1 /* Extended regular expression matching and search library.
2 Copyright (C) 2002-2013 Free Software Foundation, Inc.
3 This file is part of the GNU C Library.
4 Contributed by Isamu Hasegawa <isamu@yamato.ibm.com>.
6 The GNU C Library is free software; you can redistribute it and/or
7 modify it under the terms of the GNU Lesser General Public
8 License as published by the Free Software Foundation; either
9 version 2.1 of the License, or (at your option) any later version.
11 The GNU C Library is distributed in the hope that it will be useful,
12 but WITHOUT ANY WARRANTY; without even the implied warranty of
13 MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
14 Lesser General Public License for more details.
16 You should have received a copy of the GNU Lesser General Public
17 License along with the GNU C Library; if not, see
18 <http://www.gnu.org/licenses/>. */
20 static reg_errcode_t
match_ctx_init (re_match_context_t
*cache
, int eflags
,
21 int n
) internal_function
;
22 static void match_ctx_clean (re_match_context_t
*mctx
) internal_function
;
23 static void match_ctx_free (re_match_context_t
*cache
) internal_function
;
24 static reg_errcode_t
match_ctx_add_entry (re_match_context_t
*cache
, int node
,
25 int str_idx
, int from
, int to
)
27 static int search_cur_bkref_entry (const re_match_context_t
*mctx
, int str_idx
)
29 static reg_errcode_t
match_ctx_add_subtop (re_match_context_t
*mctx
, int node
,
30 int str_idx
) internal_function
;
31 static re_sub_match_last_t
* match_ctx_add_sublast (re_sub_match_top_t
*subtop
,
32 int node
, int str_idx
)
34 static void sift_ctx_init (re_sift_context_t
*sctx
, re_dfastate_t
**sifted_sts
,
35 re_dfastate_t
**limited_sts
, int last_node
,
38 static reg_errcode_t
re_search_internal (const regex_t
*preg
,
39 const char *string
, int length
,
40 int start
, int range
, int stop
,
41 size_t nmatch
, regmatch_t pmatch
[],
42 int eflags
) internal_function
;
43 static int re_search_2_stub (struct re_pattern_buffer
*bufp
,
44 const char *string1
, int length1
,
45 const char *string2
, int length2
,
46 int start
, int range
, struct re_registers
*regs
,
47 int stop
, int ret_len
) internal_function
;
48 static int re_search_stub (struct re_pattern_buffer
*bufp
,
49 const char *string
, int length
, int start
,
50 int range
, int stop
, struct re_registers
*regs
,
51 int ret_len
) internal_function
;
52 static unsigned re_copy_regs (struct re_registers
*regs
, regmatch_t
*pmatch
,
53 int nregs
, int regs_allocated
) internal_function
;
54 static reg_errcode_t
prune_impossible_nodes (re_match_context_t
*mctx
)
56 static int check_matching (re_match_context_t
*mctx
, int fl_longest_match
,
57 int *p_match_first
) internal_function
;
58 static int check_halt_state_context (const re_match_context_t
*mctx
,
59 const re_dfastate_t
*state
, int idx
)
61 static void update_regs (const re_dfa_t
*dfa
, regmatch_t
*pmatch
,
62 regmatch_t
*prev_idx_match
, int cur_node
,
63 int cur_idx
, int nmatch
) internal_function
;
64 static reg_errcode_t
push_fail_stack (struct re_fail_stack_t
*fs
,
65 int str_idx
, int dest_node
, int nregs
,
67 re_node_set
*eps_via_nodes
)
69 static reg_errcode_t
set_regs (const regex_t
*preg
,
70 const re_match_context_t
*mctx
,
71 size_t nmatch
, regmatch_t
*pmatch
,
72 int fl_backtrack
) internal_function
;
73 static reg_errcode_t
free_fail_stack_return (struct re_fail_stack_t
*fs
)
77 static int sift_states_iter_mb (const re_match_context_t
*mctx
,
78 re_sift_context_t
*sctx
,
79 int node_idx
, int str_idx
, int max_str_idx
)
81 #endif /* RE_ENABLE_I18N */
82 static reg_errcode_t
sift_states_backward (const re_match_context_t
*mctx
,
83 re_sift_context_t
*sctx
)
85 static reg_errcode_t
build_sifted_states (const re_match_context_t
*mctx
,
86 re_sift_context_t
*sctx
, int str_idx
,
87 re_node_set
*cur_dest
)
89 static reg_errcode_t
update_cur_sifted_state (const re_match_context_t
*mctx
,
90 re_sift_context_t
*sctx
,
92 re_node_set
*dest_nodes
)
94 static reg_errcode_t
add_epsilon_src_nodes (const re_dfa_t
*dfa
,
95 re_node_set
*dest_nodes
,
96 const re_node_set
*candidates
)
98 static int check_dst_limits (const re_match_context_t
*mctx
,
100 int dst_node
, int dst_idx
, int src_node
,
101 int src_idx
) internal_function
;
102 static int check_dst_limits_calc_pos_1 (const re_match_context_t
*mctx
,
103 int boundaries
, int subexp_idx
,
104 int from_node
, int bkref_idx
)
106 static int check_dst_limits_calc_pos (const re_match_context_t
*mctx
,
107 int limit
, int subexp_idx
,
108 int node
, int str_idx
,
109 int bkref_idx
) internal_function
;
110 static reg_errcode_t
check_subexp_limits (const re_dfa_t
*dfa
,
111 re_node_set
*dest_nodes
,
112 const re_node_set
*candidates
,
114 struct re_backref_cache_entry
*bkref_ents
,
115 int str_idx
) internal_function
;
116 static reg_errcode_t
sift_states_bkref (const re_match_context_t
*mctx
,
117 re_sift_context_t
*sctx
,
118 int str_idx
, const re_node_set
*candidates
)
120 static reg_errcode_t
merge_state_array (const re_dfa_t
*dfa
,
122 re_dfastate_t
**src
, int num
)
124 static re_dfastate_t
*find_recover_state (reg_errcode_t
*err
,
125 re_match_context_t
*mctx
) internal_function
;
126 static re_dfastate_t
*transit_state (reg_errcode_t
*err
,
127 re_match_context_t
*mctx
,
128 re_dfastate_t
*state
) internal_function
;
129 static re_dfastate_t
*merge_state_with_log (reg_errcode_t
*err
,
130 re_match_context_t
*mctx
,
131 re_dfastate_t
*next_state
)
133 static reg_errcode_t
check_subexp_matching_top (re_match_context_t
*mctx
,
134 re_node_set
*cur_nodes
,
135 int str_idx
) internal_function
;
137 static re_dfastate_t
*transit_state_sb (reg_errcode_t
*err
,
138 re_match_context_t
*mctx
,
139 re_dfastate_t
*pstate
)
142 #ifdef RE_ENABLE_I18N
143 static reg_errcode_t
transit_state_mb (re_match_context_t
*mctx
,
144 re_dfastate_t
*pstate
)
146 #endif /* RE_ENABLE_I18N */
147 static reg_errcode_t
transit_state_bkref (re_match_context_t
*mctx
,
148 const re_node_set
*nodes
)
150 static reg_errcode_t
get_subexp (re_match_context_t
*mctx
,
151 int bkref_node
, int bkref_str_idx
)
153 static reg_errcode_t
get_subexp_sub (re_match_context_t
*mctx
,
154 const re_sub_match_top_t
*sub_top
,
155 re_sub_match_last_t
*sub_last
,
156 int bkref_node
, int bkref_str
)
158 static int find_subexp_node (const re_dfa_t
*dfa
, const re_node_set
*nodes
,
159 int subexp_idx
, int type
) internal_function
;
160 static reg_errcode_t
check_arrival (re_match_context_t
*mctx
,
161 state_array_t
*path
, int top_node
,
162 int top_str
, int last_node
, int last_str
,
163 int type
) internal_function
;
164 static reg_errcode_t
check_arrival_add_next_nodes (re_match_context_t
*mctx
,
166 re_node_set
*cur_nodes
,
167 re_node_set
*next_nodes
)
169 static reg_errcode_t
check_arrival_expand_ecl (const re_dfa_t
*dfa
,
170 re_node_set
*cur_nodes
,
171 int ex_subexp
, int type
)
173 static reg_errcode_t
check_arrival_expand_ecl_sub (const re_dfa_t
*dfa
,
174 re_node_set
*dst_nodes
,
175 int target
, int ex_subexp
,
176 int type
) internal_function
;
177 static reg_errcode_t
expand_bkref_cache (re_match_context_t
*mctx
,
178 re_node_set
*cur_nodes
, int cur_str
,
179 int subexp_num
, int type
)
181 static int build_trtable (const re_dfa_t
*dfa
,
182 re_dfastate_t
*state
) internal_function
;
183 #ifdef RE_ENABLE_I18N
184 static int check_node_accept_bytes (const re_dfa_t
*dfa
, int node_idx
,
185 const re_string_t
*input
, int idx
)
188 static unsigned int find_collation_sequence_value (const unsigned char *mbs
,
192 #endif /* RE_ENABLE_I18N */
193 static int group_nodes_into_DFAstates (const re_dfa_t
*dfa
,
194 const re_dfastate_t
*state
,
195 re_node_set
*states_node
,
196 bitset_t
*states_ch
) internal_function
;
197 static int check_node_accept (const re_match_context_t
*mctx
,
198 const re_token_t
*node
, int idx
)
200 static reg_errcode_t
extend_buffers (re_match_context_t
*mctx
)
203 /* Entry point for POSIX code. */
205 /* regexec searches for a given pattern, specified by PREG, in the
208 If NMATCH is zero or REG_NOSUB was set in the cflags argument to
209 `regcomp', we ignore PMATCH. Otherwise, we assume PMATCH has at
210 least NMATCH elements, and we set them to the offsets of the
211 corresponding matched substrings.
213 EFLAGS specifies `execution flags' which affect matching: if
214 REG_NOTBOL is set, then ^ does not match at the beginning of the
215 string; if REG_NOTEOL is set, then $ does not match at the end.
217 We return 0 if we find a match and REG_NOMATCH if not. */
220 regexec (preg
, string
, nmatch
, pmatch
, eflags
)
221 const regex_t
*__restrict preg
;
222 const char *__restrict string
;
229 re_dfa_t
*dfa
= (re_dfa_t
*) preg
->buffer
;
231 if (eflags
& ~(REG_NOTBOL
| REG_NOTEOL
| REG_STARTEND
))
234 if (eflags
& REG_STARTEND
)
236 start
= pmatch
[0].rm_so
;
237 length
= pmatch
[0].rm_eo
;
242 length
= strlen (string
);
245 __libc_lock_lock (dfa
->lock
);
247 err
= re_search_internal (preg
, string
, length
, start
, length
- start
,
248 length
, 0, NULL
, eflags
);
250 err
= re_search_internal (preg
, string
, length
, start
, length
- start
,
251 length
, nmatch
, pmatch
, eflags
);
252 __libc_lock_unlock (dfa
->lock
);
253 return err
!= REG_NOERROR
;
257 # include <shlib-compat.h>
258 versioned_symbol (libc
, __regexec
, regexec
, GLIBC_2_3_4
);
260 # if SHLIB_COMPAT (libc, GLIBC_2_0, GLIBC_2_3_4)
261 __typeof__ (__regexec
) __compat_regexec
;
264 attribute_compat_text_section
265 __compat_regexec (const regex_t
*__restrict preg
,
266 const char *__restrict string
, size_t nmatch
,
267 regmatch_t pmatch
[], int eflags
)
269 return regexec (preg
, string
, nmatch
, pmatch
,
270 eflags
& (REG_NOTBOL
| REG_NOTEOL
));
272 compat_symbol (libc
, __compat_regexec
, regexec
, GLIBC_2_0
);
276 /* Entry points for GNU code. */
278 /* re_match, re_search, re_match_2, re_search_2
280 The former two functions operate on STRING with length LENGTH,
281 while the later two operate on concatenation of STRING1 and STRING2
282 with lengths LENGTH1 and LENGTH2, respectively.
284 re_match() matches the compiled pattern in BUFP against the string,
285 starting at index START.
287 re_search() first tries matching at index START, then it tries to match
288 starting from index START + 1, and so on. The last start position tried
289 is START + RANGE. (Thus RANGE = 0 forces re_search to operate the same
292 The parameter STOP of re_{match,search}_2 specifies that no match exceeding
293 the first STOP characters of the concatenation of the strings should be
296 If REGS is not NULL, and BUFP->no_sub is not set, the offsets of the match
297 and all groups is stroed in REGS. (For the "_2" variants, the offsets are
298 computed relative to the concatenation, not relative to the individual
301 On success, re_match* functions return the length of the match, re_search*
302 return the position of the start of the match. Return value -1 means no
303 match was found and -2 indicates an internal error. */
306 re_match (bufp
, string
, length
, start
, regs
)
307 struct re_pattern_buffer
*bufp
;
310 struct re_registers
*regs
;
312 return re_search_stub (bufp
, string
, length
, start
, 0, length
, regs
, 1);
315 weak_alias (__re_match
, re_match
)
319 re_search (bufp
, string
, length
, start
, range
, regs
)
320 struct re_pattern_buffer
*bufp
;
322 int length
, start
, range
;
323 struct re_registers
*regs
;
325 return re_search_stub (bufp
, string
, length
, start
, range
, length
, regs
, 0);
328 weak_alias (__re_search
, re_search
)
332 re_match_2 (bufp
, string1
, length1
, string2
, length2
, start
, regs
, stop
)
333 struct re_pattern_buffer
*bufp
;
334 const char *string1
, *string2
;
335 int length1
, length2
, start
, stop
;
336 struct re_registers
*regs
;
338 return re_search_2_stub (bufp
, string1
, length1
, string2
, length2
,
339 start
, 0, regs
, stop
, 1);
342 weak_alias (__re_match_2
, re_match_2
)
346 re_search_2 (bufp
, string1
, length1
, string2
, length2
, start
, range
, regs
, stop
)
347 struct re_pattern_buffer
*bufp
;
348 const char *string1
, *string2
;
349 int length1
, length2
, start
, range
, stop
;
350 struct re_registers
*regs
;
352 return re_search_2_stub (bufp
, string1
, length1
, string2
, length2
,
353 start
, range
, regs
, stop
, 0);
356 weak_alias (__re_search_2
, re_search_2
)
360 re_search_2_stub (bufp
, string1
, length1
, string2
, length2
, start
, range
, regs
,
362 struct re_pattern_buffer
*bufp
;
363 const char *string1
, *string2
;
364 int length1
, length2
, start
, range
, stop
, ret_len
;
365 struct re_registers
*regs
;
369 int len
= length1
+ length2
;
372 if (BE (length1
< 0 || length2
< 0 || stop
< 0 || len
< length1
, 0))
375 /* Concatenate the strings. */
379 s
= re_malloc (char, len
);
381 if (BE (s
== NULL
, 0))
384 memcpy (__mempcpy (s
, string1
, length1
), string2
, length2
);
386 memcpy (s
, string1
, length1
);
387 memcpy (s
+ length1
, string2
, length2
);
396 rval
= re_search_stub (bufp
, str
, len
, start
, range
, stop
, regs
, ret_len
);
401 /* The parameters have the same meaning as those of re_search.
402 Additional parameters:
403 If RET_LEN is nonzero the length of the match is returned (re_match style);
404 otherwise the position of the match is returned. */
407 re_search_stub (bufp
, string
, length
, start
, range
, stop
, regs
, ret_len
)
408 struct re_pattern_buffer
*bufp
;
410 int length
, start
, range
, stop
, ret_len
;
411 struct re_registers
*regs
;
413 reg_errcode_t result
;
417 re_dfa_t
*dfa
= (re_dfa_t
*) bufp
->buffer
;
419 /* Check for out-of-range. */
420 if (BE (start
< 0 || start
> length
, 0))
422 if (BE (start
+ range
> length
, 0))
423 range
= length
- start
;
424 else if (BE (start
+ range
< 0, 0))
427 __libc_lock_lock (dfa
->lock
);
429 eflags
|= (bufp
->not_bol
) ? REG_NOTBOL
: 0;
430 eflags
|= (bufp
->not_eol
) ? REG_NOTEOL
: 0;
432 /* Compile fastmap if we haven't yet. */
433 if (range
> 0 && bufp
->fastmap
!= NULL
&& !bufp
->fastmap_accurate
)
434 re_compile_fastmap (bufp
);
436 if (BE (bufp
->no_sub
, 0))
439 /* We need at least 1 register. */
442 else if (BE (bufp
->regs_allocated
== REGS_FIXED
&&
443 regs
->num_regs
< bufp
->re_nsub
+ 1, 0))
445 nregs
= regs
->num_regs
;
446 if (BE (nregs
< 1, 0))
448 /* Nothing can be copied to regs. */
454 nregs
= bufp
->re_nsub
+ 1;
455 pmatch
= re_malloc (regmatch_t
, nregs
);
456 if (BE (pmatch
== NULL
, 0))
462 result
= re_search_internal (bufp
, string
, length
, start
, range
, stop
,
463 nregs
, pmatch
, eflags
);
467 /* I hope we needn't fill ther regs with -1's when no match was found. */
468 if (result
!= REG_NOERROR
)
470 else if (regs
!= NULL
)
472 /* If caller wants register contents data back, copy them. */
473 bufp
->regs_allocated
= re_copy_regs (regs
, pmatch
, nregs
,
474 bufp
->regs_allocated
);
475 if (BE (bufp
->regs_allocated
== REGS_UNALLOCATED
, 0))
479 if (BE (rval
== 0, 1))
483 assert (pmatch
[0].rm_so
== start
);
484 rval
= pmatch
[0].rm_eo
- start
;
487 rval
= pmatch
[0].rm_so
;
491 __libc_lock_unlock (dfa
->lock
);
496 re_copy_regs (regs
, pmatch
, nregs
, regs_allocated
)
497 struct re_registers
*regs
;
499 int nregs
, regs_allocated
;
501 int rval
= REGS_REALLOCATE
;
503 int need_regs
= nregs
+ 1;
504 /* We need one extra element beyond `num_regs' for the `-1' marker GNU code
507 /* Have the register data arrays been allocated? */
508 if (regs_allocated
== REGS_UNALLOCATED
)
509 { /* No. So allocate them with malloc. */
510 regs
->start
= re_malloc (regoff_t
, need_regs
);
511 if (BE (regs
->start
== NULL
, 0))
512 return REGS_UNALLOCATED
;
513 regs
->end
= re_malloc (regoff_t
, need_regs
);
514 if (BE (regs
->end
== NULL
, 0))
516 re_free (regs
->start
);
517 return REGS_UNALLOCATED
;
519 regs
->num_regs
= need_regs
;
521 else if (regs_allocated
== REGS_REALLOCATE
)
522 { /* Yes. If we need more elements than were already
523 allocated, reallocate them. If we need fewer, just
525 if (BE (need_regs
> regs
->num_regs
, 0))
527 regoff_t
*new_start
= re_realloc (regs
->start
, regoff_t
, need_regs
);
529 if (BE (new_start
== NULL
, 0))
530 return REGS_UNALLOCATED
;
531 new_end
= re_realloc (regs
->end
, regoff_t
, need_regs
);
532 if (BE (new_end
== NULL
, 0))
535 return REGS_UNALLOCATED
;
537 regs
->start
= new_start
;
539 regs
->num_regs
= need_regs
;
544 assert (regs_allocated
== REGS_FIXED
);
545 /* This function may not be called with REGS_FIXED and nregs too big. */
546 assert (regs
->num_regs
>= nregs
);
551 for (i
= 0; i
< nregs
; ++i
)
553 regs
->start
[i
] = pmatch
[i
].rm_so
;
554 regs
->end
[i
] = pmatch
[i
].rm_eo
;
556 for ( ; i
< regs
->num_regs
; ++i
)
557 regs
->start
[i
] = regs
->end
[i
] = -1;
562 /* Set REGS to hold NUM_REGS registers, storing them in STARTS and
563 ENDS. Subsequent matches using PATTERN_BUFFER and REGS will use
564 this memory for recording register information. STARTS and ENDS
565 must be allocated using the malloc library routine, and must each
566 be at least NUM_REGS * sizeof (regoff_t) bytes long.
568 If NUM_REGS == 0, then subsequent matches should allocate their own
571 Unless this function is called, the first search or match using
572 PATTERN_BUFFER will allocate its own register data, without
573 freeing the old data. */
576 re_set_registers (bufp
, regs
, num_regs
, starts
, ends
)
577 struct re_pattern_buffer
*bufp
;
578 struct re_registers
*regs
;
580 regoff_t
*starts
, *ends
;
584 bufp
->regs_allocated
= REGS_REALLOCATE
;
585 regs
->num_regs
= num_regs
;
586 regs
->start
= starts
;
591 bufp
->regs_allocated
= REGS_UNALLOCATED
;
593 regs
->start
= regs
->end
= (regoff_t
*) 0;
597 weak_alias (__re_set_registers
, re_set_registers
)
600 /* Entry points compatible with 4.2 BSD regex library. We don't define
601 them unless specifically requested. */
603 #if defined _REGEX_RE_COMP || defined _LIBC
611 return 0 == regexec (&re_comp_buf
, s
, 0, NULL
, 0);
613 #endif /* _REGEX_RE_COMP */
615 /* Internal entry point. */
617 /* Searches for a compiled pattern PREG in the string STRING, whose
618 length is LENGTH. NMATCH, PMATCH, and EFLAGS have the same
619 mingings with regexec. START, and RANGE have the same meanings
621 Return REG_NOERROR if we find a match, and REG_NOMATCH if not,
622 otherwise return the error code.
623 Note: We assume front end functions already check ranges.
624 (START + RANGE >= 0 && START + RANGE <= LENGTH) */
627 __attribute_warn_unused_result__
628 re_search_internal (preg
, string
, length
, start
, range
, stop
, nmatch
, pmatch
,
632 int length
, start
, range
, stop
, eflags
;
637 const re_dfa_t
*dfa
= (const re_dfa_t
*) preg
->buffer
;
638 int left_lim
, right_lim
, incr
;
639 int fl_longest_match
, match_first
, match_kind
, match_last
= -1;
642 #if defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L)
643 re_match_context_t mctx
= { .dfa
= dfa
};
645 re_match_context_t mctx
;
647 char *fastmap
= (preg
->fastmap
!= NULL
&& preg
->fastmap_accurate
648 && range
&& !preg
->can_be_null
) ? preg
->fastmap
: NULL
;
649 RE_TRANSLATE_TYPE t
= preg
->translate
;
651 #if !(defined _LIBC || (defined __STDC_VERSION__ && __STDC_VERSION__ >= 199901L))
652 memset (&mctx
, '\0', sizeof (re_match_context_t
));
656 extra_nmatch
= (nmatch
> preg
->re_nsub
) ? nmatch
- (preg
->re_nsub
+ 1) : 0;
657 nmatch
-= extra_nmatch
;
659 /* Check if the DFA haven't been compiled. */
660 if (BE (preg
->used
== 0 || dfa
->init_state
== NULL
661 || dfa
->init_state_word
== NULL
|| dfa
->init_state_nl
== NULL
662 || dfa
->init_state_begbuf
== NULL
, 0))
666 /* We assume front-end functions already check them. */
667 assert (start
+ range
>= 0 && start
+ range
<= length
);
670 /* If initial states with non-begbuf contexts have no elements,
671 the regex must be anchored. If preg->newline_anchor is set,
672 we'll never use init_state_nl, so do not check it. */
673 if (dfa
->init_state
->nodes
.nelem
== 0
674 && dfa
->init_state_word
->nodes
.nelem
== 0
675 && (dfa
->init_state_nl
->nodes
.nelem
== 0
676 || !preg
->newline_anchor
))
678 if (start
!= 0 && start
+ range
!= 0)
683 /* We must check the longest matching, if nmatch > 0. */
684 fl_longest_match
= (nmatch
!= 0 || dfa
->nbackref
);
686 err
= re_string_allocate (&mctx
.input
, string
, length
, dfa
->nodes_len
+ 1,
687 preg
->translate
, preg
->syntax
& RE_ICASE
, dfa
);
688 if (BE (err
!= REG_NOERROR
, 0))
690 mctx
.input
.stop
= stop
;
691 mctx
.input
.raw_stop
= stop
;
692 mctx
.input
.newline_anchor
= preg
->newline_anchor
;
694 err
= match_ctx_init (&mctx
, eflags
, dfa
->nbackref
* 2);
695 if (BE (err
!= REG_NOERROR
, 0))
698 /* We will log all the DFA states through which the dfa pass,
699 if nmatch > 1, or this dfa has "multibyte node", which is a
700 back-reference or a node which can accept multibyte character or
701 multi character collating element. */
702 if (nmatch
> 1 || dfa
->has_mb_node
)
704 /* Avoid overflow. */
705 if (BE (SIZE_MAX
/ sizeof (re_dfastate_t
*) <= mctx
.input
.bufs_len
, 0))
711 mctx
.state_log
= re_malloc (re_dfastate_t
*, mctx
.input
.bufs_len
+ 1);
712 if (BE (mctx
.state_log
== NULL
, 0))
719 mctx
.state_log
= NULL
;
722 mctx
.input
.tip_context
= (eflags
& REG_NOTBOL
) ? CONTEXT_BEGBUF
723 : CONTEXT_NEWLINE
| CONTEXT_BEGBUF
;
725 /* Check incrementally whether of not the input string match. */
726 incr
= (range
< 0) ? -1 : 1;
727 left_lim
= (range
< 0) ? start
+ range
: start
;
728 right_lim
= (range
< 0) ? start
: start
+ range
;
729 sb
= dfa
->mb_cur_max
== 1;
732 ? ((sb
|| !(preg
->syntax
& RE_ICASE
|| t
) ? 4 : 0)
733 | (range
>= 0 ? 2 : 0)
734 | (t
!= NULL
? 1 : 0))
737 for (;; match_first
+= incr
)
740 if (match_first
< left_lim
|| right_lim
< match_first
)
743 /* Advance as rapidly as possible through the string, until we
744 find a plausible place to start matching. This may be done
745 with varying efficiency, so there are various possibilities:
746 only the most common of them are specialized, in order to
747 save on code size. We use a switch statement for speed. */
755 /* Fastmap with single-byte translation, match forward. */
756 while (BE (match_first
< right_lim
, 1)
757 && !fastmap
[t
[(unsigned char) string
[match_first
]]])
759 goto forward_match_found_start_or_reached_end
;
762 /* Fastmap without translation, match forward. */
763 while (BE (match_first
< right_lim
, 1)
764 && !fastmap
[(unsigned char) string
[match_first
]])
767 forward_match_found_start_or_reached_end
:
768 if (BE (match_first
== right_lim
, 0))
770 ch
= match_first
>= length
771 ? 0 : (unsigned char) string
[match_first
];
772 if (!fastmap
[t
? t
[ch
] : ch
])
779 /* Fastmap without multi-byte translation, match backwards. */
780 while (match_first
>= left_lim
)
782 ch
= match_first
>= length
783 ? 0 : (unsigned char) string
[match_first
];
784 if (fastmap
[t
? t
[ch
] : ch
])
788 if (match_first
< left_lim
)
793 /* In this case, we can't determine easily the current byte,
794 since it might be a component byte of a multibyte
795 character. Then we use the constructed buffer instead. */
798 /* If MATCH_FIRST is out of the valid range, reconstruct the
800 unsigned int offset
= match_first
- mctx
.input
.raw_mbs_idx
;
801 if (BE (offset
>= (unsigned int) mctx
.input
.valid_raw_len
, 0))
803 err
= re_string_reconstruct (&mctx
.input
, match_first
,
805 if (BE (err
!= REG_NOERROR
, 0))
808 offset
= match_first
- mctx
.input
.raw_mbs_idx
;
810 /* If MATCH_FIRST is out of the buffer, leave it as '\0'.
811 Note that MATCH_FIRST must not be smaller than 0. */
812 ch
= (match_first
>= length
813 ? 0 : re_string_byte_at (&mctx
.input
, offset
));
817 if (match_first
< left_lim
|| match_first
> right_lim
)
826 /* Reconstruct the buffers so that the matcher can assume that
827 the matching starts from the beginning of the buffer. */
828 err
= re_string_reconstruct (&mctx
.input
, match_first
, eflags
);
829 if (BE (err
!= REG_NOERROR
, 0))
832 #ifdef RE_ENABLE_I18N
833 /* Don't consider this char as a possible match start if it part,
834 yet isn't the head, of a multibyte character. */
835 if (!sb
&& !re_string_first_byte (&mctx
.input
, 0))
839 /* It seems to be appropriate one, then use the matcher. */
840 /* We assume that the matching starts from 0. */
841 mctx
.state_log_top
= mctx
.nbkref_ents
= mctx
.max_mb_elem_len
= 0;
842 match_last
= check_matching (&mctx
, fl_longest_match
,
843 range
>= 0 ? &match_first
: NULL
);
844 if (match_last
!= -1)
846 if (BE (match_last
== -2, 0))
853 mctx
.match_last
= match_last
;
854 if ((!preg
->no_sub
&& nmatch
> 1) || dfa
->nbackref
)
856 re_dfastate_t
*pstate
= mctx
.state_log
[match_last
];
857 mctx
.last_node
= check_halt_state_context (&mctx
, pstate
,
860 if ((!preg
->no_sub
&& nmatch
> 1 && dfa
->has_plural_match
)
863 err
= prune_impossible_nodes (&mctx
);
864 if (err
== REG_NOERROR
)
866 if (BE (err
!= REG_NOMATCH
, 0))
871 break; /* We found a match. */
875 match_ctx_clean (&mctx
);
879 assert (match_last
!= -1);
880 assert (err
== REG_NOERROR
);
883 /* Set pmatch[] if we need. */
888 /* Initialize registers. */
889 for (reg_idx
= 1; reg_idx
< nmatch
; ++reg_idx
)
890 pmatch
[reg_idx
].rm_so
= pmatch
[reg_idx
].rm_eo
= -1;
892 /* Set the points where matching start/end. */
894 pmatch
[0].rm_eo
= mctx
.match_last
;
896 if (!preg
->no_sub
&& nmatch
> 1)
898 err
= set_regs (preg
, &mctx
, nmatch
, pmatch
,
899 dfa
->has_plural_match
&& dfa
->nbackref
> 0);
900 if (BE (err
!= REG_NOERROR
, 0))
904 /* At last, add the offset to the each registers, since we slided
905 the buffers so that we could assume that the matching starts
907 for (reg_idx
= 0; reg_idx
< nmatch
; ++reg_idx
)
908 if (pmatch
[reg_idx
].rm_so
!= -1)
910 #ifdef RE_ENABLE_I18N
911 if (BE (mctx
.input
.offsets_needed
!= 0, 0))
913 pmatch
[reg_idx
].rm_so
=
914 (pmatch
[reg_idx
].rm_so
== mctx
.input
.valid_len
915 ? mctx
.input
.valid_raw_len
916 : mctx
.input
.offsets
[pmatch
[reg_idx
].rm_so
]);
917 pmatch
[reg_idx
].rm_eo
=
918 (pmatch
[reg_idx
].rm_eo
== mctx
.input
.valid_len
919 ? mctx
.input
.valid_raw_len
920 : mctx
.input
.offsets
[pmatch
[reg_idx
].rm_eo
]);
923 assert (mctx
.input
.offsets_needed
== 0);
925 pmatch
[reg_idx
].rm_so
+= match_first
;
926 pmatch
[reg_idx
].rm_eo
+= match_first
;
928 for (reg_idx
= 0; reg_idx
< extra_nmatch
; ++reg_idx
)
930 pmatch
[nmatch
+ reg_idx
].rm_so
= -1;
931 pmatch
[nmatch
+ reg_idx
].rm_eo
= -1;
935 for (reg_idx
= 0; reg_idx
+ 1 < nmatch
; reg_idx
++)
936 if (dfa
->subexp_map
[reg_idx
] != reg_idx
)
938 pmatch
[reg_idx
+ 1].rm_so
939 = pmatch
[dfa
->subexp_map
[reg_idx
] + 1].rm_so
;
940 pmatch
[reg_idx
+ 1].rm_eo
941 = pmatch
[dfa
->subexp_map
[reg_idx
] + 1].rm_eo
;
946 re_free (mctx
.state_log
);
948 match_ctx_free (&mctx
);
949 re_string_destruct (&mctx
.input
);
954 __attribute_warn_unused_result__
955 prune_impossible_nodes (mctx
)
956 re_match_context_t
*mctx
;
958 const re_dfa_t
*const dfa
= mctx
->dfa
;
959 int halt_node
, match_last
;
961 re_dfastate_t
**sifted_states
;
962 re_dfastate_t
**lim_states
= NULL
;
963 re_sift_context_t sctx
;
965 assert (mctx
->state_log
!= NULL
);
967 match_last
= mctx
->match_last
;
968 halt_node
= mctx
->last_node
;
970 /* Avoid overflow. */
971 if (BE (SIZE_MAX
/ sizeof (re_dfastate_t
*) <= match_last
, 0))
974 sifted_states
= re_malloc (re_dfastate_t
*, match_last
+ 1);
975 if (BE (sifted_states
== NULL
, 0))
982 lim_states
= re_malloc (re_dfastate_t
*, match_last
+ 1);
983 if (BE (lim_states
== NULL
, 0))
990 memset (lim_states
, '\0',
991 sizeof (re_dfastate_t
*) * (match_last
+ 1));
992 sift_ctx_init (&sctx
, sifted_states
, lim_states
, halt_node
,
994 ret
= sift_states_backward (mctx
, &sctx
);
995 re_node_set_free (&sctx
.limits
);
996 if (BE (ret
!= REG_NOERROR
, 0))
998 if (sifted_states
[0] != NULL
|| lim_states
[0] != NULL
)
1008 } while (mctx
->state_log
[match_last
] == NULL
1009 || !mctx
->state_log
[match_last
]->halt
);
1010 halt_node
= check_halt_state_context (mctx
,
1011 mctx
->state_log
[match_last
],
1014 ret
= merge_state_array (dfa
, sifted_states
, lim_states
,
1016 re_free (lim_states
);
1018 if (BE (ret
!= REG_NOERROR
, 0))
1023 sift_ctx_init (&sctx
, sifted_states
, lim_states
, halt_node
, match_last
);
1024 ret
= sift_states_backward (mctx
, &sctx
);
1025 re_node_set_free (&sctx
.limits
);
1026 if (BE (ret
!= REG_NOERROR
, 0))
1028 if (sifted_states
[0] == NULL
)
1034 re_free (mctx
->state_log
);
1035 mctx
->state_log
= sifted_states
;
1036 sifted_states
= NULL
;
1037 mctx
->last_node
= halt_node
;
1038 mctx
->match_last
= match_last
;
1041 re_free (sifted_states
);
1042 re_free (lim_states
);
1046 /* Acquire an initial state and return it.
1047 We must select appropriate initial state depending on the context,
1048 since initial states may have constraints like "\<", "^", etc.. */
1050 static inline re_dfastate_t
*
1051 __attribute ((always_inline
)) internal_function
1052 acquire_init_state_context (reg_errcode_t
*err
, const re_match_context_t
*mctx
,
1055 const re_dfa_t
*const dfa
= mctx
->dfa
;
1056 if (dfa
->init_state
->has_constraint
)
1058 unsigned int context
;
1059 context
= re_string_context_at (&mctx
->input
, idx
- 1, mctx
->eflags
);
1060 if (IS_WORD_CONTEXT (context
))
1061 return dfa
->init_state_word
;
1062 else if (IS_ORDINARY_CONTEXT (context
))
1063 return dfa
->init_state
;
1064 else if (IS_BEGBUF_CONTEXT (context
) && IS_NEWLINE_CONTEXT (context
))
1065 return dfa
->init_state_begbuf
;
1066 else if (IS_NEWLINE_CONTEXT (context
))
1067 return dfa
->init_state_nl
;
1068 else if (IS_BEGBUF_CONTEXT (context
))
1070 /* It is relatively rare case, then calculate on demand. */
1071 return re_acquire_state_context (err
, dfa
,
1072 dfa
->init_state
->entrance_nodes
,
1076 /* Must not happen? */
1077 return dfa
->init_state
;
1080 return dfa
->init_state
;
1083 /* Check whether the regular expression match input string INPUT or not,
1084 and return the index where the matching end, return -1 if not match,
1085 or return -2 in case of an error.
1086 FL_LONGEST_MATCH means we want the POSIX longest matching.
1087 If P_MATCH_FIRST is not NULL, and the match fails, it is set to the
1088 next place where we may want to try matching.
1089 Note that the matcher assume that the maching starts from the current
1090 index of the buffer. */
1093 internal_function __attribute_warn_unused_result__
1094 check_matching (re_match_context_t
*mctx
, int fl_longest_match
,
1097 const re_dfa_t
*const dfa
= mctx
->dfa
;
1100 int match_last
= -1;
1101 int cur_str_idx
= re_string_cur_idx (&mctx
->input
);
1102 re_dfastate_t
*cur_state
;
1103 int at_init_state
= p_match_first
!= NULL
;
1104 int next_start_idx
= cur_str_idx
;
1107 cur_state
= acquire_init_state_context (&err
, mctx
, cur_str_idx
);
1108 /* An initial state must not be NULL (invalid). */
1109 if (BE (cur_state
== NULL
, 0))
1111 assert (err
== REG_ESPACE
);
1115 if (mctx
->state_log
!= NULL
)
1117 mctx
->state_log
[cur_str_idx
] = cur_state
;
1119 /* Check OP_OPEN_SUBEXP in the initial state in case that we use them
1120 later. E.g. Processing back references. */
1121 if (BE (dfa
->nbackref
, 0))
1124 err
= check_subexp_matching_top (mctx
, &cur_state
->nodes
, 0);
1125 if (BE (err
!= REG_NOERROR
, 0))
1128 if (cur_state
->has_backref
)
1130 err
= transit_state_bkref (mctx
, &cur_state
->nodes
);
1131 if (BE (err
!= REG_NOERROR
, 0))
1137 /* If the RE accepts NULL string. */
1138 if (BE (cur_state
->halt
, 0))
1140 if (!cur_state
->has_constraint
1141 || check_halt_state_context (mctx
, cur_state
, cur_str_idx
))
1143 if (!fl_longest_match
)
1147 match_last
= cur_str_idx
;
1153 while (!re_string_eoi (&mctx
->input
))
1155 re_dfastate_t
*old_state
= cur_state
;
1156 int next_char_idx
= re_string_cur_idx (&mctx
->input
) + 1;
1158 if ((BE (next_char_idx
>= mctx
->input
.bufs_len
, 0)
1159 && mctx
->input
.bufs_len
< mctx
->input
.len
)
1160 || (BE (next_char_idx
>= mctx
->input
.valid_len
, 0)
1161 && mctx
->input
.valid_len
< mctx
->input
.len
))
1163 err
= extend_buffers (mctx
);
1164 if (BE (err
!= REG_NOERROR
, 0))
1166 assert (err
== REG_ESPACE
);
1171 cur_state
= transit_state (&err
, mctx
, cur_state
);
1172 if (mctx
->state_log
!= NULL
)
1173 cur_state
= merge_state_with_log (&err
, mctx
, cur_state
);
1175 if (cur_state
== NULL
)
1177 /* Reached the invalid state or an error. Try to recover a valid
1178 state using the state log, if available and if we have not
1179 already found a valid (even if not the longest) match. */
1180 if (BE (err
!= REG_NOERROR
, 0))
1183 if (mctx
->state_log
== NULL
1184 || (match
&& !fl_longest_match
)
1185 || (cur_state
= find_recover_state (&err
, mctx
)) == NULL
)
1189 if (BE (at_init_state
, 0))
1191 if (old_state
== cur_state
)
1192 next_start_idx
= next_char_idx
;
1197 if (cur_state
->halt
)
1199 /* Reached a halt state.
1200 Check the halt state can satisfy the current context. */
1201 if (!cur_state
->has_constraint
1202 || check_halt_state_context (mctx
, cur_state
,
1203 re_string_cur_idx (&mctx
->input
)))
1205 /* We found an appropriate halt state. */
1206 match_last
= re_string_cur_idx (&mctx
->input
);
1209 /* We found a match, do not modify match_first below. */
1210 p_match_first
= NULL
;
1211 if (!fl_longest_match
)
1218 *p_match_first
+= next_start_idx
;
1223 /* Check NODE match the current context. */
1227 check_halt_node_context (const re_dfa_t
*dfa
, int node
, unsigned int context
)
1229 re_token_type_t type
= dfa
->nodes
[node
].type
;
1230 unsigned int constraint
= dfa
->nodes
[node
].constraint
;
1231 if (type
!= END_OF_RE
)
1235 if (NOT_SATISFY_NEXT_CONSTRAINT (constraint
, context
))
1240 /* Check the halt state STATE match the current context.
1241 Return 0 if not match, if the node, STATE has, is a halt node and
1242 match the context, return the node. */
1246 check_halt_state_context (const re_match_context_t
*mctx
,
1247 const re_dfastate_t
*state
, int idx
)
1250 unsigned int context
;
1252 assert (state
->halt
);
1254 context
= re_string_context_at (&mctx
->input
, idx
, mctx
->eflags
);
1255 for (i
= 0; i
< state
->nodes
.nelem
; ++i
)
1256 if (check_halt_node_context (mctx
->dfa
, state
->nodes
.elems
[i
], context
))
1257 return state
->nodes
.elems
[i
];
1261 /* Compute the next node to which "NFA" transit from NODE("NFA" is a NFA
1262 corresponding to the DFA).
1263 Return the destination node, and update EPS_VIA_NODES, return -1 in case
1268 proceed_next_node (const re_match_context_t
*mctx
, int nregs
, regmatch_t
*regs
,
1269 int *pidx
, int node
, re_node_set
*eps_via_nodes
,
1270 struct re_fail_stack_t
*fs
)
1272 const re_dfa_t
*const dfa
= mctx
->dfa
;
1274 if (IS_EPSILON_NODE (dfa
->nodes
[node
].type
))
1276 re_node_set
*cur_nodes
= &mctx
->state_log
[*pidx
]->nodes
;
1277 re_node_set
*edests
= &dfa
->edests
[node
];
1279 err
= re_node_set_insert (eps_via_nodes
, node
);
1280 if (BE (err
< 0, 0))
1282 /* Pick up a valid destination, or return -1 if none is found. */
1283 for (dest_node
= -1, i
= 0; i
< edests
->nelem
; ++i
)
1285 int candidate
= edests
->elems
[i
];
1286 if (!re_node_set_contains (cur_nodes
, candidate
))
1288 if (dest_node
== -1)
1289 dest_node
= candidate
;
1293 /* In order to avoid infinite loop like "(a*)*", return the second
1294 epsilon-transition if the first was already considered. */
1295 if (re_node_set_contains (eps_via_nodes
, dest_node
))
1298 /* Otherwise, push the second epsilon-transition on the fail stack. */
1300 && push_fail_stack (fs
, *pidx
, candidate
, nregs
, regs
,
1304 /* We know we are going to exit. */
1313 re_token_type_t type
= dfa
->nodes
[node
].type
;
1315 #ifdef RE_ENABLE_I18N
1316 if (dfa
->nodes
[node
].accept_mb
)
1317 naccepted
= check_node_accept_bytes (dfa
, node
, &mctx
->input
, *pidx
);
1319 #endif /* RE_ENABLE_I18N */
1320 if (type
== OP_BACK_REF
)
1322 int subexp_idx
= dfa
->nodes
[node
].opr
.idx
+ 1;
1323 naccepted
= regs
[subexp_idx
].rm_eo
- regs
[subexp_idx
].rm_so
;
1326 if (regs
[subexp_idx
].rm_so
== -1 || regs
[subexp_idx
].rm_eo
== -1)
1330 char *buf
= (char *) re_string_get_buffer (&mctx
->input
);
1331 if (memcmp (buf
+ regs
[subexp_idx
].rm_so
, buf
+ *pidx
,
1340 err
= re_node_set_insert (eps_via_nodes
, node
);
1341 if (BE (err
< 0, 0))
1343 dest_node
= dfa
->edests
[node
].elems
[0];
1344 if (re_node_set_contains (&mctx
->state_log
[*pidx
]->nodes
,
1351 || check_node_accept (mctx
, dfa
->nodes
+ node
, *pidx
))
1353 int dest_node
= dfa
->nexts
[node
];
1354 *pidx
= (naccepted
== 0) ? *pidx
+ 1 : *pidx
+ naccepted
;
1355 if (fs
&& (*pidx
> mctx
->match_last
|| mctx
->state_log
[*pidx
] == NULL
1356 || !re_node_set_contains (&mctx
->state_log
[*pidx
]->nodes
,
1359 re_node_set_empty (eps_via_nodes
);
1366 static reg_errcode_t
1367 internal_function __attribute_warn_unused_result__
1368 push_fail_stack (struct re_fail_stack_t
*fs
, int str_idx
, int dest_node
,
1369 int nregs
, regmatch_t
*regs
, re_node_set
*eps_via_nodes
)
1372 int num
= fs
->num
++;
1373 if (fs
->num
== fs
->alloc
)
1375 struct re_fail_stack_ent_t
*new_array
;
1376 new_array
= realloc (fs
->stack
, (sizeof (struct re_fail_stack_ent_t
)
1378 if (new_array
== NULL
)
1381 fs
->stack
= new_array
;
1383 fs
->stack
[num
].idx
= str_idx
;
1384 fs
->stack
[num
].node
= dest_node
;
1385 fs
->stack
[num
].regs
= re_malloc (regmatch_t
, nregs
);
1386 if (fs
->stack
[num
].regs
== NULL
)
1388 memcpy (fs
->stack
[num
].regs
, regs
, sizeof (regmatch_t
) * nregs
);
1389 err
= re_node_set_init_copy (&fs
->stack
[num
].eps_via_nodes
, eps_via_nodes
);
1395 pop_fail_stack (struct re_fail_stack_t
*fs
, int *pidx
, int nregs
,
1396 regmatch_t
*regs
, re_node_set
*eps_via_nodes
)
1398 int num
= --fs
->num
;
1400 *pidx
= fs
->stack
[num
].idx
;
1401 memcpy (regs
, fs
->stack
[num
].regs
, sizeof (regmatch_t
) * nregs
);
1402 re_node_set_free (eps_via_nodes
);
1403 re_free (fs
->stack
[num
].regs
);
1404 *eps_via_nodes
= fs
->stack
[num
].eps_via_nodes
;
1405 return fs
->stack
[num
].node
;
1408 /* Set the positions where the subexpressions are starts/ends to registers
1410 Note: We assume that pmatch[0] is already set, and
1411 pmatch[i].rm_so == pmatch[i].rm_eo == -1 for 0 < i < nmatch. */
1413 static reg_errcode_t
1414 internal_function __attribute_warn_unused_result__
1415 set_regs (const regex_t
*preg
, const re_match_context_t
*mctx
, size_t nmatch
,
1416 regmatch_t
*pmatch
, int fl_backtrack
)
1418 const re_dfa_t
*dfa
= (const re_dfa_t
*) preg
->buffer
;
1420 re_node_set eps_via_nodes
;
1421 struct re_fail_stack_t
*fs
;
1422 struct re_fail_stack_t fs_body
= { 0, 2, NULL
};
1423 regmatch_t
*prev_idx_match
;
1424 int prev_idx_match_malloced
= 0;
1427 assert (nmatch
> 1);
1428 assert (mctx
->state_log
!= NULL
);
1433 fs
->stack
= re_malloc (struct re_fail_stack_ent_t
, fs
->alloc
);
1434 if (fs
->stack
== NULL
)
1440 cur_node
= dfa
->init_node
;
1441 re_node_set_init_empty (&eps_via_nodes
);
1443 if (__libc_use_alloca (nmatch
* sizeof (regmatch_t
)))
1444 prev_idx_match
= (regmatch_t
*) alloca (nmatch
* sizeof (regmatch_t
));
1447 prev_idx_match
= re_malloc (regmatch_t
, nmatch
);
1448 if (prev_idx_match
== NULL
)
1450 free_fail_stack_return (fs
);
1453 prev_idx_match_malloced
= 1;
1455 memcpy (prev_idx_match
, pmatch
, sizeof (regmatch_t
) * nmatch
);
1457 for (idx
= pmatch
[0].rm_so
; idx
<= pmatch
[0].rm_eo
;)
1459 update_regs (dfa
, pmatch
, prev_idx_match
, cur_node
, idx
, nmatch
);
1461 if (idx
== pmatch
[0].rm_eo
&& cur_node
== mctx
->last_node
)
1466 for (reg_idx
= 0; reg_idx
< nmatch
; ++reg_idx
)
1467 if (pmatch
[reg_idx
].rm_so
> -1 && pmatch
[reg_idx
].rm_eo
== -1)
1469 if (reg_idx
== nmatch
)
1471 re_node_set_free (&eps_via_nodes
);
1472 if (prev_idx_match_malloced
)
1473 re_free (prev_idx_match
);
1474 return free_fail_stack_return (fs
);
1476 cur_node
= pop_fail_stack (fs
, &idx
, nmatch
, pmatch
,
1481 re_node_set_free (&eps_via_nodes
);
1482 if (prev_idx_match_malloced
)
1483 re_free (prev_idx_match
);
1488 /* Proceed to next node. */
1489 cur_node
= proceed_next_node (mctx
, nmatch
, pmatch
, &idx
, cur_node
,
1490 &eps_via_nodes
, fs
);
1492 if (BE (cur_node
< 0, 0))
1494 if (BE (cur_node
== -2, 0))
1496 re_node_set_free (&eps_via_nodes
);
1497 if (prev_idx_match_malloced
)
1498 re_free (prev_idx_match
);
1499 free_fail_stack_return (fs
);
1503 cur_node
= pop_fail_stack (fs
, &idx
, nmatch
, pmatch
,
1507 re_node_set_free (&eps_via_nodes
);
1508 if (prev_idx_match_malloced
)
1509 re_free (prev_idx_match
);
1514 re_node_set_free (&eps_via_nodes
);
1515 if (prev_idx_match_malloced
)
1516 re_free (prev_idx_match
);
1517 return free_fail_stack_return (fs
);
1520 static reg_errcode_t
1522 free_fail_stack_return (struct re_fail_stack_t
*fs
)
1527 for (fs_idx
= 0; fs_idx
< fs
->num
; ++fs_idx
)
1529 re_node_set_free (&fs
->stack
[fs_idx
].eps_via_nodes
);
1530 re_free (fs
->stack
[fs_idx
].regs
);
1532 re_free (fs
->stack
);
1539 update_regs (const re_dfa_t
*dfa
, regmatch_t
*pmatch
,
1540 regmatch_t
*prev_idx_match
, int cur_node
, int cur_idx
, int nmatch
)
1542 int type
= dfa
->nodes
[cur_node
].type
;
1543 if (type
== OP_OPEN_SUBEXP
)
1545 int reg_num
= dfa
->nodes
[cur_node
].opr
.idx
+ 1;
1547 /* We are at the first node of this sub expression. */
1548 if (reg_num
< nmatch
)
1550 pmatch
[reg_num
].rm_so
= cur_idx
;
1551 pmatch
[reg_num
].rm_eo
= -1;
1554 else if (type
== OP_CLOSE_SUBEXP
)
1556 int reg_num
= dfa
->nodes
[cur_node
].opr
.idx
+ 1;
1557 if (reg_num
< nmatch
)
1559 /* We are at the last node of this sub expression. */
1560 if (pmatch
[reg_num
].rm_so
< cur_idx
)
1562 pmatch
[reg_num
].rm_eo
= cur_idx
;
1563 /* This is a non-empty match or we are not inside an optional
1564 subexpression. Accept this right away. */
1565 memcpy (prev_idx_match
, pmatch
, sizeof (regmatch_t
) * nmatch
);
1569 if (dfa
->nodes
[cur_node
].opt_subexp
1570 && prev_idx_match
[reg_num
].rm_so
!= -1)
1571 /* We transited through an empty match for an optional
1572 subexpression, like (a?)*, and this is not the subexp's
1573 first match. Copy back the old content of the registers
1574 so that matches of an inner subexpression are undone as
1575 well, like in ((a?))*. */
1576 memcpy (pmatch
, prev_idx_match
, sizeof (regmatch_t
) * nmatch
);
1578 /* We completed a subexpression, but it may be part of
1579 an optional one, so do not update PREV_IDX_MATCH. */
1580 pmatch
[reg_num
].rm_eo
= cur_idx
;
1586 /* This function checks the STATE_LOG from the SCTX->last_str_idx to 0
1587 and sift the nodes in each states according to the following rules.
1588 Updated state_log will be wrote to STATE_LOG.
1590 Rules: We throw away the Node `a' in the STATE_LOG[STR_IDX] if...
1591 1. When STR_IDX == MATCH_LAST(the last index in the state_log):
1592 If `a' isn't the LAST_NODE and `a' can't epsilon transit to
1593 the LAST_NODE, we throw away the node `a'.
1594 2. When 0 <= STR_IDX < MATCH_LAST and `a' accepts
1595 string `s' and transit to `b':
1596 i. If 'b' isn't in the STATE_LOG[STR_IDX+strlen('s')], we throw
1598 ii. If 'b' is in the STATE_LOG[STR_IDX+strlen('s')] but 'b' is
1599 thrown away, we throw away the node `a'.
1600 3. When 0 <= STR_IDX < MATCH_LAST and 'a' epsilon transit to 'b':
1601 i. If 'b' isn't in the STATE_LOG[STR_IDX], we throw away the
1603 ii. If 'b' is in the STATE_LOG[STR_IDX] but 'b' is thrown away,
1604 we throw away the node `a'. */
1606 #define STATE_NODE_CONTAINS(state,node) \
1607 ((state) != NULL && re_node_set_contains (&(state)->nodes, node))
1609 static reg_errcode_t
1611 sift_states_backward (const re_match_context_t
*mctx
, re_sift_context_t
*sctx
)
1615 int str_idx
= sctx
->last_str_idx
;
1616 re_node_set cur_dest
;
1619 assert (mctx
->state_log
!= NULL
&& mctx
->state_log
[str_idx
] != NULL
);
1622 /* Build sifted state_log[str_idx]. It has the nodes which can epsilon
1623 transit to the last_node and the last_node itself. */
1624 err
= re_node_set_init_1 (&cur_dest
, sctx
->last_node
);
1625 if (BE (err
!= REG_NOERROR
, 0))
1627 err
= update_cur_sifted_state (mctx
, sctx
, str_idx
, &cur_dest
);
1628 if (BE (err
!= REG_NOERROR
, 0))
1631 /* Then check each states in the state_log. */
1634 /* Update counters. */
1635 null_cnt
= (sctx
->sifted_states
[str_idx
] == NULL
) ? null_cnt
+ 1 : 0;
1636 if (null_cnt
> mctx
->max_mb_elem_len
)
1638 memset (sctx
->sifted_states
, '\0',
1639 sizeof (re_dfastate_t
*) * str_idx
);
1640 re_node_set_free (&cur_dest
);
1643 re_node_set_empty (&cur_dest
);
1646 if (mctx
->state_log
[str_idx
])
1648 err
= build_sifted_states (mctx
, sctx
, str_idx
, &cur_dest
);
1649 if (BE (err
!= REG_NOERROR
, 0))
1653 /* Add all the nodes which satisfy the following conditions:
1654 - It can epsilon transit to a node in CUR_DEST.
1656 And update state_log. */
1657 err
= update_cur_sifted_state (mctx
, sctx
, str_idx
, &cur_dest
);
1658 if (BE (err
!= REG_NOERROR
, 0))
1663 re_node_set_free (&cur_dest
);
1667 static reg_errcode_t
1668 internal_function __attribute_warn_unused_result__
1669 build_sifted_states (const re_match_context_t
*mctx
, re_sift_context_t
*sctx
,
1670 int str_idx
, re_node_set
*cur_dest
)
1672 const re_dfa_t
*const dfa
= mctx
->dfa
;
1673 const re_node_set
*cur_src
= &mctx
->state_log
[str_idx
]->non_eps_nodes
;
1676 /* Then build the next sifted state.
1677 We build the next sifted state on `cur_dest', and update
1678 `sifted_states[str_idx]' with `cur_dest'.
1680 `cur_dest' is the sifted state from `state_log[str_idx + 1]'.
1681 `cur_src' points the node_set of the old `state_log[str_idx]'
1682 (with the epsilon nodes pre-filtered out). */
1683 for (i
= 0; i
< cur_src
->nelem
; i
++)
1685 int prev_node
= cur_src
->elems
[i
];
1690 re_token_type_t type
= dfa
->nodes
[prev_node
].type
;
1691 assert (!IS_EPSILON_NODE (type
));
1693 #ifdef RE_ENABLE_I18N
1694 /* If the node may accept `multi byte'. */
1695 if (dfa
->nodes
[prev_node
].accept_mb
)
1696 naccepted
= sift_states_iter_mb (mctx
, sctx
, prev_node
,
1697 str_idx
, sctx
->last_str_idx
);
1698 #endif /* RE_ENABLE_I18N */
1700 /* We don't check backreferences here.
1701 See update_cur_sifted_state(). */
1703 && check_node_accept (mctx
, dfa
->nodes
+ prev_node
, str_idx
)
1704 && STATE_NODE_CONTAINS (sctx
->sifted_states
[str_idx
+ 1],
1705 dfa
->nexts
[prev_node
]))
1711 if (sctx
->limits
.nelem
)
1713 int to_idx
= str_idx
+ naccepted
;
1714 if (check_dst_limits (mctx
, &sctx
->limits
,
1715 dfa
->nexts
[prev_node
], to_idx
,
1716 prev_node
, str_idx
))
1719 ret
= re_node_set_insert (cur_dest
, prev_node
);
1720 if (BE (ret
== -1, 0))
1727 /* Helper functions. */
1729 static reg_errcode_t
1731 clean_state_log_if_needed (re_match_context_t
*mctx
, int next_state_log_idx
)
1733 int top
= mctx
->state_log_top
;
1735 if ((next_state_log_idx
>= mctx
->input
.bufs_len
1736 && mctx
->input
.bufs_len
< mctx
->input
.len
)
1737 || (next_state_log_idx
>= mctx
->input
.valid_len
1738 && mctx
->input
.valid_len
< mctx
->input
.len
))
1741 err
= extend_buffers (mctx
);
1742 if (BE (err
!= REG_NOERROR
, 0))
1746 if (top
< next_state_log_idx
)
1748 memset (mctx
->state_log
+ top
+ 1, '\0',
1749 sizeof (re_dfastate_t
*) * (next_state_log_idx
- top
));
1750 mctx
->state_log_top
= next_state_log_idx
;
1755 static reg_errcode_t
1757 merge_state_array (const re_dfa_t
*dfa
, re_dfastate_t
**dst
,
1758 re_dfastate_t
**src
, int num
)
1762 for (st_idx
= 0; st_idx
< num
; ++st_idx
)
1764 if (dst
[st_idx
] == NULL
)
1765 dst
[st_idx
] = src
[st_idx
];
1766 else if (src
[st_idx
] != NULL
)
1768 re_node_set merged_set
;
1769 err
= re_node_set_init_union (&merged_set
, &dst
[st_idx
]->nodes
,
1770 &src
[st_idx
]->nodes
);
1771 if (BE (err
!= REG_NOERROR
, 0))
1773 dst
[st_idx
] = re_acquire_state (&err
, dfa
, &merged_set
);
1774 re_node_set_free (&merged_set
);
1775 if (BE (err
!= REG_NOERROR
, 0))
1782 static reg_errcode_t
1784 update_cur_sifted_state (const re_match_context_t
*mctx
,
1785 re_sift_context_t
*sctx
, int str_idx
,
1786 re_node_set
*dest_nodes
)
1788 const re_dfa_t
*const dfa
= mctx
->dfa
;
1789 reg_errcode_t err
= REG_NOERROR
;
1790 const re_node_set
*candidates
;
1791 candidates
= ((mctx
->state_log
[str_idx
] == NULL
) ? NULL
1792 : &mctx
->state_log
[str_idx
]->nodes
);
1794 if (dest_nodes
->nelem
== 0)
1795 sctx
->sifted_states
[str_idx
] = NULL
;
1800 /* At first, add the nodes which can epsilon transit to a node in
1802 err
= add_epsilon_src_nodes (dfa
, dest_nodes
, candidates
);
1803 if (BE (err
!= REG_NOERROR
, 0))
1806 /* Then, check the limitations in the current sift_context. */
1807 if (sctx
->limits
.nelem
)
1809 err
= check_subexp_limits (dfa
, dest_nodes
, candidates
, &sctx
->limits
,
1810 mctx
->bkref_ents
, str_idx
);
1811 if (BE (err
!= REG_NOERROR
, 0))
1816 sctx
->sifted_states
[str_idx
] = re_acquire_state (&err
, dfa
, dest_nodes
);
1817 if (BE (err
!= REG_NOERROR
, 0))
1821 if (candidates
&& mctx
->state_log
[str_idx
]->has_backref
)
1823 err
= sift_states_bkref (mctx
, sctx
, str_idx
, candidates
);
1824 if (BE (err
!= REG_NOERROR
, 0))
1830 static reg_errcode_t
1831 internal_function __attribute_warn_unused_result__
1832 add_epsilon_src_nodes (const re_dfa_t
*dfa
, re_node_set
*dest_nodes
,
1833 const re_node_set
*candidates
)
1835 reg_errcode_t err
= REG_NOERROR
;
1838 re_dfastate_t
*state
= re_acquire_state (&err
, dfa
, dest_nodes
);
1839 if (BE (err
!= REG_NOERROR
, 0))
1842 if (!state
->inveclosure
.alloc
)
1844 err
= re_node_set_alloc (&state
->inveclosure
, dest_nodes
->nelem
);
1845 if (BE (err
!= REG_NOERROR
, 0))
1847 for (i
= 0; i
< dest_nodes
->nelem
; i
++)
1849 err
= re_node_set_merge (&state
->inveclosure
,
1850 dfa
->inveclosures
+ dest_nodes
->elems
[i
]);
1851 if (BE (err
!= REG_NOERROR
, 0))
1855 return re_node_set_add_intersect (dest_nodes
, candidates
,
1856 &state
->inveclosure
);
1859 static reg_errcode_t
1861 sub_epsilon_src_nodes (const re_dfa_t
*dfa
, int node
, re_node_set
*dest_nodes
,
1862 const re_node_set
*candidates
)
1866 re_node_set
*inv_eclosure
= dfa
->inveclosures
+ node
;
1867 re_node_set except_nodes
;
1868 re_node_set_init_empty (&except_nodes
);
1869 for (ecl_idx
= 0; ecl_idx
< inv_eclosure
->nelem
; ++ecl_idx
)
1871 int cur_node
= inv_eclosure
->elems
[ecl_idx
];
1872 if (cur_node
== node
)
1874 if (IS_EPSILON_NODE (dfa
->nodes
[cur_node
].type
))
1876 int edst1
= dfa
->edests
[cur_node
].elems
[0];
1877 int edst2
= ((dfa
->edests
[cur_node
].nelem
> 1)
1878 ? dfa
->edests
[cur_node
].elems
[1] : -1);
1879 if ((!re_node_set_contains (inv_eclosure
, edst1
)
1880 && re_node_set_contains (dest_nodes
, edst1
))
1882 && !re_node_set_contains (inv_eclosure
, edst2
)
1883 && re_node_set_contains (dest_nodes
, edst2
)))
1885 err
= re_node_set_add_intersect (&except_nodes
, candidates
,
1886 dfa
->inveclosures
+ cur_node
);
1887 if (BE (err
!= REG_NOERROR
, 0))
1889 re_node_set_free (&except_nodes
);
1895 for (ecl_idx
= 0; ecl_idx
< inv_eclosure
->nelem
; ++ecl_idx
)
1897 int cur_node
= inv_eclosure
->elems
[ecl_idx
];
1898 if (!re_node_set_contains (&except_nodes
, cur_node
))
1900 int idx
= re_node_set_contains (dest_nodes
, cur_node
) - 1;
1901 re_node_set_remove_at (dest_nodes
, idx
);
1904 re_node_set_free (&except_nodes
);
1910 check_dst_limits (const re_match_context_t
*mctx
, re_node_set
*limits
,
1911 int dst_node
, int dst_idx
, int src_node
, int src_idx
)
1913 const re_dfa_t
*const dfa
= mctx
->dfa
;
1914 int lim_idx
, src_pos
, dst_pos
;
1916 int dst_bkref_idx
= search_cur_bkref_entry (mctx
, dst_idx
);
1917 int src_bkref_idx
= search_cur_bkref_entry (mctx
, src_idx
);
1918 for (lim_idx
= 0; lim_idx
< limits
->nelem
; ++lim_idx
)
1921 struct re_backref_cache_entry
*ent
;
1922 ent
= mctx
->bkref_ents
+ limits
->elems
[lim_idx
];
1923 subexp_idx
= dfa
->nodes
[ent
->node
].opr
.idx
;
1925 dst_pos
= check_dst_limits_calc_pos (mctx
, limits
->elems
[lim_idx
],
1926 subexp_idx
, dst_node
, dst_idx
,
1928 src_pos
= check_dst_limits_calc_pos (mctx
, limits
->elems
[lim_idx
],
1929 subexp_idx
, src_node
, src_idx
,
1933 <src> <dst> ( <subexp> )
1934 ( <subexp> ) <src> <dst>
1935 ( <subexp1> <src> <subexp2> <dst> <subexp3> ) */
1936 if (src_pos
== dst_pos
)
1937 continue; /* This is unrelated limitation. */
1946 check_dst_limits_calc_pos_1 (const re_match_context_t
*mctx
, int boundaries
,
1947 int subexp_idx
, int from_node
, int bkref_idx
)
1949 const re_dfa_t
*const dfa
= mctx
->dfa
;
1950 const re_node_set
*eclosures
= dfa
->eclosures
+ from_node
;
1953 /* Else, we are on the boundary: examine the nodes on the epsilon
1955 for (node_idx
= 0; node_idx
< eclosures
->nelem
; ++node_idx
)
1957 int node
= eclosures
->elems
[node_idx
];
1958 switch (dfa
->nodes
[node
].type
)
1961 if (bkref_idx
!= -1)
1963 struct re_backref_cache_entry
*ent
= mctx
->bkref_ents
+ bkref_idx
;
1968 if (ent
->node
!= node
)
1971 if (subexp_idx
< BITSET_WORD_BITS
1972 && !(ent
->eps_reachable_subexps_map
1973 & ((bitset_word_t
) 1 << subexp_idx
)))
1976 /* Recurse trying to reach the OP_OPEN_SUBEXP and
1977 OP_CLOSE_SUBEXP cases below. But, if the
1978 destination node is the same node as the source
1979 node, don't recurse because it would cause an
1980 infinite loop: a regex that exhibits this behavior
1982 dst
= dfa
->edests
[node
].elems
[0];
1983 if (dst
== from_node
)
1987 else /* if (boundaries & 2) */
1992 check_dst_limits_calc_pos_1 (mctx
, boundaries
, subexp_idx
,
1994 if (cpos
== -1 /* && (boundaries & 1) */)
1996 if (cpos
== 0 && (boundaries
& 2))
1999 if (subexp_idx
< BITSET_WORD_BITS
)
2000 ent
->eps_reachable_subexps_map
2001 &= ~((bitset_word_t
) 1 << subexp_idx
);
2003 while (ent
++->more
);
2007 case OP_OPEN_SUBEXP
:
2008 if ((boundaries
& 1) && subexp_idx
== dfa
->nodes
[node
].opr
.idx
)
2012 case OP_CLOSE_SUBEXP
:
2013 if ((boundaries
& 2) && subexp_idx
== dfa
->nodes
[node
].opr
.idx
)
2022 return (boundaries
& 2) ? 1 : 0;
2027 check_dst_limits_calc_pos (const re_match_context_t
*mctx
, int limit
,
2028 int subexp_idx
, int from_node
, int str_idx
,
2031 struct re_backref_cache_entry
*lim
= mctx
->bkref_ents
+ limit
;
2034 /* If we are outside the range of the subexpression, return -1 or 1. */
2035 if (str_idx
< lim
->subexp_from
)
2038 if (lim
->subexp_to
< str_idx
)
2041 /* If we are within the subexpression, return 0. */
2042 boundaries
= (str_idx
== lim
->subexp_from
);
2043 boundaries
|= (str_idx
== lim
->subexp_to
) << 1;
2044 if (boundaries
== 0)
2047 /* Else, examine epsilon closure. */
2048 return check_dst_limits_calc_pos_1 (mctx
, boundaries
, subexp_idx
,
2049 from_node
, bkref_idx
);
2052 /* Check the limitations of sub expressions LIMITS, and remove the nodes
2053 which are against limitations from DEST_NODES. */
2055 static reg_errcode_t
2057 check_subexp_limits (const re_dfa_t
*dfa
, re_node_set
*dest_nodes
,
2058 const re_node_set
*candidates
, re_node_set
*limits
,
2059 struct re_backref_cache_entry
*bkref_ents
, int str_idx
)
2062 int node_idx
, lim_idx
;
2064 for (lim_idx
= 0; lim_idx
< limits
->nelem
; ++lim_idx
)
2067 struct re_backref_cache_entry
*ent
;
2068 ent
= bkref_ents
+ limits
->elems
[lim_idx
];
2070 if (str_idx
<= ent
->subexp_from
|| ent
->str_idx
< str_idx
)
2071 continue; /* This is unrelated limitation. */
2073 subexp_idx
= dfa
->nodes
[ent
->node
].opr
.idx
;
2074 if (ent
->subexp_to
== str_idx
)
2078 for (node_idx
= 0; node_idx
< dest_nodes
->nelem
; ++node_idx
)
2080 int node
= dest_nodes
->elems
[node_idx
];
2081 re_token_type_t type
= dfa
->nodes
[node
].type
;
2082 if (type
== OP_OPEN_SUBEXP
2083 && subexp_idx
== dfa
->nodes
[node
].opr
.idx
)
2085 else if (type
== OP_CLOSE_SUBEXP
2086 && subexp_idx
== dfa
->nodes
[node
].opr
.idx
)
2090 /* Check the limitation of the open subexpression. */
2091 /* Note that (ent->subexp_to = str_idx != ent->subexp_from). */
2094 err
= sub_epsilon_src_nodes (dfa
, ops_node
, dest_nodes
,
2096 if (BE (err
!= REG_NOERROR
, 0))
2100 /* Check the limitation of the close subexpression. */
2102 for (node_idx
= 0; node_idx
< dest_nodes
->nelem
; ++node_idx
)
2104 int node
= dest_nodes
->elems
[node_idx
];
2105 if (!re_node_set_contains (dfa
->inveclosures
+ node
,
2107 && !re_node_set_contains (dfa
->eclosures
+ node
,
2110 /* It is against this limitation.
2111 Remove it form the current sifted state. */
2112 err
= sub_epsilon_src_nodes (dfa
, node
, dest_nodes
,
2114 if (BE (err
!= REG_NOERROR
, 0))
2120 else /* (ent->subexp_to != str_idx) */
2122 for (node_idx
= 0; node_idx
< dest_nodes
->nelem
; ++node_idx
)
2124 int node
= dest_nodes
->elems
[node_idx
];
2125 re_token_type_t type
= dfa
->nodes
[node
].type
;
2126 if (type
== OP_CLOSE_SUBEXP
|| type
== OP_OPEN_SUBEXP
)
2128 if (subexp_idx
!= dfa
->nodes
[node
].opr
.idx
)
2130 /* It is against this limitation.
2131 Remove it form the current sifted state. */
2132 err
= sub_epsilon_src_nodes (dfa
, node
, dest_nodes
,
2134 if (BE (err
!= REG_NOERROR
, 0))
2143 static reg_errcode_t
2144 internal_function __attribute_warn_unused_result__
2145 sift_states_bkref (const re_match_context_t
*mctx
, re_sift_context_t
*sctx
,
2146 int str_idx
, const re_node_set
*candidates
)
2148 const re_dfa_t
*const dfa
= mctx
->dfa
;
2151 re_sift_context_t local_sctx
;
2152 int first_idx
= search_cur_bkref_entry (mctx
, str_idx
);
2154 if (first_idx
== -1)
2157 local_sctx
.sifted_states
= NULL
; /* Mark that it hasn't been initialized. */
2159 for (node_idx
= 0; node_idx
< candidates
->nelem
; ++node_idx
)
2162 re_token_type_t type
;
2163 struct re_backref_cache_entry
*entry
;
2164 node
= candidates
->elems
[node_idx
];
2165 type
= dfa
->nodes
[node
].type
;
2166 /* Avoid infinite loop for the REs like "()\1+". */
2167 if (node
== sctx
->last_node
&& str_idx
== sctx
->last_str_idx
)
2169 if (type
!= OP_BACK_REF
)
2172 entry
= mctx
->bkref_ents
+ first_idx
;
2173 enabled_idx
= first_idx
;
2180 re_dfastate_t
*cur_state
;
2182 if (entry
->node
!= node
)
2184 subexp_len
= entry
->subexp_to
- entry
->subexp_from
;
2185 to_idx
= str_idx
+ subexp_len
;
2186 dst_node
= (subexp_len
? dfa
->nexts
[node
]
2187 : dfa
->edests
[node
].elems
[0]);
2189 if (to_idx
> sctx
->last_str_idx
2190 || sctx
->sifted_states
[to_idx
] == NULL
2191 || !STATE_NODE_CONTAINS (sctx
->sifted_states
[to_idx
], dst_node
)
2192 || check_dst_limits (mctx
, &sctx
->limits
, node
,
2193 str_idx
, dst_node
, to_idx
))
2196 if (local_sctx
.sifted_states
== NULL
)
2199 err
= re_node_set_init_copy (&local_sctx
.limits
, &sctx
->limits
);
2200 if (BE (err
!= REG_NOERROR
, 0))
2203 local_sctx
.last_node
= node
;
2204 local_sctx
.last_str_idx
= str_idx
;
2205 ret
= re_node_set_insert (&local_sctx
.limits
, enabled_idx
);
2206 if (BE (ret
< 0, 0))
2211 cur_state
= local_sctx
.sifted_states
[str_idx
];
2212 err
= sift_states_backward (mctx
, &local_sctx
);
2213 if (BE (err
!= REG_NOERROR
, 0))
2215 if (sctx
->limited_states
!= NULL
)
2217 err
= merge_state_array (dfa
, sctx
->limited_states
,
2218 local_sctx
.sifted_states
,
2220 if (BE (err
!= REG_NOERROR
, 0))
2223 local_sctx
.sifted_states
[str_idx
] = cur_state
;
2224 re_node_set_remove (&local_sctx
.limits
, enabled_idx
);
2226 /* mctx->bkref_ents may have changed, reload the pointer. */
2227 entry
= mctx
->bkref_ents
+ enabled_idx
;
2229 while (enabled_idx
++, entry
++->more
);
2233 if (local_sctx
.sifted_states
!= NULL
)
2235 re_node_set_free (&local_sctx
.limits
);
2242 #ifdef RE_ENABLE_I18N
2245 sift_states_iter_mb (const re_match_context_t
*mctx
, re_sift_context_t
*sctx
,
2246 int node_idx
, int str_idx
, int max_str_idx
)
2248 const re_dfa_t
*const dfa
= mctx
->dfa
;
2250 /* Check the node can accept `multi byte'. */
2251 naccepted
= check_node_accept_bytes (dfa
, node_idx
, &mctx
->input
, str_idx
);
2252 if (naccepted
> 0 && str_idx
+ naccepted
<= max_str_idx
&&
2253 !STATE_NODE_CONTAINS (sctx
->sifted_states
[str_idx
+ naccepted
],
2254 dfa
->nexts
[node_idx
]))
2255 /* The node can't accept the `multi byte', or the
2256 destination was already thrown away, then the node
2257 could't accept the current input `multi byte'. */
2259 /* Otherwise, it is sure that the node could accept
2260 `naccepted' bytes input. */
2263 #endif /* RE_ENABLE_I18N */
2266 /* Functions for state transition. */
2268 /* Return the next state to which the current state STATE will transit by
2269 accepting the current input byte, and update STATE_LOG if necessary.
2270 If STATE can accept a multibyte char/collating element/back reference
2271 update the destination of STATE_LOG. */
2273 static re_dfastate_t
*
2274 internal_function __attribute_warn_unused_result__
2275 transit_state (reg_errcode_t
*err
, re_match_context_t
*mctx
,
2276 re_dfastate_t
*state
)
2278 re_dfastate_t
**trtable
;
2281 #ifdef RE_ENABLE_I18N
2282 /* If the current state can accept multibyte. */
2283 if (BE (state
->accept_mb
, 0))
2285 *err
= transit_state_mb (mctx
, state
);
2286 if (BE (*err
!= REG_NOERROR
, 0))
2289 #endif /* RE_ENABLE_I18N */
2291 /* Then decide the next state with the single byte. */
2294 /* don't use transition table */
2295 return transit_state_sb (err
, mctx
, state
);
2298 /* Use transition table */
2299 ch
= re_string_fetch_byte (&mctx
->input
);
2302 trtable
= state
->trtable
;
2303 if (BE (trtable
!= NULL
, 1))
2306 trtable
= state
->word_trtable
;
2307 if (BE (trtable
!= NULL
, 1))
2309 unsigned int context
;
2311 = re_string_context_at (&mctx
->input
,
2312 re_string_cur_idx (&mctx
->input
) - 1,
2314 if (IS_WORD_CONTEXT (context
))
2315 return trtable
[ch
+ SBC_MAX
];
2320 if (!build_trtable (mctx
->dfa
, state
))
2326 /* Retry, we now have a transition table. */
2330 /* Update the state_log if we need */
2333 merge_state_with_log (reg_errcode_t
*err
, re_match_context_t
*mctx
,
2334 re_dfastate_t
*next_state
)
2336 const re_dfa_t
*const dfa
= mctx
->dfa
;
2337 int cur_idx
= re_string_cur_idx (&mctx
->input
);
2339 if (cur_idx
> mctx
->state_log_top
)
2341 mctx
->state_log
[cur_idx
] = next_state
;
2342 mctx
->state_log_top
= cur_idx
;
2344 else if (mctx
->state_log
[cur_idx
] == 0)
2346 mctx
->state_log
[cur_idx
] = next_state
;
2350 re_dfastate_t
*pstate
;
2351 unsigned int context
;
2352 re_node_set next_nodes
, *log_nodes
, *table_nodes
= NULL
;
2353 /* If (state_log[cur_idx] != 0), it implies that cur_idx is
2354 the destination of a multibyte char/collating element/
2355 back reference. Then the next state is the union set of
2356 these destinations and the results of the transition table. */
2357 pstate
= mctx
->state_log
[cur_idx
];
2358 log_nodes
= pstate
->entrance_nodes
;
2359 if (next_state
!= NULL
)
2361 table_nodes
= next_state
->entrance_nodes
;
2362 *err
= re_node_set_init_union (&next_nodes
, table_nodes
,
2364 if (BE (*err
!= REG_NOERROR
, 0))
2368 next_nodes
= *log_nodes
;
2369 /* Note: We already add the nodes of the initial state,
2370 then we don't need to add them here. */
2372 context
= re_string_context_at (&mctx
->input
,
2373 re_string_cur_idx (&mctx
->input
) - 1,
2375 next_state
= mctx
->state_log
[cur_idx
]
2376 = re_acquire_state_context (err
, dfa
, &next_nodes
, context
);
2377 /* We don't need to check errors here, since the return value of
2378 this function is next_state and ERR is already set. */
2380 if (table_nodes
!= NULL
)
2381 re_node_set_free (&next_nodes
);
2384 if (BE (dfa
->nbackref
, 0) && next_state
!= NULL
)
2386 /* Check OP_OPEN_SUBEXP in the current state in case that we use them
2387 later. We must check them here, since the back references in the
2388 next state might use them. */
2389 *err
= check_subexp_matching_top (mctx
, &next_state
->nodes
,
2391 if (BE (*err
!= REG_NOERROR
, 0))
2394 /* If the next state has back references. */
2395 if (next_state
->has_backref
)
2397 *err
= transit_state_bkref (mctx
, &next_state
->nodes
);
2398 if (BE (*err
!= REG_NOERROR
, 0))
2400 next_state
= mctx
->state_log
[cur_idx
];
2407 /* Skip bytes in the input that correspond to part of a
2408 multi-byte match, then look in the log for a state
2409 from which to restart matching. */
2412 find_recover_state (reg_errcode_t
*err
, re_match_context_t
*mctx
)
2414 re_dfastate_t
*cur_state
;
2417 int max
= mctx
->state_log_top
;
2418 int cur_str_idx
= re_string_cur_idx (&mctx
->input
);
2422 if (++cur_str_idx
> max
)
2424 re_string_skip_bytes (&mctx
->input
, 1);
2426 while (mctx
->state_log
[cur_str_idx
] == NULL
);
2428 cur_state
= merge_state_with_log (err
, mctx
, NULL
);
2430 while (*err
== REG_NOERROR
&& cur_state
== NULL
);
2434 /* Helper functions for transit_state. */
2436 /* From the node set CUR_NODES, pick up the nodes whose types are
2437 OP_OPEN_SUBEXP and which have corresponding back references in the regular
2438 expression. And register them to use them later for evaluating the
2439 correspoding back references. */
2441 static reg_errcode_t
2443 check_subexp_matching_top (re_match_context_t
*mctx
, re_node_set
*cur_nodes
,
2446 const re_dfa_t
*const dfa
= mctx
->dfa
;
2450 /* TODO: This isn't efficient.
2451 Because there might be more than one nodes whose types are
2452 OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all
2455 for (node_idx
= 0; node_idx
< cur_nodes
->nelem
; ++node_idx
)
2457 int node
= cur_nodes
->elems
[node_idx
];
2458 if (dfa
->nodes
[node
].type
== OP_OPEN_SUBEXP
2459 && dfa
->nodes
[node
].opr
.idx
< BITSET_WORD_BITS
2460 && (dfa
->used_bkref_map
2461 & ((bitset_word_t
) 1 << dfa
->nodes
[node
].opr
.idx
)))
2463 err
= match_ctx_add_subtop (mctx
, node
, str_idx
);
2464 if (BE (err
!= REG_NOERROR
, 0))
2472 /* Return the next state to which the current state STATE will transit by
2473 accepting the current input byte. */
2475 static re_dfastate_t
*
2476 transit_state_sb (reg_errcode_t
*err
, re_match_context_t
*mctx
,
2477 re_dfastate_t
*state
)
2479 const re_dfa_t
*const dfa
= mctx
->dfa
;
2480 re_node_set next_nodes
;
2481 re_dfastate_t
*next_state
;
2482 int node_cnt
, cur_str_idx
= re_string_cur_idx (&mctx
->input
);
2483 unsigned int context
;
2485 *err
= re_node_set_alloc (&next_nodes
, state
->nodes
.nelem
+ 1);
2486 if (BE (*err
!= REG_NOERROR
, 0))
2488 for (node_cnt
= 0; node_cnt
< state
->nodes
.nelem
; ++node_cnt
)
2490 int cur_node
= state
->nodes
.elems
[node_cnt
];
2491 if (check_node_accept (mctx
, dfa
->nodes
+ cur_node
, cur_str_idx
))
2493 *err
= re_node_set_merge (&next_nodes
,
2494 dfa
->eclosures
+ dfa
->nexts
[cur_node
]);
2495 if (BE (*err
!= REG_NOERROR
, 0))
2497 re_node_set_free (&next_nodes
);
2502 context
= re_string_context_at (&mctx
->input
, cur_str_idx
, mctx
->eflags
);
2503 next_state
= re_acquire_state_context (err
, dfa
, &next_nodes
, context
);
2504 /* We don't need to check errors here, since the return value of
2505 this function is next_state and ERR is already set. */
2507 re_node_set_free (&next_nodes
);
2508 re_string_skip_bytes (&mctx
->input
, 1);
2513 #ifdef RE_ENABLE_I18N
2514 static reg_errcode_t
2516 transit_state_mb (re_match_context_t
*mctx
, re_dfastate_t
*pstate
)
2518 const re_dfa_t
*const dfa
= mctx
->dfa
;
2522 for (i
= 0; i
< pstate
->nodes
.nelem
; ++i
)
2524 re_node_set dest_nodes
, *new_nodes
;
2525 int cur_node_idx
= pstate
->nodes
.elems
[i
];
2526 int naccepted
, dest_idx
;
2527 unsigned int context
;
2528 re_dfastate_t
*dest_state
;
2530 if (!dfa
->nodes
[cur_node_idx
].accept_mb
)
2533 if (dfa
->nodes
[cur_node_idx
].constraint
)
2535 context
= re_string_context_at (&mctx
->input
,
2536 re_string_cur_idx (&mctx
->input
),
2538 if (NOT_SATISFY_NEXT_CONSTRAINT (dfa
->nodes
[cur_node_idx
].constraint
,
2543 /* How many bytes the node can accept? */
2544 naccepted
= check_node_accept_bytes (dfa
, cur_node_idx
, &mctx
->input
,
2545 re_string_cur_idx (&mctx
->input
));
2549 /* The node can accepts `naccepted' bytes. */
2550 dest_idx
= re_string_cur_idx (&mctx
->input
) + naccepted
;
2551 mctx
->max_mb_elem_len
= ((mctx
->max_mb_elem_len
< naccepted
) ? naccepted
2552 : mctx
->max_mb_elem_len
);
2553 err
= clean_state_log_if_needed (mctx
, dest_idx
);
2554 if (BE (err
!= REG_NOERROR
, 0))
2557 assert (dfa
->nexts
[cur_node_idx
] != -1);
2559 new_nodes
= dfa
->eclosures
+ dfa
->nexts
[cur_node_idx
];
2561 dest_state
= mctx
->state_log
[dest_idx
];
2562 if (dest_state
== NULL
)
2563 dest_nodes
= *new_nodes
;
2566 err
= re_node_set_init_union (&dest_nodes
,
2567 dest_state
->entrance_nodes
, new_nodes
);
2568 if (BE (err
!= REG_NOERROR
, 0))
2571 context
= re_string_context_at (&mctx
->input
, dest_idx
- 1,
2573 mctx
->state_log
[dest_idx
]
2574 = re_acquire_state_context (&err
, dfa
, &dest_nodes
, context
);
2575 if (dest_state
!= NULL
)
2576 re_node_set_free (&dest_nodes
);
2577 if (BE (mctx
->state_log
[dest_idx
] == NULL
&& err
!= REG_NOERROR
, 0))
2582 #endif /* RE_ENABLE_I18N */
2584 static reg_errcode_t
2586 transit_state_bkref (re_match_context_t
*mctx
, const re_node_set
*nodes
)
2588 const re_dfa_t
*const dfa
= mctx
->dfa
;
2591 int cur_str_idx
= re_string_cur_idx (&mctx
->input
);
2593 for (i
= 0; i
< nodes
->nelem
; ++i
)
2595 int dest_str_idx
, prev_nelem
, bkc_idx
;
2596 int node_idx
= nodes
->elems
[i
];
2597 unsigned int context
;
2598 const re_token_t
*node
= dfa
->nodes
+ node_idx
;
2599 re_node_set
*new_dest_nodes
;
2601 /* Check whether `node' is a backreference or not. */
2602 if (node
->type
!= OP_BACK_REF
)
2605 if (node
->constraint
)
2607 context
= re_string_context_at (&mctx
->input
, cur_str_idx
,
2609 if (NOT_SATISFY_NEXT_CONSTRAINT (node
->constraint
, context
))
2613 /* `node' is a backreference.
2614 Check the substring which the substring matched. */
2615 bkc_idx
= mctx
->nbkref_ents
;
2616 err
= get_subexp (mctx
, node_idx
, cur_str_idx
);
2617 if (BE (err
!= REG_NOERROR
, 0))
2620 /* And add the epsilon closures (which is `new_dest_nodes') of
2621 the backreference to appropriate state_log. */
2623 assert (dfa
->nexts
[node_idx
] != -1);
2625 for (; bkc_idx
< mctx
->nbkref_ents
; ++bkc_idx
)
2628 re_dfastate_t
*dest_state
;
2629 struct re_backref_cache_entry
*bkref_ent
;
2630 bkref_ent
= mctx
->bkref_ents
+ bkc_idx
;
2631 if (bkref_ent
->node
!= node_idx
|| bkref_ent
->str_idx
!= cur_str_idx
)
2633 subexp_len
= bkref_ent
->subexp_to
- bkref_ent
->subexp_from
;
2634 new_dest_nodes
= (subexp_len
== 0
2635 ? dfa
->eclosures
+ dfa
->edests
[node_idx
].elems
[0]
2636 : dfa
->eclosures
+ dfa
->nexts
[node_idx
]);
2637 dest_str_idx
= (cur_str_idx
+ bkref_ent
->subexp_to
2638 - bkref_ent
->subexp_from
);
2639 context
= re_string_context_at (&mctx
->input
, dest_str_idx
- 1,
2641 dest_state
= mctx
->state_log
[dest_str_idx
];
2642 prev_nelem
= ((mctx
->state_log
[cur_str_idx
] == NULL
) ? 0
2643 : mctx
->state_log
[cur_str_idx
]->nodes
.nelem
);
2644 /* Add `new_dest_node' to state_log. */
2645 if (dest_state
== NULL
)
2647 mctx
->state_log
[dest_str_idx
]
2648 = re_acquire_state_context (&err
, dfa
, new_dest_nodes
,
2650 if (BE (mctx
->state_log
[dest_str_idx
] == NULL
2651 && err
!= REG_NOERROR
, 0))
2656 re_node_set dest_nodes
;
2657 err
= re_node_set_init_union (&dest_nodes
,
2658 dest_state
->entrance_nodes
,
2660 if (BE (err
!= REG_NOERROR
, 0))
2662 re_node_set_free (&dest_nodes
);
2665 mctx
->state_log
[dest_str_idx
]
2666 = re_acquire_state_context (&err
, dfa
, &dest_nodes
, context
);
2667 re_node_set_free (&dest_nodes
);
2668 if (BE (mctx
->state_log
[dest_str_idx
] == NULL
2669 && err
!= REG_NOERROR
, 0))
2672 /* We need to check recursively if the backreference can epsilon
2675 && mctx
->state_log
[cur_str_idx
]->nodes
.nelem
> prev_nelem
)
2677 err
= check_subexp_matching_top (mctx
, new_dest_nodes
,
2679 if (BE (err
!= REG_NOERROR
, 0))
2681 err
= transit_state_bkref (mctx
, new_dest_nodes
);
2682 if (BE (err
!= REG_NOERROR
, 0))
2692 /* Enumerate all the candidates which the backreference BKREF_NODE can match
2693 at BKREF_STR_IDX, and register them by match_ctx_add_entry().
2694 Note that we might collect inappropriate candidates here.
2695 However, the cost of checking them strictly here is too high, then we
2696 delay these checking for prune_impossible_nodes(). */
2698 static reg_errcode_t
2699 internal_function __attribute_warn_unused_result__
2700 get_subexp (re_match_context_t
*mctx
, int bkref_node
, int bkref_str_idx
)
2702 const re_dfa_t
*const dfa
= mctx
->dfa
;
2703 int subexp_num
, sub_top_idx
;
2704 const char *buf
= (const char *) re_string_get_buffer (&mctx
->input
);
2705 /* Return if we have already checked BKREF_NODE at BKREF_STR_IDX. */
2706 int cache_idx
= search_cur_bkref_entry (mctx
, bkref_str_idx
);
2707 if (cache_idx
!= -1)
2709 const struct re_backref_cache_entry
*entry
2710 = mctx
->bkref_ents
+ cache_idx
;
2712 if (entry
->node
== bkref_node
)
2713 return REG_NOERROR
; /* We already checked it. */
2714 while (entry
++->more
);
2717 subexp_num
= dfa
->nodes
[bkref_node
].opr
.idx
;
2719 /* For each sub expression */
2720 for (sub_top_idx
= 0; sub_top_idx
< mctx
->nsub_tops
; ++sub_top_idx
)
2723 re_sub_match_top_t
*sub_top
= mctx
->sub_tops
[sub_top_idx
];
2724 re_sub_match_last_t
*sub_last
;
2725 int sub_last_idx
, sl_str
, bkref_str_off
;
2727 if (dfa
->nodes
[sub_top
->node
].opr
.idx
!= subexp_num
)
2728 continue; /* It isn't related. */
2730 sl_str
= sub_top
->str_idx
;
2731 bkref_str_off
= bkref_str_idx
;
2732 /* At first, check the last node of sub expressions we already
2734 for (sub_last_idx
= 0; sub_last_idx
< sub_top
->nlasts
; ++sub_last_idx
)
2737 sub_last
= sub_top
->lasts
[sub_last_idx
];
2738 sl_str_diff
= sub_last
->str_idx
- sl_str
;
2739 /* The matched string by the sub expression match with the substring
2740 at the back reference? */
2741 if (sl_str_diff
> 0)
2743 if (BE (bkref_str_off
+ sl_str_diff
> mctx
->input
.valid_len
, 0))
2745 /* Not enough chars for a successful match. */
2746 if (bkref_str_off
+ sl_str_diff
> mctx
->input
.len
)
2749 err
= clean_state_log_if_needed (mctx
,
2752 if (BE (err
!= REG_NOERROR
, 0))
2754 buf
= (const char *) re_string_get_buffer (&mctx
->input
);
2756 if (memcmp (buf
+ bkref_str_off
, buf
+ sl_str
, sl_str_diff
) != 0)
2757 /* We don't need to search this sub expression any more. */
2760 bkref_str_off
+= sl_str_diff
;
2761 sl_str
+= sl_str_diff
;
2762 err
= get_subexp_sub (mctx
, sub_top
, sub_last
, bkref_node
,
2765 /* Reload buf, since the preceding call might have reallocated
2767 buf
= (const char *) re_string_get_buffer (&mctx
->input
);
2769 if (err
== REG_NOMATCH
)
2771 if (BE (err
!= REG_NOERROR
, 0))
2775 if (sub_last_idx
< sub_top
->nlasts
)
2777 if (sub_last_idx
> 0)
2779 /* Then, search for the other last nodes of the sub expression. */
2780 for (; sl_str
<= bkref_str_idx
; ++sl_str
)
2782 int cls_node
, sl_str_off
;
2783 const re_node_set
*nodes
;
2784 sl_str_off
= sl_str
- sub_top
->str_idx
;
2785 /* The matched string by the sub expression match with the substring
2786 at the back reference? */
2789 if (BE (bkref_str_off
>= mctx
->input
.valid_len
, 0))
2791 /* If we are at the end of the input, we cannot match. */
2792 if (bkref_str_off
>= mctx
->input
.len
)
2795 err
= extend_buffers (mctx
);
2796 if (BE (err
!= REG_NOERROR
, 0))
2799 buf
= (const char *) re_string_get_buffer (&mctx
->input
);
2801 if (buf
[bkref_str_off
++] != buf
[sl_str
- 1])
2802 break; /* We don't need to search this sub expression
2805 if (mctx
->state_log
[sl_str
] == NULL
)
2807 /* Does this state have a ')' of the sub expression? */
2808 nodes
= &mctx
->state_log
[sl_str
]->nodes
;
2809 cls_node
= find_subexp_node (dfa
, nodes
, subexp_num
,
2813 if (sub_top
->path
== NULL
)
2815 sub_top
->path
= calloc (sizeof (state_array_t
),
2816 sl_str
- sub_top
->str_idx
+ 1);
2817 if (sub_top
->path
== NULL
)
2820 /* Can the OP_OPEN_SUBEXP node arrive the OP_CLOSE_SUBEXP node
2821 in the current context? */
2822 err
= check_arrival (mctx
, sub_top
->path
, sub_top
->node
,
2823 sub_top
->str_idx
, cls_node
, sl_str
,
2825 if (err
== REG_NOMATCH
)
2827 if (BE (err
!= REG_NOERROR
, 0))
2829 sub_last
= match_ctx_add_sublast (sub_top
, cls_node
, sl_str
);
2830 if (BE (sub_last
== NULL
, 0))
2832 err
= get_subexp_sub (mctx
, sub_top
, sub_last
, bkref_node
,
2834 if (err
== REG_NOMATCH
)
2841 /* Helper functions for get_subexp(). */
2843 /* Check SUB_LAST can arrive to the back reference BKREF_NODE at BKREF_STR.
2844 If it can arrive, register the sub expression expressed with SUB_TOP
2847 static reg_errcode_t
2849 get_subexp_sub (re_match_context_t
*mctx
, const re_sub_match_top_t
*sub_top
,
2850 re_sub_match_last_t
*sub_last
, int bkref_node
, int bkref_str
)
2854 /* Can the subexpression arrive the back reference? */
2855 err
= check_arrival (mctx
, &sub_last
->path
, sub_last
->node
,
2856 sub_last
->str_idx
, bkref_node
, bkref_str
,
2858 if (err
!= REG_NOERROR
)
2860 err
= match_ctx_add_entry (mctx
, bkref_node
, bkref_str
, sub_top
->str_idx
,
2862 if (BE (err
!= REG_NOERROR
, 0))
2864 to_idx
= bkref_str
+ sub_last
->str_idx
- sub_top
->str_idx
;
2865 return clean_state_log_if_needed (mctx
, to_idx
);
2868 /* Find the first node which is '(' or ')' and whose index is SUBEXP_IDX.
2869 Search '(' if FL_OPEN, or search ')' otherwise.
2870 TODO: This function isn't efficient...
2871 Because there might be more than one nodes whose types are
2872 OP_OPEN_SUBEXP and whose index is SUBEXP_IDX, we must check all
2878 find_subexp_node (const re_dfa_t
*dfa
, const re_node_set
*nodes
,
2879 int subexp_idx
, int type
)
2882 for (cls_idx
= 0; cls_idx
< nodes
->nelem
; ++cls_idx
)
2884 int cls_node
= nodes
->elems
[cls_idx
];
2885 const re_token_t
*node
= dfa
->nodes
+ cls_node
;
2886 if (node
->type
== type
2887 && node
->opr
.idx
== subexp_idx
)
2893 /* Check whether the node TOP_NODE at TOP_STR can arrive to the node
2894 LAST_NODE at LAST_STR. We record the path onto PATH since it will be
2896 Return REG_NOERROR if it can arrive, or REG_NOMATCH otherwise. */
2898 static reg_errcode_t
2899 internal_function __attribute_warn_unused_result__
2900 check_arrival (re_match_context_t
*mctx
, state_array_t
*path
, int top_node
,
2901 int top_str
, int last_node
, int last_str
, int type
)
2903 const re_dfa_t
*const dfa
= mctx
->dfa
;
2904 reg_errcode_t err
= REG_NOERROR
;
2905 int subexp_num
, backup_cur_idx
, str_idx
, null_cnt
;
2906 re_dfastate_t
*cur_state
= NULL
;
2907 re_node_set
*cur_nodes
, next_nodes
;
2908 re_dfastate_t
**backup_state_log
;
2909 unsigned int context
;
2911 subexp_num
= dfa
->nodes
[top_node
].opr
.idx
;
2912 /* Extend the buffer if we need. */
2913 if (BE (path
->alloc
< last_str
+ mctx
->max_mb_elem_len
+ 1, 0))
2915 re_dfastate_t
**new_array
;
2916 int old_alloc
= path
->alloc
;
2917 path
->alloc
+= last_str
+ mctx
->max_mb_elem_len
+ 1;
2918 new_array
= re_realloc (path
->array
, re_dfastate_t
*, path
->alloc
);
2919 if (BE (new_array
== NULL
, 0))
2921 path
->alloc
= old_alloc
;
2924 path
->array
= new_array
;
2925 memset (new_array
+ old_alloc
, '\0',
2926 sizeof (re_dfastate_t
*) * (path
->alloc
- old_alloc
));
2929 str_idx
= path
->next_idx
?: top_str
;
2931 /* Temporary modify MCTX. */
2932 backup_state_log
= mctx
->state_log
;
2933 backup_cur_idx
= mctx
->input
.cur_idx
;
2934 mctx
->state_log
= path
->array
;
2935 mctx
->input
.cur_idx
= str_idx
;
2937 /* Setup initial node set. */
2938 context
= re_string_context_at (&mctx
->input
, str_idx
- 1, mctx
->eflags
);
2939 if (str_idx
== top_str
)
2941 err
= re_node_set_init_1 (&next_nodes
, top_node
);
2942 if (BE (err
!= REG_NOERROR
, 0))
2944 err
= check_arrival_expand_ecl (dfa
, &next_nodes
, subexp_num
, type
);
2945 if (BE (err
!= REG_NOERROR
, 0))
2947 re_node_set_free (&next_nodes
);
2953 cur_state
= mctx
->state_log
[str_idx
];
2954 if (cur_state
&& cur_state
->has_backref
)
2956 err
= re_node_set_init_copy (&next_nodes
, &cur_state
->nodes
);
2957 if (BE (err
!= REG_NOERROR
, 0))
2961 re_node_set_init_empty (&next_nodes
);
2963 if (str_idx
== top_str
|| (cur_state
&& cur_state
->has_backref
))
2965 if (next_nodes
.nelem
)
2967 err
= expand_bkref_cache (mctx
, &next_nodes
, str_idx
,
2969 if (BE (err
!= REG_NOERROR
, 0))
2971 re_node_set_free (&next_nodes
);
2975 cur_state
= re_acquire_state_context (&err
, dfa
, &next_nodes
, context
);
2976 if (BE (cur_state
== NULL
&& err
!= REG_NOERROR
, 0))
2978 re_node_set_free (&next_nodes
);
2981 mctx
->state_log
[str_idx
] = cur_state
;
2984 for (null_cnt
= 0; str_idx
< last_str
&& null_cnt
<= mctx
->max_mb_elem_len
;)
2986 re_node_set_empty (&next_nodes
);
2987 if (mctx
->state_log
[str_idx
+ 1])
2989 err
= re_node_set_merge (&next_nodes
,
2990 &mctx
->state_log
[str_idx
+ 1]->nodes
);
2991 if (BE (err
!= REG_NOERROR
, 0))
2993 re_node_set_free (&next_nodes
);
2999 err
= check_arrival_add_next_nodes (mctx
, str_idx
,
3000 &cur_state
->non_eps_nodes
,
3002 if (BE (err
!= REG_NOERROR
, 0))
3004 re_node_set_free (&next_nodes
);
3009 if (next_nodes
.nelem
)
3011 err
= check_arrival_expand_ecl (dfa
, &next_nodes
, subexp_num
, type
);
3012 if (BE (err
!= REG_NOERROR
, 0))
3014 re_node_set_free (&next_nodes
);
3017 err
= expand_bkref_cache (mctx
, &next_nodes
, str_idx
,
3019 if (BE (err
!= REG_NOERROR
, 0))
3021 re_node_set_free (&next_nodes
);
3025 context
= re_string_context_at (&mctx
->input
, str_idx
- 1, mctx
->eflags
);
3026 cur_state
= re_acquire_state_context (&err
, dfa
, &next_nodes
, context
);
3027 if (BE (cur_state
== NULL
&& err
!= REG_NOERROR
, 0))
3029 re_node_set_free (&next_nodes
);
3032 mctx
->state_log
[str_idx
] = cur_state
;
3033 null_cnt
= cur_state
== NULL
? null_cnt
+ 1 : 0;
3035 re_node_set_free (&next_nodes
);
3036 cur_nodes
= (mctx
->state_log
[last_str
] == NULL
? NULL
3037 : &mctx
->state_log
[last_str
]->nodes
);
3038 path
->next_idx
= str_idx
;
3041 mctx
->state_log
= backup_state_log
;
3042 mctx
->input
.cur_idx
= backup_cur_idx
;
3044 /* Then check the current node set has the node LAST_NODE. */
3045 if (cur_nodes
!= NULL
&& re_node_set_contains (cur_nodes
, last_node
))
3051 /* Helper functions for check_arrival. */
3053 /* Calculate the destination nodes of CUR_NODES at STR_IDX, and append them
3055 TODO: This function is similar to the functions transit_state*(),
3056 however this function has many additional works.
3057 Can't we unify them? */
3059 static reg_errcode_t
3060 internal_function __attribute_warn_unused_result__
3061 check_arrival_add_next_nodes (re_match_context_t
*mctx
, int str_idx
,
3062 re_node_set
*cur_nodes
, re_node_set
*next_nodes
)
3064 const re_dfa_t
*const dfa
= mctx
->dfa
;
3067 reg_errcode_t err
= REG_NOERROR
;
3068 re_node_set union_set
;
3069 re_node_set_init_empty (&union_set
);
3070 for (cur_idx
= 0; cur_idx
< cur_nodes
->nelem
; ++cur_idx
)
3073 int cur_node
= cur_nodes
->elems
[cur_idx
];
3075 re_token_type_t type
= dfa
->nodes
[cur_node
].type
;
3076 assert (!IS_EPSILON_NODE (type
));
3078 #ifdef RE_ENABLE_I18N
3079 /* If the node may accept `multi byte'. */
3080 if (dfa
->nodes
[cur_node
].accept_mb
)
3082 naccepted
= check_node_accept_bytes (dfa
, cur_node
, &mctx
->input
,
3086 re_dfastate_t
*dest_state
;
3087 int next_node
= dfa
->nexts
[cur_node
];
3088 int next_idx
= str_idx
+ naccepted
;
3089 dest_state
= mctx
->state_log
[next_idx
];
3090 re_node_set_empty (&union_set
);
3093 err
= re_node_set_merge (&union_set
, &dest_state
->nodes
);
3094 if (BE (err
!= REG_NOERROR
, 0))
3096 re_node_set_free (&union_set
);
3100 result
= re_node_set_insert (&union_set
, next_node
);
3101 if (BE (result
< 0, 0))
3103 re_node_set_free (&union_set
);
3106 mctx
->state_log
[next_idx
] = re_acquire_state (&err
, dfa
,
3108 if (BE (mctx
->state_log
[next_idx
] == NULL
3109 && err
!= REG_NOERROR
, 0))
3111 re_node_set_free (&union_set
);
3116 #endif /* RE_ENABLE_I18N */
3118 || check_node_accept (mctx
, dfa
->nodes
+ cur_node
, str_idx
))
3120 result
= re_node_set_insert (next_nodes
, dfa
->nexts
[cur_node
]);
3121 if (BE (result
< 0, 0))
3123 re_node_set_free (&union_set
);
3128 re_node_set_free (&union_set
);
3132 /* For all the nodes in CUR_NODES, add the epsilon closures of them to
3133 CUR_NODES, however exclude the nodes which are:
3134 - inside the sub expression whose number is EX_SUBEXP, if FL_OPEN.
3135 - out of the sub expression whose number is EX_SUBEXP, if !FL_OPEN.
3138 static reg_errcode_t
3140 check_arrival_expand_ecl (const re_dfa_t
*dfa
, re_node_set
*cur_nodes
,
3141 int ex_subexp
, int type
)
3144 int idx
, outside_node
;
3145 re_node_set new_nodes
;
3147 assert (cur_nodes
->nelem
);
3149 err
= re_node_set_alloc (&new_nodes
, cur_nodes
->nelem
);
3150 if (BE (err
!= REG_NOERROR
, 0))
3152 /* Create a new node set NEW_NODES with the nodes which are epsilon
3153 closures of the node in CUR_NODES. */
3155 for (idx
= 0; idx
< cur_nodes
->nelem
; ++idx
)
3157 int cur_node
= cur_nodes
->elems
[idx
];
3158 const re_node_set
*eclosure
= dfa
->eclosures
+ cur_node
;
3159 outside_node
= find_subexp_node (dfa
, eclosure
, ex_subexp
, type
);
3160 if (outside_node
== -1)
3162 /* There are no problematic nodes, just merge them. */
3163 err
= re_node_set_merge (&new_nodes
, eclosure
);
3164 if (BE (err
!= REG_NOERROR
, 0))
3166 re_node_set_free (&new_nodes
);
3172 /* There are problematic nodes, re-calculate incrementally. */
3173 err
= check_arrival_expand_ecl_sub (dfa
, &new_nodes
, cur_node
,
3175 if (BE (err
!= REG_NOERROR
, 0))
3177 re_node_set_free (&new_nodes
);
3182 re_node_set_free (cur_nodes
);
3183 *cur_nodes
= new_nodes
;
3187 /* Helper function for check_arrival_expand_ecl.
3188 Check incrementally the epsilon closure of TARGET, and if it isn't
3189 problematic append it to DST_NODES. */
3191 static reg_errcode_t
3192 internal_function __attribute_warn_unused_result__
3193 check_arrival_expand_ecl_sub (const re_dfa_t
*dfa
, re_node_set
*dst_nodes
,
3194 int target
, int ex_subexp
, int type
)
3197 for (cur_node
= target
; !re_node_set_contains (dst_nodes
, cur_node
);)
3201 if (dfa
->nodes
[cur_node
].type
== type
3202 && dfa
->nodes
[cur_node
].opr
.idx
== ex_subexp
)
3204 if (type
== OP_CLOSE_SUBEXP
)
3206 err
= re_node_set_insert (dst_nodes
, cur_node
);
3207 if (BE (err
== -1, 0))
3212 err
= re_node_set_insert (dst_nodes
, cur_node
);
3213 if (BE (err
== -1, 0))
3215 if (dfa
->edests
[cur_node
].nelem
== 0)
3217 if (dfa
->edests
[cur_node
].nelem
== 2)
3219 err
= check_arrival_expand_ecl_sub (dfa
, dst_nodes
,
3220 dfa
->edests
[cur_node
].elems
[1],
3222 if (BE (err
!= REG_NOERROR
, 0))
3225 cur_node
= dfa
->edests
[cur_node
].elems
[0];
3231 /* For all the back references in the current state, calculate the
3232 destination of the back references by the appropriate entry
3233 in MCTX->BKREF_ENTS. */
3235 static reg_errcode_t
3236 internal_function __attribute_warn_unused_result__
3237 expand_bkref_cache (re_match_context_t
*mctx
, re_node_set
*cur_nodes
,
3238 int cur_str
, int subexp_num
, int type
)
3240 const re_dfa_t
*const dfa
= mctx
->dfa
;
3242 int cache_idx_start
= search_cur_bkref_entry (mctx
, cur_str
);
3243 struct re_backref_cache_entry
*ent
;
3245 if (cache_idx_start
== -1)
3249 ent
= mctx
->bkref_ents
+ cache_idx_start
;
3252 int to_idx
, next_node
;
3254 /* Is this entry ENT is appropriate? */
3255 if (!re_node_set_contains (cur_nodes
, ent
->node
))
3258 to_idx
= cur_str
+ ent
->subexp_to
- ent
->subexp_from
;
3259 /* Calculate the destination of the back reference, and append it
3260 to MCTX->STATE_LOG. */
3261 if (to_idx
== cur_str
)
3263 /* The backreference did epsilon transit, we must re-check all the
3264 node in the current state. */
3265 re_node_set new_dests
;
3266 reg_errcode_t err2
, err3
;
3267 next_node
= dfa
->edests
[ent
->node
].elems
[0];
3268 if (re_node_set_contains (cur_nodes
, next_node
))
3270 err
= re_node_set_init_1 (&new_dests
, next_node
);
3271 err2
= check_arrival_expand_ecl (dfa
, &new_dests
, subexp_num
, type
);
3272 err3
= re_node_set_merge (cur_nodes
, &new_dests
);
3273 re_node_set_free (&new_dests
);
3274 if (BE (err
!= REG_NOERROR
|| err2
!= REG_NOERROR
3275 || err3
!= REG_NOERROR
, 0))
3277 err
= (err
!= REG_NOERROR
? err
3278 : (err2
!= REG_NOERROR
? err2
: err3
));
3281 /* TODO: It is still inefficient... */
3286 re_node_set union_set
;
3287 next_node
= dfa
->nexts
[ent
->node
];
3288 if (mctx
->state_log
[to_idx
])
3291 if (re_node_set_contains (&mctx
->state_log
[to_idx
]->nodes
,
3294 err
= re_node_set_init_copy (&union_set
,
3295 &mctx
->state_log
[to_idx
]->nodes
);
3296 ret
= re_node_set_insert (&union_set
, next_node
);
3297 if (BE (err
!= REG_NOERROR
|| ret
< 0, 0))
3299 re_node_set_free (&union_set
);
3300 err
= err
!= REG_NOERROR
? err
: REG_ESPACE
;
3306 err
= re_node_set_init_1 (&union_set
, next_node
);
3307 if (BE (err
!= REG_NOERROR
, 0))
3310 mctx
->state_log
[to_idx
] = re_acquire_state (&err
, dfa
, &union_set
);
3311 re_node_set_free (&union_set
);
3312 if (BE (mctx
->state_log
[to_idx
] == NULL
3313 && err
!= REG_NOERROR
, 0))
3317 while (ent
++->more
);
3321 /* Build transition table for the state.
3322 Return 1 if succeeded, otherwise return NULL. */
3326 build_trtable (const re_dfa_t
*dfa
, re_dfastate_t
*state
)
3329 int i
, j
, ch
, need_word_trtable
= 0;
3330 bitset_word_t elem
, mask
;
3331 bool dests_node_malloced
= false;
3332 bool dest_states_malloced
= false;
3333 int ndests
; /* Number of the destination states from `state'. */
3334 re_dfastate_t
**trtable
;
3335 re_dfastate_t
**dest_states
= NULL
, **dest_states_word
, **dest_states_nl
;
3336 re_node_set follows
, *dests_node
;
3338 bitset_t acceptable
;
3342 re_node_set dests_node
[SBC_MAX
];
3343 bitset_t dests_ch
[SBC_MAX
];
3346 /* We build DFA states which corresponds to the destination nodes
3347 from `state'. `dests_node[i]' represents the nodes which i-th
3348 destination state contains, and `dests_ch[i]' represents the
3349 characters which i-th destination state accepts. */
3350 if (__libc_use_alloca (sizeof (struct dests_alloc
)))
3351 dests_alloc
= (struct dests_alloc
*) alloca (sizeof (struct dests_alloc
));
3354 dests_alloc
= re_malloc (struct dests_alloc
, 1);
3355 if (BE (dests_alloc
== NULL
, 0))
3357 dests_node_malloced
= true;
3359 dests_node
= dests_alloc
->dests_node
;
3360 dests_ch
= dests_alloc
->dests_ch
;
3362 /* Initialize transiton table. */
3363 state
->word_trtable
= state
->trtable
= NULL
;
3365 /* At first, group all nodes belonging to `state' into several
3367 ndests
= group_nodes_into_DFAstates (dfa
, state
, dests_node
, dests_ch
);
3368 if (BE (ndests
<= 0, 0))
3370 if (dests_node_malloced
)
3372 /* Return 0 in case of an error, 1 otherwise. */
3375 state
->trtable
= (re_dfastate_t
**)
3376 calloc (sizeof (re_dfastate_t
*), SBC_MAX
);
3377 if (BE (state
->trtable
== NULL
, 0))
3384 err
= re_node_set_alloc (&follows
, ndests
+ 1);
3385 if (BE (err
!= REG_NOERROR
, 0))
3388 /* Avoid arithmetic overflow in size calculation. */
3389 if (BE ((((SIZE_MAX
- (sizeof (re_node_set
) + sizeof (bitset_t
)) * SBC_MAX
)
3390 / (3 * sizeof (re_dfastate_t
*)))
3395 if (__libc_use_alloca ((sizeof (re_node_set
) + sizeof (bitset_t
)) * SBC_MAX
3396 + ndests
* 3 * sizeof (re_dfastate_t
*)))
3397 dest_states
= (re_dfastate_t
**)
3398 alloca (ndests
* 3 * sizeof (re_dfastate_t
*));
3401 dest_states
= (re_dfastate_t
**)
3402 malloc (ndests
* 3 * sizeof (re_dfastate_t
*));
3403 if (BE (dest_states
== NULL
, 0))
3406 if (dest_states_malloced
)
3408 re_node_set_free (&follows
);
3409 for (i
= 0; i
< ndests
; ++i
)
3410 re_node_set_free (dests_node
+ i
);
3411 if (dests_node_malloced
)
3415 dest_states_malloced
= true;
3417 dest_states_word
= dest_states
+ ndests
;
3418 dest_states_nl
= dest_states_word
+ ndests
;
3419 bitset_empty (acceptable
);
3421 /* Then build the states for all destinations. */
3422 for (i
= 0; i
< ndests
; ++i
)
3425 re_node_set_empty (&follows
);
3426 /* Merge the follows of this destination states. */
3427 for (j
= 0; j
< dests_node
[i
].nelem
; ++j
)
3429 next_node
= dfa
->nexts
[dests_node
[i
].elems
[j
]];
3430 if (next_node
!= -1)
3432 err
= re_node_set_merge (&follows
, dfa
->eclosures
+ next_node
);
3433 if (BE (err
!= REG_NOERROR
, 0))
3437 dest_states
[i
] = re_acquire_state_context (&err
, dfa
, &follows
, 0);
3438 if (BE (dest_states
[i
] == NULL
&& err
!= REG_NOERROR
, 0))
3440 /* If the new state has context constraint,
3441 build appropriate states for these contexts. */
3442 if (dest_states
[i
]->has_constraint
)
3444 dest_states_word
[i
] = re_acquire_state_context (&err
, dfa
, &follows
,
3446 if (BE (dest_states_word
[i
] == NULL
&& err
!= REG_NOERROR
, 0))
3449 if (dest_states
[i
] != dest_states_word
[i
] && dfa
->mb_cur_max
> 1)
3450 need_word_trtable
= 1;
3452 dest_states_nl
[i
] = re_acquire_state_context (&err
, dfa
, &follows
,
3454 if (BE (dest_states_nl
[i
] == NULL
&& err
!= REG_NOERROR
, 0))
3459 dest_states_word
[i
] = dest_states
[i
];
3460 dest_states_nl
[i
] = dest_states
[i
];
3462 bitset_merge (acceptable
, dests_ch
[i
]);
3465 if (!BE (need_word_trtable
, 0))
3467 /* We don't care about whether the following character is a word
3468 character, or we are in a single-byte character set so we can
3469 discern by looking at the character code: allocate a
3470 256-entry transition table. */
3471 trtable
= state
->trtable
=
3472 (re_dfastate_t
**) calloc (sizeof (re_dfastate_t
*), SBC_MAX
);
3473 if (BE (trtable
== NULL
, 0))
3476 /* For all characters ch...: */
3477 for (i
= 0; i
< BITSET_WORDS
; ++i
)
3478 for (ch
= i
* BITSET_WORD_BITS
, elem
= acceptable
[i
], mask
= 1;
3480 mask
<<= 1, elem
>>= 1, ++ch
)
3481 if (BE (elem
& 1, 0))
3483 /* There must be exactly one destination which accepts
3484 character ch. See group_nodes_into_DFAstates. */
3485 for (j
= 0; (dests_ch
[j
][i
] & mask
) == 0; ++j
)
3488 /* j-th destination accepts the word character ch. */
3489 if (dfa
->word_char
[i
] & mask
)
3490 trtable
[ch
] = dest_states_word
[j
];
3492 trtable
[ch
] = dest_states
[j
];
3497 /* We care about whether the following character is a word
3498 character, and we are in a multi-byte character set: discern
3499 by looking at the character code: build two 256-entry
3500 transition tables, one starting at trtable[0] and one
3501 starting at trtable[SBC_MAX]. */
3502 trtable
= state
->word_trtable
=
3503 (re_dfastate_t
**) calloc (sizeof (re_dfastate_t
*), 2 * SBC_MAX
);
3504 if (BE (trtable
== NULL
, 0))
3507 /* For all characters ch...: */
3508 for (i
= 0; i
< BITSET_WORDS
; ++i
)
3509 for (ch
= i
* BITSET_WORD_BITS
, elem
= acceptable
[i
], mask
= 1;
3511 mask
<<= 1, elem
>>= 1, ++ch
)
3512 if (BE (elem
& 1, 0))
3514 /* There must be exactly one destination which accepts
3515 character ch. See group_nodes_into_DFAstates. */
3516 for (j
= 0; (dests_ch
[j
][i
] & mask
) == 0; ++j
)
3519 /* j-th destination accepts the word character ch. */
3520 trtable
[ch
] = dest_states
[j
];
3521 trtable
[ch
+ SBC_MAX
] = dest_states_word
[j
];
3526 if (bitset_contain (acceptable
, NEWLINE_CHAR
))
3528 /* The current state accepts newline character. */
3529 for (j
= 0; j
< ndests
; ++j
)
3530 if (bitset_contain (dests_ch
[j
], NEWLINE_CHAR
))
3532 /* k-th destination accepts newline character. */
3533 trtable
[NEWLINE_CHAR
] = dest_states_nl
[j
];
3534 if (need_word_trtable
)
3535 trtable
[NEWLINE_CHAR
+ SBC_MAX
] = dest_states_nl
[j
];
3536 /* There must be only one destination which accepts
3537 newline. See group_nodes_into_DFAstates. */
3542 if (dest_states_malloced
)
3545 re_node_set_free (&follows
);
3546 for (i
= 0; i
< ndests
; ++i
)
3547 re_node_set_free (dests_node
+ i
);
3549 if (dests_node_malloced
)
3555 /* Group all nodes belonging to STATE into several destinations.
3556 Then for all destinations, set the nodes belonging to the destination
3557 to DESTS_NODE[i] and set the characters accepted by the destination
3558 to DEST_CH[i]. This function return the number of destinations. */
3562 group_nodes_into_DFAstates (const re_dfa_t
*dfa
, const re_dfastate_t
*state
,
3563 re_node_set
*dests_node
, bitset_t
*dests_ch
)
3568 int ndests
; /* Number of the destinations from `state'. */
3569 bitset_t accepts
; /* Characters a node can accept. */
3570 const re_node_set
*cur_nodes
= &state
->nodes
;
3571 bitset_empty (accepts
);
3574 /* For all the nodes belonging to `state', */
3575 for (i
= 0; i
< cur_nodes
->nelem
; ++i
)
3577 re_token_t
*node
= &dfa
->nodes
[cur_nodes
->elems
[i
]];
3578 re_token_type_t type
= node
->type
;
3579 unsigned int constraint
= node
->constraint
;
3581 /* Enumerate all single byte character this node can accept. */
3582 if (type
== CHARACTER
)
3583 bitset_set (accepts
, node
->opr
.c
);
3584 else if (type
== SIMPLE_BRACKET
)
3586 bitset_merge (accepts
, node
->opr
.sbcset
);
3588 else if (type
== OP_PERIOD
)
3590 #ifdef RE_ENABLE_I18N
3591 if (dfa
->mb_cur_max
> 1)
3592 bitset_merge (accepts
, dfa
->sb_char
);
3595 bitset_set_all (accepts
);
3596 if (!(dfa
->syntax
& RE_DOT_NEWLINE
))
3597 bitset_clear (accepts
, '\n');
3598 if (dfa
->syntax
& RE_DOT_NOT_NULL
)
3599 bitset_clear (accepts
, '\0');
3601 #ifdef RE_ENABLE_I18N
3602 else if (type
== OP_UTF8_PERIOD
)
3604 memset (accepts
, '\xff', sizeof (bitset_t
) / 2);
3605 if (!(dfa
->syntax
& RE_DOT_NEWLINE
))
3606 bitset_clear (accepts
, '\n');
3607 if (dfa
->syntax
& RE_DOT_NOT_NULL
)
3608 bitset_clear (accepts
, '\0');
3614 /* Check the `accepts' and sift the characters which are not
3615 match it the context. */
3618 if (constraint
& NEXT_NEWLINE_CONSTRAINT
)
3620 bool accepts_newline
= bitset_contain (accepts
, NEWLINE_CHAR
);
3621 bitset_empty (accepts
);
3622 if (accepts_newline
)
3623 bitset_set (accepts
, NEWLINE_CHAR
);
3627 if (constraint
& NEXT_ENDBUF_CONSTRAINT
)
3629 bitset_empty (accepts
);
3633 if (constraint
& NEXT_WORD_CONSTRAINT
)
3635 bitset_word_t any_set
= 0;
3636 if (type
== CHARACTER
&& !node
->word_char
)
3638 bitset_empty (accepts
);
3641 #ifdef RE_ENABLE_I18N
3642 if (dfa
->mb_cur_max
> 1)
3643 for (j
= 0; j
< BITSET_WORDS
; ++j
)
3644 any_set
|= (accepts
[j
] &= (dfa
->word_char
[j
] | ~dfa
->sb_char
[j
]));
3647 for (j
= 0; j
< BITSET_WORDS
; ++j
)
3648 any_set
|= (accepts
[j
] &= dfa
->word_char
[j
]);
3652 if (constraint
& NEXT_NOTWORD_CONSTRAINT
)
3654 bitset_word_t any_set
= 0;
3655 if (type
== CHARACTER
&& node
->word_char
)
3657 bitset_empty (accepts
);
3660 #ifdef RE_ENABLE_I18N
3661 if (dfa
->mb_cur_max
> 1)
3662 for (j
= 0; j
< BITSET_WORDS
; ++j
)
3663 any_set
|= (accepts
[j
] &= ~(dfa
->word_char
[j
] & dfa
->sb_char
[j
]));
3666 for (j
= 0; j
< BITSET_WORDS
; ++j
)
3667 any_set
|= (accepts
[j
] &= ~dfa
->word_char
[j
]);
3673 /* Then divide `accepts' into DFA states, or create a new
3674 state. Above, we make sure that accepts is not empty. */
3675 for (j
= 0; j
< ndests
; ++j
)
3677 bitset_t intersec
; /* Intersection sets, see below. */
3679 /* Flags, see below. */
3680 bitset_word_t has_intersec
, not_subset
, not_consumed
;
3682 /* Optimization, skip if this state doesn't accept the character. */
3683 if (type
== CHARACTER
&& !bitset_contain (dests_ch
[j
], node
->opr
.c
))
3686 /* Enumerate the intersection set of this state and `accepts'. */
3688 for (k
= 0; k
< BITSET_WORDS
; ++k
)
3689 has_intersec
|= intersec
[k
] = accepts
[k
] & dests_ch
[j
][k
];
3690 /* And skip if the intersection set is empty. */
3694 /* Then check if this state is a subset of `accepts'. */
3695 not_subset
= not_consumed
= 0;
3696 for (k
= 0; k
< BITSET_WORDS
; ++k
)
3698 not_subset
|= remains
[k
] = ~accepts
[k
] & dests_ch
[j
][k
];
3699 not_consumed
|= accepts
[k
] = accepts
[k
] & ~dests_ch
[j
][k
];
3702 /* If this state isn't a subset of `accepts', create a
3703 new group state, which has the `remains'. */
3706 bitset_copy (dests_ch
[ndests
], remains
);
3707 bitset_copy (dests_ch
[j
], intersec
);
3708 err
= re_node_set_init_copy (dests_node
+ ndests
, &dests_node
[j
]);
3709 if (BE (err
!= REG_NOERROR
, 0))
3714 /* Put the position in the current group. */
3715 result
= re_node_set_insert (&dests_node
[j
], cur_nodes
->elems
[i
]);
3716 if (BE (result
< 0, 0))
3719 /* If all characters are consumed, go to next node. */
3723 /* Some characters remain, create a new group. */
3726 bitset_copy (dests_ch
[ndests
], accepts
);
3727 err
= re_node_set_init_1 (dests_node
+ ndests
, cur_nodes
->elems
[i
]);
3728 if (BE (err
!= REG_NOERROR
, 0))
3731 bitset_empty (accepts
);
3736 for (j
= 0; j
< ndests
; ++j
)
3737 re_node_set_free (dests_node
+ j
);
3741 #ifdef RE_ENABLE_I18N
3742 /* Check how many bytes the node `dfa->nodes[node_idx]' accepts.
3743 Return the number of the bytes the node accepts.
3744 STR_IDX is the current index of the input string.
3746 This function handles the nodes which can accept one character, or
3747 one collating element like '.', '[a-z]', opposite to the other nodes
3748 can only accept one byte. */
3752 check_node_accept_bytes (const re_dfa_t
*dfa
, int node_idx
,
3753 const re_string_t
*input
, int str_idx
)
3755 const re_token_t
*node
= dfa
->nodes
+ node_idx
;
3756 int char_len
, elem_len
;
3759 if (BE (node
->type
== OP_UTF8_PERIOD
, 0))
3761 unsigned char c
= re_string_byte_at (input
, str_idx
), d
;
3762 if (BE (c
< 0xc2, 1))
3765 if (str_idx
+ 2 > input
->len
)
3768 d
= re_string_byte_at (input
, str_idx
+ 1);
3770 return (d
< 0x80 || d
> 0xbf) ? 0 : 2;
3774 if (c
== 0xe0 && d
< 0xa0)
3780 if (c
== 0xf0 && d
< 0x90)
3786 if (c
== 0xf8 && d
< 0x88)
3792 if (c
== 0xfc && d
< 0x84)
3798 if (str_idx
+ char_len
> input
->len
)
3801 for (i
= 1; i
< char_len
; ++i
)
3803 d
= re_string_byte_at (input
, str_idx
+ i
);
3804 if (d
< 0x80 || d
> 0xbf)
3810 char_len
= re_string_char_size_at (input
, str_idx
);
3811 if (node
->type
== OP_PERIOD
)
3815 /* FIXME: I don't think this if is needed, as both '\n'
3816 and '\0' are char_len == 1. */
3817 /* '.' accepts any one character except the following two cases. */
3818 if ((!(dfa
->syntax
& RE_DOT_NEWLINE
) &&
3819 re_string_byte_at (input
, str_idx
) == '\n') ||
3820 ((dfa
->syntax
& RE_DOT_NOT_NULL
) &&
3821 re_string_byte_at (input
, str_idx
) == '\0'))
3826 elem_len
= re_string_elem_size_at (input
, str_idx
);
3827 if ((elem_len
<= 1 && char_len
<= 1) || char_len
== 0)
3830 if (node
->type
== COMPLEX_BRACKET
)
3832 const re_charset_t
*cset
= node
->opr
.mbcset
;
3834 const unsigned char *pin
3835 = ((const unsigned char *) re_string_get_buffer (input
) + str_idx
);
3840 wchar_t wc
= ((cset
->nranges
|| cset
->nchar_classes
|| cset
->nmbchars
)
3841 ? re_string_wchar_at (input
, str_idx
) : 0);
3843 /* match with multibyte character? */
3844 for (i
= 0; i
< cset
->nmbchars
; ++i
)
3845 if (wc
== cset
->mbchars
[i
])
3847 match_len
= char_len
;
3848 goto check_node_accept_bytes_match
;
3850 /* match with character_class? */
3851 for (i
= 0; i
< cset
->nchar_classes
; ++i
)
3853 wctype_t wt
= cset
->char_classes
[i
];
3854 if (__iswctype (wc
, wt
))
3856 match_len
= char_len
;
3857 goto check_node_accept_bytes_match
;
3862 nrules
= _NL_CURRENT_WORD (LC_COLLATE
, _NL_COLLATE_NRULES
);
3865 unsigned int in_collseq
= 0;
3866 const int32_t *table
, *indirect
;
3867 const unsigned char *weights
, *extra
;
3868 const char *collseqwc
;
3869 /* This #include defines a local function! */
3870 # include <locale/weight.h>
3872 /* match with collating_symbol? */
3873 if (cset
->ncoll_syms
)
3874 extra
= (const unsigned char *)
3875 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_SYMB_EXTRAMB
);
3876 for (i
= 0; i
< cset
->ncoll_syms
; ++i
)
3878 const unsigned char *coll_sym
= extra
+ cset
->coll_syms
[i
];
3879 /* Compare the length of input collating element and
3880 the length of current collating element. */
3881 if (*coll_sym
!= elem_len
)
3883 /* Compare each bytes. */
3884 for (j
= 0; j
< *coll_sym
; j
++)
3885 if (pin
[j
] != coll_sym
[1 + j
])
3889 /* Match if every bytes is equal. */
3891 goto check_node_accept_bytes_match
;
3897 if (elem_len
<= char_len
)
3899 collseqwc
= _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_COLLSEQWC
);
3900 in_collseq
= __collseq_table_lookup (collseqwc
, wc
);
3903 in_collseq
= find_collation_sequence_value (pin
, elem_len
);
3905 /* match with range expression? */
3906 for (i
= 0; i
< cset
->nranges
; ++i
)
3907 if (cset
->range_starts
[i
] <= in_collseq
3908 && in_collseq
<= cset
->range_ends
[i
])
3910 match_len
= elem_len
;
3911 goto check_node_accept_bytes_match
;
3914 /* match with equivalence_class? */
3915 if (cset
->nequiv_classes
)
3917 const unsigned char *cp
= pin
;
3918 table
= (const int32_t *)
3919 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_TABLEMB
);
3920 weights
= (const unsigned char *)
3921 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_WEIGHTMB
);
3922 extra
= (const unsigned char *)
3923 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_EXTRAMB
);
3924 indirect
= (const int32_t *)
3925 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_INDIRECTMB
);
3926 int32_t idx
= findidx (&cp
, elem_len
);
3928 for (i
= 0; i
< cset
->nequiv_classes
; ++i
)
3930 int32_t equiv_class_idx
= cset
->equiv_classes
[i
];
3931 size_t weight_len
= weights
[idx
& 0xffffff];
3932 if (weight_len
== weights
[equiv_class_idx
& 0xffffff]
3933 && (idx
>> 24) == (equiv_class_idx
>> 24))
3938 equiv_class_idx
&= 0xffffff;
3940 while (cnt
<= weight_len
3941 && (weights
[equiv_class_idx
+ 1 + cnt
]
3942 == weights
[idx
+ 1 + cnt
]))
3944 if (cnt
> weight_len
)
3946 match_len
= elem_len
;
3947 goto check_node_accept_bytes_match
;
3956 /* match with range expression? */
3958 wchar_t cmp_buf
[] = {L
'\0', L
'\0', wc
, L
'\0', L
'\0', L
'\0'};
3960 wchar_t cmp_buf
[] = {L
'\0', L
'\0', L
'\0', L
'\0', L
'\0', L
'\0'};
3963 for (i
= 0; i
< cset
->nranges
; ++i
)
3965 cmp_buf
[0] = cset
->range_starts
[i
];
3966 cmp_buf
[4] = cset
->range_ends
[i
];
3967 if (wcscoll (cmp_buf
, cmp_buf
+ 2) <= 0
3968 && wcscoll (cmp_buf
+ 2, cmp_buf
+ 4) <= 0)
3970 match_len
= char_len
;
3971 goto check_node_accept_bytes_match
;
3975 check_node_accept_bytes_match
:
3976 if (!cset
->non_match
)
3983 return (elem_len
> char_len
) ? elem_len
: char_len
;
3992 find_collation_sequence_value (const unsigned char *mbs
, size_t mbs_len
)
3994 uint32_t nrules
= _NL_CURRENT_WORD (LC_COLLATE
, _NL_COLLATE_NRULES
);
3999 /* No valid character. Match it as a single byte character. */
4000 const unsigned char *collseq
= (const unsigned char *)
4001 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_COLLSEQMB
);
4002 return collseq
[mbs
[0]];
4009 const unsigned char *extra
= (const unsigned char *)
4010 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_SYMB_EXTRAMB
);
4011 int32_t extrasize
= (const unsigned char *)
4012 _NL_CURRENT (LC_COLLATE
, _NL_COLLATE_SYMB_EXTRAMB
+ 1) - extra
;
4014 for (idx
= 0; idx
< extrasize
;)
4016 int mbs_cnt
, found
= 0;
4017 int32_t elem_mbs_len
;
4018 /* Skip the name of collating element name. */
4019 idx
= idx
+ extra
[idx
] + 1;
4020 elem_mbs_len
= extra
[idx
++];
4021 if (mbs_len
== elem_mbs_len
)
4023 for (mbs_cnt
= 0; mbs_cnt
< elem_mbs_len
; ++mbs_cnt
)
4024 if (extra
[idx
+ mbs_cnt
] != mbs
[mbs_cnt
])
4026 if (mbs_cnt
== elem_mbs_len
)
4027 /* Found the entry. */
4030 /* Skip the byte sequence of the collating element. */
4031 idx
+= elem_mbs_len
;
4032 /* Adjust for the alignment. */
4033 idx
= (idx
+ 3) & ~3;
4034 /* Skip the collation sequence value. */
4035 idx
+= sizeof (uint32_t);
4036 /* Skip the wide char sequence of the collating element. */
4037 idx
= idx
+ sizeof (uint32_t) * (*(int32_t *) (extra
+ idx
) + 1);
4038 /* If we found the entry, return the sequence value. */
4040 return *(uint32_t *) (extra
+ idx
);
4041 /* Skip the collation sequence value. */
4042 idx
+= sizeof (uint32_t);
4048 #endif /* RE_ENABLE_I18N */
4050 /* Check whether the node accepts the byte which is IDX-th
4051 byte of the INPUT. */
4055 check_node_accept (const re_match_context_t
*mctx
, const re_token_t
*node
,
4059 ch
= re_string_byte_at (&mctx
->input
, idx
);
4063 if (node
->opr
.c
!= ch
)
4067 case SIMPLE_BRACKET
:
4068 if (!bitset_contain (node
->opr
.sbcset
, ch
))
4072 #ifdef RE_ENABLE_I18N
4073 case OP_UTF8_PERIOD
:
4079 if ((ch
== '\n' && !(mctx
->dfa
->syntax
& RE_DOT_NEWLINE
))
4080 || (ch
== '\0' && (mctx
->dfa
->syntax
& RE_DOT_NOT_NULL
)))
4088 if (node
->constraint
)
4090 /* The node has constraints. Check whether the current context
4091 satisfies the constraints. */
4092 unsigned int context
= re_string_context_at (&mctx
->input
, idx
,
4094 if (NOT_SATISFY_NEXT_CONSTRAINT (node
->constraint
, context
))
4101 /* Extend the buffers, if the buffers have run out. */
4103 static reg_errcode_t
4104 internal_function __attribute_warn_unused_result__
4105 extend_buffers (re_match_context_t
*mctx
)
4108 re_string_t
*pstr
= &mctx
->input
;
4110 /* Avoid overflow. */
4111 if (BE (INT_MAX
/ 2 / sizeof (re_dfastate_t
*) <= pstr
->bufs_len
, 0))
4114 /* Double the lengthes of the buffers. */
4115 ret
= re_string_realloc_buffers (pstr
, MIN (pstr
->len
, pstr
->bufs_len
* 2));
4116 if (BE (ret
!= REG_NOERROR
, 0))
4119 if (mctx
->state_log
!= NULL
)
4121 /* And double the length of state_log. */
4122 /* XXX We have no indication of the size of this buffer. If this
4123 allocation fail we have no indication that the state_log array
4124 does not have the right size. */
4125 re_dfastate_t
**new_array
= re_realloc (mctx
->state_log
, re_dfastate_t
*,
4126 pstr
->bufs_len
+ 1);
4127 if (BE (new_array
== NULL
, 0))
4129 mctx
->state_log
= new_array
;
4132 /* Then reconstruct the buffers. */
4135 #ifdef RE_ENABLE_I18N
4136 if (pstr
->mb_cur_max
> 1)
4138 ret
= build_wcs_upper_buffer (pstr
);
4139 if (BE (ret
!= REG_NOERROR
, 0))
4143 #endif /* RE_ENABLE_I18N */
4144 build_upper_buffer (pstr
);
4148 #ifdef RE_ENABLE_I18N
4149 if (pstr
->mb_cur_max
> 1)
4150 build_wcs_buffer (pstr
);
4152 #endif /* RE_ENABLE_I18N */
4154 if (pstr
->trans
!= NULL
)
4155 re_string_translate_buffer (pstr
);
4162 /* Functions for matching context. */
4164 /* Initialize MCTX. */
4166 static reg_errcode_t
4167 internal_function __attribute_warn_unused_result__
4168 match_ctx_init (re_match_context_t
*mctx
, int eflags
, int n
)
4170 mctx
->eflags
= eflags
;
4171 mctx
->match_last
= -1;
4174 mctx
->bkref_ents
= re_malloc (struct re_backref_cache_entry
, n
);
4175 mctx
->sub_tops
= re_malloc (re_sub_match_top_t
*, n
);
4176 if (BE (mctx
->bkref_ents
== NULL
|| mctx
->sub_tops
== NULL
, 0))
4179 /* Already zero-ed by the caller.
4181 mctx->bkref_ents = NULL;
4182 mctx->nbkref_ents = 0;
4183 mctx->nsub_tops = 0; */
4184 mctx
->abkref_ents
= n
;
4185 mctx
->max_mb_elem_len
= 1;
4186 mctx
->asub_tops
= n
;
4190 /* Clean the entries which depend on the current input in MCTX.
4191 This function must be invoked when the matcher changes the start index
4192 of the input, or changes the input string. */
4196 match_ctx_clean (re_match_context_t
*mctx
)
4199 for (st_idx
= 0; st_idx
< mctx
->nsub_tops
; ++st_idx
)
4202 re_sub_match_top_t
*top
= mctx
->sub_tops
[st_idx
];
4203 for (sl_idx
= 0; sl_idx
< top
->nlasts
; ++sl_idx
)
4205 re_sub_match_last_t
*last
= top
->lasts
[sl_idx
];
4206 re_free (last
->path
.array
);
4209 re_free (top
->lasts
);
4212 re_free (top
->path
->array
);
4213 re_free (top
->path
);
4218 mctx
->nsub_tops
= 0;
4219 mctx
->nbkref_ents
= 0;
4222 /* Free all the memory associated with MCTX. */
4226 match_ctx_free (re_match_context_t
*mctx
)
4228 /* First, free all the memory associated with MCTX->SUB_TOPS. */
4229 match_ctx_clean (mctx
);
4230 re_free (mctx
->sub_tops
);
4231 re_free (mctx
->bkref_ents
);
4234 /* Add a new backreference entry to MCTX.
4235 Note that we assume that caller never call this function with duplicate
4236 entry, and call with STR_IDX which isn't smaller than any existing entry.
4239 static reg_errcode_t
4240 internal_function __attribute_warn_unused_result__
4241 match_ctx_add_entry (re_match_context_t
*mctx
, int node
, int str_idx
, int from
,
4244 if (mctx
->nbkref_ents
>= mctx
->abkref_ents
)
4246 struct re_backref_cache_entry
* new_entry
;
4247 new_entry
= re_realloc (mctx
->bkref_ents
, struct re_backref_cache_entry
,
4248 mctx
->abkref_ents
* 2);
4249 if (BE (new_entry
== NULL
, 0))
4251 re_free (mctx
->bkref_ents
);
4254 mctx
->bkref_ents
= new_entry
;
4255 memset (mctx
->bkref_ents
+ mctx
->nbkref_ents
, '\0',
4256 sizeof (struct re_backref_cache_entry
) * mctx
->abkref_ents
);
4257 mctx
->abkref_ents
*= 2;
4259 if (mctx
->nbkref_ents
> 0
4260 && mctx
->bkref_ents
[mctx
->nbkref_ents
- 1].str_idx
== str_idx
)
4261 mctx
->bkref_ents
[mctx
->nbkref_ents
- 1].more
= 1;
4263 mctx
->bkref_ents
[mctx
->nbkref_ents
].node
= node
;
4264 mctx
->bkref_ents
[mctx
->nbkref_ents
].str_idx
= str_idx
;
4265 mctx
->bkref_ents
[mctx
->nbkref_ents
].subexp_from
= from
;
4266 mctx
->bkref_ents
[mctx
->nbkref_ents
].subexp_to
= to
;
4268 /* This is a cache that saves negative results of check_dst_limits_calc_pos.
4269 If bit N is clear, means that this entry won't epsilon-transition to
4270 an OP_OPEN_SUBEXP or OP_CLOSE_SUBEXP for the N+1-th subexpression. If
4271 it is set, check_dst_limits_calc_pos_1 will recurse and try to find one
4274 A backreference does not epsilon-transition unless it is empty, so set
4275 to all zeros if FROM != TO. */
4276 mctx
->bkref_ents
[mctx
->nbkref_ents
].eps_reachable_subexps_map
4277 = (from
== to
? ~0 : 0);
4279 mctx
->bkref_ents
[mctx
->nbkref_ents
++].more
= 0;
4280 if (mctx
->max_mb_elem_len
< to
- from
)
4281 mctx
->max_mb_elem_len
= to
- from
;
4285 /* Search for the first entry which has the same str_idx, or -1 if none is
4286 found. Note that MCTX->BKREF_ENTS is already sorted by MCTX->STR_IDX. */
4290 search_cur_bkref_entry (const re_match_context_t
*mctx
, int str_idx
)
4292 int left
, right
, mid
, last
;
4293 last
= right
= mctx
->nbkref_ents
;
4294 for (left
= 0; left
< right
;)
4296 mid
= (left
+ right
) / 2;
4297 if (mctx
->bkref_ents
[mid
].str_idx
< str_idx
)
4302 if (left
< last
&& mctx
->bkref_ents
[left
].str_idx
== str_idx
)
4308 /* Register the node NODE, whose type is OP_OPEN_SUBEXP, and which matches
4311 static reg_errcode_t
4312 internal_function __attribute_warn_unused_result__
4313 match_ctx_add_subtop (re_match_context_t
*mctx
, int node
, int str_idx
)
4316 assert (mctx
->sub_tops
!= NULL
);
4317 assert (mctx
->asub_tops
> 0);
4319 if (BE (mctx
->nsub_tops
== mctx
->asub_tops
, 0))
4321 int new_asub_tops
= mctx
->asub_tops
* 2;
4322 re_sub_match_top_t
**new_array
= re_realloc (mctx
->sub_tops
,
4323 re_sub_match_top_t
*,
4325 if (BE (new_array
== NULL
, 0))
4327 mctx
->sub_tops
= new_array
;
4328 mctx
->asub_tops
= new_asub_tops
;
4330 mctx
->sub_tops
[mctx
->nsub_tops
] = calloc (1, sizeof (re_sub_match_top_t
));
4331 if (BE (mctx
->sub_tops
[mctx
->nsub_tops
] == NULL
, 0))
4333 mctx
->sub_tops
[mctx
->nsub_tops
]->node
= node
;
4334 mctx
->sub_tops
[mctx
->nsub_tops
++]->str_idx
= str_idx
;
4338 /* Register the node NODE, whose type is OP_CLOSE_SUBEXP, and which matches
4339 at STR_IDX, whose corresponding OP_OPEN_SUBEXP is SUB_TOP. */
4341 static re_sub_match_last_t
*
4343 match_ctx_add_sublast (re_sub_match_top_t
*subtop
, int node
, int str_idx
)
4345 re_sub_match_last_t
*new_entry
;
4346 if (BE (subtop
->nlasts
== subtop
->alasts
, 0))
4348 int new_alasts
= 2 * subtop
->alasts
+ 1;
4349 re_sub_match_last_t
**new_array
= re_realloc (subtop
->lasts
,
4350 re_sub_match_last_t
*,
4352 if (BE (new_array
== NULL
, 0))
4354 subtop
->lasts
= new_array
;
4355 subtop
->alasts
= new_alasts
;
4357 new_entry
= calloc (1, sizeof (re_sub_match_last_t
));
4358 if (BE (new_entry
!= NULL
, 1))
4360 subtop
->lasts
[subtop
->nlasts
] = new_entry
;
4361 new_entry
->node
= node
;
4362 new_entry
->str_idx
= str_idx
;
4370 sift_ctx_init (re_sift_context_t
*sctx
, re_dfastate_t
**sifted_sts
,
4371 re_dfastate_t
**limited_sts
, int last_node
, int last_str_idx
)
4373 sctx
->sifted_states
= sifted_sts
;
4374 sctx
->limited_states
= limited_sts
;
4375 sctx
->last_node
= last_node
;
4376 sctx
->last_str_idx
= last_str_idx
;
4377 re_node_set_init_empty (&sctx
->limits
);