE (sbcset == NULL || mbcset == NULL, 0)) if (BE (sbcset == NULL || mbcset == NULL, 0))
#else /* not RE_ENABLE_I18N */
if (BE (sbcset == NULL, 0))
#endif /* not RE_ENABLE_I18N */
{ {
*err = REG_ESPACE; *err = REG_ESPACE;
return NULL; return NULL;
@ -3093,26 +3208,35 @@ build_word_op (dfa, not, err)
if (not) if (not)
{ {
#ifdef RE_ENABLE_I18N
int i; int i;
mbcset->non_match = 1;
/* /*
if (syntax & RE_HAT_LISTS_NOT_NEWLINE) if (syntax & RE_HAT_LISTS_NOT_NEWLINE)
bitset_set(cset->sbcset, '\0'); bitset_set(cset->sbcset, '\0');
*/ */
#ifdef RE_ENABLE_I18N mbcset->non_match = 1;
if (MB_CUR_MAX > 1) if (MB_CUR_MAX > 1)
for (i = 0; i < SBC_MAX; ++i) for (i = 0; i < SBC_MAX; ++i)
if (__btowc (i) == WEOF) if (__btowc (i) == WEOF)
bitset_set (sbcset, i); bitset_set (sbcset, i);
#endif /* RE_ENABLE_I18N */ #else /* not RE_ENABLE_I18N */
non_match = 1;
#endif /* not RE_ENABLE_I18N */
} }
/* We don't care the syntax in this case. */ /* We don't care the syntax in this case. */
ret = build_charclass (mbcset, sbcset, &alloc, "alpha", 0); ret = build_charclass (sbcset,
#ifdef RE_ENABLE_I18N
mbcset, &alloc,
#endif /* RE_ENABLE_I18N */
"alpha", 0);
if (BE (ret != REG_NOERROR, 0)) if (BE (ret != REG_NOERROR, 0))
{ {
re_free (sbcset); re_free (sbcset);
#ifdef RE_ENABLE_I18N
free_charset (mbcset); free_charset (mbcset);
#endif /* RE_ENABLE_I18N */
*err = REG_ESPACE; *err = REG_ESPACE;
return NULL; return NULL;
} }
@ -3120,7 +3244,11 @@ build_word_op (dfa, not, err)
bitset_set (sbcset, '_'); bitset_set (sbcset, '_');
/* If it is non-matching list. */ /* If it is non-matching list. */
#ifdef RE_ENABLE_I18N
if (mbcset->non_match) if (mbcset->non_match)
#else /* not RE_ENABLE_I18N */
if (non_match)
#endif /* not RE_ENABLE_I18N */
bitset_not (sbcset); bitset_not (sbcset);
/* Build a tree for simple bracket. */ /* Build a tree for simple bracket. */
@ -3131,6 +3259,7 @@ build_word_op (dfa, not, err)
if (BE (new_idx == -1 || tree == NULL, 0)) if (BE (new_idx == -1 || tree == NULL, 0))
goto build_word_op_espace; goto build_word_op_espace;
#ifdef RE_ENABLE_I18N
if (MB_CUR_MAX > 1) if (MB_CUR_MAX > 1)
{ {
re_token_t alt_token; re_token_t alt_token;
@ -3155,9 +3284,15 @@ build_word_op (dfa, not, err)
free_charset (mbcset); free_charset (mbcset);
return tree; return tree;
} }
#else /* not RE_ENABLE_I18N */
return tree;
#endif /* not RE_ENABLE_I18N */
build_word_op_espace: build_word_op_espace:
re_free (sbcset); re_free (sbcset);
#ifdef RE_ENABLE_I18N
free_charset (mbcset); free_charset (mbcset);
#endif /* RE_ENABLE_I18N */
*err = REG_ESPACE; *err = REG_ESPACE;
return NULL; return NULL;
} }
@ -3190,19 +3325,21 @@ fetch_number (input, token, syntax)
return num; return num;
} }
#ifdef RE_ENABLE_I18N
static void static void
free_charset (re_charset_t *cset) free_charset (re_charset_t *cset)
{ {
re_free (cset->mbchars); re_free (cset->mbchars);
#ifdef _LIBC # ifdef _LIBC
re_free (cset->coll_syms); re_free (cset->coll_syms);
re_free (cset->equiv_classes); re_free (cset->equiv_classes);
re_free (cset->range_starts); re_free (cset->range_starts);
re_free (cset->range_ends); re_free (cset->range_ends);
#endif # endif
re_free (cset->char_classes); re_free (cset->char_classes);
re_free (cset); re_free (cset);
} }
#endif /* RE_ENABLE_I18N */
/* Functions for binary tree operation. */ /* Functions for binary tree operation. */

1
posix/regex.c
View File

@ -45,6 +45,7 @@
# define BE(expr, val) __builtin_expect (expr, val) # define BE(expr, val) __builtin_expect (expr, val)
#else #else
# define BE(expr, val) (expr) # define BE(expr, val) (expr)
# define inline
#endif #endif
#include "regcomp.c" #include "regcomp.c"

3
posix/regex.h
View File

@ -189,7 +189,8 @@ extern reg_syntax_t re_syntax_options;
#define RE_SYNTAX_GNU_AWK \ #define RE_SYNTAX_GNU_AWK \
((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \ ((RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS | RE_DEBUG) \
& ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS)) & ~(RE_DOT_NOT_NULL | RE_INTERVALS | RE_CONTEXT_INDEP_OPS \
| RE_CONTEXT_INVALID_OPS ))
#define RE_SYNTAX_POSIX_AWK \ #define RE_SYNTAX_POSIX_AWK \
(RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \ (RE_SYNTAX_POSIX_EXTENDED | RE_BACKSLASH_ESCAPE_IN_LISTS \

10
posix/regex_internal.c
View File

@ -169,7 +169,7 @@ re_string_realloc_buffers (pstr, new_buf_len)
#ifdef RE_ENABLE_I18N #ifdef RE_ENABLE_I18N
if (MB_CUR_MAX > 1) if (MB_CUR_MAX > 1)
{ {
pstr->wcs = re_realloc (pstr->wcs, wchar_t, new_buf_len); pstr->wcs = re_realloc (pstr->wcs, wint_t, new_buf_len);
if (BE (pstr->wcs == NULL, 0)) if (BE (pstr->wcs == NULL, 0))
return REG_ESPACE; return REG_ESPACE;
} }
@ -436,7 +436,7 @@ re_string_reconstruct (pstr, idx, eflags, newline)
#ifdef RE_ENABLE_I18N #ifdef RE_ENABLE_I18N
if (MB_CUR_MAX > 1) if (MB_CUR_MAX > 1)
memmove (pstr->wcs, pstr->wcs + offset, memmove (pstr->wcs, pstr->wcs + offset,
(pstr->valid_len - offset) * sizeof (wchar_t)); (pstr->valid_len - offset) * sizeof (wint_t));
#endif /* RE_ENABLE_I18N */ #endif /* RE_ENABLE_I18N */
if (MBS_ALLOCATED (pstr)) if (MBS_ALLOCATED (pstr))
memmove (pstr->mbs, pstr->mbs + offset, memmove (pstr->mbs, pstr->mbs + offset,
@ -1166,9 +1166,11 @@ create_ci_newstate (dfa, nodes, hash)
/* If the state has the halt node, the state is a halt state. */ /* If the state has the halt node, the state is a halt state. */
else if (type == END_OF_RE) else if (type == END_OF_RE)
newstate->halt = 1; newstate->halt = 1;
#ifdef RE_ENABLE_I18N
else if (type == COMPLEX_BRACKET else if (type == COMPLEX_BRACKET
|| (type == OP_PERIOD && MB_CUR_MAX > 1)) || (type == OP_PERIOD && MB_CUR_MAX > 1))
newstate->accept_mb = 1; newstate->accept_mb = 1;
#endif /* RE_ENABLE_I18N */
else if (type == OP_BACK_REF) else if (type == OP_BACK_REF)
newstate->has_backref = 1; newstate->has_backref = 1;
else if (type == ANCHOR || OP_CONTEXT_NODE) else if (type == ANCHOR || OP_CONTEXT_NODE)
@ -1213,9 +1215,11 @@ create_cd_newstate (dfa, nodes, context, hash)
/* If the state has the halt node, the state is a halt state. */ /* If the state has the halt node, the state is a halt state. */
else if (type == END_OF_RE) else if (type == END_OF_RE)
newstate->halt = 1; newstate->halt = 1;
#ifdef RE_ENABLE_I18N
else if (type == COMPLEX_BRACKET else if (type == COMPLEX_BRACKET
|| (type == OP_PERIOD && MB_CUR_MAX > 1)) || (type == OP_PERIOD && MB_CUR_MAX > 1))
newstate->accept_mb = 1; newstate->accept_mb = 1;
#endif /* RE_ENABLE_I18N */
else if (type == OP_BACK_REF) else if (type == OP_BACK_REF)
newstate->has_backref = 1; newstate->has_backref = 1;
else if (type == ANCHOR) else if (type == ANCHOR)
@ -1228,9 +1232,11 @@ create_cd_newstate (dfa, nodes, context, hash)
newstate->halt = 1; newstate->halt = 1;
else if (ctype == OP_BACK_REF) else if (ctype == OP_BACK_REF)
newstate->has_backref = 1; newstate->has_backref = 1;
#ifdef RE_ENABLE_I18N
else if (ctype == COMPLEX_BRACKET else if (ctype == COMPLEX_BRACKET
|| (type == OP_PERIOD && MB_CUR_MAX > 1)) || (type == OP_PERIOD && MB_CUR_MAX > 1))
newstate->accept_mb = 1; newstate->accept_mb = 1;
#endif /* RE_ENABLE_I18N */
} }
if (constraint) if (constraint)

22
posix/regex_internal.h
View File

@ -119,7 +119,9 @@ typedef enum
ALT, ALT,
SUBEXP, SUBEXP,
SIMPLE_BRACKET, SIMPLE_BRACKET,
#ifdef RE_ENABLE_I18N
COMPLEX_BRACKET, COMPLEX_BRACKET,
#endif /* RE_ENABLE_I18N */
/* Node type, These are used by token, node, tree. */ /* Node type, These are used by token, node, tree. */
OP_PERIOD, OP_PERIOD,
@ -137,6 +139,7 @@ typedef enum
END_OF_RE_TOKEN_T END_OF_RE_TOKEN_T
} re_token_type_t; } re_token_type_t;
#ifdef RE_ENABLE_I18N
typedef struct typedef struct
{ {
/* If this character set is the non-matching list. */ /* If this character set is the non-matching list. */
@ -147,31 +150,32 @@ typedef struct
int nmbchars; int nmbchars;
/* Collating symbols. */ /* Collating symbols. */
#ifdef _LIBC # ifdef _LIBC
int32_t *coll_syms; int32_t *coll_syms;
#endif # endif
int ncoll_syms; int ncoll_syms;
/* Equivalence classes. */ /* Equivalence classes. */
#ifdef _LIBC # ifdef _LIBC
int32_t *equiv_classes; int32_t *equiv_classes;
#endif # endif
int nequiv_classes; int nequiv_classes;
/* Range expressions. */ /* Range expressions. */
#ifdef _LIBC # ifdef _LIBC
uint32_t *range_starts; uint32_t *range_starts;
uint32_t *range_ends; uint32_t *range_ends;
#else /* not _LIBC */ # else /* not _LIBC */
wchar_t *range_starts; wchar_t *range_starts;
wchar_t *range_ends; wchar_t *range_ends;
#endif /* not _LIBC */ # endif /* not _LIBC */
int nranges; int nranges;
/* Character classes. */ /* Character classes. */
wctype_t *char_classes; wctype_t *char_classes;
int nchar_classes; int nchar_classes;
} re_charset_t; } re_charset_t;
#endif /* RE_ENABLE_I18N */
typedef struct typedef struct
{ {
@ -180,7 +184,9 @@ typedef struct
{ {
unsigned char c; /* for CHARACTER */ unsigned char c; /* for CHARACTER */
re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */ re_bitset_ptr_t sbcset; /* for SIMPLE_BRACKET */
#ifdef RE_ENABLE_I18N
re_charset_t *mbcset; /* for COMPLEX_BRACKET */ re_charset_t *mbcset; /* for COMPLEX_BRACKET */
#endif /* RE_ENABLE_I18N */
int idx; /* for BACK_REF */ int idx; /* for BACK_REF */
re_context_type ctx_type; /* for ANCHOR */ re_context_type ctx_type; /* for ANCHOR */
struct struct
@ -221,7 +227,7 @@ struct re_string_t
unsigned char *mbs_case; unsigned char *mbs_case;
#ifdef RE_ENABLE_I18N #ifdef RE_ENABLE_I18N
/* Store the wide character string which is corresponding to MBS. */ /* Store the wide character string which is corresponding to MBS. */
wchar_t *wcs; wint_t *wcs;
mbstate_t cur_state; mbstate_t cur_state;
#endif #endif
/* The length of the valid characters in the buffers. */ /* The length of the valid characters in the buffers. */