5 * 'A fair jaw-cracker dwarf-language must be.' --Samwise Gamgee
7 * [p.285 of _The Lord of the Rings_, II/iii: "The Ring Goes South"]
10 /* This file contains functions for compiling a regular expression. See
11 * also regexec.c which funnily enough, contains functions for executing
12 * a regular expression.
14 * This file is also copied at build time to ext/re/re_comp.c, where
15 * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
16 * This causes the main functions to be compiled under new names and with
17 * debugging support added, which makes "use re 'debug'" work.
20 /* NOTE: this is derived from Henry Spencer's regexp code, and should not
21 * confused with the original package (see point 3 below). Thanks, Henry!
24 /* Additional note: this code is very heavily munged from Henry's version
25 * in places. In some spots I've traded clarity for efficiency, so don't
26 * blame Henry for some of the lack of readability.
29 /* The names of the functions have been changed from regcomp and
30 * regexec to pregcomp and pregexec in order to avoid conflicts
31 * with the POSIX routines of the same names.
34 #ifdef PERL_EXT_RE_BUILD
39 * pregcomp and pregexec -- regsub and regerror are not used in perl
41 * Copyright (c) 1986 by University of Toronto.
42 * Written by Henry Spencer. Not derived from licensed software.
44 * Permission is granted to anyone to use this software for any
45 * purpose on any computer system, and to redistribute it freely,
46 * subject to the following restrictions:
48 * 1. The author is not responsible for the consequences of use of
49 * this software, no matter how awful, even if they arise
52 * 2. The origin of this software must not be misrepresented, either
53 * by explicit claim or by omission.
55 * 3. Altered versions must be plainly marked as such, and must not
56 * be misrepresented as being the original software.
59 **** Alterations to Henry's code are...
61 **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
62 **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
63 **** by Larry Wall and others
65 **** You may distribute under the terms of either the GNU General Public
66 **** License or the Artistic License, as specified in the README file.
69 * Beware that some of this code is subtly aware of the way operator
70 * precedence is structured in regular expressions. Serious changes in
71 * regular-expression syntax might require a total rethink.
74 #define PERL_IN_REGCOMP_C
77 #ifndef PERL_IN_XSUB_RE
82 #ifdef PERL_IN_XSUB_RE
93 # if defined(BUGGY_MSC6)
94 /* MSC 6.00A breaks on op/regexp.t test 85 unless we turn this off */
95 # pragma optimize("a",off)
96 /* But MSC 6.00A is happy with 'w', for aliases only across function calls*/
97 # pragma optimize("w",on )
98 # endif /* BUGGY_MSC6 */
102 #define STATIC static
105 typedef struct RExC_state_t {
106 U32 flags; /* are we folding, multilining? */
107 char *precomp; /* uncompiled string. */
108 REGEXP *rx_sv; /* The SV that is the regexp. */
109 regexp *rx; /* perl core regexp structure */
110 regexp_internal *rxi; /* internal data for regexp object pprivate field */
111 char *start; /* Start of input for compile */
112 char *end; /* End of input for compile */
113 char *parse; /* Input-scan pointer. */
114 I32 whilem_seen; /* number of WHILEM in this expr */
115 regnode *emit_start; /* Start of emitted-code area */
116 regnode *emit_bound; /* First regnode outside of the allocated space */
117 regnode *emit; /* Code-emit pointer; ®dummy = don't = compiling */
118 I32 naughty; /* How bad is this pattern? */
119 I32 sawback; /* Did we see \1, ...? */
121 I32 size; /* Code size. */
122 I32 npar; /* Capture buffer count, (OPEN). */
123 I32 cpar; /* Capture buffer count, (CLOSE). */
124 I32 nestroot; /* root parens we are in - used by accept */
128 regnode **open_parens; /* pointers to open parens */
129 regnode **close_parens; /* pointers to close parens */
130 regnode *opend; /* END node in program */
131 I32 utf8; /* whether the pattern is utf8 or not */
132 I32 orig_utf8; /* whether the pattern was originally in utf8 */
133 /* XXX use this for future optimisation of case
134 * where pattern must be upgraded to utf8. */
135 HV *charnames; /* cache of named sequences */
136 HV *paren_names; /* Paren names */
138 regnode **recurse; /* Recurse regops */
139 I32 recurse_count; /* Number of recurse regops */
141 char *starttry; /* -Dr: where regtry was called. */
142 #define RExC_starttry (pRExC_state->starttry)
145 const char *lastparse;
147 AV *paren_name_list; /* idx -> name */
148 #define RExC_lastparse (pRExC_state->lastparse)
149 #define RExC_lastnum (pRExC_state->lastnum)
150 #define RExC_paren_name_list (pRExC_state->paren_name_list)
154 #define RExC_flags (pRExC_state->flags)
155 #define RExC_precomp (pRExC_state->precomp)
156 #define RExC_rx_sv (pRExC_state->rx_sv)
157 #define RExC_rx (pRExC_state->rx)
158 #define RExC_rxi (pRExC_state->rxi)
159 #define RExC_start (pRExC_state->start)
160 #define RExC_end (pRExC_state->end)
161 #define RExC_parse (pRExC_state->parse)
162 #define RExC_whilem_seen (pRExC_state->whilem_seen)
163 #ifdef RE_TRACK_PATTERN_OFFSETS
164 #define RExC_offsets (pRExC_state->rxi->u.offsets) /* I am not like the others */
166 #define RExC_emit (pRExC_state->emit)
167 #define RExC_emit_start (pRExC_state->emit_start)
168 #define RExC_emit_bound (pRExC_state->emit_bound)
169 #define RExC_naughty (pRExC_state->naughty)
170 #define RExC_sawback (pRExC_state->sawback)
171 #define RExC_seen (pRExC_state->seen)
172 #define RExC_size (pRExC_state->size)
173 #define RExC_npar (pRExC_state->npar)
174 #define RExC_nestroot (pRExC_state->nestroot)
175 #define RExC_extralen (pRExC_state->extralen)
176 #define RExC_seen_zerolen (pRExC_state->seen_zerolen)
177 #define RExC_seen_evals (pRExC_state->seen_evals)
178 #define RExC_utf8 (pRExC_state->utf8)
179 #define RExC_orig_utf8 (pRExC_state->orig_utf8)
180 #define RExC_charnames (pRExC_state->charnames)
181 #define RExC_open_parens (pRExC_state->open_parens)
182 #define RExC_close_parens (pRExC_state->close_parens)
183 #define RExC_opend (pRExC_state->opend)
184 #define RExC_paren_names (pRExC_state->paren_names)
185 #define RExC_recurse (pRExC_state->recurse)
186 #define RExC_recurse_count (pRExC_state->recurse_count)
189 #define ISMULT1(c) ((c) == '*' || (c) == '+' || (c) == '?')
190 #define ISMULT2(s) ((*s) == '*' || (*s) == '+' || (*s) == '?' || \
191 ((*s) == '{' && regcurly(s)))
194 #undef SPSTART /* dratted cpp namespace... */
197 * Flags to be passed up and down.
199 #define WORST 0 /* Worst case. */
200 #define HASWIDTH 0x01 /* Known to match non-null strings. */
201 #define SIMPLE 0x02 /* Simple enough to be STAR/PLUS operand. */
202 #define SPSTART 0x04 /* Starts with * or +. */
203 #define TRYAGAIN 0x08 /* Weeded out a declaration. */
204 #define POSTPONED 0x10 /* (?1),(?&name), (??{...}) or similar */
206 #define REG_NODE_NUM(x) ((x) ? (int)((x)-RExC_emit_start) : -1)
208 /* whether trie related optimizations are enabled */
209 #if PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION
210 #define TRIE_STUDY_OPT
211 #define FULL_TRIE_STUDY
217 #define PBYTE(u8str,paren) ((U8*)(u8str))[(paren) >> 3]
218 #define PBITVAL(paren) (1 << ((paren) & 7))
219 #define PAREN_TEST(u8str,paren) ( PBYTE(u8str,paren) & PBITVAL(paren))
220 #define PAREN_SET(u8str,paren) PBYTE(u8str,paren) |= PBITVAL(paren)
221 #define PAREN_UNSET(u8str,paren) PBYTE(u8str,paren) &= (~PBITVAL(paren))
224 /* About scan_data_t.
226 During optimisation we recurse through the regexp program performing
227 various inplace (keyhole style) optimisations. In addition study_chunk
228 and scan_commit populate this data structure with information about
229 what strings MUST appear in the pattern. We look for the longest
230 string that must appear for at a fixed location, and we look for the
231 longest string that may appear at a floating location. So for instance
236 Both 'FOO' and 'A' are fixed strings. Both 'B' and 'BAR' are floating
237 strings (because they follow a .* construct). study_chunk will identify
238 both FOO and BAR as being the longest fixed and floating strings respectively.
240 The strings can be composites, for instance
244 will result in a composite fixed substring 'foo'.
246 For each string some basic information is maintained:
248 - offset or min_offset
249 This is the position the string must appear at, or not before.
250 It also implicitly (when combined with minlenp) tells us how many
251 character must match before the string we are searching.
252 Likewise when combined with minlenp and the length of the string
253 tells us how many characters must appear after the string we have
257 Only used for floating strings. This is the rightmost point that
258 the string can appear at. Ifset to I32 max it indicates that the
259 string can occur infinitely far to the right.
262 A pointer to the minimum length of the pattern that the string
263 was found inside. This is important as in the case of positive
264 lookahead or positive lookbehind we can have multiple patterns
269 The minimum length of the pattern overall is 3, the minimum length
270 of the lookahead part is 3, but the minimum length of the part that
271 will actually match is 1. So 'FOO's minimum length is 3, but the
272 minimum length for the F is 1. This is important as the minimum length
273 is used to determine offsets in front of and behind the string being
274 looked for. Since strings can be composites this is the length of the
275 pattern at the time it was commited with a scan_commit. Note that
276 the length is calculated by study_chunk, so that the minimum lengths
277 are not known until the full pattern has been compiled, thus the
278 pointer to the value.
282 In the case of lookbehind the string being searched for can be
283 offset past the start point of the final matching string.
284 If this value was just blithely removed from the min_offset it would
285 invalidate some of the calculations for how many chars must match
286 before or after (as they are derived from min_offset and minlen and
287 the length of the string being searched for).
288 When the final pattern is compiled and the data is moved from the
289 scan_data_t structure into the regexp structure the information
290 about lookbehind is factored in, with the information that would
291 have been lost precalculated in the end_shift field for the
294 The fields pos_min and pos_delta are used to store the minimum offset
295 and the delta to the maximum offset at the current point in the pattern.
299 typedef struct scan_data_t {
300 /*I32 len_min; unused */
301 /*I32 len_delta; unused */
305 I32 last_end; /* min value, <0 unless valid. */
308 SV **longest; /* Either &l_fixed, or &l_float. */
309 SV *longest_fixed; /* longest fixed string found in pattern */
310 I32 offset_fixed; /* offset where it starts */
311 I32 *minlen_fixed; /* pointer to the minlen relevent to the string */
312 I32 lookbehind_fixed; /* is the position of the string modfied by LB */
313 SV *longest_float; /* longest floating string found in pattern */
314 I32 offset_float_min; /* earliest point in string it can appear */
315 I32 offset_float_max; /* latest point in string it can appear */
316 I32 *minlen_float; /* pointer to the minlen relevent to the string */
317 I32 lookbehind_float; /* is the position of the string modified by LB */
321 struct regnode_charclass_class *start_class;
325 * Forward declarations for pregcomp()'s friends.
328 static const scan_data_t zero_scan_data =
329 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ,0};
331 #define SF_BEFORE_EOL (SF_BEFORE_SEOL|SF_BEFORE_MEOL)
332 #define SF_BEFORE_SEOL 0x0001
333 #define SF_BEFORE_MEOL 0x0002
334 #define SF_FIX_BEFORE_EOL (SF_FIX_BEFORE_SEOL|SF_FIX_BEFORE_MEOL)
335 #define SF_FL_BEFORE_EOL (SF_FL_BEFORE_SEOL|SF_FL_BEFORE_MEOL)
338 # define SF_FIX_SHIFT_EOL (0+2)
339 # define SF_FL_SHIFT_EOL (0+4)
341 # define SF_FIX_SHIFT_EOL (+2)
342 # define SF_FL_SHIFT_EOL (+4)
345 #define SF_FIX_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FIX_SHIFT_EOL)
346 #define SF_FIX_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FIX_SHIFT_EOL)
348 #define SF_FL_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FL_SHIFT_EOL)
349 #define SF_FL_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FL_SHIFT_EOL) /* 0x20 */
350 #define SF_IS_INF 0x0040
351 #define SF_HAS_PAR 0x0080
352 #define SF_IN_PAR 0x0100
353 #define SF_HAS_EVAL 0x0200
354 #define SCF_DO_SUBSTR 0x0400
355 #define SCF_DO_STCLASS_AND 0x0800
356 #define SCF_DO_STCLASS_OR 0x1000
357 #define SCF_DO_STCLASS (SCF_DO_STCLASS_AND|SCF_DO_STCLASS_OR)
358 #define SCF_WHILEM_VISITED_POS 0x2000
360 #define SCF_TRIE_RESTUDY 0x4000 /* Do restudy? */
361 #define SCF_SEEN_ACCEPT 0x8000
363 #define UTF (RExC_utf8 != 0)
364 #define LOC ((RExC_flags & RXf_PMf_LOCALE) != 0)
365 #define FOLD ((RExC_flags & RXf_PMf_FOLD) != 0)
367 #define OOB_UNICODE 12345678
368 #define OOB_NAMEDCLASS -1
370 #define CHR_SVLEN(sv) (UTF ? sv_len_utf8(sv) : SvCUR(sv))
371 #define CHR_DIST(a,b) (UTF ? utf8_distance(a,b) : a - b)
374 /* length of regex to show in messages that don't mark a position within */
375 #define RegexLengthToShowInErrorMessages 127
378 * If MARKER[12] are adjusted, be sure to adjust the constants at the top
379 * of t/op/regmesg.t, the tests in t/op/re_tests, and those in
380 * op/pragma/warn/regcomp.
382 #define MARKER1 "<-- HERE" /* marker as it appears in the description */
383 #define MARKER2 " <-- HERE " /* marker as it appears within the regex */
385 #define REPORT_LOCATION " in regex; marked by " MARKER1 " in m/%.*s" MARKER2 "%s/"
388 * Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
389 * arg. Show regex, up to a maximum length. If it's too long, chop and add
392 #define _FAIL(code) STMT_START { \
393 const char *ellipses = ""; \
394 IV len = RExC_end - RExC_precomp; \
397 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
398 if (len > RegexLengthToShowInErrorMessages) { \
399 /* chop 10 shorter than the max, to ensure meaning of "..." */ \
400 len = RegexLengthToShowInErrorMessages - 10; \
406 #define FAIL(msg) _FAIL( \
407 Perl_croak(aTHX_ "%s in regex m/%.*s%s/", \
408 msg, (int)len, RExC_precomp, ellipses))
410 #define FAIL2(msg,arg) _FAIL( \
411 Perl_croak(aTHX_ msg " in regex m/%.*s%s/", \
412 arg, (int)len, RExC_precomp, ellipses))
415 * Simple_vFAIL -- like FAIL, but marks the current location in the scan
417 #define Simple_vFAIL(m) STMT_START { \
418 const IV offset = RExC_parse - RExC_precomp; \
419 Perl_croak(aTHX_ "%s" REPORT_LOCATION, \
420 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
424 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL()
426 #define vFAIL(m) STMT_START { \
428 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
433 * Like Simple_vFAIL(), but accepts two arguments.
435 #define Simple_vFAIL2(m,a1) STMT_START { \
436 const IV offset = RExC_parse - RExC_precomp; \
437 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, \
438 (int)offset, RExC_precomp, RExC_precomp + offset); \
442 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL2().
444 #define vFAIL2(m,a1) STMT_START { \
446 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
447 Simple_vFAIL2(m, a1); \
452 * Like Simple_vFAIL(), but accepts three arguments.
454 #define Simple_vFAIL3(m, a1, a2) STMT_START { \
455 const IV offset = RExC_parse - RExC_precomp; \
456 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, \
457 (int)offset, RExC_precomp, RExC_precomp + offset); \
461 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL3().
463 #define vFAIL3(m,a1,a2) STMT_START { \
465 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
466 Simple_vFAIL3(m, a1, a2); \
470 * Like Simple_vFAIL(), but accepts four arguments.
472 #define Simple_vFAIL4(m, a1, a2, a3) STMT_START { \
473 const IV offset = RExC_parse - RExC_precomp; \
474 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, \
475 (int)offset, RExC_precomp, RExC_precomp + offset); \
478 #define vWARN(loc,m) STMT_START { \
479 const IV offset = loc - RExC_precomp; \
480 Perl_warner(aTHX_ packWARN(WARN_REGEXP), "%s" REPORT_LOCATION, \
481 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
484 #define vWARNdep(loc,m) STMT_START { \
485 const IV offset = loc - RExC_precomp; \
486 Perl_warner(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
487 "%s" REPORT_LOCATION, \
488 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
492 #define vWARN2(loc, m, a1) STMT_START { \
493 const IV offset = loc - RExC_precomp; \
494 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
495 a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
498 #define vWARN3(loc, m, a1, a2) STMT_START { \
499 const IV offset = loc - RExC_precomp; \
500 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
501 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
504 #define vWARN4(loc, m, a1, a2, a3) STMT_START { \
505 const IV offset = loc - RExC_precomp; \
506 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
507 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
510 #define vWARN5(loc, m, a1, a2, a3, a4) STMT_START { \
511 const IV offset = loc - RExC_precomp; \
512 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
513 a1, a2, a3, a4, (int)offset, RExC_precomp, RExC_precomp + offset); \
517 /* Allow for side effects in s */
518 #define REGC(c,s) STMT_START { \
519 if (!SIZE_ONLY) *(s) = (c); else (void)(s); \
522 /* Macros for recording node offsets. 20001227 mjd@plover.com
523 * Nodes are numbered 1, 2, 3, 4. Node #n's position is recorded in
524 * element 2*n-1 of the array. Element #2n holds the byte length node #n.
525 * Element 0 holds the number n.
526 * Position is 1 indexed.
528 #ifndef RE_TRACK_PATTERN_OFFSETS
529 #define Set_Node_Offset_To_R(node,byte)
530 #define Set_Node_Offset(node,byte)
531 #define Set_Cur_Node_Offset
532 #define Set_Node_Length_To_R(node,len)
533 #define Set_Node_Length(node,len)
534 #define Set_Node_Cur_Length(node)
535 #define Node_Offset(n)
536 #define Node_Length(n)
537 #define Set_Node_Offset_Length(node,offset,len)
538 #define ProgLen(ri) ri->u.proglen
539 #define SetProgLen(ri,x) ri->u.proglen = x
541 #define ProgLen(ri) ri->u.offsets[0]
542 #define SetProgLen(ri,x) ri->u.offsets[0] = x
543 #define Set_Node_Offset_To_R(node,byte) STMT_START { \
545 MJD_OFFSET_DEBUG(("** (%d) offset of node %d is %d.\n", \
546 __LINE__, (int)(node), (int)(byte))); \
548 Perl_croak(aTHX_ "value of node is %d in Offset macro", (int)(node)); \
550 RExC_offsets[2*(node)-1] = (byte); \
555 #define Set_Node_Offset(node,byte) \
556 Set_Node_Offset_To_R((node)-RExC_emit_start, (byte)-RExC_start)
557 #define Set_Cur_Node_Offset Set_Node_Offset(RExC_emit, RExC_parse)
559 #define Set_Node_Length_To_R(node,len) STMT_START { \
561 MJD_OFFSET_DEBUG(("** (%d) size of node %d is %d.\n", \
562 __LINE__, (int)(node), (int)(len))); \
564 Perl_croak(aTHX_ "value of node is %d in Length macro", (int)(node)); \
566 RExC_offsets[2*(node)] = (len); \
571 #define Set_Node_Length(node,len) \
572 Set_Node_Length_To_R((node)-RExC_emit_start, len)
573 #define Set_Cur_Node_Length(len) Set_Node_Length(RExC_emit, len)
574 #define Set_Node_Cur_Length(node) \
575 Set_Node_Length(node, RExC_parse - parse_start)
577 /* Get offsets and lengths */
578 #define Node_Offset(n) (RExC_offsets[2*((n)-RExC_emit_start)-1])
579 #define Node_Length(n) (RExC_offsets[2*((n)-RExC_emit_start)])
581 #define Set_Node_Offset_Length(node,offset,len) STMT_START { \
582 Set_Node_Offset_To_R((node)-RExC_emit_start, (offset)); \
583 Set_Node_Length_To_R((node)-RExC_emit_start, (len)); \
587 #if PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS
588 #define EXPERIMENTAL_INPLACESCAN
589 #endif /*RE_TRACK_PATTERN_OFFSETS*/
591 #define DEBUG_STUDYDATA(str,data,depth) \
592 DEBUG_OPTIMISE_MORE_r(if(data){ \
593 PerlIO_printf(Perl_debug_log, \
594 "%*s" str "Pos:%"IVdf"/%"IVdf \
595 " Flags: 0x%"UVXf" Whilem_c: %"IVdf" Lcp: %"IVdf" %s", \
596 (int)(depth)*2, "", \
597 (IV)((data)->pos_min), \
598 (IV)((data)->pos_delta), \
599 (UV)((data)->flags), \
600 (IV)((data)->whilem_c), \
601 (IV)((data)->last_closep ? *((data)->last_closep) : -1), \
602 is_inf ? "INF " : "" \
604 if ((data)->last_found) \
605 PerlIO_printf(Perl_debug_log, \
606 "Last:'%s' %"IVdf":%"IVdf"/%"IVdf" %sFixed:'%s' @ %"IVdf \
607 " %sFloat: '%s' @ %"IVdf"/%"IVdf"", \
608 SvPVX_const((data)->last_found), \
609 (IV)((data)->last_end), \
610 (IV)((data)->last_start_min), \
611 (IV)((data)->last_start_max), \
612 ((data)->longest && \
613 (data)->longest==&((data)->longest_fixed)) ? "*" : "", \
614 SvPVX_const((data)->longest_fixed), \
615 (IV)((data)->offset_fixed), \
616 ((data)->longest && \
617 (data)->longest==&((data)->longest_float)) ? "*" : "", \
618 SvPVX_const((data)->longest_float), \
619 (IV)((data)->offset_float_min), \
620 (IV)((data)->offset_float_max) \
622 PerlIO_printf(Perl_debug_log,"\n"); \
625 static void clear_re(pTHX_ void *r);
627 /* Mark that we cannot extend a found fixed substring at this point.
628 Update the longest found anchored substring and the longest found
629 floating substrings if needed. */
632 S_scan_commit(pTHX_ const RExC_state_t *pRExC_state, scan_data_t *data, I32 *minlenp, int is_inf)
634 const STRLEN l = CHR_SVLEN(data->last_found);
635 const STRLEN old_l = CHR_SVLEN(*data->longest);
636 GET_RE_DEBUG_FLAGS_DECL;
638 PERL_ARGS_ASSERT_SCAN_COMMIT;
640 if ((l >= old_l) && ((l > old_l) || (data->flags & SF_BEFORE_EOL))) {
641 SvSetMagicSV(*data->longest, data->last_found);
642 if (*data->longest == data->longest_fixed) {
643 data->offset_fixed = l ? data->last_start_min : data->pos_min;
644 if (data->flags & SF_BEFORE_EOL)
646 |= ((data->flags & SF_BEFORE_EOL) << SF_FIX_SHIFT_EOL);
648 data->flags &= ~SF_FIX_BEFORE_EOL;
649 data->minlen_fixed=minlenp;
650 data->lookbehind_fixed=0;
652 else { /* *data->longest == data->longest_float */
653 data->offset_float_min = l ? data->last_start_min : data->pos_min;
654 data->offset_float_max = (l
655 ? data->last_start_max
656 : data->pos_min + data->pos_delta);
657 if (is_inf || (U32)data->offset_float_max > (U32)I32_MAX)
658 data->offset_float_max = I32_MAX;
659 if (data->flags & SF_BEFORE_EOL)
661 |= ((data->flags & SF_BEFORE_EOL) << SF_FL_SHIFT_EOL);
663 data->flags &= ~SF_FL_BEFORE_EOL;
664 data->minlen_float=minlenp;
665 data->lookbehind_float=0;
668 SvCUR_set(data->last_found, 0);
670 SV * const sv = data->last_found;
671 if (SvUTF8(sv) && SvMAGICAL(sv)) {
672 MAGIC * const mg = mg_find(sv, PERL_MAGIC_utf8);
678 data->flags &= ~SF_BEFORE_EOL;
679 DEBUG_STUDYDATA("commit: ",data,0);
682 /* Can match anything (initialization) */
684 S_cl_anything(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
686 PERL_ARGS_ASSERT_CL_ANYTHING;
688 ANYOF_CLASS_ZERO(cl);
689 ANYOF_BITMAP_SETALL(cl);
690 cl->flags = ANYOF_EOS|ANYOF_UNICODE_ALL;
692 cl->flags |= ANYOF_LOCALE;
695 /* Can match anything (initialization) */
697 S_cl_is_anything(const struct regnode_charclass_class *cl)
701 PERL_ARGS_ASSERT_CL_IS_ANYTHING;
703 for (value = 0; value <= ANYOF_MAX; value += 2)
704 if (ANYOF_CLASS_TEST(cl, value) && ANYOF_CLASS_TEST(cl, value + 1))
706 if (!(cl->flags & ANYOF_UNICODE_ALL))
708 if (!ANYOF_BITMAP_TESTALLSET((const void*)cl))
713 /* Can match anything (initialization) */
715 S_cl_init(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
717 PERL_ARGS_ASSERT_CL_INIT;
719 Zero(cl, 1, struct regnode_charclass_class);
721 cl_anything(pRExC_state, cl);
725 S_cl_init_zero(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
727 PERL_ARGS_ASSERT_CL_INIT_ZERO;
729 Zero(cl, 1, struct regnode_charclass_class);
731 cl_anything(pRExC_state, cl);
733 cl->flags |= ANYOF_LOCALE;
736 /* 'And' a given class with another one. Can create false positives */
737 /* We assume that cl is not inverted */
739 S_cl_and(struct regnode_charclass_class *cl,
740 const struct regnode_charclass_class *and_with)
742 PERL_ARGS_ASSERT_CL_AND;
744 assert(and_with->type == ANYOF);
745 if (!(and_with->flags & ANYOF_CLASS)
746 && !(cl->flags & ANYOF_CLASS)
747 && (and_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
748 && !(and_with->flags & ANYOF_FOLD)
749 && !(cl->flags & ANYOF_FOLD)) {
752 if (and_with->flags & ANYOF_INVERT)
753 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
754 cl->bitmap[i] &= ~and_with->bitmap[i];
756 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
757 cl->bitmap[i] &= and_with->bitmap[i];
758 } /* XXXX: logic is complicated otherwise, leave it along for a moment. */
759 if (!(and_with->flags & ANYOF_EOS))
760 cl->flags &= ~ANYOF_EOS;
762 if (cl->flags & ANYOF_UNICODE_ALL && and_with->flags & ANYOF_UNICODE &&
763 !(and_with->flags & ANYOF_INVERT)) {
764 cl->flags &= ~ANYOF_UNICODE_ALL;
765 cl->flags |= ANYOF_UNICODE;
766 ARG_SET(cl, ARG(and_with));
768 if (!(and_with->flags & ANYOF_UNICODE_ALL) &&
769 !(and_with->flags & ANYOF_INVERT))
770 cl->flags &= ~ANYOF_UNICODE_ALL;
771 if (!(and_with->flags & (ANYOF_UNICODE|ANYOF_UNICODE_ALL)) &&
772 !(and_with->flags & ANYOF_INVERT))
773 cl->flags &= ~ANYOF_UNICODE;
776 /* 'OR' a given class with another one. Can create false positives */
777 /* We assume that cl is not inverted */
779 S_cl_or(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, const struct regnode_charclass_class *or_with)
781 PERL_ARGS_ASSERT_CL_OR;
783 if (or_with->flags & ANYOF_INVERT) {
785 * (B1 | CL1) | (!B2 & !CL2) = (B1 | !B2 & !CL2) | (CL1 | (!B2 & !CL2))
786 * <= (B1 | !B2) | (CL1 | !CL2)
787 * which is wasteful if CL2 is small, but we ignore CL2:
788 * (B1 | CL1) | (!B2 & !CL2) <= (B1 | CL1) | !B2 = (B1 | !B2) | CL1
789 * XXXX Can we handle case-fold? Unclear:
790 * (OK1(i) | OK1(i')) | !(OK1(i) | OK1(i')) =
791 * (OK1(i) | OK1(i')) | (!OK1(i) & !OK1(i'))
793 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
794 && !(or_with->flags & ANYOF_FOLD)
795 && !(cl->flags & ANYOF_FOLD) ) {
798 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
799 cl->bitmap[i] |= ~or_with->bitmap[i];
800 } /* XXXX: logic is complicated otherwise */
802 cl_anything(pRExC_state, cl);
805 /* (B1 | CL1) | (B2 | CL2) = (B1 | B2) | (CL1 | CL2)) */
806 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
807 && (!(or_with->flags & ANYOF_FOLD)
808 || (cl->flags & ANYOF_FOLD)) ) {
811 /* OR char bitmap and class bitmap separately */
812 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
813 cl->bitmap[i] |= or_with->bitmap[i];
814 if (or_with->flags & ANYOF_CLASS) {
815 for (i = 0; i < ANYOF_CLASSBITMAP_SIZE; i++)
816 cl->classflags[i] |= or_with->classflags[i];
817 cl->flags |= ANYOF_CLASS;
820 else { /* XXXX: logic is complicated, leave it along for a moment. */
821 cl_anything(pRExC_state, cl);
824 if (or_with->flags & ANYOF_EOS)
825 cl->flags |= ANYOF_EOS;
827 if (cl->flags & ANYOF_UNICODE && or_with->flags & ANYOF_UNICODE &&
828 ARG(cl) != ARG(or_with)) {
829 cl->flags |= ANYOF_UNICODE_ALL;
830 cl->flags &= ~ANYOF_UNICODE;
832 if (or_with->flags & ANYOF_UNICODE_ALL) {
833 cl->flags |= ANYOF_UNICODE_ALL;
834 cl->flags &= ~ANYOF_UNICODE;
838 #define TRIE_LIST_ITEM(state,idx) (trie->states[state].trans.list)[ idx ]
839 #define TRIE_LIST_CUR(state) ( TRIE_LIST_ITEM( state, 0 ).forid )
840 #define TRIE_LIST_LEN(state) ( TRIE_LIST_ITEM( state, 0 ).newstate )
841 #define TRIE_LIST_USED(idx) ( trie->states[state].trans.list ? (TRIE_LIST_CUR( idx ) - 1) : 0 )
846 dump_trie(trie,widecharmap,revcharmap)
847 dump_trie_interim_list(trie,widecharmap,revcharmap,next_alloc)
848 dump_trie_interim_table(trie,widecharmap,revcharmap,next_alloc)
850 These routines dump out a trie in a somewhat readable format.
851 The _interim_ variants are used for debugging the interim
852 tables that are used to generate the final compressed
853 representation which is what dump_trie expects.
855 Part of the reason for their existance is to provide a form
856 of documentation as to how the different representations function.
861 Dumps the final compressed table form of the trie to Perl_debug_log.
862 Used for debugging make_trie().
866 S_dump_trie(pTHX_ const struct _reg_trie_data *trie, HV *widecharmap,
867 AV *revcharmap, U32 depth)
870 SV *sv=sv_newmortal();
871 int colwidth= widecharmap ? 6 : 4;
872 GET_RE_DEBUG_FLAGS_DECL;
874 PERL_ARGS_ASSERT_DUMP_TRIE;
876 PerlIO_printf( Perl_debug_log, "%*sChar : %-6s%-6s%-4s ",
877 (int)depth * 2 + 2,"",
878 "Match","Base","Ofs" );
880 for( state = 0 ; state < trie->uniquecharcount ; state++ ) {
881 SV ** const tmp = av_fetch( revcharmap, state, 0);
883 PerlIO_printf( Perl_debug_log, "%*s",
885 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
886 PL_colors[0], PL_colors[1],
887 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
888 PERL_PV_ESCAPE_FIRSTCHAR
893 PerlIO_printf( Perl_debug_log, "\n%*sState|-----------------------",
894 (int)depth * 2 + 2,"");
896 for( state = 0 ; state < trie->uniquecharcount ; state++ )
897 PerlIO_printf( Perl_debug_log, "%.*s", colwidth, "--------");
898 PerlIO_printf( Perl_debug_log, "\n");
900 for( state = 1 ; state < trie->statecount ; state++ ) {
901 const U32 base = trie->states[ state ].trans.base;
903 PerlIO_printf( Perl_debug_log, "%*s#%4"UVXf"|", (int)depth * 2 + 2,"", (UV)state);
905 if ( trie->states[ state ].wordnum ) {
906 PerlIO_printf( Perl_debug_log, " W%4X", trie->states[ state ].wordnum );
908 PerlIO_printf( Perl_debug_log, "%6s", "" );
911 PerlIO_printf( Perl_debug_log, " @%4"UVXf" ", (UV)base );
916 while( ( base + ofs < trie->uniquecharcount ) ||
917 ( base + ofs - trie->uniquecharcount < trie->lasttrans
918 && trie->trans[ base + ofs - trie->uniquecharcount ].check != state))
921 PerlIO_printf( Perl_debug_log, "+%2"UVXf"[ ", (UV)ofs);
923 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
924 if ( ( base + ofs >= trie->uniquecharcount ) &&
925 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
926 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
928 PerlIO_printf( Perl_debug_log, "%*"UVXf,
930 (UV)trie->trans[ base + ofs - trie->uniquecharcount ].next );
932 PerlIO_printf( Perl_debug_log, "%*s",colwidth," ." );
936 PerlIO_printf( Perl_debug_log, "]");
939 PerlIO_printf( Perl_debug_log, "\n" );
943 Dumps a fully constructed but uncompressed trie in list form.
944 List tries normally only are used for construction when the number of
945 possible chars (trie->uniquecharcount) is very high.
946 Used for debugging make_trie().
949 S_dump_trie_interim_list(pTHX_ const struct _reg_trie_data *trie,
950 HV *widecharmap, AV *revcharmap, U32 next_alloc,
954 SV *sv=sv_newmortal();
955 int colwidth= widecharmap ? 6 : 4;
956 GET_RE_DEBUG_FLAGS_DECL;
958 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_LIST;
960 /* print out the table precompression. */
961 PerlIO_printf( Perl_debug_log, "%*sState :Word | Transition Data\n%*s%s",
962 (int)depth * 2 + 2,"", (int)depth * 2 + 2,"",
963 "------:-----+-----------------\n" );
965 for( state=1 ; state < next_alloc ; state ++ ) {
968 PerlIO_printf( Perl_debug_log, "%*s %4"UVXf" :",
969 (int)depth * 2 + 2,"", (UV)state );
970 if ( ! trie->states[ state ].wordnum ) {
971 PerlIO_printf( Perl_debug_log, "%5s| ","");
973 PerlIO_printf( Perl_debug_log, "W%4x| ",
974 trie->states[ state ].wordnum
977 for( charid = 1 ; charid <= TRIE_LIST_USED( state ) ; charid++ ) {
978 SV ** const tmp = av_fetch( revcharmap, TRIE_LIST_ITEM(state,charid).forid, 0);
980 PerlIO_printf( Perl_debug_log, "%*s:%3X=%4"UVXf" | ",
982 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
983 PL_colors[0], PL_colors[1],
984 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
985 PERL_PV_ESCAPE_FIRSTCHAR
987 TRIE_LIST_ITEM(state,charid).forid,
988 (UV)TRIE_LIST_ITEM(state,charid).newstate
991 PerlIO_printf(Perl_debug_log, "\n%*s| ",
992 (int)((depth * 2) + 14), "");
995 PerlIO_printf( Perl_debug_log, "\n");
1000 Dumps a fully constructed but uncompressed trie in table form.
1001 This is the normal DFA style state transition table, with a few
1002 twists to facilitate compression later.
1003 Used for debugging make_trie().
1006 S_dump_trie_interim_table(pTHX_ const struct _reg_trie_data *trie,
1007 HV *widecharmap, AV *revcharmap, U32 next_alloc,
1012 SV *sv=sv_newmortal();
1013 int colwidth= widecharmap ? 6 : 4;
1014 GET_RE_DEBUG_FLAGS_DECL;
1016 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_TABLE;
1019 print out the table precompression so that we can do a visual check
1020 that they are identical.
1023 PerlIO_printf( Perl_debug_log, "%*sChar : ",(int)depth * 2 + 2,"" );
1025 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1026 SV ** const tmp = av_fetch( revcharmap, charid, 0);
1028 PerlIO_printf( Perl_debug_log, "%*s",
1030 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1031 PL_colors[0], PL_colors[1],
1032 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1033 PERL_PV_ESCAPE_FIRSTCHAR
1039 PerlIO_printf( Perl_debug_log, "\n%*sState+-",(int)depth * 2 + 2,"" );
1041 for( charid=0 ; charid < trie->uniquecharcount ; charid++ ) {
1042 PerlIO_printf( Perl_debug_log, "%.*s", colwidth,"--------");
1045 PerlIO_printf( Perl_debug_log, "\n" );
1047 for( state=1 ; state < next_alloc ; state += trie->uniquecharcount ) {
1049 PerlIO_printf( Perl_debug_log, "%*s%4"UVXf" : ",
1050 (int)depth * 2 + 2,"",
1051 (UV)TRIE_NODENUM( state ) );
1053 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1054 UV v=(UV)SAFE_TRIE_NODENUM( trie->trans[ state + charid ].next );
1056 PerlIO_printf( Perl_debug_log, "%*"UVXf, colwidth, v );
1058 PerlIO_printf( Perl_debug_log, "%*s", colwidth, "." );
1060 if ( ! trie->states[ TRIE_NODENUM( state ) ].wordnum ) {
1061 PerlIO_printf( Perl_debug_log, " (%4"UVXf")\n", (UV)trie->trans[ state ].check );
1063 PerlIO_printf( Perl_debug_log, " (%4"UVXf") W%4X\n", (UV)trie->trans[ state ].check,
1064 trie->states[ TRIE_NODENUM( state ) ].wordnum );
1071 /* make_trie(startbranch,first,last,tail,word_count,flags,depth)
1072 startbranch: the first branch in the whole branch sequence
1073 first : start branch of sequence of branch-exact nodes.
1074 May be the same as startbranch
1075 last : Thing following the last branch.
1076 May be the same as tail.
1077 tail : item following the branch sequence
1078 count : words in the sequence
1079 flags : currently the OP() type we will be building one of /EXACT(|F|Fl)/
1080 depth : indent depth
1082 Inplace optimizes a sequence of 2 or more Branch-Exact nodes into a TRIE node.
1084 A trie is an N'ary tree where the branches are determined by digital
1085 decomposition of the key. IE, at the root node you look up the 1st character and
1086 follow that branch repeat until you find the end of the branches. Nodes can be
1087 marked as "accepting" meaning they represent a complete word. Eg:
1091 would convert into the following structure. Numbers represent states, letters
1092 following numbers represent valid transitions on the letter from that state, if
1093 the number is in square brackets it represents an accepting state, otherwise it
1094 will be in parenthesis.
1096 +-h->+-e->[3]-+-r->(8)-+-s->[9]
1100 (1) +-i->(6)-+-s->[7]
1102 +-s->(3)-+-h->(4)-+-e->[5]
1104 Accept Word Mapping: 3=>1 (he),5=>2 (she), 7=>3 (his), 9=>4 (hers)
1106 This shows that when matching against the string 'hers' we will begin at state 1
1107 read 'h' and move to state 2, read 'e' and move to state 3 which is accepting,
1108 then read 'r' and go to state 8 followed by 's' which takes us to state 9 which
1109 is also accepting. Thus we know that we can match both 'he' and 'hers' with a
1110 single traverse. We store a mapping from accepting to state to which word was
1111 matched, and then when we have multiple possibilities we try to complete the
1112 rest of the regex in the order in which they occured in the alternation.
1114 The only prior NFA like behaviour that would be changed by the TRIE support is
1115 the silent ignoring of duplicate alternations which are of the form:
1117 / (DUPE|DUPE) X? (?{ ... }) Y /x
1119 Thus EVAL blocks follwing a trie may be called a different number of times with
1120 and without the optimisation. With the optimisations dupes will be silently
1121 ignored. This inconsistant behaviour of EVAL type nodes is well established as
1122 the following demonstrates:
1124 'words'=~/(word|word|word)(?{ print $1 })[xyz]/
1126 which prints out 'word' three times, but
1128 'words'=~/(word|word|word)(?{ print $1 })S/
1130 which doesnt print it out at all. This is due to other optimisations kicking in.
1132 Example of what happens on a structural level:
1134 The regexp /(ac|ad|ab)+/ will produce the folowing debug output:
1136 1: CURLYM[1] {1,32767}(18)
1147 This would be optimizable with startbranch=5, first=5, last=16, tail=16
1148 and should turn into:
1150 1: CURLYM[1] {1,32767}(18)
1152 [Words:3 Chars Stored:6 Unique Chars:4 States:5 NCP:1]
1160 Cases where tail != last would be like /(?foo|bar)baz/:
1170 which would be optimizable with startbranch=1, first=1, last=7, tail=8
1171 and would end up looking like:
1174 [Words:2 Chars Stored:6 Unique Chars:5 States:7 NCP:1]
1181 d = uvuni_to_utf8_flags(d, uv, 0);
1183 is the recommended Unicode-aware way of saying
1188 #define TRIE_STORE_REVCHAR \
1191 SV *zlopp = newSV(2); \
1192 unsigned char *flrbbbbb = (unsigned char *) SvPVX(zlopp); \
1193 unsigned const char *const kapow = uvuni_to_utf8(flrbbbbb, uvc & 0xFF); \
1194 SvCUR_set(zlopp, kapow - flrbbbbb); \
1197 av_push(revcharmap, zlopp); \
1199 char ooooff = (char)uvc; \
1200 av_push(revcharmap, newSVpvn(&ooooff, 1)); \
1204 #define TRIE_READ_CHAR STMT_START { \
1208 if ( foldlen > 0 ) { \
1209 uvc = utf8n_to_uvuni( scan, UTF8_MAXLEN, &len, uniflags ); \
1214 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1215 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1216 foldlen -= UNISKIP( uvc ); \
1217 scan = foldbuf + UNISKIP( uvc ); \
1220 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1230 #define TRIE_LIST_PUSH(state,fid,ns) STMT_START { \
1231 if ( TRIE_LIST_CUR( state ) >=TRIE_LIST_LEN( state ) ) { \
1232 U32 ging = TRIE_LIST_LEN( state ) *= 2; \
1233 Renew( trie->states[ state ].trans.list, ging, reg_trie_trans_le ); \
1235 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).forid = fid; \
1236 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).newstate = ns; \
1237 TRIE_LIST_CUR( state )++; \
1240 #define TRIE_LIST_NEW(state) STMT_START { \
1241 Newxz( trie->states[ state ].trans.list, \
1242 4, reg_trie_trans_le ); \
1243 TRIE_LIST_CUR( state ) = 1; \
1244 TRIE_LIST_LEN( state ) = 4; \
1247 #define TRIE_HANDLE_WORD(state) STMT_START { \
1248 U16 dupe= trie->states[ state ].wordnum; \
1249 regnode * const noper_next = regnext( noper ); \
1251 if (trie->wordlen) \
1252 trie->wordlen[ curword ] = wordlen; \
1254 /* store the word for dumping */ \
1256 if (OP(noper) != NOTHING) \
1257 tmp = newSVpvn_utf8(STRING(noper), STR_LEN(noper), UTF); \
1259 tmp = newSVpvn_utf8( "", 0, UTF ); \
1260 av_push( trie_words, tmp ); \
1265 if ( noper_next < tail ) { \
1267 trie->jump = (U16 *) PerlMemShared_calloc( word_count + 1, sizeof(U16) ); \
1268 trie->jump[curword] = (U16)(noper_next - convert); \
1270 jumper = noper_next; \
1272 nextbranch= regnext(cur); \
1276 /* So it's a dupe. This means we need to maintain a */\
1277 /* linked-list from the first to the next. */\
1278 /* we only allocate the nextword buffer when there */\
1279 /* a dupe, so first time we have to do the allocation */\
1280 if (!trie->nextword) \
1281 trie->nextword = (U16 *) \
1282 PerlMemShared_calloc( word_count + 1, sizeof(U16)); \
1283 while ( trie->nextword[dupe] ) \
1284 dupe= trie->nextword[dupe]; \
1285 trie->nextword[dupe]= curword; \
1287 /* we haven't inserted this word yet. */ \
1288 trie->states[ state ].wordnum = curword; \
1293 #define TRIE_TRANS_STATE(state,base,ucharcount,charid,special) \
1294 ( ( base + charid >= ucharcount \
1295 && base + charid < ubound \
1296 && state == trie->trans[ base - ucharcount + charid ].check \
1297 && trie->trans[ base - ucharcount + charid ].next ) \
1298 ? trie->trans[ base - ucharcount + charid ].next \
1299 : ( state==1 ? special : 0 ) \
1303 #define MADE_JUMP_TRIE 2
1304 #define MADE_EXACT_TRIE 4
1307 S_make_trie(pTHX_ RExC_state_t *pRExC_state, regnode *startbranch, regnode *first, regnode *last, regnode *tail, U32 word_count, U32 flags, U32 depth)
1310 /* first pass, loop through and scan words */
1311 reg_trie_data *trie;
1312 HV *widecharmap = NULL;
1313 AV *revcharmap = newAV();
1315 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1320 regnode *jumper = NULL;
1321 regnode *nextbranch = NULL;
1322 regnode *convert = NULL;
1323 /* we just use folder as a flag in utf8 */
1324 const U8 * const folder = ( flags == EXACTF
1326 : ( flags == EXACTFL
1333 const U32 data_slot = add_data( pRExC_state, 4, "tuuu" );
1334 AV *trie_words = NULL;
1335 /* along with revcharmap, this only used during construction but both are
1336 * useful during debugging so we store them in the struct when debugging.
1339 const U32 data_slot = add_data( pRExC_state, 2, "tu" );
1340 STRLEN trie_charcount=0;
1342 SV *re_trie_maxbuff;
1343 GET_RE_DEBUG_FLAGS_DECL;
1345 PERL_ARGS_ASSERT_MAKE_TRIE;
1347 PERL_UNUSED_ARG(depth);
1350 trie = (reg_trie_data *) PerlMemShared_calloc( 1, sizeof(reg_trie_data) );
1352 trie->startstate = 1;
1353 trie->wordcount = word_count;
1354 RExC_rxi->data->data[ data_slot ] = (void*)trie;
1355 trie->charmap = (U16 *) PerlMemShared_calloc( 256, sizeof(U16) );
1356 if (!(UTF && folder))
1357 trie->bitmap = (char *) PerlMemShared_calloc( ANYOF_BITMAP_SIZE, 1 );
1359 trie_words = newAV();
1362 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
1363 if (!SvIOK(re_trie_maxbuff)) {
1364 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
1367 PerlIO_printf( Perl_debug_log,
1368 "%*smake_trie start==%d, first==%d, last==%d, tail==%d depth=%d\n",
1369 (int)depth * 2 + 2, "",
1370 REG_NODE_NUM(startbranch),REG_NODE_NUM(first),
1371 REG_NODE_NUM(last), REG_NODE_NUM(tail),
1375 /* Find the node we are going to overwrite */
1376 if ( first == startbranch && OP( last ) != BRANCH ) {
1377 /* whole branch chain */
1380 /* branch sub-chain */
1381 convert = NEXTOPER( first );
1384 /* -- First loop and Setup --
1386 We first traverse the branches and scan each word to determine if it
1387 contains widechars, and how many unique chars there are, this is
1388 important as we have to build a table with at least as many columns as we
1391 We use an array of integers to represent the character codes 0..255
1392 (trie->charmap) and we use a an HV* to store Unicode characters. We use the
1393 native representation of the character value as the key and IV's for the
1396 *TODO* If we keep track of how many times each character is used we can
1397 remap the columns so that the table compression later on is more
1398 efficient in terms of memory by ensuring most common value is in the
1399 middle and the least common are on the outside. IMO this would be better
1400 than a most to least common mapping as theres a decent chance the most
1401 common letter will share a node with the least common, meaning the node
1402 will not be compressable. With a middle is most common approach the worst
1403 case is when we have the least common nodes twice.
1407 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1408 regnode * const noper = NEXTOPER( cur );
1409 const U8 *uc = (U8*)STRING( noper );
1410 const U8 * const e = uc + STR_LEN( noper );
1412 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1413 const U8 *scan = (U8*)NULL;
1414 U32 wordlen = 0; /* required init */
1416 bool set_bit = trie->bitmap ? 1 : 0; /*store the first char in the bitmap?*/
1418 if (OP(noper) == NOTHING) {
1422 if ( set_bit ) /* bitmap only alloced when !(UTF&&Folding) */
1423 TRIE_BITMAP_SET(trie,*uc); /* store the raw first byte
1424 regardless of encoding */
1426 for ( ; uc < e ; uc += len ) {
1427 TRIE_CHARCOUNT(trie)++;
1431 if ( !trie->charmap[ uvc ] ) {
1432 trie->charmap[ uvc ]=( ++trie->uniquecharcount );
1434 trie->charmap[ folder[ uvc ] ] = trie->charmap[ uvc ];
1438 /* store the codepoint in the bitmap, and if its ascii
1439 also store its folded equivelent. */
1440 TRIE_BITMAP_SET(trie,uvc);
1442 /* store the folded codepoint */
1443 if ( folder ) TRIE_BITMAP_SET(trie,folder[ uvc ]);
1446 /* store first byte of utf8 representation of
1447 codepoints in the 127 < uvc < 256 range */
1448 if (127 < uvc && uvc < 192) {
1449 TRIE_BITMAP_SET(trie,194);
1450 } else if (191 < uvc ) {
1451 TRIE_BITMAP_SET(trie,195);
1452 /* && uvc < 256 -- we know uvc is < 256 already */
1455 set_bit = 0; /* We've done our bit :-) */
1460 widecharmap = newHV();
1462 svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 1 );
1465 Perl_croak( aTHX_ "error creating/fetching widecharmap entry for 0x%"UVXf, uvc );
1467 if ( !SvTRUE( *svpp ) ) {
1468 sv_setiv( *svpp, ++trie->uniquecharcount );
1473 if( cur == first ) {
1476 } else if (chars < trie->minlen) {
1478 } else if (chars > trie->maxlen) {
1482 } /* end first pass */
1483 DEBUG_TRIE_COMPILE_r(
1484 PerlIO_printf( Perl_debug_log, "%*sTRIE(%s): W:%d C:%d Uq:%d Min:%d Max:%d\n",
1485 (int)depth * 2 + 2,"",
1486 ( widecharmap ? "UTF8" : "NATIVE" ), (int)word_count,
1487 (int)TRIE_CHARCOUNT(trie), trie->uniquecharcount,
1488 (int)trie->minlen, (int)trie->maxlen )
1490 trie->wordlen = (U32 *) PerlMemShared_calloc( word_count, sizeof(U32) );
1493 We now know what we are dealing with in terms of unique chars and
1494 string sizes so we can calculate how much memory a naive
1495 representation using a flat table will take. If it's over a reasonable
1496 limit (as specified by ${^RE_TRIE_MAXBUF}) we use a more memory
1497 conservative but potentially much slower representation using an array
1500 At the end we convert both representations into the same compressed
1501 form that will be used in regexec.c for matching with. The latter
1502 is a form that cannot be used to construct with but has memory
1503 properties similar to the list form and access properties similar
1504 to the table form making it both suitable for fast searches and
1505 small enough that its feasable to store for the duration of a program.
1507 See the comment in the code where the compressed table is produced
1508 inplace from the flat tabe representation for an explanation of how
1509 the compression works.
1514 if ( (IV)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1) > SvIV(re_trie_maxbuff) ) {
1516 Second Pass -- Array Of Lists Representation
1518 Each state will be represented by a list of charid:state records
1519 (reg_trie_trans_le) the first such element holds the CUR and LEN
1520 points of the allocated array. (See defines above).
1522 We build the initial structure using the lists, and then convert
1523 it into the compressed table form which allows faster lookups
1524 (but cant be modified once converted).
1527 STRLEN transcount = 1;
1529 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1530 "%*sCompiling trie using list compiler\n",
1531 (int)depth * 2 + 2, ""));
1533 trie->states = (reg_trie_state *)
1534 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1535 sizeof(reg_trie_state) );
1539 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1541 regnode * const noper = NEXTOPER( cur );
1542 U8 *uc = (U8*)STRING( noper );
1543 const U8 * const e = uc + STR_LEN( noper );
1544 U32 state = 1; /* required init */
1545 U16 charid = 0; /* sanity init */
1546 U8 *scan = (U8*)NULL; /* sanity init */
1547 STRLEN foldlen = 0; /* required init */
1548 U32 wordlen = 0; /* required init */
1549 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1551 if (OP(noper) != NOTHING) {
1552 for ( ; uc < e ; uc += len ) {
1557 charid = trie->charmap[ uvc ];
1559 SV** const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1563 charid=(U16)SvIV( *svpp );
1566 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1573 if ( !trie->states[ state ].trans.list ) {
1574 TRIE_LIST_NEW( state );
1576 for ( check = 1; check <= TRIE_LIST_USED( state ); check++ ) {
1577 if ( TRIE_LIST_ITEM( state, check ).forid == charid ) {
1578 newstate = TRIE_LIST_ITEM( state, check ).newstate;
1583 newstate = next_alloc++;
1584 TRIE_LIST_PUSH( state, charid, newstate );
1589 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1593 TRIE_HANDLE_WORD(state);
1595 } /* end second pass */
1597 /* next alloc is the NEXT state to be allocated */
1598 trie->statecount = next_alloc;
1599 trie->states = (reg_trie_state *)
1600 PerlMemShared_realloc( trie->states,
1602 * sizeof(reg_trie_state) );
1604 /* and now dump it out before we compress it */
1605 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_list(trie, widecharmap,
1606 revcharmap, next_alloc,
1610 trie->trans = (reg_trie_trans *)
1611 PerlMemShared_calloc( transcount, sizeof(reg_trie_trans) );
1618 for( state=1 ; state < next_alloc ; state ++ ) {
1622 DEBUG_TRIE_COMPILE_MORE_r(
1623 PerlIO_printf( Perl_debug_log, "tp: %d zp: %d ",tp,zp)
1627 if (trie->states[state].trans.list) {
1628 U16 minid=TRIE_LIST_ITEM( state, 1).forid;
1632 for( idx = 2 ; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1633 const U16 forid = TRIE_LIST_ITEM( state, idx).forid;
1634 if ( forid < minid ) {
1636 } else if ( forid > maxid ) {
1640 if ( transcount < tp + maxid - minid + 1) {
1642 trie->trans = (reg_trie_trans *)
1643 PerlMemShared_realloc( trie->trans,
1645 * sizeof(reg_trie_trans) );
1646 Zero( trie->trans + (transcount / 2), transcount / 2 , reg_trie_trans );
1648 base = trie->uniquecharcount + tp - minid;
1649 if ( maxid == minid ) {
1651 for ( ; zp < tp ; zp++ ) {
1652 if ( ! trie->trans[ zp ].next ) {
1653 base = trie->uniquecharcount + zp - minid;
1654 trie->trans[ zp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1655 trie->trans[ zp ].check = state;
1661 trie->trans[ tp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1662 trie->trans[ tp ].check = state;
1667 for ( idx=1; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1668 const U32 tid = base - trie->uniquecharcount + TRIE_LIST_ITEM( state, idx ).forid;
1669 trie->trans[ tid ].next = TRIE_LIST_ITEM( state, idx ).newstate;
1670 trie->trans[ tid ].check = state;
1672 tp += ( maxid - minid + 1 );
1674 Safefree(trie->states[ state ].trans.list);
1677 DEBUG_TRIE_COMPILE_MORE_r(
1678 PerlIO_printf( Perl_debug_log, " base: %d\n",base);
1681 trie->states[ state ].trans.base=base;
1683 trie->lasttrans = tp + 1;
1687 Second Pass -- Flat Table Representation.
1689 we dont use the 0 slot of either trans[] or states[] so we add 1 to each.
1690 We know that we will need Charcount+1 trans at most to store the data
1691 (one row per char at worst case) So we preallocate both structures
1692 assuming worst case.
1694 We then construct the trie using only the .next slots of the entry
1697 We use the .check field of the first entry of the node temporarily to
1698 make compression both faster and easier by keeping track of how many non
1699 zero fields are in the node.
1701 Since trans are numbered from 1 any 0 pointer in the table is a FAIL
1704 There are two terms at use here: state as a TRIE_NODEIDX() which is a
1705 number representing the first entry of the node, and state as a
1706 TRIE_NODENUM() which is the trans number. state 1 is TRIE_NODEIDX(1) and
1707 TRIE_NODENUM(1), state 2 is TRIE_NODEIDX(2) and TRIE_NODENUM(3) if there
1708 are 2 entrys per node. eg:
1716 The table is internally in the right hand, idx form. However as we also
1717 have to deal with the states array which is indexed by nodenum we have to
1718 use TRIE_NODENUM() to convert.
1721 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1722 "%*sCompiling trie using table compiler\n",
1723 (int)depth * 2 + 2, ""));
1725 trie->trans = (reg_trie_trans *)
1726 PerlMemShared_calloc( ( TRIE_CHARCOUNT(trie) + 1 )
1727 * trie->uniquecharcount + 1,
1728 sizeof(reg_trie_trans) );
1729 trie->states = (reg_trie_state *)
1730 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1731 sizeof(reg_trie_state) );
1732 next_alloc = trie->uniquecharcount + 1;
1735 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1737 regnode * const noper = NEXTOPER( cur );
1738 const U8 *uc = (U8*)STRING( noper );
1739 const U8 * const e = uc + STR_LEN( noper );
1741 U32 state = 1; /* required init */
1743 U16 charid = 0; /* sanity init */
1744 U32 accept_state = 0; /* sanity init */
1745 U8 *scan = (U8*)NULL; /* sanity init */
1747 STRLEN foldlen = 0; /* required init */
1748 U32 wordlen = 0; /* required init */
1749 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1751 if ( OP(noper) != NOTHING ) {
1752 for ( ; uc < e ; uc += len ) {
1757 charid = trie->charmap[ uvc ];
1759 SV* const * const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1760 charid = svpp ? (U16)SvIV(*svpp) : 0;
1764 if ( !trie->trans[ state + charid ].next ) {
1765 trie->trans[ state + charid ].next = next_alloc;
1766 trie->trans[ state ].check++;
1767 next_alloc += trie->uniquecharcount;
1769 state = trie->trans[ state + charid ].next;
1771 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1773 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1776 accept_state = TRIE_NODENUM( state );
1777 TRIE_HANDLE_WORD(accept_state);
1779 } /* end second pass */
1781 /* and now dump it out before we compress it */
1782 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_table(trie, widecharmap,
1784 next_alloc, depth+1));
1788 * Inplace compress the table.*
1790 For sparse data sets the table constructed by the trie algorithm will
1791 be mostly 0/FAIL transitions or to put it another way mostly empty.
1792 (Note that leaf nodes will not contain any transitions.)
1794 This algorithm compresses the tables by eliminating most such
1795 transitions, at the cost of a modest bit of extra work during lookup:
1797 - Each states[] entry contains a .base field which indicates the
1798 index in the state[] array wheres its transition data is stored.
1800 - If .base is 0 there are no valid transitions from that node.
1802 - If .base is nonzero then charid is added to it to find an entry in
1805 -If trans[states[state].base+charid].check!=state then the
1806 transition is taken to be a 0/Fail transition. Thus if there are fail
1807 transitions at the front of the node then the .base offset will point
1808 somewhere inside the previous nodes data (or maybe even into a node
1809 even earlier), but the .check field determines if the transition is
1813 The following process inplace converts the table to the compressed
1814 table: We first do not compress the root node 1,and mark its all its
1815 .check pointers as 1 and set its .base pointer as 1 as well. This
1816 allows to do a DFA construction from the compressed table later, and
1817 ensures that any .base pointers we calculate later are greater than
1820 - We set 'pos' to indicate the first entry of the second node.
1822 - We then iterate over the columns of the node, finding the first and
1823 last used entry at l and m. We then copy l..m into pos..(pos+m-l),
1824 and set the .check pointers accordingly, and advance pos
1825 appropriately and repreat for the next node. Note that when we copy
1826 the next pointers we have to convert them from the original
1827 NODEIDX form to NODENUM form as the former is not valid post
1830 - If a node has no transitions used we mark its base as 0 and do not
1831 advance the pos pointer.
1833 - If a node only has one transition we use a second pointer into the
1834 structure to fill in allocated fail transitions from other states.
1835 This pointer is independent of the main pointer and scans forward
1836 looking for null transitions that are allocated to a state. When it
1837 finds one it writes the single transition into the "hole". If the
1838 pointer doesnt find one the single transition is appended as normal.
1840 - Once compressed we can Renew/realloc the structures to release the
1843 See "Table-Compression Methods" in sec 3.9 of the Red Dragon,
1844 specifically Fig 3.47 and the associated pseudocode.
1848 const U32 laststate = TRIE_NODENUM( next_alloc );
1851 trie->statecount = laststate;
1853 for ( state = 1 ; state < laststate ; state++ ) {
1855 const U32 stateidx = TRIE_NODEIDX( state );
1856 const U32 o_used = trie->trans[ stateidx ].check;
1857 U32 used = trie->trans[ stateidx ].check;
1858 trie->trans[ stateidx ].check = 0;
1860 for ( charid = 0 ; used && charid < trie->uniquecharcount ; charid++ ) {
1861 if ( flag || trie->trans[ stateidx + charid ].next ) {
1862 if ( trie->trans[ stateidx + charid ].next ) {
1864 for ( ; zp < pos ; zp++ ) {
1865 if ( ! trie->trans[ zp ].next ) {
1869 trie->states[ state ].trans.base = zp + trie->uniquecharcount - charid ;
1870 trie->trans[ zp ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1871 trie->trans[ zp ].check = state;
1872 if ( ++zp > pos ) pos = zp;
1879 trie->states[ state ].trans.base = pos + trie->uniquecharcount - charid ;
1881 trie->trans[ pos ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1882 trie->trans[ pos ].check = state;
1887 trie->lasttrans = pos + 1;
1888 trie->states = (reg_trie_state *)
1889 PerlMemShared_realloc( trie->states, laststate
1890 * sizeof(reg_trie_state) );
1891 DEBUG_TRIE_COMPILE_MORE_r(
1892 PerlIO_printf( Perl_debug_log,
1893 "%*sAlloc: %d Orig: %"IVdf" elements, Final:%"IVdf". Savings of %%%5.2f\n",
1894 (int)depth * 2 + 2,"",
1895 (int)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1 ),
1898 ( ( next_alloc - pos ) * 100 ) / (double)next_alloc );
1901 } /* end table compress */
1903 DEBUG_TRIE_COMPILE_MORE_r(
1904 PerlIO_printf(Perl_debug_log, "%*sStatecount:%"UVxf" Lasttrans:%"UVxf"\n",
1905 (int)depth * 2 + 2, "",
1906 (UV)trie->statecount,
1907 (UV)trie->lasttrans)
1909 /* resize the trans array to remove unused space */
1910 trie->trans = (reg_trie_trans *)
1911 PerlMemShared_realloc( trie->trans, trie->lasttrans
1912 * sizeof(reg_trie_trans) );
1914 /* and now dump out the compressed format */
1915 DEBUG_TRIE_COMPILE_r(dump_trie(trie, widecharmap, revcharmap, depth+1));
1917 { /* Modify the program and insert the new TRIE node*/
1918 U8 nodetype =(U8)(flags & 0xFF);
1922 regnode *optimize = NULL;
1923 #ifdef RE_TRACK_PATTERN_OFFSETS
1926 U32 mjd_nodelen = 0;
1927 #endif /* RE_TRACK_PATTERN_OFFSETS */
1928 #endif /* DEBUGGING */
1930 This means we convert either the first branch or the first Exact,
1931 depending on whether the thing following (in 'last') is a branch
1932 or not and whther first is the startbranch (ie is it a sub part of
1933 the alternation or is it the whole thing.)
1934 Assuming its a sub part we conver the EXACT otherwise we convert
1935 the whole branch sequence, including the first.
1937 /* Find the node we are going to overwrite */
1938 if ( first != startbranch || OP( last ) == BRANCH ) {
1939 /* branch sub-chain */
1940 NEXT_OFF( first ) = (U16)(last - first);
1941 #ifdef RE_TRACK_PATTERN_OFFSETS
1943 mjd_offset= Node_Offset((convert));
1944 mjd_nodelen= Node_Length((convert));
1947 /* whole branch chain */
1949 #ifdef RE_TRACK_PATTERN_OFFSETS
1952 const regnode *nop = NEXTOPER( convert );
1953 mjd_offset= Node_Offset((nop));
1954 mjd_nodelen= Node_Length((nop));
1958 PerlIO_printf(Perl_debug_log, "%*sMJD offset:%"UVuf" MJD length:%"UVuf"\n",
1959 (int)depth * 2 + 2, "",
1960 (UV)mjd_offset, (UV)mjd_nodelen)
1963 /* But first we check to see if there is a common prefix we can
1964 split out as an EXACT and put in front of the TRIE node. */
1965 trie->startstate= 1;
1966 if ( trie->bitmap && !widecharmap && !trie->jump ) {
1968 for ( state = 1 ; state < trie->statecount-1 ; state++ ) {
1972 const U32 base = trie->states[ state ].trans.base;
1974 if ( trie->states[state].wordnum )
1977 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
1978 if ( ( base + ofs >= trie->uniquecharcount ) &&
1979 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
1980 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
1982 if ( ++count > 1 ) {
1983 SV **tmp = av_fetch( revcharmap, ofs, 0);
1984 const U8 *ch = (U8*)SvPV_nolen_const( *tmp );
1985 if ( state == 1 ) break;
1987 Zero(trie->bitmap, ANYOF_BITMAP_SIZE, char);
1989 PerlIO_printf(Perl_debug_log,
1990 "%*sNew Start State=%"UVuf" Class: [",
1991 (int)depth * 2 + 2, "",
1994 SV ** const tmp = av_fetch( revcharmap, idx, 0);
1995 const U8 * const ch = (U8*)SvPV_nolen_const( *tmp );
1997 TRIE_BITMAP_SET(trie,*ch);
1999 TRIE_BITMAP_SET(trie, folder[ *ch ]);
2001 PerlIO_printf(Perl_debug_log, "%s", (char*)ch)
2005 TRIE_BITMAP_SET(trie,*ch);
2007 TRIE_BITMAP_SET(trie,folder[ *ch ]);
2008 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"%s", ch));
2014 SV **tmp = av_fetch( revcharmap, idx, 0);
2016 char *ch = SvPV( *tmp, len );
2018 SV *sv=sv_newmortal();
2019 PerlIO_printf( Perl_debug_log,
2020 "%*sPrefix State: %"UVuf" Idx:%"UVuf" Char='%s'\n",
2021 (int)depth * 2 + 2, "",
2023 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 6,
2024 PL_colors[0], PL_colors[1],
2025 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
2026 PERL_PV_ESCAPE_FIRSTCHAR
2031 OP( convert ) = nodetype;
2032 str=STRING(convert);
2035 STR_LEN(convert) += len;
2041 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"]\n"));
2047 regnode *n = convert+NODE_SZ_STR(convert);
2048 NEXT_OFF(convert) = NODE_SZ_STR(convert);
2049 trie->startstate = state;
2050 trie->minlen -= (state - 1);
2051 trie->maxlen -= (state - 1);
2053 /* At least the UNICOS C compiler choked on this
2054 * being argument to DEBUG_r(), so let's just have
2057 #ifdef PERL_EXT_RE_BUILD
2063 regnode *fix = convert;
2064 U32 word = trie->wordcount;
2066 Set_Node_Offset_Length(convert, mjd_offset, state - 1);
2067 while( ++fix < n ) {
2068 Set_Node_Offset_Length(fix, 0, 0);
2071 SV ** const tmp = av_fetch( trie_words, word, 0 );
2073 if ( STR_LEN(convert) <= SvCUR(*tmp) )
2074 sv_chop(*tmp, SvPV_nolen(*tmp) + STR_LEN(convert));
2076 sv_chop(*tmp, SvPV_nolen(*tmp) + SvCUR(*tmp));
2084 NEXT_OFF(convert) = (U16)(tail - convert);
2085 DEBUG_r(optimize= n);
2091 if ( trie->maxlen ) {
2092 NEXT_OFF( convert ) = (U16)(tail - convert);
2093 ARG_SET( convert, data_slot );
2094 /* Store the offset to the first unabsorbed branch in
2095 jump[0], which is otherwise unused by the jump logic.
2096 We use this when dumping a trie and during optimisation. */
2098 trie->jump[0] = (U16)(nextbranch - convert);
2101 if ( !trie->states[trie->startstate].wordnum && trie->bitmap &&
2102 ( (char *)jumper - (char *)convert) >= (int)sizeof(struct regnode_charclass) )
2104 OP( convert ) = TRIEC;
2105 Copy(trie->bitmap, ((struct regnode_charclass *)convert)->bitmap, ANYOF_BITMAP_SIZE, char);
2106 PerlMemShared_free(trie->bitmap);
2109 OP( convert ) = TRIE;
2111 /* store the type in the flags */
2112 convert->flags = nodetype;
2116 + regarglen[ OP( convert ) ];
2118 /* XXX We really should free up the resource in trie now,
2119 as we won't use them - (which resources?) dmq */
2121 /* needed for dumping*/
2122 DEBUG_r(if (optimize) {
2123 regnode *opt = convert;
2125 while ( ++opt < optimize) {
2126 Set_Node_Offset_Length(opt,0,0);
2129 Try to clean up some of the debris left after the
2132 while( optimize < jumper ) {
2133 mjd_nodelen += Node_Length((optimize));
2134 OP( optimize ) = OPTIMIZED;
2135 Set_Node_Offset_Length(optimize,0,0);
2138 Set_Node_Offset_Length(convert,mjd_offset,mjd_nodelen);
2140 } /* end node insert */
2141 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, convert);
2142 RExC_rxi->data->data[ data_slot + 1 ] = (void*)widecharmap;
2144 RExC_rxi->data->data[ data_slot + TRIE_WORDS_OFFSET ] = (void*)trie_words;
2145 RExC_rxi->data->data[ data_slot + 3 ] = (void*)revcharmap;
2147 SvREFCNT_dec(revcharmap);
2151 : trie->startstate>1
2157 S_make_trie_failtable(pTHX_ RExC_state_t *pRExC_state, regnode *source, regnode *stclass, U32 depth)
2159 /* The Trie is constructed and compressed now so we can build a fail array now if its needed
2161 This is basically the Aho-Corasick algorithm. Its from exercise 3.31 and 3.32 in the
2162 "Red Dragon" -- Compilers, principles, techniques, and tools. Aho, Sethi, Ullman 1985/88
2165 We find the fail state for each state in the trie, this state is the longest proper
2166 suffix of the current states 'word' that is also a proper prefix of another word in our
2167 trie. State 1 represents the word '' and is the thus the default fail state. This allows
2168 the DFA not to have to restart after its tried and failed a word at a given point, it
2169 simply continues as though it had been matching the other word in the first place.
2171 'abcdgu'=~/abcdefg|cdgu/
2172 When we get to 'd' we are still matching the first word, we would encounter 'g' which would
2173 fail, which would bring use to the state representing 'd' in the second word where we would
2174 try 'g' and succeed, prodceding to match 'cdgu'.
2176 /* add a fail transition */
2177 const U32 trie_offset = ARG(source);
2178 reg_trie_data *trie=(reg_trie_data *)RExC_rxi->data->data[trie_offset];
2180 const U32 ucharcount = trie->uniquecharcount;
2181 const U32 numstates = trie->statecount;
2182 const U32 ubound = trie->lasttrans + ucharcount;
2186 U32 base = trie->states[ 1 ].trans.base;
2189 const U32 data_slot = add_data( pRExC_state, 1, "T" );
2190 GET_RE_DEBUG_FLAGS_DECL;
2192 PERL_ARGS_ASSERT_MAKE_TRIE_FAILTABLE;
2194 PERL_UNUSED_ARG(depth);
2198 ARG_SET( stclass, data_slot );
2199 aho = (reg_ac_data *) PerlMemShared_calloc( 1, sizeof(reg_ac_data) );
2200 RExC_rxi->data->data[ data_slot ] = (void*)aho;
2201 aho->trie=trie_offset;
2202 aho->states=(reg_trie_state *)PerlMemShared_malloc( numstates * sizeof(reg_trie_state) );
2203 Copy( trie->states, aho->states, numstates, reg_trie_state );
2204 Newxz( q, numstates, U32);
2205 aho->fail = (U32 *) PerlMemShared_calloc( numstates, sizeof(U32) );
2208 /* initialize fail[0..1] to be 1 so that we always have
2209 a valid final fail state */
2210 fail[ 0 ] = fail[ 1 ] = 1;
2212 for ( charid = 0; charid < ucharcount ; charid++ ) {
2213 const U32 newstate = TRIE_TRANS_STATE( 1, base, ucharcount, charid, 0 );
2215 q[ q_write ] = newstate;
2216 /* set to point at the root */
2217 fail[ q[ q_write++ ] ]=1;
2220 while ( q_read < q_write) {
2221 const U32 cur = q[ q_read++ % numstates ];
2222 base = trie->states[ cur ].trans.base;
2224 for ( charid = 0 ; charid < ucharcount ; charid++ ) {
2225 const U32 ch_state = TRIE_TRANS_STATE( cur, base, ucharcount, charid, 1 );
2227 U32 fail_state = cur;
2230 fail_state = fail[ fail_state ];
2231 fail_base = aho->states[ fail_state ].trans.base;
2232 } while ( !TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 ) );
2234 fail_state = TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 );
2235 fail[ ch_state ] = fail_state;
2236 if ( !aho->states[ ch_state ].wordnum && aho->states[ fail_state ].wordnum )
2238 aho->states[ ch_state ].wordnum = aho->states[ fail_state ].wordnum;
2240 q[ q_write++ % numstates] = ch_state;
2244 /* restore fail[0..1] to 0 so that we "fall out" of the AC loop
2245 when we fail in state 1, this allows us to use the
2246 charclass scan to find a valid start char. This is based on the principle
2247 that theres a good chance the string being searched contains lots of stuff
2248 that cant be a start char.
2250 fail[ 0 ] = fail[ 1 ] = 0;
2251 DEBUG_TRIE_COMPILE_r({
2252 PerlIO_printf(Perl_debug_log,
2253 "%*sStclass Failtable (%"UVuf" states): 0",
2254 (int)(depth * 2), "", (UV)numstates
2256 for( q_read=1; q_read<numstates; q_read++ ) {
2257 PerlIO_printf(Perl_debug_log, ", %"UVuf, (UV)fail[q_read]);
2259 PerlIO_printf(Perl_debug_log, "\n");
2262 /*RExC_seen |= REG_SEEN_TRIEDFA;*/
2267 * There are strange code-generation bugs caused on sparc64 by gcc-2.95.2.
2268 * These need to be revisited when a newer toolchain becomes available.
2270 #if defined(__sparc64__) && defined(__GNUC__)
2271 # if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
2272 # undef SPARC64_GCC_WORKAROUND
2273 # define SPARC64_GCC_WORKAROUND 1
2277 #define DEBUG_PEEP(str,scan,depth) \
2278 DEBUG_OPTIMISE_r({if (scan){ \
2279 SV * const mysv=sv_newmortal(); \
2280 regnode *Next = regnext(scan); \
2281 regprop(RExC_rx, mysv, scan); \
2282 PerlIO_printf(Perl_debug_log, "%*s" str ">%3d: %s (%d)\n", \
2283 (int)depth*2, "", REG_NODE_NUM(scan), SvPV_nolen_const(mysv),\
2284 Next ? (REG_NODE_NUM(Next)) : 0 ); \
2291 #define JOIN_EXACT(scan,min,flags) \
2292 if (PL_regkind[OP(scan)] == EXACT) \
2293 join_exact(pRExC_state,(scan),(min),(flags),NULL,depth+1)
2296 S_join_exact(pTHX_ RExC_state_t *pRExC_state, regnode *scan, I32 *min, U32 flags,regnode *val, U32 depth) {
2297 /* Merge several consecutive EXACTish nodes into one. */
2298 regnode *n = regnext(scan);
2300 regnode *next = scan + NODE_SZ_STR(scan);
2304 regnode *stop = scan;
2305 GET_RE_DEBUG_FLAGS_DECL;
2307 PERL_UNUSED_ARG(depth);
2310 PERL_ARGS_ASSERT_JOIN_EXACT;
2311 #ifndef EXPERIMENTAL_INPLACESCAN
2312 PERL_UNUSED_ARG(flags);
2313 PERL_UNUSED_ARG(val);
2315 DEBUG_PEEP("join",scan,depth);
2317 /* Skip NOTHING, merge EXACT*. */
2319 ( PL_regkind[OP(n)] == NOTHING ||
2320 (stringok && (OP(n) == OP(scan))))
2322 && NEXT_OFF(scan) + NEXT_OFF(n) < I16_MAX) {
2324 if (OP(n) == TAIL || n > next)
2326 if (PL_regkind[OP(n)] == NOTHING) {
2327 DEBUG_PEEP("skip:",n,depth);
2328 NEXT_OFF(scan) += NEXT_OFF(n);
2329 next = n + NODE_STEP_REGNODE;
2336 else if (stringok) {
2337 const unsigned int oldl = STR_LEN(scan);
2338 regnode * const nnext = regnext(n);
2340 DEBUG_PEEP("merg",n,depth);
2343 if (oldl + STR_LEN(n) > U8_MAX)
2345 NEXT_OFF(scan) += NEXT_OFF(n);
2346 STR_LEN(scan) += STR_LEN(n);
2347 next = n + NODE_SZ_STR(n);
2348 /* Now we can overwrite *n : */
2349 Move(STRING(n), STRING(scan) + oldl, STR_LEN(n), char);
2357 #ifdef EXPERIMENTAL_INPLACESCAN
2358 if (flags && !NEXT_OFF(n)) {
2359 DEBUG_PEEP("atch", val, depth);
2360 if (reg_off_by_arg[OP(n)]) {
2361 ARG_SET(n, val - n);
2364 NEXT_OFF(n) = val - n;
2371 if (UTF && ( OP(scan) == EXACTF ) && ( STR_LEN(scan) >= 6 ) ) {
2373 Two problematic code points in Unicode casefolding of EXACT nodes:
2375 U+0390 - GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
2376 U+03B0 - GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
2382 U+03B9 U+0308 U+0301 0xCE 0xB9 0xCC 0x88 0xCC 0x81
2383 U+03C5 U+0308 U+0301 0xCF 0x85 0xCC 0x88 0xCC 0x81
2385 This means that in case-insensitive matching (or "loose matching",
2386 as Unicode calls it), an EXACTF of length six (the UTF-8 encoded byte
2387 length of the above casefolded versions) can match a target string
2388 of length two (the byte length of UTF-8 encoded U+0390 or U+03B0).
2389 This would rather mess up the minimum length computation.
2391 What we'll do is to look for the tail four bytes, and then peek
2392 at the preceding two bytes to see whether we need to decrease
2393 the minimum length by four (six minus two).
2395 Thanks to the design of UTF-8, there cannot be false matches:
2396 A sequence of valid UTF-8 bytes cannot be a subsequence of
2397 another valid sequence of UTF-8 bytes.
2400 char * const s0 = STRING(scan), *s, *t;
2401 char * const s1 = s0 + STR_LEN(scan) - 1;
2402 char * const s2 = s1 - 4;
2403 #ifdef EBCDIC /* RD tunifold greek 0390 and 03B0 */
2404 const char t0[] = "\xaf\x49\xaf\x42";
2406 const char t0[] = "\xcc\x88\xcc\x81";
2408 const char * const t1 = t0 + 3;
2411 s < s2 && (t = ninstr(s, s1, t0, t1));
2414 if (((U8)t[-1] == 0x68 && (U8)t[-2] == 0xB4) ||
2415 ((U8)t[-1] == 0x46 && (U8)t[-2] == 0xB5))
2417 if (((U8)t[-1] == 0xB9 && (U8)t[-2] == 0xCE) ||
2418 ((U8)t[-1] == 0x85 && (U8)t[-2] == 0xCF))
2426 n = scan + NODE_SZ_STR(scan);
2428 if (PL_regkind[OP(n)] != NOTHING || OP(n) == NOTHING) {
2435 DEBUG_OPTIMISE_r(if (merged){DEBUG_PEEP("finl",scan,depth)});
2439 /* REx optimizer. Converts nodes into quickier variants "in place".
2440 Finds fixed substrings. */
2442 /* Stops at toplevel WHILEM as well as at "last". At end *scanp is set
2443 to the position after last scanned or to NULL. */
2445 #define INIT_AND_WITHP \
2446 assert(!and_withp); \
2447 Newx(and_withp,1,struct regnode_charclass_class); \
2448 SAVEFREEPV(and_withp)
2450 /* this is a chain of data about sub patterns we are processing that
2451 need to be handled seperately/specially in study_chunk. Its so
2452 we can simulate recursion without losing state. */
2454 typedef struct scan_frame {
2455 regnode *last; /* last node to process in this frame */
2456 regnode *next; /* next node to process when last is reached */
2457 struct scan_frame *prev; /*previous frame*/
2458 I32 stop; /* what stopparen do we use */
2462 #define SCAN_COMMIT(s, data, m) scan_commit(s, data, m, is_inf)
2464 #define CASE_SYNST_FNC(nAmE) \
2466 if (flags & SCF_DO_STCLASS_AND) { \
2467 for (value = 0; value < 256; value++) \
2468 if (!is_ ## nAmE ## _cp(value)) \
2469 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2472 for (value = 0; value < 256; value++) \
2473 if (is_ ## nAmE ## _cp(value)) \
2474 ANYOF_BITMAP_SET(data->start_class, value); \
2478 if (flags & SCF_DO_STCLASS_AND) { \
2479 for (value = 0; value < 256; value++) \
2480 if (is_ ## nAmE ## _cp(value)) \
2481 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2484 for (value = 0; value < 256; value++) \
2485 if (!is_ ## nAmE ## _cp(value)) \
2486 ANYOF_BITMAP_SET(data->start_class, value); \
2493 S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
2494 I32 *minlenp, I32 *deltap,
2499 struct regnode_charclass_class *and_withp,
2500 U32 flags, U32 depth)
2501 /* scanp: Start here (read-write). */
2502 /* deltap: Write maxlen-minlen here. */
2503 /* last: Stop before this one. */
2504 /* data: string data about the pattern */
2505 /* stopparen: treat close N as END */
2506 /* recursed: which subroutines have we recursed into */
2507 /* and_withp: Valid if flags & SCF_DO_STCLASS_OR */
2510 I32 min = 0, pars = 0, code;
2511 regnode *scan = *scanp, *next;
2513 int is_inf = (flags & SCF_DO_SUBSTR) && (data->flags & SF_IS_INF);
2514 int is_inf_internal = 0; /* The studied chunk is infinite */
2515 I32 is_par = OP(scan) == OPEN ? ARG(scan) : 0;
2516 scan_data_t data_fake;
2517 SV *re_trie_maxbuff = NULL;
2518 regnode *first_non_open = scan;
2519 I32 stopmin = I32_MAX;
2520 scan_frame *frame = NULL;
2521 GET_RE_DEBUG_FLAGS_DECL;
2523 PERL_ARGS_ASSERT_STUDY_CHUNK;
2526 StructCopy(&zero_scan_data, &data_fake, scan_data_t);
2530 while (first_non_open && OP(first_non_open) == OPEN)
2531 first_non_open=regnext(first_non_open);
2536 while ( scan && OP(scan) != END && scan < last ){
2537 /* Peephole optimizer: */
2538 DEBUG_STUDYDATA("Peep:", data,depth);
2539 DEBUG_PEEP("Peep",scan,depth);
2540 JOIN_EXACT(scan,&min,0);
2542 /* Follow the next-chain of the current node and optimize
2543 away all the NOTHINGs from it. */
2544 if (OP(scan) != CURLYX) {
2545 const int max = (reg_off_by_arg[OP(scan)]
2547 /* I32 may be smaller than U16 on CRAYs! */
2548 : (I32_MAX < U16_MAX ? I32_MAX : U16_MAX));
2549 int off = (reg_off_by_arg[OP(scan)] ? ARG(scan) : NEXT_OFF(scan));
2553 /* Skip NOTHING and LONGJMP. */
2554 while ((n = regnext(n))
2555 && ((PL_regkind[OP(n)] == NOTHING && (noff = NEXT_OFF(n)))
2556 || ((OP(n) == LONGJMP) && (noff = ARG(n))))
2557 && off + noff < max)
2559 if (reg_off_by_arg[OP(scan)])
2562 NEXT_OFF(scan) = off;
2567 /* The principal pseudo-switch. Cannot be a switch, since we
2568 look into several different things. */
2569 if (OP(scan) == BRANCH || OP(scan) == BRANCHJ
2570 || OP(scan) == IFTHEN) {
2571 next = regnext(scan);
2573 /* demq: the op(next)==code check is to see if we have "branch-branch" AFAICT */
2575 if (OP(next) == code || code == IFTHEN) {
2576 /* NOTE - There is similar code to this block below for handling
2577 TRIE nodes on a re-study. If you change stuff here check there
2579 I32 max1 = 0, min1 = I32_MAX, num = 0;
2580 struct regnode_charclass_class accum;
2581 regnode * const startbranch=scan;
2583 if (flags & SCF_DO_SUBSTR)
2584 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot merge strings after this. */
2585 if (flags & SCF_DO_STCLASS)
2586 cl_init_zero(pRExC_state, &accum);
2588 while (OP(scan) == code) {
2589 I32 deltanext, minnext, f = 0, fake;
2590 struct regnode_charclass_class this_class;
2593 data_fake.flags = 0;
2595 data_fake.whilem_c = data->whilem_c;
2596 data_fake.last_closep = data->last_closep;
2599 data_fake.last_closep = &fake;
2601 data_fake.pos_delta = delta;
2602 next = regnext(scan);
2603 scan = NEXTOPER(scan);
2605 scan = NEXTOPER(scan);
2606 if (flags & SCF_DO_STCLASS) {
2607 cl_init(pRExC_state, &this_class);
2608 data_fake.start_class = &this_class;
2609 f = SCF_DO_STCLASS_AND;
2611 if (flags & SCF_WHILEM_VISITED_POS)
2612 f |= SCF_WHILEM_VISITED_POS;
2614 /* we suppose the run is continuous, last=next...*/
2615 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
2617 stopparen, recursed, NULL, f,depth+1);
2620 if (max1 < minnext + deltanext)
2621 max1 = minnext + deltanext;
2622 if (deltanext == I32_MAX)
2623 is_inf = is_inf_internal = 1;
2625 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
2627 if (data_fake.flags & SCF_SEEN_ACCEPT) {
2628 if ( stopmin > minnext)
2629 stopmin = min + min1;
2630 flags &= ~SCF_DO_SUBSTR;
2632 data->flags |= SCF_SEEN_ACCEPT;
2635 if (data_fake.flags & SF_HAS_EVAL)
2636 data->flags |= SF_HAS_EVAL;
2637 data->whilem_c = data_fake.whilem_c;
2639 if (flags & SCF_DO_STCLASS)
2640 cl_or(pRExC_state, &accum, &this_class);
2642 if (code == IFTHEN && num < 2) /* Empty ELSE branch */
2644 if (flags & SCF_DO_SUBSTR) {
2645 data->pos_min += min1;
2646 data->pos_delta += max1 - min1;
2647 if (max1 != min1 || is_inf)
2648 data->longest = &(data->longest_float);
2651 delta += max1 - min1;
2652 if (flags & SCF_DO_STCLASS_OR) {
2653 cl_or(pRExC_state, data->start_class, &accum);
2655 cl_and(data->start_class, and_withp);
2656 flags &= ~SCF_DO_STCLASS;
2659 else if (flags & SCF_DO_STCLASS_AND) {
2661 cl_and(data->start_class, &accum);
2662 flags &= ~SCF_DO_STCLASS;
2665 /* Switch to OR mode: cache the old value of
2666 * data->start_class */
2668 StructCopy(data->start_class, and_withp,
2669 struct regnode_charclass_class);
2670 flags &= ~SCF_DO_STCLASS_AND;
2671 StructCopy(&accum, data->start_class,
2672 struct regnode_charclass_class);
2673 flags |= SCF_DO_STCLASS_OR;
2674 data->start_class->flags |= ANYOF_EOS;
2678 if (PERL_ENABLE_TRIE_OPTIMISATION && OP( startbranch ) == BRANCH ) {
2681 Assuming this was/is a branch we are dealing with: 'scan' now
2682 points at the item that follows the branch sequence, whatever
2683 it is. We now start at the beginning of the sequence and look
2690 which would be constructed from a pattern like /A|LIST|OF|WORDS/
2692 If we can find such a subseqence we need to turn the first
2693 element into a trie and then add the subsequent branch exact
2694 strings to the trie.
2698 1. patterns where the whole set of branch can be converted.
2700 2. patterns where only a subset can be converted.
2702 In case 1 we can replace the whole set with a single regop
2703 for the trie. In case 2 we need to keep the start and end
2706 'BRANCH EXACT; BRANCH EXACT; BRANCH X'
2707 becomes BRANCH TRIE; BRANCH X;
2709 There is an additional case, that being where there is a
2710 common prefix, which gets split out into an EXACT like node
2711 preceding the TRIE node.
2713 If x(1..n)==tail then we can do a simple trie, if not we make
2714 a "jump" trie, such that when we match the appropriate word
2715 we "jump" to the appopriate tail node. Essentailly we turn
2716 a nested if into a case structure of sorts.
2721 if (!re_trie_maxbuff) {
2722 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
2723 if (!SvIOK(re_trie_maxbuff))
2724 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
2726 if ( SvIV(re_trie_maxbuff)>=0 ) {
2728 regnode *first = (regnode *)NULL;
2729 regnode *last = (regnode *)NULL;
2730 regnode *tail = scan;
2735 SV * const mysv = sv_newmortal(); /* for dumping */
2737 /* var tail is used because there may be a TAIL
2738 regop in the way. Ie, the exacts will point to the
2739 thing following the TAIL, but the last branch will
2740 point at the TAIL. So we advance tail. If we
2741 have nested (?:) we may have to move through several
2745 while ( OP( tail ) == TAIL ) {
2746 /* this is the TAIL generated by (?:) */
2747 tail = regnext( tail );
2752 regprop(RExC_rx, mysv, tail );
2753 PerlIO_printf( Perl_debug_log, "%*s%s%s\n",
2754 (int)depth * 2 + 2, "",
2755 "Looking for TRIE'able sequences. Tail node is: ",
2756 SvPV_nolen_const( mysv )
2762 step through the branches, cur represents each
2763 branch, noper is the first thing to be matched
2764 as part of that branch and noper_next is the
2765 regnext() of that node. if noper is an EXACT
2766 and noper_next is the same as scan (our current
2767 position in the regex) then the EXACT branch is
2768 a possible optimization target. Once we have
2769 two or more consequetive such branches we can
2770 create a trie of the EXACT's contents and stich
2771 it in place. If the sequence represents all of
2772 the branches we eliminate the whole thing and
2773 replace it with a single TRIE. If it is a
2774 subsequence then we need to stitch it in. This
2775 means the first branch has to remain, and needs
2776 to be repointed at the item on the branch chain
2777 following the last branch optimized. This could
2778 be either a BRANCH, in which case the
2779 subsequence is internal, or it could be the
2780 item following the branch sequence in which
2781 case the subsequence is at the end.
2785 /* dont use tail as the end marker for this traverse */
2786 for ( cur = startbranch ; cur != scan ; cur = regnext( cur ) ) {
2787 regnode * const noper = NEXTOPER( cur );
2788 #if defined(DEBUGGING) || defined(NOJUMPTRIE)
2789 regnode * const noper_next = regnext( noper );
2793 regprop(RExC_rx, mysv, cur);
2794 PerlIO_printf( Perl_debug_log, "%*s- %s (%d)",
2795 (int)depth * 2 + 2,"", SvPV_nolen_const( mysv ), REG_NODE_NUM(cur) );
2797 regprop(RExC_rx, mysv, noper);
2798 PerlIO_printf( Perl_debug_log, " -> %s",
2799 SvPV_nolen_const(mysv));
2802 regprop(RExC_rx, mysv, noper_next );
2803 PerlIO_printf( Perl_debug_log,"\t=> %s\t",
2804 SvPV_nolen_const(mysv));
2806 PerlIO_printf( Perl_debug_log, "(First==%d,Last==%d,Cur==%d)\n",
2807 REG_NODE_NUM(first), REG_NODE_NUM(last), REG_NODE_NUM(cur) );
2809 if ( (((first && optype!=NOTHING) ? OP( noper ) == optype
2810 : PL_regkind[ OP( noper ) ] == EXACT )
2811 || OP(noper) == NOTHING )
2813 && noper_next == tail
2818 if ( !first || optype == NOTHING ) {
2819 if (!first) first = cur;
2820 optype = OP( noper );
2826 Currently we assume that the trie can handle unicode and ascii
2827 matches fold cased matches. If this proves true then the following
2828 define will prevent tries in this situation.
2830 #define TRIE_TYPE_IS_SAFE (UTF || optype==EXACT)
2832 #define TRIE_TYPE_IS_SAFE 1
2833 if ( last && TRIE_TYPE_IS_SAFE ) {
2834 make_trie( pRExC_state,
2835 startbranch, first, cur, tail, count,
2838 if ( PL_regkind[ OP( noper ) ] == EXACT
2840 && noper_next == tail
2845 optype = OP( noper );
2855 regprop(RExC_rx, mysv, cur);
2856 PerlIO_printf( Perl_debug_log,
2857 "%*s- %s (%d) <SCAN FINISHED>\n", (int)depth * 2 + 2,
2858 "", SvPV_nolen_const( mysv ),REG_NODE_NUM(cur));
2862 if ( last && TRIE_TYPE_IS_SAFE ) {
2863 made= make_trie( pRExC_state, startbranch, first, scan, tail, count, optype, depth+1 );
2864 #ifdef TRIE_STUDY_OPT
2865 if ( ((made == MADE_EXACT_TRIE &&
2866 startbranch == first)
2867 || ( first_non_open == first )) &&
2869 flags |= SCF_TRIE_RESTUDY;
2870 if ( startbranch == first
2873 RExC_seen &=~REG_TOP_LEVEL_BRANCHES;
2883 else if ( code == BRANCHJ ) { /* single branch is optimized. */
2884 scan = NEXTOPER(NEXTOPER(scan));
2885 } else /* single branch is optimized. */
2886 scan = NEXTOPER(scan);
2888 } else if (OP(scan) == SUSPEND || OP(scan) == GOSUB || OP(scan) == GOSTART) {
2889 scan_frame *newframe = NULL;
2894 if (OP(scan) != SUSPEND) {
2895 /* set the pointer */
2896 if (OP(scan) == GOSUB) {
2898 RExC_recurse[ARG2L(scan)] = scan;
2899 start = RExC_open_parens[paren-1];
2900 end = RExC_close_parens[paren-1];
2903 start = RExC_rxi->program + 1;
2907 Newxz(recursed, (((RExC_npar)>>3) +1), U8);
2908 SAVEFREEPV(recursed);
2910 if (!PAREN_TEST(recursed,paren+1)) {
2911 PAREN_SET(recursed,paren+1);
2912 Newx(newframe,1,scan_frame);
2914 if (flags & SCF_DO_SUBSTR) {
2915 SCAN_COMMIT(pRExC_state,data,minlenp);
2916 data->longest = &(data->longest_float);
2918 is_inf = is_inf_internal = 1;
2919 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
2920 cl_anything(pRExC_state, data->start_class);
2921 flags &= ~SCF_DO_STCLASS;
2924 Newx(newframe,1,scan_frame);
2927 end = regnext(scan);
2932 SAVEFREEPV(newframe);
2933 newframe->next = regnext(scan);
2934 newframe->last = last;
2935 newframe->stop = stopparen;
2936 newframe->prev = frame;
2946 else if (OP(scan) == EXACT) {
2947 I32 l = STR_LEN(scan);
2950 const U8 * const s = (U8*)STRING(scan);
2951 l = utf8_length(s, s + l);
2952 uc = utf8_to_uvchr(s, NULL);
2954 uc = *((U8*)STRING(scan));
2957 if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
2958 /* The code below prefers earlier match for fixed
2959 offset, later match for variable offset. */
2960 if (data->last_end == -1) { /* Update the start info. */
2961 data->last_start_min = data->pos_min;
2962 data->last_start_max = is_inf
2963 ? I32_MAX : data->pos_min + data->pos_delta;
2965 sv_catpvn(data->last_found, STRING(scan), STR_LEN(scan));
2967 SvUTF8_on(data->last_found);
2969 SV * const sv = data->last_found;
2970 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
2971 mg_find(sv, PERL_MAGIC_utf8) : NULL;
2972 if (mg && mg->mg_len >= 0)
2973 mg->mg_len += utf8_length((U8*)STRING(scan),
2974 (U8*)STRING(scan)+STR_LEN(scan));
2976 data->last_end = data->pos_min + l;
2977 data->pos_min += l; /* As in the first entry. */
2978 data->flags &= ~SF_BEFORE_EOL;
2980 if (flags & SCF_DO_STCLASS_AND) {
2981 /* Check whether it is compatible with what we know already! */
2985 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
2986 && !ANYOF_BITMAP_TEST(data->start_class, uc)
2987 && (!(data->start_class->flags & ANYOF_FOLD)
2988 || !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
2991 ANYOF_CLASS_ZERO(data->start_class);
2992 ANYOF_BITMAP_ZERO(data->start_class);
2994 ANYOF_BITMAP_SET(data->start_class, uc);
2995 data->start_class->flags &= ~ANYOF_EOS;
2997 data->start_class->flags &= ~ANYOF_UNICODE_ALL;
2999 else if (flags & SCF_DO_STCLASS_OR) {
3000 /* false positive possible if the class is case-folded */
3002 ANYOF_BITMAP_SET(data->start_class, uc);
3004 data->start_class->flags |= ANYOF_UNICODE_ALL;
3005 data->start_class->flags &= ~ANYOF_EOS;
3006 cl_and(data->start_class, and_withp);
3008 flags &= ~SCF_DO_STCLASS;
3010 else if (PL_regkind[OP(scan)] == EXACT) { /* But OP != EXACT! */
3011 I32 l = STR_LEN(scan);
3012 UV uc = *((U8*)STRING(scan));
3014 /* Search for fixed substrings supports EXACT only. */
3015 if (flags & SCF_DO_SUBSTR) {
3017 SCAN_COMMIT(pRExC_state, data, minlenp);
3020 const U8 * const s = (U8 *)STRING(scan);
3021 l = utf8_length(s, s + l);
3022 uc = utf8_to_uvchr(s, NULL);
3025 if (flags & SCF_DO_SUBSTR)
3027 if (flags & SCF_DO_STCLASS_AND) {
3028 /* Check whether it is compatible with what we know already! */
3032 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3033 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3034 && !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
3036 ANYOF_CLASS_ZERO(data->start_class);
3037 ANYOF_BITMAP_ZERO(data->start_class);
3039 ANYOF_BITMAP_SET(data->start_class, uc);
3040 data->start_class->flags &= ~ANYOF_EOS;
3041 data->start_class->flags |= ANYOF_FOLD;
3042 if (OP(scan) == EXACTFL)
3043 data->start_class->flags |= ANYOF_LOCALE;
3046 else if (flags & SCF_DO_STCLASS_OR) {
3047 if (data->start_class->flags & ANYOF_FOLD) {
3048 /* false positive possible if the class is case-folded.
3049 Assume that the locale settings are the same... */
3051 ANYOF_BITMAP_SET(data->start_class, uc);
3052 data->start_class->flags &= ~ANYOF_EOS;
3054 cl_and(data->start_class, and_withp);
3056 flags &= ~SCF_DO_STCLASS;
3058 else if (strchr((const char*)PL_varies,OP(scan))) {
3059 I32 mincount, maxcount, minnext, deltanext, fl = 0;
3060 I32 f = flags, pos_before = 0;
3061 regnode * const oscan = scan;
3062 struct regnode_charclass_class this_class;
3063 struct regnode_charclass_class *oclass = NULL;
3064 I32 next_is_eval = 0;
3066 switch (PL_regkind[OP(scan)]) {
3067 case WHILEM: /* End of (?:...)* . */
3068 scan = NEXTOPER(scan);
3071 if (flags & (SCF_DO_SUBSTR | SCF_DO_STCLASS)) {
3072 next = NEXTOPER(scan);
3073 if (OP(next) == EXACT || (flags & SCF_DO_STCLASS)) {
3075 maxcount = REG_INFTY;
3076 next = regnext(scan);
3077 scan = NEXTOPER(scan);
3081 if (flags & SCF_DO_SUBSTR)
3086 if (flags & SCF_DO_STCLASS) {
3088 maxcount = REG_INFTY;
3089 next = regnext(scan);
3090 scan = NEXTOPER(scan);
3093 is_inf = is_inf_internal = 1;
3094 scan = regnext(scan);
3095 if (flags & SCF_DO_SUBSTR) {
3096 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot extend fixed substrings */
3097 data->longest = &(data->longest_float);
3099 goto optimize_curly_tail;
3101 if (stopparen>0 && (OP(scan)==CURLYN || OP(scan)==CURLYM)
3102 && (scan->flags == stopparen))
3107 mincount = ARG1(scan);
3108 maxcount = ARG2(scan);
3110 next = regnext(scan);
3111 if (OP(scan) == CURLYX) {
3112 I32 lp = (data ? *(data->last_closep) : 0);
3113 scan->flags = ((lp <= (I32)U8_MAX) ? (U8)lp : U8_MAX);
3115 scan = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
3116 next_is_eval = (OP(scan) == EVAL);
3118 if (flags & SCF_DO_SUBSTR) {
3119 if (mincount == 0) SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot extend fixed substrings */
3120 pos_before = data->pos_min;
3124 data->flags &= ~(SF_HAS_PAR|SF_IN_PAR|SF_HAS_EVAL);
3126 data->flags |= SF_IS_INF;
3128 if (flags & SCF_DO_STCLASS) {
3129 cl_init(pRExC_state, &this_class);
3130 oclass = data->start_class;
3131 data->start_class = &this_class;
3132 f |= SCF_DO_STCLASS_AND;
3133 f &= ~SCF_DO_STCLASS_OR;
3135 /* These are the cases when once a subexpression
3136 fails at a particular position, it cannot succeed
3137 even after backtracking at the enclosing scope.
3139 XXXX what if minimal match and we are at the
3140 initial run of {n,m}? */
3141 if ((mincount != maxcount - 1) && (maxcount != REG_INFTY))
3142 f &= ~SCF_WHILEM_VISITED_POS;
3144 /* This will finish on WHILEM, setting scan, or on NULL: */
3145 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
3146 last, data, stopparen, recursed, NULL,
3148 ? (f & ~SCF_DO_SUBSTR) : f),depth+1);
3150 if (flags & SCF_DO_STCLASS)
3151 data->start_class = oclass;
3152 if (mincount == 0 || minnext == 0) {
3153 if (flags & SCF_DO_STCLASS_OR) {
3154 cl_or(pRExC_state, data->start_class, &this_class);
3156 else if (flags & SCF_DO_STCLASS_AND) {
3157 /* Switch to OR mode: cache the old value of
3158 * data->start_class */
3160 StructCopy(data->start_class, and_withp,
3161 struct regnode_charclass_class);
3162 flags &= ~SCF_DO_STCLASS_AND;
3163 StructCopy(&this_class, data->start_class,
3164 struct regnode_charclass_class);
3165 flags |= SCF_DO_STCLASS_OR;
3166 data->start_class->flags |= ANYOF_EOS;
3168 } else { /* Non-zero len */
3169 if (flags & SCF_DO_STCLASS_OR) {
3170 cl_or(pRExC_state, data->start_class, &this_class);
3171 cl_and(data->start_class, and_withp);
3173 else if (flags & SCF_DO_STCLASS_AND)
3174 cl_and(data->start_class, &this_class);
3175 flags &= ~SCF_DO_STCLASS;
3177 if (!scan) /* It was not CURLYX, but CURLY. */
3179 if ( /* ? quantifier ok, except for (?{ ... }) */
3180 (next_is_eval || !(mincount == 0 && maxcount == 1))
3181 && (minnext == 0) && (deltanext == 0)
3182 && data && !(data->flags & (SF_HAS_PAR|SF_IN_PAR))
3183 && maxcount <= REG_INFTY/3 /* Complement check for big count */
3184 && ckWARN(WARN_REGEXP))
3187 "Quantifier unexpected on zero-length expression");
3190 min += minnext * mincount;
3191 is_inf_internal |= ((maxcount == REG_INFTY
3192 && (minnext + deltanext) > 0)
3193 || deltanext == I32_MAX);
3194 is_inf |= is_inf_internal;
3195 delta += (minnext + deltanext) * maxcount - minnext * mincount;
3197 /* Try powerful optimization CURLYX => CURLYN. */
3198 if ( OP(oscan) == CURLYX && data
3199 && data->flags & SF_IN_PAR
3200 && !(data->flags & SF_HAS_EVAL)
3201 && !deltanext && minnext == 1 ) {
3202 /* Try to optimize to CURLYN. */
3203 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS;
3204 regnode * const nxt1 = nxt;
3211 if (!strchr((const char*)PL_simple,OP(nxt))
3212 && !(PL_regkind[OP(nxt)] == EXACT
3213 && STR_LEN(nxt) == 1))
3219 if (OP(nxt) != CLOSE)
3221 if (RExC_open_parens) {
3222 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3223 RExC_close_parens[ARG(nxt1)-1]=nxt+2; /*close->while*/
3225 /* Now we know that nxt2 is the only contents: */
3226 oscan->flags = (U8)ARG(nxt);
3228 OP(nxt1) = NOTHING; /* was OPEN. */
3231 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3232 NEXT_OFF(nxt1+ 1) = 0; /* just for consistancy. */
3233 NEXT_OFF(nxt2) = 0; /* just for consistancy with CURLY. */
3234 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3235 OP(nxt + 1) = OPTIMIZED; /* was count. */
3236 NEXT_OFF(nxt+ 1) = 0; /* just for consistancy. */
3241 /* Try optimization CURLYX => CURLYM. */
3242 if ( OP(oscan) == CURLYX && data
3243 && !(data->flags & SF_HAS_PAR)
3244 && !(data->flags & SF_HAS_EVAL)
3245 && !deltanext /* atom is fixed width */
3246 && minnext != 0 /* CURLYM can't handle zero width */
3248 /* XXXX How to optimize if data == 0? */
3249 /* Optimize to a simpler form. */
3250 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN */
3254 while ( (nxt2 = regnext(nxt)) /* skip over embedded stuff*/
3255 && (OP(nxt2) != WHILEM))
3257 OP(nxt2) = SUCCEED; /* Whas WHILEM */
3258 /* Need to optimize away parenths. */
3259 if (data->flags & SF_IN_PAR) {
3260 /* Set the parenth number. */
3261 regnode *nxt1 = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN*/
3263 if (OP(nxt) != CLOSE)
3264 FAIL("Panic opt close");
3265 oscan->flags = (U8)ARG(nxt);
3266 if (RExC_open_parens) {
3267 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3268 RExC_close_parens[ARG(nxt1)-1]=nxt2+1; /*close->NOTHING*/
3270 OP(nxt1) = OPTIMIZED; /* was OPEN. */
3271 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3274 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3275 OP(nxt + 1) = OPTIMIZED; /* was count. */
3276 NEXT_OFF(nxt1 + 1) = 0; /* just for consistancy. */
3277 NEXT_OFF(nxt + 1) = 0; /* just for consistancy. */
3280 while ( nxt1 && (OP(nxt1) != WHILEM)) {
3281 regnode *nnxt = regnext(nxt1);
3284 if (reg_off_by_arg[OP(nxt1)])
3285 ARG_SET(nxt1, nxt2 - nxt1);
3286 else if (nxt2 - nxt1 < U16_MAX)
3287 NEXT_OFF(nxt1) = nxt2 - nxt1;
3289 OP(nxt) = NOTHING; /* Cannot beautify */
3294 /* Optimize again: */
3295 study_chunk(pRExC_state, &nxt1, minlenp, &deltanext, nxt,
3296 NULL, stopparen, recursed, NULL, 0,depth+1);
3301 else if ((OP(oscan) == CURLYX)
3302 && (flags & SCF_WHILEM_VISITED_POS)
3303 /* See the comment on a similar expression above.
3304 However, this time it not a subexpression
3305 we care about, but the expression itself. */
3306 && (maxcount == REG_INFTY)
3307 && data && ++data->whilem_c < 16) {
3308 /* This stays as CURLYX, we can put the count/of pair. */
3309 /* Find WHILEM (as in regexec.c) */
3310 regnode *nxt = oscan + NEXT_OFF(oscan);
3312 if (OP(PREVOPER(nxt)) == NOTHING) /* LONGJMP */
3314 PREVOPER(nxt)->flags = (U8)(data->whilem_c
3315 | (RExC_whilem_seen << 4)); /* On WHILEM */
3317 if (data && fl & (SF_HAS_PAR|SF_IN_PAR))
3319 if (flags & SCF_DO_SUBSTR) {
3320 SV *last_str = NULL;
3321 int counted = mincount != 0;
3323 if (data->last_end > 0 && mincount != 0) { /* Ends with a string. */
3324 #if defined(SPARC64_GCC_WORKAROUND)
3327 const char *s = NULL;
3330 if (pos_before >= data->last_start_min)
3333 b = data->last_start_min;
3336 s = SvPV_const(data->last_found, l);
3337 old = b - data->last_start_min;
3340 I32 b = pos_before >= data->last_start_min
3341 ? pos_before : data->last_start_min;
3343 const char * const s = SvPV_const(data->last_found, l);
3344 I32 old = b - data->last_start_min;
3348 old = utf8_hop((U8*)s, old) - (U8*)s;
3351 /* Get the added string: */
3352 last_str = newSVpvn_utf8(s + old, l, UTF);
3353 if (deltanext == 0 && pos_before == b) {
3354 /* What was added is a constant string */
3356 SvGROW(last_str, (mincount * l) + 1);
3357 repeatcpy(SvPVX(last_str) + l,
3358 SvPVX_const(last_str), l, mincount - 1);
3359 SvCUR_set(last_str, SvCUR(last_str) * mincount);
3360 /* Add additional parts. */
3361 SvCUR_set(data->last_found,
3362 SvCUR(data->last_found) - l);
3363 sv_catsv(data->last_found, last_str);
3365 SV * sv = data->last_found;
3367 SvUTF8(sv) && SvMAGICAL(sv) ?
3368 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3369 if (mg && mg->mg_len >= 0)
3370 mg->mg_len += CHR_SVLEN(last_str) - l;
3372 data->last_end += l * (mincount - 1);
3375 /* start offset must point into the last copy */
3376 data->last_start_min += minnext * (mincount - 1);
3377 data->last_start_max += is_inf ? I32_MAX
3378 : (maxcount - 1) * (minnext + data->pos_delta);
3381 /* It is counted once already... */
3382 data->pos_min += minnext * (mincount - counted);
3383 data->pos_delta += - counted * deltanext +
3384 (minnext + deltanext) * maxcount - minnext * mincount;
3385 if (mincount != maxcount) {
3386 /* Cannot extend fixed substrings found inside
3388 SCAN_COMMIT(pRExC_state,data,minlenp);
3389 if (mincount && last_str) {
3390 SV * const sv = data->last_found;
3391 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3392 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3396 sv_setsv(sv, last_str);
3397 data->last_end = data->pos_min;
3398 data->last_start_min =
3399 data->pos_min - CHR_SVLEN(last_str);
3400 data->last_start_max = is_inf
3402 : data->pos_min + data->pos_delta
3403 - CHR_SVLEN(last_str);
3405 data->longest = &(data->longest_float);
3407 SvREFCNT_dec(last_str);
3409 if (data && (fl & SF_HAS_EVAL))
3410 data->flags |= SF_HAS_EVAL;
3411 optimize_curly_tail:
3412 if (OP(oscan) != CURLYX) {
3413 while (PL_regkind[OP(next = regnext(oscan))] == NOTHING
3415 NEXT_OFF(oscan) += NEXT_OFF(next);
3418 default: /* REF and CLUMP only? */
3419 if (flags & SCF_DO_SUBSTR) {
3420 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3421 data->longest = &(data->longest_float);
3423 is_inf = is_inf_internal = 1;
3424 if (flags & SCF_DO_STCLASS_OR)
3425 cl_anything(pRExC_state, data->start_class);
3426 flags &= ~SCF_DO_STCLASS;
3430 else if (OP(scan) == LNBREAK) {
3431 if (flags & SCF_DO_STCLASS) {
3433 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3434 if (flags & SCF_DO_STCLASS_AND) {
3435 for (value = 0; value < 256; value++)
3436 if (!is_VERTWS_cp(value))
3437 ANYOF_BITMAP_CLEAR(data->start_class, value);
3440 for (value = 0; value < 256; value++)
3441 if (is_VERTWS_cp(value))
3442 ANYOF_BITMAP_SET(data->start_class, value);
3444 if (flags & SCF_DO_STCLASS_OR)
3445 cl_and(data->start_class, and_withp);
3446 flags &= ~SCF_DO_STCLASS;
3450 if (flags & SCF_DO_SUBSTR) {
3451 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3453 data->pos_delta += 1;
3454 data->longest = &(data->longest_float);
3458 else if (OP(scan) == FOLDCHAR) {
3459 int d = ARG(scan)==0xDF ? 1 : 2;
3460 flags &= ~SCF_DO_STCLASS;
3463 if (flags & SCF_DO_SUBSTR) {
3464 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3466 data->pos_delta += d;
3467 data->longest = &(data->longest_float);
3470 else if (strchr((const char*)PL_simple,OP(scan))) {
3473 if (flags & SCF_DO_SUBSTR) {
3474 SCAN_COMMIT(pRExC_state,data,minlenp);
3478 if (flags & SCF_DO_STCLASS) {
3479 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3481 /* Some of the logic below assumes that switching
3482 locale on will only add false positives. */
3483 switch (PL_regkind[OP(scan)]) {
3487 /* Perl_croak(aTHX_ "panic: unexpected simple REx opcode %d", OP(scan)); */
3488 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3489 cl_anything(pRExC_state, data->start_class);
3492 if (OP(scan) == SANY)
3494 if (flags & SCF_DO_STCLASS_OR) { /* Everything but \n */
3495 value = (ANYOF_BITMAP_TEST(data->start_class,'\n')
3496 || (data->start_class->flags & ANYOF_CLASS));
3497 cl_anything(pRExC_state, data->start_class);
3499 if (flags & SCF_DO_STCLASS_AND || !value)
3500 ANYOF_BITMAP_CLEAR(data->start_class,'\n');
3503 if (flags & SCF_DO_STCLASS_AND)
3504 cl_and(data->start_class,
3505 (struct regnode_charclass_class*)scan);
3507 cl_or(pRExC_state, data->start_class,
3508 (struct regnode_charclass_class*)scan);
3511 if (flags & SCF_DO_STCLASS_AND) {
3512 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3513 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3514 for (value = 0; value < 256; value++)
3515 if (!isALNUM(value))
3516 ANYOF_BITMAP_CLEAR(data->start_class, value);
3520 if (data->start_class->flags & ANYOF_LOCALE)
3521 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3523 for (value = 0; value < 256; value++)
3525 ANYOF_BITMAP_SET(data->start_class, value);
3530 if (flags & SCF_DO_STCLASS_AND) {
3531 if (data->start_class->flags & ANYOF_LOCALE)
3532 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3535 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3536 data->start_class->flags |= ANYOF_LOCALE;
3540 if (flags & SCF_DO_STCLASS_AND) {
3541 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3542 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3543 for (value = 0; value < 256; value++)
3545 ANYOF_BITMAP_CLEAR(data->start_class, value);
3549 if (data->start_class->flags & ANYOF_LOCALE)
3550 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3552 for (value = 0; value < 256; value++)
3553 if (!isALNUM(value))
3554 ANYOF_BITMAP_SET(data->start_class, value);
3559 if (flags & SCF_DO_STCLASS_AND) {
3560 if (data->start_class->flags & ANYOF_LOCALE)
3561 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3564 data->start_class->flags |= ANYOF_LOCALE;
3565 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3569 if (flags & SCF_DO_STCLASS_AND) {
3570 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3571 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3572 for (value = 0; value < 256; value++)
3573 if (!isSPACE(value))
3574 ANYOF_BITMAP_CLEAR(data->start_class, value);
3578 if (data->start_class->flags & ANYOF_LOCALE)
3579 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3581 for (value = 0; value < 256; value++)
3583 ANYOF_BITMAP_SET(data->start_class, value);
3588 if (flags & SCF_DO_STCLASS_AND) {
3589 if (data->start_class->flags & ANYOF_LOCALE)
3590 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3593 data->start_class->flags |= ANYOF_LOCALE;
3594 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3598 if (flags & SCF_DO_STCLASS_AND) {
3599 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3600 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3601 for (value = 0; value < 256; value++)
3603 ANYOF_BITMAP_CLEAR(data->start_class, value);
3607 if (data->start_class->flags & ANYOF_LOCALE)
3608 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3610 for (value = 0; value < 256; value++)
3611 if (!isSPACE(value))
3612 ANYOF_BITMAP_SET(data->start_class, value);
3617 if (flags & SCF_DO_STCLASS_AND) {
3618 if (data->start_class->flags & ANYOF_LOCALE) {
3619 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3620 for (value = 0; value < 256; value++)
3621 if (!isSPACE(value))
3622 ANYOF_BITMAP_CLEAR(data->start_class, value);
3626 data->start_class->flags |= ANYOF_LOCALE;
3627 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3631 if (flags & SCF_DO_STCLASS_AND) {
3632 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NDIGIT);
3633 for (value = 0; value < 256; value++)
3634 if (!isDIGIT(value))
3635 ANYOF_BITMAP_CLEAR(data->start_class, value);
3638 if (data->start_class->flags & ANYOF_LOCALE)
3639 ANYOF_CLASS_SET(data->start_class,ANYOF_DIGIT);
3641 for (value = 0; value < 256; value++)
3643 ANYOF_BITMAP_SET(data->start_class, value);
3648 if (flags & SCF_DO_STCLASS_AND) {
3649 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_DIGIT);
3650 for (value = 0; value < 256; value++)
3652 ANYOF_BITMAP_CLEAR(data->start_class, value);
3655 if (data->start_class->flags & ANYOF_LOCALE)
3656 ANYOF_CLASS_SET(data->start_class,ANYOF_NDIGIT);
3658 for (value = 0; value < 256; value++)
3659 if (!isDIGIT(value))
3660 ANYOF_BITMAP_SET(data->start_class, value);
3664 CASE_SYNST_FNC(VERTWS);
3665 CASE_SYNST_FNC(HORIZWS);
3668 if (flags & SCF_DO_STCLASS_OR)
3669 cl_and(data->start_class, and_withp);
3670 flags &= ~SCF_DO_STCLASS;
3673 else if (PL_regkind[OP(scan)] == EOL && flags & SCF_DO_SUBSTR) {
3674 data->flags |= (OP(scan) == MEOL
3678 else if ( PL_regkind[OP(scan)] == BRANCHJ
3679 /* Lookbehind, or need to calculate parens/evals/stclass: */
3680 && (scan->flags || data || (flags & SCF_DO_STCLASS))
3681 && (OP(scan) == IFMATCH || OP(scan) == UNLESSM)) {
3682 if ( !PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3683 || OP(scan) == UNLESSM )
3685 /* Negative Lookahead/lookbehind
3686 In this case we can't do fixed string optimisation.
3689 I32 deltanext, minnext, fake = 0;
3691 struct regnode_charclass_class intrnl;
3694 data_fake.flags = 0;
3696 data_fake.whilem_c = data->whilem_c;
3697 data_fake.last_closep = data->last_closep;
3700 data_fake.last_closep = &fake;
3701 data_fake.pos_delta = delta;
3702 if ( flags & SCF_DO_STCLASS && !scan->flags
3703 && OP(scan) == IFMATCH ) { /* Lookahead */
3704 cl_init(pRExC_state, &intrnl);
3705 data_fake.start_class = &intrnl;
3706 f |= SCF_DO_STCLASS_AND;
3708 if (flags & SCF_WHILEM_VISITED_POS)
3709 f |= SCF_WHILEM_VISITED_POS;
3710 next = regnext(scan);
3711 nscan = NEXTOPER(NEXTOPER(scan));
3712 minnext = study_chunk(pRExC_state, &nscan, minlenp, &deltanext,
3713 last, &data_fake, stopparen, recursed, NULL, f, depth+1);
3716 FAIL("Variable length lookbehind not implemented");
3718 else if (minnext > (I32)U8_MAX) {
3719 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3721 scan->flags = (U8)minnext;
3724 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3726 if (data_fake.flags & SF_HAS_EVAL)
3727 data->flags |= SF_HAS_EVAL;
3728 data->whilem_c = data_fake.whilem_c;
3730 if (f & SCF_DO_STCLASS_AND) {
3731 if (flags & SCF_DO_STCLASS_OR) {
3732 /* OR before, AND after: ideally we would recurse with
3733 * data_fake to get the AND applied by study of the
3734 * remainder of the pattern, and then derecurse;
3735 * *** HACK *** for now just treat as "no information".
3736 * See [perl #56690].
3738 cl_init(pRExC_state, data->start_class);
3740 /* AND before and after: combine and continue */
3741 const int was = (data->start_class->flags & ANYOF_EOS);
3743 cl_and(data->start_class, &intrnl);
3745 data->start_class->flags |= ANYOF_EOS;
3749 #if PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3751 /* Positive Lookahead/lookbehind
3752 In this case we can do fixed string optimisation,
3753 but we must be careful about it. Note in the case of
3754 lookbehind the positions will be offset by the minimum
3755 length of the pattern, something we won't know about
3756 until after the recurse.
3758 I32 deltanext, fake = 0;
3760 struct regnode_charclass_class intrnl;
3762 /* We use SAVEFREEPV so that when the full compile
3763 is finished perl will clean up the allocated
3764 minlens when its all done. This was we don't
3765 have to worry about freeing them when we know
3766 they wont be used, which would be a pain.
3769 Newx( minnextp, 1, I32 );
3770 SAVEFREEPV(minnextp);
3773 StructCopy(data, &data_fake, scan_data_t);
3774 if ((flags & SCF_DO_SUBSTR) && data->last_found) {
3777 SCAN_COMMIT(pRExC_state, &data_fake,minlenp);
3778 data_fake.last_found=newSVsv(data->last_found);
3782 data_fake.last_closep = &fake;
3783 data_fake.flags = 0;
3784 data_fake.pos_delta = delta;
3786 data_fake.flags |= SF_IS_INF;
3787 if ( flags & SCF_DO_STCLASS && !scan->flags
3788 && OP(scan) == IFMATCH ) { /* Lookahead */
3789 cl_init(pRExC_state, &intrnl);
3790 data_fake.start_class = &intrnl;
3791 f |= SCF_DO_STCLASS_AND;
3793 if (flags & SCF_WHILEM_VISITED_POS)
3794 f |= SCF_WHILEM_VISITED_POS;
3795 next = regnext(scan);
3796 nscan = NEXTOPER(NEXTOPER(scan));
3798 *minnextp = study_chunk(pRExC_state, &nscan, minnextp, &deltanext,
3799 last, &data_fake, stopparen, recursed, NULL, f,depth+1);
3802 FAIL("Variable length lookbehind not implemented");
3804 else if (*minnextp > (I32)U8_MAX) {
3805 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3807 scan->flags = (U8)*minnextp;
3812 if (f & SCF_DO_STCLASS_AND) {
3813 const int was = (data->start_class->flags & ANYOF_EOS);
3815 cl_and(data->start_class, &intrnl);
3817 data->start_class->flags |= ANYOF_EOS;
3820 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3822 if (data_fake.flags & SF_HAS_EVAL)
3823 data->flags |= SF_HAS_EVAL;
3824 data->whilem_c = data_fake.whilem_c;
3825 if ((flags & SCF_DO_SUBSTR) && data_fake.last_found) {
3826 if (RExC_rx->minlen<*minnextp)
3827 RExC_rx->minlen=*minnextp;
3828 SCAN_COMMIT(pRExC_state, &data_fake, minnextp);
3829 SvREFCNT_dec(data_fake.last_found);
3831 if ( data_fake.minlen_fixed != minlenp )
3833 data->offset_fixed= data_fake.offset_fixed;
3834 data->minlen_fixed= data_fake.minlen_fixed;
3835 data->lookbehind_fixed+= scan->flags;
3837 if ( data_fake.minlen_float != minlenp )
3839 data->minlen_float= data_fake.minlen_float;
3840 data->offset_float_min=data_fake.offset_float_min;
3841 data->offset_float_max=data_fake.offset_float_max;
3842 data->lookbehind_float+= scan->flags;
3851 else if (OP(scan) == OPEN) {
3852 if (stopparen != (I32)ARG(scan))
3855 else if (OP(scan) == CLOSE) {
3856 if (stopparen == (I32)ARG(scan)) {
3859 if ((I32)ARG(scan) == is_par) {
3860 next = regnext(scan);
3862 if ( next && (OP(next) != WHILEM) && next < last)
3863 is_par = 0; /* Disable optimization */
3866 *(data->last_closep) = ARG(scan);
3868 else if (OP(scan) == EVAL) {
3870 data->flags |= SF_HAS_EVAL;
3872 else if ( PL_regkind[OP(scan)] == ENDLIKE ) {
3873 if (flags & SCF_DO_SUBSTR) {
3874 SCAN_COMMIT(pRExC_state,data,minlenp);
3875 flags &= ~SCF_DO_SUBSTR;
3877 if (data && OP(scan)==ACCEPT) {
3878 data->flags |= SCF_SEEN_ACCEPT;
3883 else if (OP(scan) == LOGICAL && scan->flags == 2) /* Embedded follows */
3885 if (flags & SCF_DO_SUBSTR) {
3886 SCAN_COMMIT(pRExC_state,data,minlenp);
3887 data->longest = &(data->longest_float);
3889 is_inf = is_inf_internal = 1;
3890 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3891 cl_anything(pRExC_state, data->start_class);
3892 flags &= ~SCF_DO_STCLASS;
3894 else if (OP(scan) == GPOS) {
3895 if (!(RExC_rx->extflags & RXf_GPOS_FLOAT) &&
3896 !(delta || is_inf || (data && data->pos_delta)))
3898 if (!(RExC_rx->extflags & RXf_ANCH) && (flags & SCF_DO_SUBSTR))
3899 RExC_rx->extflags |= RXf_ANCH_GPOS;
3900 if (RExC_rx->gofs < (U32)min)
3901 RExC_rx->gofs = min;
3903 RExC_rx->extflags |= RXf_GPOS_FLOAT;
3907 #ifdef TRIE_STUDY_OPT
3908 #ifdef FULL_TRIE_STUDY
3909 else if (PL_regkind[OP(scan)] == TRIE) {
3910 /* NOTE - There is similar code to this block above for handling
3911 BRANCH nodes on the initial study. If you change stuff here
3913 regnode *trie_node= scan;
3914 regnode *tail= regnext(scan);
3915 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
3916 I32 max1 = 0, min1 = I32_MAX;
3917 struct regnode_charclass_class accum;
3919 if (flags & SCF_DO_SUBSTR) /* XXXX Add !SUSPEND? */
3920 SCAN_COMMIT(pRExC_state, data,minlenp); /* Cannot merge strings after this. */
3921 if (flags & SCF_DO_STCLASS)
3922 cl_init_zero(pRExC_state, &accum);
3928 const regnode *nextbranch= NULL;
3931 for ( word=1 ; word <= trie->wordcount ; word++)
3933 I32 deltanext=0, minnext=0, f = 0, fake;
3934 struct regnode_charclass_class this_class;
3936 data_fake.flags = 0;
3938 data_fake.whilem_c = data->whilem_c;
3939 data_fake.last_closep = data->last_closep;
3942 data_fake.last_closep = &fake;
3943 data_fake.pos_delta = delta;
3944 if (flags & SCF_DO_STCLASS) {
3945 cl_init(pRExC_state, &this_class);
3946 data_fake.start_class = &this_class;
3947 f = SCF_DO_STCLASS_AND;
3949 if (flags & SCF_WHILEM_VISITED_POS)
3950 f |= SCF_WHILEM_VISITED_POS;
3952 if (trie->jump[word]) {
3954 nextbranch = trie_node + trie->jump[0];
3955 scan= trie_node + trie->jump[word];
3956 /* We go from the jump point to the branch that follows
3957 it. Note this means we need the vestigal unused branches
3958 even though they arent otherwise used.
3960 minnext = study_chunk(pRExC_state, &scan, minlenp,
3961 &deltanext, (regnode *)nextbranch, &data_fake,
3962 stopparen, recursed, NULL, f,depth+1);
3964 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
3965 nextbranch= regnext((regnode*)nextbranch);
3967 if (min1 > (I32)(minnext + trie->minlen))
3968 min1 = minnext + trie->minlen;
3969 if (max1 < (I32)(minnext + deltanext + trie->maxlen))
3970 max1 = minnext + deltanext + trie->maxlen;
3971 if (deltanext == I32_MAX)
3972 is_inf = is_inf_internal = 1;
3974 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3976 if (data_fake.flags & SCF_SEEN_ACCEPT) {
3977 if ( stopmin > min + min1)
3978 stopmin = min + min1;
3979 flags &= ~SCF_DO_SUBSTR;
3981 data->flags |= SCF_SEEN_ACCEPT;
3984 if (data_fake.flags & SF_HAS_EVAL)
3985 data->flags |= SF_HAS_EVAL;
3986 data->whilem_c = data_fake.whilem_c;
3988 if (flags & SCF_DO_STCLASS)
3989 cl_or(pRExC_state, &accum, &this_class);
3992 if (flags & SCF_DO_SUBSTR) {
3993 data->pos_min += min1;
3994 data->pos_delta += max1 - min1;
3995 if (max1 != min1 || is_inf)
3996 data->longest = &(data->longest_float);
3999 delta += max1 - min1;
4000 if (flags & SCF_DO_STCLASS_OR) {
4001 cl_or(pRExC_state, data->start_class, &accum);
4003 cl_and(data->start_class, and_withp);
4004 flags &= ~SCF_DO_STCLASS;
4007 else if (flags & SCF_DO_STCLASS_AND) {
4009 cl_and(data->start_class, &accum);
4010 flags &= ~SCF_DO_STCLASS;
4013 /* Switch to OR mode: cache the old value of
4014 * data->start_class */
4016 StructCopy(data->start_class, and_withp,
4017 struct regnode_charclass_class);
4018 flags &= ~SCF_DO_STCLASS_AND;
4019 StructCopy(&accum, data->start_class,
4020 struct regnode_charclass_class);
4021 flags |= SCF_DO_STCLASS_OR;
4022 data->start_class->flags |= ANYOF_EOS;
4029 else if (PL_regkind[OP(scan)] == TRIE) {
4030 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
4033 min += trie->minlen;
4034 delta += (trie->maxlen - trie->minlen);
4035 flags &= ~SCF_DO_STCLASS; /* xxx */
4036 if (flags & SCF_DO_SUBSTR) {
4037 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
4038 data->pos_min += trie->minlen;
4039 data->pos_delta += (trie->maxlen - trie->minlen);
4040 if (trie->maxlen != trie->minlen)
4041 data->longest = &(data->longest_float);
4043 if (trie->jump) /* no more substrings -- for now /grr*/
4044 flags &= ~SCF_DO_SUBSTR;
4046 #endif /* old or new */
4047 #endif /* TRIE_STUDY_OPT */
4049 /* Else: zero-length, ignore. */
4050 scan = regnext(scan);
4055 stopparen = frame->stop;
4056 frame = frame->prev;
4057 goto fake_study_recurse;
4062 DEBUG_STUDYDATA("pre-fin:",data,depth);
4065 *deltap = is_inf_internal ? I32_MAX : delta;
4066 if (flags & SCF_DO_SUBSTR && is_inf)
4067 data->pos_delta = I32_MAX - data->pos_min;
4068 if (is_par > (I32)U8_MAX)
4070 if (is_par && pars==1 && data) {
4071 data->flags |= SF_IN_PAR;
4072 data->flags &= ~SF_HAS_PAR;
4074 else if (pars && data) {
4075 data->flags |= SF_HAS_PAR;
4076 data->flags &= ~SF_IN_PAR;
4078 if (flags & SCF_DO_STCLASS_OR)
4079 cl_and(data->start_class, and_withp);
4080 if (flags & SCF_TRIE_RESTUDY)
4081 data->flags |= SCF_TRIE_RESTUDY;
4083 DEBUG_STUDYDATA("post-fin:",data,depth);
4085 return min < stopmin ? min : stopmin;
4089 S_add_data(RExC_state_t *pRExC_state, U32 n, const char *s)
4091 U32 count = RExC_rxi->data ? RExC_rxi->data->count : 0;
4093 PERL_ARGS_ASSERT_ADD_DATA;
4095 Renewc(RExC_rxi->data,
4096 sizeof(*RExC_rxi->data) + sizeof(void*) * (count + n - 1),
4097 char, struct reg_data);
4099 Renew(RExC_rxi->data->what, count + n, U8);
4101 Newx(RExC_rxi->data->what, n, U8);
4102 RExC_rxi->data->count = count + n;
4103 Copy(s, RExC_rxi->data->what + count, n, U8);
4107 /*XXX: todo make this not included in a non debugging perl */
4108 #ifndef PERL_IN_XSUB_RE
4110 Perl_reginitcolors(pTHX)
4113 const char * const s = PerlEnv_getenv("PERL_RE_COLORS");
4115 char *t = savepv(s);
4119 t = strchr(t, '\t');
4125 PL_colors[i] = t = (char *)"";
4130 PL_colors[i++] = (char *)"";
4137 #ifdef TRIE_STUDY_OPT
4138 #define CHECK_RESTUDY_GOTO \
4140 (data.flags & SCF_TRIE_RESTUDY) \
4144 #define CHECK_RESTUDY_GOTO
4148 - pregcomp - compile a regular expression into internal code
4150 * We can't allocate space until we know how big the compiled form will be,
4151 * but we can't compile it (and thus know how big it is) until we've got a
4152 * place to put the code. So we cheat: we compile it twice, once with code
4153 * generation turned off and size counting turned on, and once "for real".
4154 * This also means that we don't allocate space until we are sure that the
4155 * thing really will compile successfully, and we never have to move the
4156 * code and thus invalidate pointers into it. (Note that it has to be in
4157 * one piece because free() must be able to free it all.) [NB: not true in perl]
4159 * Beware that the optimization-preparation code in here knows about some
4160 * of the structure of the compiled regexp. [I'll say.]
4165 #ifndef PERL_IN_XSUB_RE
4166 #define RE_ENGINE_PTR &reh_regexp_engine
4168 extern const struct regexp_engine my_reg_engine;
4169 #define RE_ENGINE_PTR &my_reg_engine
4172 #ifndef PERL_IN_XSUB_RE
4174 Perl_pregcomp(pTHX_ SV * const pattern, const U32 flags)
4177 HV * const table = GvHV(PL_hintgv);
4179 PERL_ARGS_ASSERT_PREGCOMP;
4181 /* Dispatch a request to compile a regexp to correct
4184 SV **ptr= hv_fetchs(table, "regcomp", FALSE);
4185 GET_RE_DEBUG_FLAGS_DECL;
4186 if (ptr && SvIOK(*ptr) && SvIV(*ptr)) {
4187 const regexp_engine *eng=INT2PTR(regexp_engine*,SvIV(*ptr));
4189 PerlIO_printf(Perl_debug_log, "Using engine %"UVxf"\n",
4192 return CALLREGCOMP_ENG(eng, pattern, flags);
4195 return Perl_re_compile(aTHX_ pattern, flags);
4200 Perl_re_compile(pTHX_ SV * const pattern, U32 pm_flags)
4205 register regexp_internal *ri;
4207 char *exp = SvPV(pattern, plen);
4208 char* xend = exp + plen;
4215 RExC_state_t RExC_state;
4216 RExC_state_t * const pRExC_state = &RExC_state;
4217 #ifdef TRIE_STUDY_OPT
4219 RExC_state_t copyRExC_state;
4221 GET_RE_DEBUG_FLAGS_DECL;
4223 PERL_ARGS_ASSERT_RE_COMPILE;
4225 DEBUG_r(if (!PL_colorset) reginitcolors());
4227 RExC_utf8 = RExC_orig_utf8 = SvUTF8(pattern);
4230 SV *dsv= sv_newmortal();
4231 RE_PV_QUOTED_DECL(s, RExC_utf8,
4232 dsv, exp, plen, 60);
4233 PerlIO_printf(Perl_debug_log, "%sCompiling REx%s %s\n",
4234 PL_colors[4],PL_colors[5],s);
4239 RExC_flags = pm_flags;
4243 RExC_seen_zerolen = *exp == '^' ? -1 : 0;
4244 RExC_seen_evals = 0;
4247 /* First pass: determine size, legality. */
4255 RExC_emit = &PL_regdummy;
4256 RExC_whilem_seen = 0;
4257 RExC_charnames = NULL;
4258 RExC_open_parens = NULL;
4259 RExC_close_parens = NULL;
4261 RExC_paren_names = NULL;
4263 RExC_paren_name_list = NULL;
4265 RExC_recurse = NULL;
4266 RExC_recurse_count = 0;
4268 #if 0 /* REGC() is (currently) a NOP at the first pass.
4269 * Clever compilers notice this and complain. --jhi */
4270 REGC((U8)REG_MAGIC, (char*)RExC_emit);
4272 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "Starting first pass (sizing)\n"));
4273 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4274 RExC_precomp = NULL;
4277 if (RExC_utf8 && !RExC_orig_utf8) {
4278 /* It's possible to write a regexp in ascii that represents Unicode
4279 codepoints outside of the byte range, such as via \x{100}. If we
4280 detect such a sequence we have to convert the entire pattern to utf8
4281 and then recompile, as our sizing calculation will have been based
4282 on 1 byte == 1 character, but we will need to use utf8 to encode
4283 at least some part of the pattern, and therefore must convert the whole
4285 XXX: somehow figure out how to make this less expensive...
4288 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log,
4289 "UTF8 mismatch! Converting to utf8 for resizing and compile\n"));
4290 exp = (char*)Perl_bytes_to_utf8(aTHX_ (U8*)exp, &len);
4292 RExC_orig_utf8 = RExC_utf8;
4294 goto redo_first_pass;
4297 PerlIO_printf(Perl_debug_log,
4298 "Required size %"IVdf" nodes\n"
4299 "Starting second pass (creation)\n",
4302 RExC_lastparse=NULL;
4304 /* Small enough for pointer-storage convention?
4305 If extralen==0, this means that we will not need long jumps. */
4306 if (RExC_size >= 0x10000L && RExC_extralen)
4307 RExC_size += RExC_extralen;
4310 if (RExC_whilem_seen > 15)
4311 RExC_whilem_seen = 15;
4313 /* Allocate space and zero-initialize. Note, the two step process
4314 of zeroing when in debug mode, thus anything assigned has to
4315 happen after that */
4316 rx = (REGEXP*) newSV_type(SVt_REGEXP);
4317 r = (struct regexp*)SvANY(rx);
4318 Newxc(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode),
4319 char, regexp_internal);
4320 if ( r == NULL || ri == NULL )
4321 FAIL("Regexp out of space");
4323 /* avoid reading uninitialized memory in DEBUGGING code in study_chunk() */
4324 Zero(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode), char);
4326 /* bulk initialize base fields with 0. */
4327 Zero(ri, sizeof(regexp_internal), char);
4330 /* non-zero initialization begins here */
4332 r->engine= RE_ENGINE_PTR;
4333 r->extflags = pm_flags;
4335 bool has_p = ((r->extflags & RXf_PMf_KEEPCOPY) == RXf_PMf_KEEPCOPY);
4336 bool has_minus = ((r->extflags & RXf_PMf_STD_PMMOD) != RXf_PMf_STD_PMMOD);
4337 bool has_runon = ((RExC_seen & REG_SEEN_RUN_ON_COMMENT)==REG_SEEN_RUN_ON_COMMENT);
4338 U16 reganch = (U16)((r->extflags & RXf_PMf_STD_PMMOD)
4339 >> RXf_PMf_STD_PMMOD_SHIFT);
4340 const char *fptr = STD_PAT_MODS; /*"msix"*/
4342 const STRLEN wraplen = plen + has_minus + has_p + has_runon
4343 + (sizeof(STD_PAT_MODS) - 1)
4344 + (sizeof("(?:)") - 1);
4346 p = sv_grow(MUTABLE_SV(rx), wraplen + 1);
4347 SvCUR_set(rx, wraplen);
4349 SvFLAGS(rx) |= SvUTF8(pattern);
4352 *p++ = KEEPCOPY_PAT_MOD; /*'p'*/
4354 char *r = p + (sizeof(STD_PAT_MODS) - 1) + has_minus - 1;
4355 char *colon = r + 1;
4358 while((ch = *fptr++)) {
4372 Copy(RExC_precomp, p, plen, char);
4373 assert ((RX_WRAPPED(rx) - p) < 16);
4374 r->pre_prefix = p - RX_WRAPPED(rx);
4383 r->nparens = RExC_npar - 1; /* set early to validate backrefs */
4385 if (RExC_seen & REG_SEEN_RECURSE) {
4386 Newxz(RExC_open_parens, RExC_npar,regnode *);
4387 SAVEFREEPV(RExC_open_parens);
4388 Newxz(RExC_close_parens,RExC_npar,regnode *);
4389 SAVEFREEPV(RExC_close_parens);
4392 /* Useful during FAIL. */
4393 #ifdef RE_TRACK_PATTERN_OFFSETS
4394 Newxz(ri->u.offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
4395 DEBUG_OFFSETS_r(PerlIO_printf(Perl_debug_log,
4396 "%s %"UVuf" bytes for offset annotations.\n",
4397 ri->u.offsets ? "Got" : "Couldn't get",
4398 (UV)((2*RExC_size+1) * sizeof(U32))));
4400 SetProgLen(ri,RExC_size);
4404 REH_CALL_COMP_BEGIN_HOOK(pRExC_state->rx);
4406 /* Second pass: emit code. */
4407 RExC_flags = pm_flags; /* don't let top level (?i) bleed */
4412 RExC_emit_start = ri->program;
4413 RExC_emit = ri->program;
4414 RExC_emit_bound = ri->program + RExC_size + 1;
4416 /* Store the count of eval-groups for security checks: */
4417 RExC_rx->seen_evals = RExC_seen_evals;
4418 REGC((U8)REG_MAGIC, (char*) RExC_emit++);
4419 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4423 /* XXXX To minimize changes to RE engine we always allocate
4424 3-units-long substrs field. */
4425 Newx(r->substrs, 1, struct reg_substr_data);
4426 if (RExC_recurse_count) {
4427 Newxz(RExC_recurse,RExC_recurse_count,regnode *);
4428 SAVEFREEPV(RExC_recurse);
4432 r->minlen = minlen = sawplus = sawopen = 0;
4433 Zero(r->substrs, 1, struct reg_substr_data);
4435 #ifdef TRIE_STUDY_OPT
4437 StructCopy(&zero_scan_data, &data, scan_data_t);
4438 copyRExC_state = RExC_state;
4441 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log,"Restudying\n"));
4443 RExC_state = copyRExC_state;
4444 if (seen & REG_TOP_LEVEL_BRANCHES)
4445 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
4447 RExC_seen &= ~REG_TOP_LEVEL_BRANCHES;
4448 if (data.last_found) {
4449 SvREFCNT_dec(data.longest_fixed);
4450 SvREFCNT_dec(data.longest_float);
4451 SvREFCNT_dec(data.last_found);
4453 StructCopy(&zero_scan_data, &data, scan_data_t);
4456 StructCopy(&zero_scan_data, &data, scan_data_t);
4459 /* Dig out information for optimizations. */
4460 r->extflags = RExC_flags; /* was pm_op */
4461 /*dmq: removed as part of de-PMOP: pm->op_pmflags = RExC_flags; */
4464 SvUTF8_on(rx); /* Unicode in it? */
4465 ri->regstclass = NULL;
4466 if (RExC_naughty >= 10) /* Probably an expensive pattern. */
4467 r->intflags |= PREGf_NAUGHTY;
4468 scan = ri->program + 1; /* First BRANCH. */
4470 /* testing for BRANCH here tells us whether there is "must appear"
4471 data in the pattern. If there is then we can use it for optimisations */
4472 if (!(RExC_seen & REG_TOP_LEVEL_BRANCHES)) { /* Only one top-level choice. */
4474 STRLEN longest_float_length, longest_fixed_length;
4475 struct regnode_charclass_class ch_class; /* pointed to by data */
4477 I32 last_close = 0; /* pointed to by data */
4478 regnode *first= scan;
4479 regnode *first_next= regnext(first);
4482 * Skip introductions and multiplicators >= 1
4483 * so that we can extract the 'meat' of the pattern that must
4484 * match in the large if() sequence following.
4485 * NOTE that EXACT is NOT covered here, as it is normally
4486 * picked up by the optimiser separately.
4488 * This is unfortunate as the optimiser isnt handling lookahead
4489 * properly currently.
4492 while ((OP(first) == OPEN && (sawopen = 1)) ||
4493 /* An OR of *one* alternative - should not happen now. */
4494 (OP(first) == BRANCH && OP(first_next) != BRANCH) ||
4495 /* for now we can't handle lookbehind IFMATCH*/
4496 (OP(first) == IFMATCH && !first->flags) ||
4497 (OP(first) == PLUS) ||
4498 (OP(first) == MINMOD) ||
4499 /* An {n,m} with n>0 */
4500 (PL_regkind[OP(first)] == CURLY && ARG1(first) > 0) ||
4501 (OP(first) == NOTHING && PL_regkind[OP(first_next)] != END ))
4504 * the only op that could be a regnode is PLUS, all the rest
4505 * will be regnode_1 or regnode_2.
4508 if (OP(first) == PLUS)
4511 first += regarglen[OP(first)];
4513 first = NEXTOPER(first);
4514 first_next= regnext(first);
4517 /* Starting-point info. */
4519 DEBUG_PEEP("first:",first,0);
4520 /* Ignore EXACT as we deal with it later. */
4521 if (PL_regkind[OP(first)] == EXACT) {
4522 if (OP(first) == EXACT)
4523 NOOP; /* Empty, get anchored substr later. */
4524 else if ((OP(first) == EXACTF || OP(first) == EXACTFL))
4525 ri->regstclass = first;
4528 else if (PL_regkind[OP(first)] == TRIE &&
4529 ((reg_trie_data *)ri->data->data[ ARG(first) ])->minlen>0)
4532 /* this can happen only on restudy */
4533 if ( OP(first) == TRIE ) {
4534 struct regnode_1 *trieop = (struct regnode_1 *)
4535 PerlMemShared_calloc(1, sizeof(struct regnode_1));
4536 StructCopy(first,trieop,struct regnode_1);
4537 trie_op=(regnode *)trieop;
4539 struct regnode_charclass *trieop = (struct regnode_charclass *)
4540 PerlMemShared_calloc(1, sizeof(struct regnode_charclass));
4541 StructCopy(first,trieop,struct regnode_charclass);
4542 trie_op=(regnode *)trieop;
4545 make_trie_failtable(pRExC_state, (regnode *)first, trie_op, 0);
4546 ri->regstclass = trie_op;
4549 else if (strchr((const char*)PL_simple,OP(first)))
4550 ri->regstclass = first;
4551 else if (PL_regkind[OP(first)] == BOUND ||
4552 PL_regkind[OP(first)] == NBOUND)
4553 ri->regstclass = first;
4554 else if (PL_regkind[OP(first)] == BOL) {
4555 r->extflags |= (OP(first) == MBOL
4557 : (OP(first) == SBOL
4560 first = NEXTOPER(first);
4563 else if (OP(first) == GPOS) {
4564 r->extflags |= RXf_ANCH_GPOS;
4565 first = NEXTOPER(first);
4568 else if ((!sawopen || !RExC_sawback) &&
4569 (OP(first) == STAR &&
4570 PL_regkind[OP(NEXTOPER(first))] == REG_ANY) &&
4571 !(r->extflags & RXf_ANCH) && !(RExC_seen & REG_SEEN_EVAL))
4573 /* turn .* into ^.* with an implied $*=1 */
4575 (OP(NEXTOPER(first)) == REG_ANY)
4578 r->extflags |= type;
4579 r->intflags |= PREGf_IMPLICIT;
4580 first = NEXTOPER(first);
4583 if (sawplus && (!sawopen || !RExC_sawback)
4584 && !(RExC_seen & REG_SEEN_EVAL)) /* May examine pos and $& */
4585 /* x+ must match at the 1st pos of run of x's */
4586 r->intflags |= PREGf_SKIP;
4588 /* Scan is after the zeroth branch, first is atomic matcher. */
4589 #ifdef TRIE_STUDY_OPT
4592 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4593 (IV)(first - scan + 1))
4597 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4598 (IV)(first - scan + 1))
4604 * If there's something expensive in the r.e., find the
4605 * longest literal string that must appear and make it the
4606 * regmust. Resolve ties in favor of later strings, since
4607 * the regstart check works with the beginning of the r.e.
4608 * and avoiding duplication strengthens checking. Not a
4609 * strong reason, but sufficient in the absence of others.
4610 * [Now we resolve ties in favor of the earlier string if
4611 * it happens that c_offset_min has been invalidated, since the
4612 * earlier string may buy us something the later one won't.]
4615 data.longest_fixed = newSVpvs("");
4616 data.longest_float = newSVpvs("");
4617 data.last_found = newSVpvs("");
4618 data.longest = &(data.longest_fixed);
4620 if (!ri->regstclass) {
4621 cl_init(pRExC_state, &ch_class);
4622 data.start_class = &ch_class;
4623 stclass_flag = SCF_DO_STCLASS_AND;
4624 } else /* XXXX Check for BOUND? */
4626 data.last_closep = &last_close;
4628 minlen = study_chunk(pRExC_state, &first, &minlen, &fake, scan + RExC_size, /* Up to end */
4629 &data, -1, NULL, NULL,
4630 SCF_DO_SUBSTR | SCF_WHILEM_VISITED_POS | stclass_flag,0);
4636 if ( RExC_npar == 1 && data.longest == &(data.longest_fixed)
4637 && data.last_start_min == 0 && data.last_end > 0
4638 && !RExC_seen_zerolen
4639 && !(RExC_seen & REG_SEEN_VERBARG)
4640 && (!(RExC_seen & REG_SEEN_GPOS) || (r->extflags & RXf_ANCH_GPOS)))
4641 r->extflags |= RXf_CHECK_ALL;
4642 scan_commit(pRExC_state, &data,&minlen,0);
4643 SvREFCNT_dec(data.last_found);
4645 /* Note that code very similar to this but for anchored string
4646 follows immediately below, changes may need to be made to both.
4649 longest_float_length = CHR_SVLEN(data.longest_float);
4650 if (longest_float_length
4651 || (data.flags & SF_FL_BEFORE_EOL
4652 && (!(data.flags & SF_FL_BEFORE_MEOL)
4653 || (RExC_flags & RXf_PMf_MULTILINE))))
4657 if (SvCUR(data.longest_fixed) /* ok to leave SvCUR */
4658 && data.offset_fixed == data.offset_float_min
4659 && SvCUR(data.longest_fixed) == SvCUR(data.longest_float))
4660 goto remove_float; /* As in (a)+. */
4662 /* copy the information about the longest float from the reg_scan_data
4663 over to the program. */
4664 if (SvUTF8(data.longest_float)) {
4665 r->float_utf8 = data.longest_float;
4666 r->float_substr = NULL;
4668 r->float_substr = data.longest_float;
4669 r->float_utf8 = NULL;
4671 /* float_end_shift is how many chars that must be matched that
4672 follow this item. We calculate it ahead of time as once the
4673 lookbehind offset is added in we lose the ability to correctly
4675 ml = data.minlen_float ? *(data.minlen_float)
4676 : (I32)longest_float_length;
4677 r->float_end_shift = ml - data.offset_float_min
4678 - longest_float_length + (SvTAIL(data.longest_float) != 0)
4679 + data.lookbehind_float;
4680 r->float_min_offset = data.offset_float_min - data.lookbehind_float;
4681 r->float_max_offset = data.offset_float_max;
4682 if (data.offset_float_max < I32_MAX) /* Don't offset infinity */
4683 r->float_max_offset -= data.lookbehind_float;
4685 t = (data.flags & SF_FL_BEFORE_EOL /* Can't have SEOL and MULTI */
4686 && (!(data.flags & SF_FL_BEFORE_MEOL)
4687 || (RExC_flags & RXf_PMf_MULTILINE)));
4688 fbm_compile(data.longest_float, t ? FBMcf_TAIL : 0);
4692 r->float_substr = r->float_utf8 = NULL;
4693 SvREFCNT_dec(data.longest_float);
4694 longest_float_length = 0;
4697 /* Note that code very similar to this but for floating string
4698 is immediately above, changes may need to be made to both.
4701 longest_fixed_length = CHR_SVLEN(data.longest_fixed);
4702 if (longest_fixed_length
4703 || (data.flags & SF_FIX_BEFORE_EOL /* Cannot have SEOL and MULTI */
4704 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4705 || (RExC_flags & RXf_PMf_MULTILINE))))
4709 /* copy the information about the longest fixed
4710 from the reg_scan_data over to the program. */
4711 if (SvUTF8(data.longest_fixed)) {
4712 r->anchored_utf8 = data.longest_fixed;
4713 r->anchored_substr = NULL;
4715 r->anchored_substr = data.longest_fixed;
4716 r->anchored_utf8 = NULL;
4718 /* fixed_end_shift is how many chars that must be matched that
4719 follow this item. We calculate it ahead of time as once the
4720 lookbehind offset is added in we lose the ability to correctly
4722 ml = data.minlen_fixed ? *(data.minlen_fixed)
4723 : (I32)longest_fixed_length;
4724 r->anchored_end_shift = ml - data.offset_fixed
4725 - longest_fixed_length + (SvTAIL(data.longest_fixed) != 0)
4726 + data.lookbehind_fixed;
4727 r->anchored_offset = data.offset_fixed - data.lookbehind_fixed;
4729 t = (data.flags & SF_FIX_BEFORE_EOL /* Can't have SEOL and MULTI */
4730 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4731 || (RExC_flags & RXf_PMf_MULTILINE)));
4732 fbm_compile(data.longest_fixed, t ? FBMcf_TAIL : 0);
4735 r->anchored_substr = r->anchored_utf8 = NULL;
4736 SvREFCNT_dec(data.longest_fixed);
4737 longest_fixed_length = 0;
4740 && (OP(ri->regstclass) == REG_ANY || OP(ri->regstclass) == SANY))
4741 ri->regstclass = NULL;
4742 if ((!(r->anchored_substr || r->anchored_utf8) || r->anchored_offset)
4744 && !(data.start_class->flags & ANYOF_EOS)
4745 && !cl_is_anything(data.start_class))
4747 const U32 n = add_data(pRExC_state, 1, "f");
4749 Newx(RExC_rxi->data->data[n], 1,
4750 struct regnode_charclass_class);
4751 StructCopy(data.start_class,
4752 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4753 struct regnode_charclass_class);
4754 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4755 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4756 DEBUG_COMPILE_r({ SV *sv = sv_newmortal();
4757 regprop(r, sv, (regnode*)data.start_class);
4758 PerlIO_printf(Perl_debug_log,
4759 "synthetic stclass \"%s\".\n",
4760 SvPVX_const(sv));});
4763 /* A temporary algorithm prefers floated substr to fixed one to dig more info. */
4764 if (longest_fixed_length > longest_float_length) {
4765 r->check_end_shift = r->anchored_end_shift;
4766 r->check_substr = r->anchored_substr;
4767 r->check_utf8 = r->anchored_utf8;
4768 r->check_offset_min = r->check_offset_max = r->anchored_offset;
4769 if (r->extflags & RXf_ANCH_SINGLE)
4770 r->extflags |= RXf_NOSCAN;
4773 r->check_end_shift = r->float_end_shift;
4774 r->check_substr = r->float_substr;
4775 r->check_utf8 = r->float_utf8;
4776 r->check_offset_min = r->float_min_offset;
4777 r->check_offset_max = r->float_max_offset;
4779 /* XXXX Currently intuiting is not compatible with ANCH_GPOS.
4780 This should be changed ASAP! */
4781 if ((r->check_substr || r->check_utf8) && !(r->extflags & RXf_ANCH_GPOS)) {
4782 r->extflags |= RXf_USE_INTUIT;
4783 if (SvTAIL(r->check_substr ? r->check_substr : r->check_utf8))
4784 r->extflags |= RXf_INTUIT_TAIL;
4786 /* XXX Unneeded? dmq (shouldn't as this is handled elsewhere)
4787 if ( (STRLEN)minlen < longest_float_length )
4788 minlen= longest_float_length;
4789 if ( (STRLEN)minlen < longest_fixed_length )
4790 minlen= longest_fixed_length;
4794 /* Several toplevels. Best we can is to set minlen. */
4796 struct regnode_charclass_class ch_class;
4799 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "\nMulti Top Level\n"));
4801 scan = ri->program + 1;
4802 cl_init(pRExC_state, &ch_class);
4803 data.start_class = &ch_class;
4804 data.last_closep = &last_close;
4807 minlen = study_chunk(pRExC_state, &scan, &minlen, &fake, scan + RExC_size,
4808 &data, -1, NULL, NULL, SCF_DO_STCLASS_AND|SCF_WHILEM_VISITED_POS,0);
4812 r->check_substr = r->check_utf8 = r->anchored_substr = r->anchored_utf8
4813 = r->float_substr = r->float_utf8 = NULL;
4814 if (!(data.start_class->flags & ANYOF_EOS)
4815 && !cl_is_anything(data.start_class))
4817 const U32 n = add_data(pRExC_state, 1, "f");
4819 Newx(RExC_rxi->data->data[n], 1,
4820 struct regnode_charclass_class);
4821 StructCopy(data.start_class,
4822 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4823 struct regnode_charclass_class);
4824 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4825 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4826 DEBUG_COMPILE_r({ SV* sv = sv_newmortal();
4827 regprop(r, sv, (regnode*)data.start_class);
4828 PerlIO_printf(Perl_debug_log,
4829 "synthetic stclass \"%s\".\n",
4830 SvPVX_const(sv));});
4834 /* Guard against an embedded (?=) or (?<=) with a longer minlen than
4835 the "real" pattern. */
4837 PerlIO_printf(Perl_debug_log,"minlen: %"IVdf" r->minlen:%"IVdf"\n",
4838 (IV)minlen, (IV)r->minlen);
4840 r->minlenret = minlen;
4841 if (r->minlen < minlen)
4844 if (RExC_seen & REG_SEEN_GPOS)
4845 r->extflags |= RXf_GPOS_SEEN;
4846 if (RExC_seen & REG_SEEN_LOOKBEHIND)
4847 r->extflags |= RXf_LOOKBEHIND_SEEN;
4848 if (RExC_seen & REG_SEEN_EVAL)
4849 r->extflags |= RXf_EVAL_SEEN;
4850 if (RExC_seen & REG_SEEN_CANY)
4851 r->extflags |= RXf_CANY_SEEN;
4852 if (RExC_seen & REG_SEEN_VERBARG)
4853 r->intflags |= PREGf_VERBARG_SEEN;
4854 if (RExC_seen & REG_SEEN_CUTGROUP)
4855 r->intflags |= PREGf_CUTGROUP_SEEN;
4856 if (RExC_paren_names)
4857 RXp_PAREN_NAMES(r) = MUTABLE_HV(SvREFCNT_inc(RExC_paren_names));
4859 RXp_PAREN_NAMES(r) = NULL;
4861 #ifdef STUPID_PATTERN_CHECKS
4862 if (RX_PRELEN(rx) == 0)
4863 r->extflags |= RXf_NULL;
4864 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
4865 /* XXX: this should happen BEFORE we compile */
4866 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
4867 else if (RX_PRELEN(rx) == 3 && memEQ("\\s+", RX_PRECOMP(rx), 3))
4868 r->extflags |= RXf_WHITE;
4869 else if (RX_PRELEN(rx) == 1 && RXp_PRECOMP(rx)[0] == '^')
4870 r->extflags |= RXf_START_ONLY;
4872 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
4873 /* XXX: this should happen BEFORE we compile */
4874 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
4876 regnode *first = ri->program + 1;
4878 U8 nop = OP(NEXTOPER(first));
4880 if (PL_regkind[fop] == NOTHING && nop == END)
4881 r->extflags |= RXf_NULL;
4882 else if (PL_regkind[fop] == BOL && nop == END)
4883 r->extflags |= RXf_START_ONLY;
4884 else if (fop == PLUS && nop ==SPACE && OP(regnext(first))==END)
4885 r->extflags |= RXf_WHITE;
4889 if (RExC_paren_names) {
4890 ri->name_list_idx = add_data( pRExC_state, 1, "p" );
4891 ri->data->data[ri->name_list_idx] = (void*)SvREFCNT_inc(RExC_paren_name_list);
4894 ri->name_list_idx = 0;
4896 if (RExC_recurse_count) {
4897 for ( ; RExC_recurse_count ; RExC_recurse_count-- ) {
4898 const regnode *scan = RExC_recurse[RExC_recurse_count-1];
4899 ARG2L_SET( scan, RExC_open_parens[ARG(scan)-1] - scan );
4902 Newxz(r->offs, RExC_npar, regexp_paren_pair);
4903 /* assume we don't need to swap parens around before we match */
4906 PerlIO_printf(Perl_debug_log,"Final program:\n");
4909 #ifdef RE_TRACK_PATTERN_OFFSETS
4910 DEBUG_OFFSETS_r(if (ri->u.offsets) {
4911 const U32 len = ri->u.offsets[0];
4913 GET_RE_DEBUG_FLAGS_DECL;
4914 PerlIO_printf(Perl_debug_log, "Offsets: [%"UVuf"]\n\t", (UV)ri->u.offsets[0]);
4915 for (i = 1; i <= len; i++) {
4916 if (ri->u.offsets[i*2-1] || ri->u.offsets[i*2])
4917 PerlIO_printf(Perl_debug_log, "%"UVuf":%"UVuf"[%"UVuf"] ",
4918 (UV)i, (UV)ri->u.offsets[i*2-1], (UV)ri->u.offsets[i*2]);
4920 PerlIO_printf(Perl_debug_log, "\n");
4926 #undef RE_ENGINE_PTR
4930 Perl_reg_named_buff(pTHX_ REGEXP * const rx, SV * const key, SV * const value,
4933 PERL_ARGS_ASSERT_REG_NAMED_BUFF;
4935 PERL_UNUSED_ARG(value);
4937 if (flags & RXapif_FETCH) {
4938 return reg_named_buff_fetch(rx, key, flags);
4939 } else if (flags & (RXapif_STORE | RXapif_DELETE | RXapif_CLEAR)) {
4940 Perl_croak(aTHX_ "%s", PL_no_modify);
4942 } else if (flags & RXapif_EXISTS) {
4943 return reg_named_buff_exists(rx, key, flags)
4946 } else if (flags & RXapif_REGNAMES) {
4947 return reg_named_buff_all(rx, flags);
4948 } else if (flags & (RXapif_SCALAR | RXapif_REGNAMES_COUNT)) {
4949 return reg_named_buff_scalar(rx, flags);
4951 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff", (int)flags);
4957 Perl_reg_named_buff_iter(pTHX_ REGEXP * const rx, const SV * const lastkey,
4960 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ITER;
4961 PERL_UNUSED_ARG(lastkey);
4963 if (flags & RXapif_FIRSTKEY)
4964 return reg_named_buff_firstkey(rx, flags);
4965 else if (flags & RXapif_NEXTKEY)
4966 return reg_named_buff_nextkey(rx, flags);
4968 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_iter", (int)flags);
4974 Perl_reg_named_buff_fetch(pTHX_ REGEXP * const r, SV * const namesv,
4977 AV *retarray = NULL;
4979 struct regexp *const rx = (struct regexp *)SvANY(r);
4981 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FETCH;
4983 if (flags & RXapif_ALL)
4986 if (rx && RXp_PAREN_NAMES(rx)) {
4987 HE *he_str = hv_fetch_ent( RXp_PAREN_NAMES(rx), namesv, 0, 0 );
4990 SV* sv_dat=HeVAL(he_str);
4991 I32 *nums=(I32*)SvPVX(sv_dat);
4992 for ( i=0; i<SvIVX(sv_dat); i++ ) {
4993 if ((I32)(rx->nparens) >= nums[i]
4994 && rx->offs[nums[i]].start != -1
4995 && rx->offs[nums[i]].end != -1)
4998 CALLREG_NUMBUF_FETCH(r,nums[i],ret);
5002 ret = newSVsv(&PL_sv_undef);
5005 av_push(retarray, ret);
5008 return newRV_noinc(MUTABLE_SV(retarray));
5015 Perl_reg_named_buff_exists(pTHX_ REGEXP * const r, SV * const key,
5018 struct regexp *const rx = (struct regexp *)SvANY(r);
5020 PERL_ARGS_ASSERT_REG_NAMED_BUFF_EXISTS;
5022 if (rx && RXp_PAREN_NAMES(rx)) {
5023 if (flags & RXapif_ALL) {
5024 return hv_exists_ent(RXp_PAREN_NAMES(rx), key, 0);
5026 SV *sv = CALLREG_NAMED_BUFF_FETCH(r, key, flags);
5040 Perl_reg_named_buff_firstkey(pTHX_ REGEXP * const r, const U32 flags)
5042 struct regexp *const rx = (struct regexp *)SvANY(r);
5044 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FIRSTKEY;
5046 if ( rx && RXp_PAREN_NAMES(rx) ) {
5047 (void)hv_iterinit(RXp_PAREN_NAMES(rx));
5049 return CALLREG_NAMED_BUFF_NEXTKEY(r, NULL, flags & ~RXapif_FIRSTKEY);
5056 Perl_reg_named_buff_nextkey(pTHX_ REGEXP * const r, const U32 flags)
5058 struct regexp *const rx = (struct regexp *)SvANY(r);
5059 GET_RE_DEBUG_FLAGS_DECL;
5061 PERL_ARGS_ASSERT_REG_NAMED_BUFF_NEXTKEY;
5063 if (rx && RXp_PAREN_NAMES(rx)) {
5064 HV *hv = RXp_PAREN_NAMES(rx);
5066 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5069 SV* sv_dat = HeVAL(temphe);
5070 I32 *nums = (I32*)SvPVX(sv_dat);
5071 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5072 if ((I32)(rx->lastparen) >= nums[i] &&
5073 rx->offs[nums[i]].start != -1 &&
5074 rx->offs[nums[i]].end != -1)
5080 if (parno || flags & RXapif_ALL) {
5081 return newSVhek(HeKEY_hek(temphe));
5089 Perl_reg_named_buff_scalar(pTHX_ REGEXP * const r, const U32 flags)
5094 struct regexp *const rx = (struct regexp *)SvANY(r);
5096 PERL_ARGS_ASSERT_REG_NAMED_BUFF_SCALAR;
5098 if (rx && RXp_PAREN_NAMES(rx)) {
5099 if (flags & (RXapif_ALL | RXapif_REGNAMES_COUNT)) {
5100 return newSViv(HvTOTALKEYS(RXp_PAREN_NAMES(rx)));
5101 } else if (flags & RXapif_ONE) {
5102 ret = CALLREG_NAMED_BUFF_ALL(r, (flags | RXapif_REGNAMES));
5103 av = MUTABLE_AV(SvRV(ret));
5104 length = av_len(av);
5106 return newSViv(length + 1);
5108 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_scalar", (int)flags);
5112 return &PL_sv_undef;
5116 Perl_reg_named_buff_all(pTHX_ REGEXP * const r, const U32 flags)
5118 struct regexp *const rx = (struct regexp *)SvANY(r);
5121 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ALL;
5123 if (rx && RXp_PAREN_NAMES(rx)) {
5124 HV *hv= RXp_PAREN_NAMES(rx);
5126 (void)hv_iterinit(hv);
5127 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5130 SV* sv_dat = HeVAL(temphe);
5131 I32 *nums = (I32*)SvPVX(sv_dat);
5132 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5133 if ((I32)(rx->lastparen) >= nums[i] &&
5134 rx->offs[nums[i]].start != -1 &&
5135 rx->offs[nums[i]].end != -1)
5141 if (parno || flags & RXapif_ALL) {
5142 av_push(av, newSVhek(HeKEY_hek(temphe)));
5147 return newRV_noinc(MUTABLE_SV(av));
5151 Perl_reg_numbered_buff_fetch(pTHX_ REGEXP * const r, const I32 paren,
5154 struct regexp *const rx = (struct regexp *)SvANY(r);
5159 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_FETCH;
5162 sv_setsv(sv,&PL_sv_undef);
5166 if (paren == RX_BUFF_IDX_PREMATCH && rx->offs[0].start != -1) {
5168 i = rx->offs[0].start;
5172 if (paren == RX_BUFF_IDX_POSTMATCH && rx->offs[0].end != -1) {
5174 s = rx->subbeg + rx->offs[0].end;
5175 i = rx->sublen - rx->offs[0].end;
5178 if ( 0 <= paren && paren <= (I32)rx->nparens &&
5179 (s1 = rx->offs[paren].start) != -1 &&
5180 (t1 = rx->offs[paren].end) != -1)
5184 s = rx->subbeg + s1;
5186 sv_setsv(sv,&PL_sv_undef);
5189 assert(rx->sublen >= (s - rx->subbeg) + i );
5191 const int oldtainted = PL_tainted;
5193 sv_setpvn(sv, s, i);
5194 PL_tainted = oldtainted;
5195 if ( (rx->extflags & RXf_CANY_SEEN)
5196 ? (RXp_MATCH_UTF8(rx)
5197 && (!i || is_utf8_string((U8*)s, i)))
5198 : (RXp_MATCH_UTF8(rx)) )
5205 if (RXp_MATCH_TAINTED(rx)) {
5206 if (SvTYPE(sv) >= SVt_PVMG) {
5207 MAGIC* const mg = SvMAGIC(sv);
5210 SvMAGIC_set(sv, mg->mg_moremagic);
5212 if ((mgt = SvMAGIC(sv))) {
5213 mg->mg_moremagic = mgt;
5214 SvMAGIC_set(sv, mg);
5224 sv_setsv(sv,&PL_sv_undef);
5230 Perl_reg_numbered_buff_store(pTHX_ REGEXP * const rx, const I32 paren,
5231 SV const * const value)
5233 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_STORE;
5235 PERL_UNUSED_ARG(rx);
5236 PERL_UNUSED_ARG(paren);
5237 PERL_UNUSED_ARG(value);
5240 Perl_croak(aTHX_ "%s", PL_no_modify);
5244 Perl_reg_numbered_buff_length(pTHX_ REGEXP * const r, const SV * const sv,
5247 struct regexp *const rx = (struct regexp *)SvANY(r);
5251 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_LENGTH;
5253 /* Some of this code was originally in C<Perl_magic_len> in F<mg.c> */
5255 /* $` / ${^PREMATCH} */
5256 case RX_BUFF_IDX_PREMATCH:
5257 if (rx->offs[0].start != -1) {
5258 i = rx->offs[0].start;
5266 /* $' / ${^POSTMATCH} */
5267 case RX_BUFF_IDX_POSTMATCH:
5268 if (rx->offs[0].end != -1) {
5269 i = rx->sublen - rx->offs[0].end;
5271 s1 = rx->offs[0].end;
5277 /* $& / ${^MATCH}, $1, $2, ... */
5279 if (paren <= (I32)rx->nparens &&
5280 (s1 = rx->offs[paren].start) != -1 &&
5281 (t1 = rx->offs[paren].end) != -1)
5286 if (ckWARN(WARN_UNINITIALIZED))
5287 report_uninit((const SV *)sv);
5292 if (i > 0 && RXp_MATCH_UTF8(rx)) {
5293 const char * const s = rx->subbeg + s1;
5298 if (is_utf8_string_loclen((U8*)s, i, &ep, &el))
5305 Perl_reg_qr_package(pTHX_ REGEXP * const rx)
5307 PERL_ARGS_ASSERT_REG_QR_PACKAGE;
5308 PERL_UNUSED_ARG(rx);
5312 return newSVpvs("Regexp");
5315 /* Scans the name of a named buffer from the pattern.
5316 * If flags is REG_RSN_RETURN_NULL returns null.
5317 * If flags is REG_RSN_RETURN_NAME returns an SV* containing the name
5318 * If flags is REG_RSN_RETURN_DATA returns the data SV* corresponding
5319 * to the parsed name as looked up in the RExC_paren_names hash.
5320 * If there is an error throws a vFAIL().. type exception.
5323 #define REG_RSN_RETURN_NULL 0
5324 #define REG_RSN_RETURN_NAME 1
5325 #define REG_RSN_RETURN_DATA 2
5328 S_reg_scan_name(pTHX_ RExC_state_t *pRExC_state, U32 flags)
5330 char *name_start = RExC_parse;
5332 PERL_ARGS_ASSERT_REG_SCAN_NAME;
5334 if (isIDFIRST_lazy_if(RExC_parse, UTF)) {
5335 /* skip IDFIRST by using do...while */
5338 RExC_parse += UTF8SKIP(RExC_parse);
5339 } while (isALNUM_utf8((U8*)RExC_parse));
5343 } while (isALNUM(*RExC_parse));
5348 = newSVpvn_flags(name_start, (int)(RExC_parse - name_start),
5349 SVs_TEMP | (UTF ? SVf_UTF8 : 0));
5350 if ( flags == REG_RSN_RETURN_NAME)
5352 else if (flags==REG_RSN_RETURN_DATA) {
5355 if ( ! sv_name ) /* should not happen*/
5356 Perl_croak(aTHX_ "panic: no svname in reg_scan_name");
5357 if (RExC_paren_names)
5358 he_str = hv_fetch_ent( RExC_paren_names, sv_name, 0, 0 );
5360 sv_dat = HeVAL(he_str);
5362 vFAIL("Reference to nonexistent named group");
5366 Perl_croak(aTHX_ "panic: bad flag in reg_scan_name");
5373 #define DEBUG_PARSE_MSG(funcname) DEBUG_PARSE_r({ \
5374 int rem=(int)(RExC_end - RExC_parse); \
5383 if (RExC_lastparse!=RExC_parse) \
5384 PerlIO_printf(Perl_debug_log," >%.*s%-*s", \
5387 iscut ? "..." : "<" \
5390 PerlIO_printf(Perl_debug_log,"%16s",""); \
5393 num = RExC_size + 1; \
5395 num=REG_NODE_NUM(RExC_emit); \
5396 if (RExC_lastnum!=num) \
5397 PerlIO_printf(Perl_debug_log,"|%4d",num); \
5399 PerlIO_printf(Perl_debug_log,"|%4s",""); \
5400 PerlIO_printf(Perl_debug_log,"|%*s%-4s", \
5401 (int)((depth*2)), "", \
5405 RExC_lastparse=RExC_parse; \
5410 #define DEBUG_PARSE(funcname) DEBUG_PARSE_r({ \
5411 DEBUG_PARSE_MSG((funcname)); \
5412 PerlIO_printf(Perl_debug_log,"%4s","\n"); \
5414 #define DEBUG_PARSE_FMT(funcname,fmt,args) DEBUG_PARSE_r({ \
5415 DEBUG_PARSE_MSG((funcname)); \
5416 PerlIO_printf(Perl_debug_log,fmt "\n",args); \
5419 - reg - regular expression, i.e. main body or parenthesized thing
5421 * Caller must absorb opening parenthesis.
5423 * Combining parenthesis handling with the base level of regular expression
5424 * is a trifle forced, but the need to tie the tails of the branches to what
5425 * follows makes it hard to avoid.
5427 #define REGTAIL(x,y,z) regtail((x),(y),(z),depth+1)
5429 #define REGTAIL_STUDY(x,y,z) regtail_study((x),(y),(z),depth+1)
5431 #define REGTAIL_STUDY(x,y,z) regtail((x),(y),(z),depth+1)
5435 S_reg(pTHX_ RExC_state_t *pRExC_state, I32 paren, I32 *flagp,U32 depth)
5436 /* paren: Parenthesized? 0=top, 1=(, inside: changed to letter. */
5439 register regnode *ret; /* Will be the head of the group. */
5440 register regnode *br;
5441 register regnode *lastbr;
5442 register regnode *ender = NULL;
5443 register I32 parno = 0;
5445 U32 oregflags = RExC_flags;
5446 bool have_branch = 0;
5448 I32 freeze_paren = 0;
5449 I32 after_freeze = 0;
5451 /* for (?g), (?gc), and (?o) warnings; warning
5452 about (?c) will warn about (?g) -- japhy */
5454 #define WASTED_O 0x01
5455 #define WASTED_G 0x02
5456 #define WASTED_C 0x04
5457 #define WASTED_GC (0x02|0x04)
5458 I32 wastedflags = 0x00;
5460 char * parse_start = RExC_parse; /* MJD */
5461 char * const oregcomp_parse = RExC_parse;
5463 GET_RE_DEBUG_FLAGS_DECL;
5465 PERL_ARGS_ASSERT_REG;
5466 DEBUG_PARSE("reg ");
5468 *flagp = 0; /* Tentatively. */
5471 /* Make an OPEN node, if parenthesized. */
5473 if ( *RExC_parse == '*') { /* (*VERB:ARG) */
5474 char *start_verb = RExC_parse;
5475 STRLEN verb_len = 0;
5476 char *start_arg = NULL;
5477 unsigned char op = 0;
5479 int internal_argval = 0; /* internal_argval is only useful if !argok */
5480 while ( *RExC_parse && *RExC_parse != ')' ) {
5481 if ( *RExC_parse == ':' ) {
5482 start_arg = RExC_parse + 1;
5488 verb_len = RExC_parse - start_verb;
5491 while ( *RExC_parse && *RExC_parse != ')' )
5493 if ( *RExC_parse != ')' )
5494 vFAIL("Unterminated verb pattern argument");
5495 if ( RExC_parse == start_arg )
5498 if ( *RExC_parse != ')' )
5499 vFAIL("Unterminated verb pattern");
5502 switch ( *start_verb ) {
5503 case 'A': /* (*ACCEPT) */
5504 if ( memEQs(start_verb,verb_len,"ACCEPT") ) {
5506 internal_argval = RExC_nestroot;
5509 case 'C': /* (*COMMIT) */
5510 if ( memEQs(start_verb,verb_len,"COMMIT") )
5513 case 'F': /* (*FAIL) */
5514 if ( verb_len==1 || memEQs(start_verb,verb_len,"FAIL") ) {
5519 case ':': /* (*:NAME) */
5520 case 'M': /* (*MARK:NAME) */
5521 if ( verb_len==0 || memEQs(start_verb,verb_len,"MARK") ) {
5526 case 'P': /* (*PRUNE) */
5527 if ( memEQs(start_verb,verb_len,"PRUNE") )
5530 case 'S': /* (*SKIP) */
5531 if ( memEQs(start_verb,verb_len,"SKIP") )
5534 case 'T': /* (*THEN) */
5535 /* [19:06] <TimToady> :: is then */
5536 if ( memEQs(start_verb,verb_len,"THEN") ) {
5538 RExC_seen |= REG_SEEN_CUTGROUP;
5544 vFAIL3("Unknown verb pattern '%.*s'",
5545 verb_len, start_verb);
5548 if ( start_arg && internal_argval ) {
5549 vFAIL3("Verb pattern '%.*s' may not have an argument",
5550 verb_len, start_verb);
5551 } else if ( argok < 0 && !start_arg ) {
5552 vFAIL3("Verb pattern '%.*s' has a mandatory argument",
5553 verb_len, start_verb);
5555 ret = reganode(pRExC_state, op, internal_argval);
5556 if ( ! internal_argval && ! SIZE_ONLY ) {
5558 SV *sv = newSVpvn( start_arg, RExC_parse - start_arg);
5559 ARG(ret) = add_data( pRExC_state, 1, "S" );
5560 RExC_rxi->data->data[ARG(ret)]=(void*)sv;
5567 if (!internal_argval)
5568 RExC_seen |= REG_SEEN_VERBARG;
5569 } else if ( start_arg ) {
5570 vFAIL3("Verb pattern '%.*s' may not have an argument",
5571 verb_len, start_verb);
5573 ret = reg_node(pRExC_state, op);
5575 nextchar(pRExC_state);
5578 if (*RExC_parse == '?') { /* (?...) */
5579 bool is_logical = 0;
5580 const char * const seqstart = RExC_parse;
5583 paren = *RExC_parse++;
5584 ret = NULL; /* For look-ahead/behind. */
5587 case 'P': /* (?P...) variants for those used to PCRE/Python */
5588 paren = *RExC_parse++;
5589 if ( paren == '<') /* (?P<...>) named capture */
5591 else if (paren == '>') { /* (?P>name) named recursion */
5592 goto named_recursion;
5594 else if (paren == '=') { /* (?P=...) named backref */
5595 /* this pretty much dupes the code for \k<NAME> in regatom(), if
5596 you change this make sure you change that */
5597 char* name_start = RExC_parse;
5599 SV *sv_dat = reg_scan_name(pRExC_state,
5600 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5601 if (RExC_parse == name_start || *RExC_parse != ')')
5602 vFAIL2("Sequence %.3s... not terminated",parse_start);
5605 num = add_data( pRExC_state, 1, "S" );
5606 RExC_rxi->data->data[num]=(void*)sv_dat;
5607 SvREFCNT_inc_simple_void(sv_dat);
5610 ret = reganode(pRExC_state,
5611 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
5615 Set_Node_Offset(ret, parse_start+1);
5616 Set_Node_Cur_Length(ret); /* MJD */
5618 nextchar(pRExC_state);
5622 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5624 case '<': /* (?<...) */
5625 if (*RExC_parse == '!')
5627 else if (*RExC_parse != '=')
5633 case '\'': /* (?'...') */
5634 name_start= RExC_parse;
5635 svname = reg_scan_name(pRExC_state,
5636 SIZE_ONLY ? /* reverse test from the others */
5637 REG_RSN_RETURN_NAME :
5638 REG_RSN_RETURN_NULL);
5639 if (RExC_parse == name_start) {
5641 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5644 if (*RExC_parse != paren)
5645 vFAIL2("Sequence (?%c... not terminated",
5646 paren=='>' ? '<' : paren);
5650 if (!svname) /* shouldnt happen */
5652 "panic: reg_scan_name returned NULL");
5653 if (!RExC_paren_names) {
5654 RExC_paren_names= newHV();
5655 sv_2mortal(MUTABLE_SV(RExC_paren_names));
5657 RExC_paren_name_list= newAV();
5658 sv_2mortal(MUTABLE_SV(RExC_paren_name_list));
5661 he_str = hv_fetch_ent( RExC_paren_names, svname, 1, 0 );
5663 sv_dat = HeVAL(he_str);
5665 /* croak baby croak */
5667 "panic: paren_name hash element allocation failed");
5668 } else if ( SvPOK(sv_dat) ) {
5669 /* (?|...) can mean we have dupes so scan to check
5670 its already been stored. Maybe a flag indicating
5671 we are inside such a construct would be useful,
5672 but the arrays are likely to be quite small, so
5673 for now we punt -- dmq */
5674 IV count = SvIV(sv_dat);
5675 I32 *pv = (I32*)SvPVX(sv_dat);
5677 for ( i = 0 ; i < count ; i++ ) {
5678 if ( pv[i] == RExC_npar ) {
5684 pv = (I32*)SvGROW(sv_dat, SvCUR(sv_dat) + sizeof(I32)+1);
5685 SvCUR_set(sv_dat, SvCUR(sv_dat) + sizeof(I32));
5686 pv[count] = RExC_npar;
5687 SvIV_set(sv_dat, SvIVX(sv_dat) + 1);
5690 (void)SvUPGRADE(sv_dat,SVt_PVNV);
5691 sv_setpvn(sv_dat, (char *)&(RExC_npar), sizeof(I32));
5693 SvIV_set(sv_dat, 1);
5696 if (!av_store(RExC_paren_name_list, RExC_npar, SvREFCNT_inc(svname)))
5697 SvREFCNT_dec(svname);
5700 /*sv_dump(sv_dat);*/
5702 nextchar(pRExC_state);
5704 goto capturing_parens;
5706 RExC_seen |= REG_SEEN_LOOKBEHIND;
5708 case '=': /* (?=...) */
5709 RExC_seen_zerolen++;
5711 case '!': /* (?!...) */
5712 RExC_seen_zerolen++;
5713 if (*RExC_parse == ')') {
5714 ret=reg_node(pRExC_state, OPFAIL);
5715 nextchar(pRExC_state);
5719 case '|': /* (?|...) */
5720 /* branch reset, behave like a (?:...) except that
5721 buffers in alternations share the same numbers */
5723 after_freeze = freeze_paren = RExC_npar;
5725 case ':': /* (?:...) */
5726 case '>': /* (?>...) */
5728 case '$': /* (?$...) */
5729 case '@': /* (?@...) */
5730 vFAIL2("Sequence (?%c...) not implemented", (int)paren);
5732 case '#': /* (?#...) */
5733 while (*RExC_parse && *RExC_parse != ')')
5735 if (*RExC_parse != ')')
5736 FAIL("Sequence (?#... not terminated");
5737 nextchar(pRExC_state);
5740 case '0' : /* (?0) */
5741 case 'R' : /* (?R) */
5742 if (*RExC_parse != ')')
5743 FAIL("Sequence (?R) not terminated");
5744 ret = reg_node(pRExC_state, GOSTART);
5745 *flagp |= POSTPONED;
5746 nextchar(pRExC_state);
5749 { /* named and numeric backreferences */
5751 case '&': /* (?&NAME) */
5752 parse_start = RExC_parse - 1;
5755 SV *sv_dat = reg_scan_name(pRExC_state,
5756 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5757 num = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5759 goto gen_recurse_regop;
5762 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5764 vFAIL("Illegal pattern");
5766 goto parse_recursion;
5768 case '-': /* (?-1) */
5769 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5770 RExC_parse--; /* rewind to let it be handled later */
5774 case '1': case '2': case '3': case '4': /* (?1) */
5775 case '5': case '6': case '7': case '8': case '9':
5778 num = atoi(RExC_parse);
5779 parse_start = RExC_parse - 1; /* MJD */
5780 if (*RExC_parse == '-')
5782 while (isDIGIT(*RExC_parse))
5784 if (*RExC_parse!=')')
5785 vFAIL("Expecting close bracket");
5788 if ( paren == '-' ) {
5790 Diagram of capture buffer numbering.
5791 Top line is the normal capture buffer numbers
5792 Botton line is the negative indexing as from
5796 /(a(x)y)(a(b(c(?-2)d)e)f)(g(h))/
5800 num = RExC_npar + num;
5803 vFAIL("Reference to nonexistent group");
5805 } else if ( paren == '+' ) {
5806 num = RExC_npar + num - 1;
5809 ret = reganode(pRExC_state, GOSUB, num);
5811 if (num > (I32)RExC_rx->nparens) {
5813 vFAIL("Reference to nonexistent group");
5815 ARG2L_SET( ret, RExC_recurse_count++);
5817 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
5818 "Recurse #%"UVuf" to %"IVdf"\n", (UV)ARG(ret), (IV)ARG2L(ret)));
5822 RExC_seen |= REG_SEEN_RECURSE;
5823 Set_Node_Length(ret, 1 + regarglen[OP(ret)]); /* MJD */
5824 Set_Node_Offset(ret, parse_start); /* MJD */
5826 *flagp |= POSTPONED;
5827 nextchar(pRExC_state);
5829 } /* named and numeric backreferences */
5832 case '?': /* (??...) */
5834 if (*RExC_parse != '{') {
5836 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5839 *flagp |= POSTPONED;
5840 paren = *RExC_parse++;
5842 case '{': /* (?{...}) */
5847 char *s = RExC_parse;
5849 RExC_seen_zerolen++;
5850 RExC_seen |= REG_SEEN_EVAL;
5851 while (count && (c = *RExC_parse)) {
5862 if (*RExC_parse != ')') {
5864 vFAIL("Sequence (?{...}) not terminated or not {}-balanced");
5868 OP_4tree *sop, *rop;
5869 SV * const sv = newSVpvn(s, RExC_parse - 1 - s);
5872 Perl_save_re_context(aTHX);
5873 rop = sv_compile_2op(sv, &sop, "re", &pad);
5874 sop->op_private |= OPpREFCOUNTED;
5875 /* re_dup will OpREFCNT_inc */
5876 OpREFCNT_set(sop, 1);
5879 n = add_data(pRExC_state, 3, "nop");
5880 RExC_rxi->data->data[n] = (void*)rop;
5881 RExC_rxi->data->data[n+1] = (void*)sop;
5882 RExC_rxi->data->data[n+2] = (void*)pad;
5885 else { /* First pass */
5886 if (PL_reginterp_cnt < ++RExC_seen_evals
5888 /* No compiled RE interpolated, has runtime
5889 components ===> unsafe. */
5890 FAIL("Eval-group not allowed at runtime, use re 'eval'");
5891 if (PL_tainting && PL_tainted)
5892 FAIL("Eval-group in insecure regular expression");
5893 #if PERL_VERSION > 8
5894 if (IN_PERL_COMPILETIME)
5899 nextchar(pRExC_state);
5901 ret = reg_node(pRExC_state, LOGICAL);
5904 REGTAIL(pRExC_state, ret, reganode(pRExC_state, EVAL, n));
5905 /* deal with the length of this later - MJD */
5908 ret = reganode(pRExC_state, EVAL, n);
5909 Set_Node_Length(ret, RExC_parse - parse_start + 1);
5910 Set_Node_Offset(ret, parse_start);
5913 case '(': /* (?(?{...})...) and (?(?=...)...) */
5916 if (RExC_parse[0] == '?') { /* (?(?...)) */
5917 if (RExC_parse[1] == '=' || RExC_parse[1] == '!'
5918 || RExC_parse[1] == '<'
5919 || RExC_parse[1] == '{') { /* Lookahead or eval. */
5922 ret = reg_node(pRExC_state, LOGICAL);
5925 REGTAIL(pRExC_state, ret, reg(pRExC_state, 1, &flag,depth+1));
5929 else if ( RExC_parse[0] == '<' /* (?(<NAME>)...) */
5930 || RExC_parse[0] == '\'' ) /* (?('NAME')...) */
5932 char ch = RExC_parse[0] == '<' ? '>' : '\'';
5933 char *name_start= RExC_parse++;
5935 SV *sv_dat=reg_scan_name(pRExC_state,
5936 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5937 if (RExC_parse == name_start || *RExC_parse != ch)
5938 vFAIL2("Sequence (?(%c... not terminated",
5939 (ch == '>' ? '<' : ch));
5942 num = add_data( pRExC_state, 1, "S" );
5943 RExC_rxi->data->data[num]=(void*)sv_dat;
5944 SvREFCNT_inc_simple_void(sv_dat);
5946 ret = reganode(pRExC_state,NGROUPP,num);
5947 goto insert_if_check_paren;
5949 else if (RExC_parse[0] == 'D' &&
5950 RExC_parse[1] == 'E' &&
5951 RExC_parse[2] == 'F' &&
5952 RExC_parse[3] == 'I' &&
5953 RExC_parse[4] == 'N' &&
5954 RExC_parse[5] == 'E')
5956 ret = reganode(pRExC_state,DEFINEP,0);
5959 goto insert_if_check_paren;
5961 else if (RExC_parse[0] == 'R') {
5964 if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
5965 parno = atoi(RExC_parse++);
5966 while (isDIGIT(*RExC_parse))
5968 } else if (RExC_parse[0] == '&') {
5971 sv_dat = reg_scan_name(pRExC_state,
5972 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5973 parno = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5975 ret = reganode(pRExC_state,INSUBP,parno);
5976 goto insert_if_check_paren;
5978 else if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
5981 parno = atoi(RExC_parse++);
5983 while (isDIGIT(*RExC_parse))
5985 ret = reganode(pRExC_state, GROUPP, parno);
5987 insert_if_check_paren:
5988 if ((c = *nextchar(pRExC_state)) != ')')
5989 vFAIL("Switch condition not recognized");
5991 REGTAIL(pRExC_state, ret, reganode(pRExC_state, IFTHEN, 0));
5992 br = regbranch(pRExC_state, &flags, 1,depth+1);
5994 br = reganode(pRExC_state, LONGJMP, 0);
5996 REGTAIL(pRExC_state, br, reganode(pRExC_state, LONGJMP, 0));
5997 c = *nextchar(pRExC_state);
6002 vFAIL("(?(DEFINE)....) does not allow branches");
6003 lastbr = reganode(pRExC_state, IFTHEN, 0); /* Fake one for optimizer. */
6004 regbranch(pRExC_state, &flags, 1,depth+1);
6005 REGTAIL(pRExC_state, ret, lastbr);
6008 c = *nextchar(pRExC_state);
6013 vFAIL("Switch (?(condition)... contains too many branches");
6014 ender = reg_node(pRExC_state, TAIL);
6015 REGTAIL(pRExC_state, br, ender);
6017 REGTAIL(pRExC_state, lastbr, ender);
6018 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender);
6021 REGTAIL(pRExC_state, ret, ender);
6022 RExC_size++; /* XXX WHY do we need this?!!
6023 For large programs it seems to be required
6024 but I can't figure out why. -- dmq*/
6028 vFAIL2("Unknown switch condition (?(%.2s", RExC_parse);
6032 RExC_parse--; /* for vFAIL to print correctly */
6033 vFAIL("Sequence (? incomplete");
6037 parse_flags: /* (?i) */
6039 U32 posflags = 0, negflags = 0;
6040 U32 *flagsp = &posflags;
6042 while (*RExC_parse) {
6043 /* && strchr("iogcmsx", *RExC_parse) */
6044 /* (?g), (?gc) and (?o) are useless here
6045 and must be globally applied -- japhy */
6046 switch (*RExC_parse) {
6047 CASE_STD_PMMOD_FLAGS_PARSE_SET(flagsp);
6048 case ONCE_PAT_MOD: /* 'o' */
6049 case GLOBAL_PAT_MOD: /* 'g' */
6050 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6051 const I32 wflagbit = *RExC_parse == 'o' ? WASTED_O : WASTED_G;
6052 if (! (wastedflags & wflagbit) ) {
6053 wastedflags |= wflagbit;
6056 "Useless (%s%c) - %suse /%c modifier",
6057 flagsp == &negflags ? "?-" : "?",
6059 flagsp == &negflags ? "don't " : "",
6066 case CONTINUE_PAT_MOD: /* 'c' */
6067 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6068 if (! (wastedflags & WASTED_C) ) {
6069 wastedflags |= WASTED_GC;
6072 "Useless (%sc) - %suse /gc modifier",
6073 flagsp == &negflags ? "?-" : "?",
6074 flagsp == &negflags ? "don't " : ""
6079 case KEEPCOPY_PAT_MOD: /* 'p' */
6080 if (flagsp == &negflags) {
6081 if (SIZE_ONLY && ckWARN(WARN_REGEXP))
6082 vWARN(RExC_parse + 1,"Useless use of (?-p)");
6084 *flagsp |= RXf_PMf_KEEPCOPY;
6088 if (flagsp == &negflags) {
6090 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6094 wastedflags = 0; /* reset so (?g-c) warns twice */
6100 RExC_flags |= posflags;
6101 RExC_flags &= ~negflags;
6103 oregflags |= posflags;
6104 oregflags &= ~negflags;
6106 nextchar(pRExC_state);
6117 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6122 }} /* one for the default block, one for the switch */
6129 ret = reganode(pRExC_state, OPEN, parno);
6132 RExC_nestroot = parno;
6133 if (RExC_seen & REG_SEEN_RECURSE
6134 && !RExC_open_parens[parno-1])
6136 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6137 "Setting open paren #%"IVdf" to %d\n",
6138 (IV)parno, REG_NODE_NUM(ret)));
6139 RExC_open_parens[parno-1]= ret;
6142 Set_Node_Length(ret, 1); /* MJD */
6143 Set_Node_Offset(ret, RExC_parse); /* MJD */
6151 /* Pick up the branches, linking them together. */
6152 parse_start = RExC_parse; /* MJD */
6153 br = regbranch(pRExC_state, &flags, 1,depth+1);
6156 if (RExC_npar > after_freeze)
6157 after_freeze = RExC_npar;
6158 RExC_npar = freeze_paren;
6161 /* branch_len = (paren != 0); */
6165 if (*RExC_parse == '|') {
6166 if (!SIZE_ONLY && RExC_extralen) {
6167 reginsert(pRExC_state, BRANCHJ, br, depth+1);
6170 reginsert(pRExC_state, BRANCH, br, depth+1);
6171 Set_Node_Length(br, paren != 0);
6172 Set_Node_Offset_To_R(br-RExC_emit_start, parse_start-RExC_start);
6176 RExC_extralen += 1; /* For BRANCHJ-BRANCH. */
6178 else if (paren == ':') {
6179 *flagp |= flags&SIMPLE;
6181 if (is_open) { /* Starts with OPEN. */
6182 REGTAIL(pRExC_state, ret, br); /* OPEN -> first. */
6184 else if (paren != '?') /* Not Conditional */
6186 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6188 while (*RExC_parse == '|') {
6189 if (!SIZE_ONLY && RExC_extralen) {
6190 ender = reganode(pRExC_state, LONGJMP,0);
6191 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender); /* Append to the previous. */
6194 RExC_extralen += 2; /* Account for LONGJMP. */
6195 nextchar(pRExC_state);
6197 if (RExC_npar > after_freeze)
6198 after_freeze = RExC_npar;
6199 RExC_npar = freeze_paren;
6201 br = regbranch(pRExC_state, &flags, 0, depth+1);
6205 REGTAIL(pRExC_state, lastbr, br); /* BRANCH -> BRANCH. */
6207 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6210 if (have_branch || paren != ':') {
6211 /* Make a closing node, and hook it on the end. */
6214 ender = reg_node(pRExC_state, TAIL);
6217 ender = reganode(pRExC_state, CLOSE, parno);
6218 if (!SIZE_ONLY && RExC_seen & REG_SEEN_RECURSE) {
6219 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6220 "Setting close paren #%"IVdf" to %d\n",
6221 (IV)parno, REG_NODE_NUM(ender)));
6222 RExC_close_parens[parno-1]= ender;
6223 if (RExC_nestroot == parno)
6226 Set_Node_Offset(ender,RExC_parse+1); /* MJD */
6227 Set_Node_Length(ender,1); /* MJD */
6233 *flagp &= ~HASWIDTH;
6236 ender = reg_node(pRExC_state, SUCCEED);
6239 ender = reg_node(pRExC_state, END);
6241 assert(!RExC_opend); /* there can only be one! */
6246 REGTAIL(pRExC_state, lastbr, ender);
6248 if (have_branch && !SIZE_ONLY) {
6250 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
6252 /* Hook the tails of the branches to the closing node. */
6253 for (br = ret; br; br = regnext(br)) {
6254 const U8 op = PL_regkind[OP(br)];
6256 REGTAIL_STUDY(pRExC_state, NEXTOPER(br), ender);
6258 else if (op == BRANCHJ) {
6259 REGTAIL_STUDY(pRExC_state, NEXTOPER(NEXTOPER(br)), ender);
6267 static const char parens[] = "=!<,>";
6269 if (paren && (p = strchr(parens, paren))) {
6270 U8 node = ((p - parens) % 2) ? UNLESSM : IFMATCH;
6271 int flag = (p - parens) > 1;
6274 node = SUSPEND, flag = 0;
6275 reginsert(pRExC_state, node,ret, depth+1);
6276 Set_Node_Cur_Length(ret);
6277 Set_Node_Offset(ret, parse_start + 1);
6279 REGTAIL_STUDY(pRExC_state, ret, reg_node(pRExC_state, TAIL));
6283 /* Check for proper termination. */
6285 RExC_flags = oregflags;
6286 if (RExC_parse >= RExC_end || *nextchar(pRExC_state) != ')') {
6287 RExC_parse = oregcomp_parse;
6288 vFAIL("Unmatched (");
6291 else if (!paren && RExC_parse < RExC_end) {
6292 if (*RExC_parse == ')') {
6294 vFAIL("Unmatched )");
6297 FAIL("Junk on end of regexp"); /* "Can't happen". */
6301 RExC_npar = after_freeze;
6306 - regbranch - one alternative of an | operator
6308 * Implements the concatenation operator.
6311 S_regbranch(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, I32 first, U32 depth)
6314 register regnode *ret;
6315 register regnode *chain = NULL;
6316 register regnode *latest;
6317 I32 flags = 0, c = 0;
6318 GET_RE_DEBUG_FLAGS_DECL;
6320 PERL_ARGS_ASSERT_REGBRANCH;
6322 DEBUG_PARSE("brnc");
6327 if (!SIZE_ONLY && RExC_extralen)
6328 ret = reganode(pRExC_state, BRANCHJ,0);
6330 ret = reg_node(pRExC_state, BRANCH);
6331 Set_Node_Length(ret, 1);
6335 if (!first && SIZE_ONLY)
6336 RExC_extralen += 1; /* BRANCHJ */
6338 *flagp = WORST; /* Tentatively. */
6341 nextchar(pRExC_state);
6342 while (RExC_parse < RExC_end && *RExC_parse != '|' && *RExC_parse != ')') {
6344 latest = regpiece(pRExC_state, &flags,depth+1);
6345 if (latest == NULL) {
6346 if (flags & TRYAGAIN)
6350 else if (ret == NULL)
6352 *flagp |= flags&(HASWIDTH|POSTPONED);
6353 if (chain == NULL) /* First piece. */
6354 *flagp |= flags&SPSTART;
6357 REGTAIL(pRExC_state, chain, latest);
6362 if (chain == NULL) { /* Loop ran zero times. */
6363 chain = reg_node(pRExC_state, NOTHING);
6368 *flagp |= flags&SIMPLE;
6375 - regpiece - something followed by possible [*+?]
6377 * Note that the branching code sequences used for ? and the general cases
6378 * of * and + are somewhat optimized: they use the same NOTHING node as
6379 * both the endmarker for their branch list and the body of the last branch.
6380 * It might seem that this node could be dispensed with entirely, but the
6381 * endmarker role is not redundant.
6384 S_regpiece(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6387 register regnode *ret;
6389 register char *next;
6391 const char * const origparse = RExC_parse;
6393 I32 max = REG_INFTY;
6395 const char *maxpos = NULL;
6396 GET_RE_DEBUG_FLAGS_DECL;
6398 PERL_ARGS_ASSERT_REGPIECE;
6400 DEBUG_PARSE("piec");
6402 ret = regatom(pRExC_state, &flags,depth+1);
6404 if (flags & TRYAGAIN)
6411 if (op == '{' && regcurly(RExC_parse)) {
6413 parse_start = RExC_parse; /* MJD */
6414 next = RExC_parse + 1;
6415 while (isDIGIT(*next) || *next == ',') {
6424 if (*next == '}') { /* got one */
6428 min = atoi(RExC_parse);
6432 maxpos = RExC_parse;
6434 if (!max && *maxpos != '0')
6435 max = REG_INFTY; /* meaning "infinity" */
6436 else if (max >= REG_INFTY)
6437 vFAIL2("Quantifier in {,} bigger than %d", REG_INFTY - 1);
6439 nextchar(pRExC_state);
6442 if ((flags&SIMPLE)) {
6443 RExC_naughty += 2 + RExC_naughty / 2;
6444 reginsert(pRExC_state, CURLY, ret, depth+1);
6445 Set_Node_Offset(ret, parse_start+1); /* MJD */
6446 Set_Node_Cur_Length(ret);
6449 regnode * const w = reg_node(pRExC_state, WHILEM);
6452 REGTAIL(pRExC_state, ret, w);
6453 if (!SIZE_ONLY && RExC_extralen) {
6454 reginsert(pRExC_state, LONGJMP,ret, depth+1);
6455 reginsert(pRExC_state, NOTHING,ret, depth+1);
6456 NEXT_OFF(ret) = 3; /* Go over LONGJMP. */
6458 reginsert(pRExC_state, CURLYX,ret, depth+1);
6460 Set_Node_Offset(ret, parse_start+1);
6461 Set_Node_Length(ret,
6462 op == '{' ? (RExC_parse - parse_start) : 1);
6464 if (!SIZE_ONLY && RExC_extralen)
6465 NEXT_OFF(ret) = 3; /* Go over NOTHING to LONGJMP. */
6466 REGTAIL(pRExC_state, ret, reg_node(pRExC_state, NOTHING));
6468 RExC_whilem_seen++, RExC_extralen += 3;
6469 RExC_naughty += 4 + RExC_naughty; /* compound interest */
6478 vFAIL("Can't do {n,m} with n > m");
6480 ARG1_SET(ret, (U16)min);
6481 ARG2_SET(ret, (U16)max);
6493 #if 0 /* Now runtime fix should be reliable. */
6495 /* if this is reinstated, don't forget to put this back into perldiag:
6497 =item Regexp *+ operand could be empty at {#} in regex m/%s/
6499 (F) The part of the regexp subject to either the * or + quantifier
6500 could match an empty string. The {#} shows in the regular
6501 expression about where the problem was discovered.
6505 if (!(flags&HASWIDTH) && op != '?')
6506 vFAIL("Regexp *+ operand could be empty");
6509 parse_start = RExC_parse;
6510 nextchar(pRExC_state);
6512 *flagp = (op != '+') ? (WORST|SPSTART|HASWIDTH) : (WORST|HASWIDTH);
6514 if (op == '*' && (flags&SIMPLE)) {
6515 reginsert(pRExC_state, STAR, ret, depth+1);
6519 else if (op == '*') {
6523 else if (op == '+' && (flags&SIMPLE)) {
6524 reginsert(pRExC_state, PLUS, ret, depth+1);
6528 else if (op == '+') {
6532 else if (op == '?') {
6537 if (!SIZE_ONLY && !(flags&(HASWIDTH|POSTPONED)) && max > REG_INFTY/3 && ckWARN(WARN_REGEXP)) {
6539 "%.*s matches null string many times",
6540 (int)(RExC_parse >= origparse ? RExC_parse - origparse : 0),
6544 if (RExC_parse < RExC_end && *RExC_parse == '?') {
6545 nextchar(pRExC_state);
6546 reginsert(pRExC_state, MINMOD, ret, depth+1);
6547 REGTAIL(pRExC_state, ret, ret + NODE_STEP_REGNODE);
6549 #ifndef REG_ALLOW_MINMOD_SUSPEND
6552 if (RExC_parse < RExC_end && *RExC_parse == '+') {
6554 nextchar(pRExC_state);
6555 ender = reg_node(pRExC_state, SUCCEED);
6556 REGTAIL(pRExC_state, ret, ender);
6557 reginsert(pRExC_state, SUSPEND, ret, depth+1);
6559 ender = reg_node(pRExC_state, TAIL);
6560 REGTAIL(pRExC_state, ret, ender);
6564 if (RExC_parse < RExC_end && ISMULT2(RExC_parse)) {
6566 vFAIL("Nested quantifiers");
6573 /* reg_namedseq(pRExC_state,UVp)
6575 This is expected to be called by a parser routine that has
6576 recognized '\N' and needs to handle the rest. RExC_parse is
6577 expected to point at the first char following the N at the time
6580 If valuep is non-null then it is assumed that we are parsing inside
6581 of a charclass definition and the first codepoint in the resolved
6582 string is returned via *valuep and the routine will return NULL.
6583 In this mode if a multichar string is returned from the charnames
6584 handler a warning will be issued, and only the first char in the
6585 sequence will be examined. If the string returned is zero length
6586 then the value of *valuep is undefined and NON-NULL will
6587 be returned to indicate failure. (This will NOT be a valid pointer
6590 If valuep is null then it is assumed that we are parsing normal text
6591 and inserts a new EXACT node into the program containing the resolved
6592 string and returns a pointer to the new node. If the string is
6593 zerolength a NOTHING node is emitted.
6595 On success RExC_parse is set to the char following the endbrace.
6596 Parsing failures will generate a fatal errorvia vFAIL(...)
6598 NOTE: We cache all results from the charnames handler locally in
6599 the RExC_charnames hash (created on first use) to prevent a charnames
6600 handler from playing silly-buggers and returning a short string and
6601 then a long string for a given pattern. Since the regexp program
6602 size is calculated during an initial parse this would result
6603 in a buffer overrun so we cache to prevent the charname result from
6604 changing during the course of the parse.
6608 S_reg_namedseq(pTHX_ RExC_state_t *pRExC_state, UV *valuep, I32 *flagp)
6610 char * name; /* start of the content of the name */
6611 char * endbrace; /* endbrace following the name */
6614 STRLEN len; /* this has various purposes throughout the code */
6615 bool cached = 0; /* if this is true then we shouldn't refcount dev sv_str */
6616 regnode *ret = NULL;
6618 PERL_ARGS_ASSERT_REG_NAMEDSEQ;
6620 if (*RExC_parse != '{' ||
6621 (*RExC_parse == '{' && RExC_parse[1]
6622 && strchr("0123456789", RExC_parse[1])))
6624 GET_RE_DEBUG_FLAGS_DECL;
6626 /* no bare \N in a charclass */
6627 vFAIL("Missing braces on \\N{}");
6629 nextchar(pRExC_state);
6630 ret = reg_node(pRExC_state, REG_ANY);
6631 *flagp |= HASWIDTH|SIMPLE;
6634 Set_Node_Length(ret, 1); /* MJD */
6637 name = RExC_parse+1;
6638 endbrace = strchr(RExC_parse, '}');
6641 vFAIL("Missing right brace on \\N{}");
6643 RExC_parse = endbrace + 1;
6646 /* RExC_parse points at the beginning brace,
6647 endbrace points at the last */
6648 if ( name[0]=='U' && name[1]=='+' ) {
6649 /* its a "Unicode hex" notation {U+89AB} */
6650 I32 fl = PERL_SCAN_ALLOW_UNDERSCORES
6651 | PERL_SCAN_DISALLOW_PREFIX
6652 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
6654 len = (STRLEN)(endbrace - name - 2);
6655 cp = grok_hex(name + 2, &len, &fl, NULL);
6656 if ( len != (STRLEN)(endbrace - name - 2) ) {
6660 if (cp > 0xff) RExC_utf8 = 1;
6665 /* Need to convert to utf8 if either: won't fit into a byte, or the re
6666 * is going to be in utf8 and the representation changes under utf8. */
6667 if (cp > 0xff || (RExC_utf8 && ! UNI_IS_INVARIANT(cp))) {
6668 U8 string[UTF8_MAXBYTES+1];
6671 tmps = uvuni_to_utf8(string, cp);
6672 sv_str = newSVpvn_utf8((char*)string, tmps - string, TRUE);
6673 } else { /* Otherwise, no need for utf8, can skip that step */
6676 sv_str= newSVpvn(&string, 1);
6679 /* fetch the charnames handler for this scope */
6680 HV * const table = GvHV(PL_hintgv);
6682 hv_fetchs(table, "charnames", FALSE) :
6684 SV *cv= cvp ? *cvp : NULL;
6687 /* create an SV with the name as argument */
6688 sv_name = newSVpvn(name, endbrace - name);
6690 if (!table || !(PL_hints & HINT_LOCALIZE_HH)) {
6691 vFAIL2("Constant(\\N{%s}) unknown: "
6692 "(possibly a missing \"use charnames ...\")",
6695 if (!cvp || !SvOK(*cvp)) { /* when $^H{charnames} = undef; */
6696 vFAIL2("Constant(\\N{%s}): "
6697 "$^H{charnames} is not defined",SvPVX(sv_name));
6702 if (!RExC_charnames) {
6703 /* make sure our cache is allocated */
6704 RExC_charnames = newHV();
6705 sv_2mortal(MUTABLE_SV(RExC_charnames));
6707 /* see if we have looked this one up before */
6708 he_str = hv_fetch_ent( RExC_charnames, sv_name, 0, 0 );
6710 sv_str = HeVAL(he_str);
6723 count= call_sv(cv, G_SCALAR);
6725 if (count == 1) { /* XXXX is this right? dmq */
6727 SvREFCNT_inc_simple_void(sv_str);
6735 if ( !sv_str || !SvOK(sv_str) ) {
6736 vFAIL2("Constant(\\N{%s}): Call to &{$^H{charnames}} "
6737 "did not return a defined value",SvPVX(sv_name));
6739 if (hv_store_ent( RExC_charnames, sv_name, sv_str, 0))
6744 char *p = SvPV(sv_str, len);
6747 if ( SvUTF8(sv_str) ) {
6748 *valuep = utf8_to_uvchr((U8*)p, &numlen);
6752 We have to turn on utf8 for high bit chars otherwise
6753 we get failures with
6755 "ss" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i
6756 "SS" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i
6758 This is different from what \x{} would do with the same
6759 codepoint, where the condition is > 0xFF.
6766 /* warn if we havent used the whole string? */
6768 if (numlen<len && SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6770 "Ignoring excess chars from \\N{%s} in character class",
6774 } else if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6776 "Ignoring zero length \\N{%s} in character class",
6781 SvREFCNT_dec(sv_name);
6783 SvREFCNT_dec(sv_str);
6784 return len ? NULL : (regnode *)&len;
6785 } else if(SvCUR(sv_str)) {
6791 char * parse_start = name-3; /* needed for the offsets */
6793 GET_RE_DEBUG_FLAGS_DECL; /* needed for the offsets */
6795 ret = reg_node(pRExC_state,
6796 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
6799 if ( RExC_utf8 && !SvUTF8(sv_str) ) {
6800 sv_utf8_upgrade(sv_str);
6801 } else if ( !RExC_utf8 && SvUTF8(sv_str) ) {
6805 p = SvPV(sv_str, len);
6807 /* len is the length written, charlen is the size the char read */
6808 for ( len = 0; p < pend; p += charlen ) {
6810 UV uvc = utf8_to_uvchr((U8*)p, &charlen);
6812 STRLEN foldlen,numlen;
6813 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
6814 uvc = toFOLD_uni(uvc, tmpbuf, &foldlen);
6815 /* Emit all the Unicode characters. */
6817 for (foldbuf = tmpbuf;
6821 uvc = utf8_to_uvchr(foldbuf, &numlen);
6823 const STRLEN unilen = reguni(pRExC_state, uvc, s);
6826 /* In EBCDIC the numlen
6827 * and unilen can differ. */
6829 if (numlen >= foldlen)
6833 break; /* "Can't happen." */
6836 const STRLEN unilen = reguni(pRExC_state, uvc, s);
6848 RExC_size += STR_SZ(len);
6851 RExC_emit += STR_SZ(len);
6853 Set_Node_Cur_Length(ret); /* MJD */
6855 nextchar(pRExC_state);
6856 } else { /* zero length */
6857 ret = reg_node(pRExC_state,NOTHING);
6860 SvREFCNT_dec(sv_str);
6863 SvREFCNT_dec(sv_name);
6873 * It returns the code point in utf8 for the value in *encp.
6874 * value: a code value in the source encoding
6875 * encp: a pointer to an Encode object
6877 * If the result from Encode is not a single character,
6878 * it returns U+FFFD (Replacement character) and sets *encp to NULL.
6881 S_reg_recode(pTHX_ const char value, SV **encp)
6884 SV * const sv = newSVpvn_flags(&value, numlen, SVs_TEMP);
6885 const char * const s = *encp ? sv_recode_to_utf8(sv, *encp) : SvPVX(sv);
6886 const STRLEN newlen = SvCUR(sv);
6887 UV uv = UNICODE_REPLACEMENT;
6889 PERL_ARGS_ASSERT_REG_RECODE;
6893 ? utf8n_to_uvchr((U8*)s, newlen, &numlen, UTF8_ALLOW_DEFAULT)
6896 if (!newlen || numlen != newlen) {
6897 uv = UNICODE_REPLACEMENT;
6905 - regatom - the lowest level
6907 Try to identify anything special at the start of the pattern. If there
6908 is, then handle it as required. This may involve generating a single regop,
6909 such as for an assertion; or it may involve recursing, such as to
6910 handle a () structure.
6912 If the string doesn't start with something special then we gobble up
6913 as much literal text as we can.
6915 Once we have been able to handle whatever type of thing started the
6916 sequence, we return.
6918 Note: we have to be careful with escapes, as they can be both literal
6919 and special, and in the case of \10 and friends can either, depending
6920 on context. Specifically there are two seperate switches for handling
6921 escape sequences, with the one for handling literal escapes requiring
6922 a dummy entry for all of the special escapes that are actually handled
6927 S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6930 register regnode *ret = NULL;
6932 char *parse_start = RExC_parse;
6933 GET_RE_DEBUG_FLAGS_DECL;
6934 DEBUG_PARSE("atom");
6935 *flagp = WORST; /* Tentatively. */
6937 PERL_ARGS_ASSERT_REGATOM;
6940 switch ((U8)*RExC_parse) {
6942 RExC_seen_zerolen++;
6943 nextchar(pRExC_state);
6944 if (RExC_flags & RXf_PMf_MULTILINE)
6945 ret = reg_node(pRExC_state, MBOL);
6946 else if (RExC_flags & RXf_PMf_SINGLELINE)
6947 ret = reg_node(pRExC_state, SBOL);
6949 ret = reg_node(pRExC_state, BOL);
6950 Set_Node_Length(ret, 1); /* MJD */
6953 nextchar(pRExC_state);
6955 RExC_seen_zerolen++;
6956 if (RExC_flags & RXf_PMf_MULTILINE)
6957 ret = reg_node(pRExC_state, MEOL);
6958 else if (RExC_flags & RXf_PMf_SINGLELINE)
6959 ret = reg_node(pRExC_state, SEOL);
6961 ret = reg_node(pRExC_state, EOL);
6962 Set_Node_Length(ret, 1); /* MJD */
6965 nextchar(pRExC_state);
6966 if (RExC_flags & RXf_PMf_SINGLELINE)
6967 ret = reg_node(pRExC_state, SANY);
6969 ret = reg_node(pRExC_state, REG_ANY);
6970 *flagp |= HASWIDTH|SIMPLE;
6972 Set_Node_Length(ret, 1); /* MJD */
6976 char * const oregcomp_parse = ++RExC_parse;
6977 ret = regclass(pRExC_state,depth+1);
6978 if (*RExC_parse != ']') {
6979 RExC_parse = oregcomp_parse;
6980 vFAIL("Unmatched [");
6982 nextchar(pRExC_state);
6983 *flagp |= HASWIDTH|SIMPLE;
6984 Set_Node_Length(ret, RExC_parse - oregcomp_parse + 1); /* MJD */
6988 nextchar(pRExC_state);
6989 ret = reg(pRExC_state, 1, &flags,depth+1);
6991 if (flags & TRYAGAIN) {
6992 if (RExC_parse == RExC_end) {
6993 /* Make parent create an empty node if needed. */
7001 *flagp |= flags&(HASWIDTH|SPSTART|SIMPLE|POSTPONED);
7005 if (flags & TRYAGAIN) {
7009 vFAIL("Internal urp");
7010 /* Supposed to be caught earlier. */
7013 if (!regcurly(RExC_parse)) {
7022 vFAIL("Quantifier follows nothing");
7030 len=0; /* silence a spurious compiler warning */
7031 if ((cp = what_len_TRICKYFOLD_safe(RExC_parse,RExC_end,UTF,len))) {
7032 *flagp |= HASWIDTH; /* could be SIMPLE too, but needs a handler in regexec.regrepeat */
7033 RExC_parse+=len-1; /* we get one from nextchar() as well. :-( */
7034 ret = reganode(pRExC_state, FOLDCHAR, cp);
7035 Set_Node_Length(ret, 1); /* MJD */
7036 nextchar(pRExC_state); /* kill whitespace under /x */
7044 This switch handles escape sequences that resolve to some kind
7045 of special regop and not to literal text. Escape sequnces that
7046 resolve to literal text are handled below in the switch marked
7049 Every entry in this switch *must* have a corresponding entry
7050 in the literal escape switch. However, the opposite is not
7051 required, as the default for this switch is to jump to the
7052 literal text handling code.
7054 switch ((U8)*++RExC_parse) {
7059 /* Special Escapes */
7061 RExC_seen_zerolen++;
7062 ret = reg_node(pRExC_state, SBOL);
7064 goto finish_meta_pat;
7066 ret = reg_node(pRExC_state, GPOS);
7067 RExC_seen |= REG_SEEN_GPOS;
7069 goto finish_meta_pat;
7071 RExC_seen_zerolen++;
7072 ret = reg_node(pRExC_state, KEEPS);
7074 /* XXX:dmq : disabling in-place substitution seems to
7075 * be necessary here to avoid cases of memory corruption, as
7076 * with: C<$_="x" x 80; s/x\K/y/> -- rgs
7078 RExC_seen |= REG_SEEN_LOOKBEHIND;
7079 goto finish_meta_pat;
7081 ret = reg_node(pRExC_state, SEOL);
7083 RExC_seen_zerolen++; /* Do not optimize RE away */
7084 goto finish_meta_pat;
7086 ret = reg_node(pRExC_state, EOS);
7088 RExC_seen_zerolen++; /* Do not optimize RE away */
7089 goto finish_meta_pat;
7091 ret = reg_node(pRExC_state, CANY);
7092 RExC_seen |= REG_SEEN_CANY;
7093 *flagp |= HASWIDTH|SIMPLE;
7094 goto finish_meta_pat;
7096 ret = reg_node(pRExC_state, CLUMP);
7098 goto finish_meta_pat;
7100 ret = reg_node(pRExC_state, (U8)(LOC ? ALNUML : ALNUM));
7101 *flagp |= HASWIDTH|SIMPLE;
7102 goto finish_meta_pat;
7104 ret = reg_node(pRExC_state, (U8)(LOC ? NALNUML : NALNUM));
7105 *flagp |= HASWIDTH|SIMPLE;
7106 goto finish_meta_pat;
7108 RExC_seen_zerolen++;
7109 RExC_seen |= REG_SEEN_LOOKBEHIND;
7110 ret = reg_node(pRExC_state, (U8)(LOC ? BOUNDL : BOUND));
7112 goto finish_meta_pat;
7114 RExC_seen_zerolen++;
7115 RExC_seen |= REG_SEEN_LOOKBEHIND;
7116 ret = reg_node(pRExC_state, (U8)(LOC ? NBOUNDL : NBOUND));
7118 goto finish_meta_pat;
7120 ret = reg_node(pRExC_state, (U8)(LOC ? SPACEL : SPACE));
7121 *flagp |= HASWIDTH|SIMPLE;
7122 goto finish_meta_pat;
7124 ret = reg_node(pRExC_state, (U8)(LOC ? NSPACEL : NSPACE));
7125 *flagp |= HASWIDTH|SIMPLE;
7126 goto finish_meta_pat;
7128 ret = reg_node(pRExC_state, DIGIT);
7129 *flagp |= HASWIDTH|SIMPLE;
7130 goto finish_meta_pat;
7132 ret = reg_node(pRExC_state, NDIGIT);
7133 *flagp |= HASWIDTH|SIMPLE;
7134 goto finish_meta_pat;
7136 ret = reg_node(pRExC_state, LNBREAK);
7137 *flagp |= HASWIDTH|SIMPLE;
7138 goto finish_meta_pat;
7140 ret = reg_node(pRExC_state, HORIZWS);
7141 *flagp |= HASWIDTH|SIMPLE;
7142 goto finish_meta_pat;
7144 ret = reg_node(pRExC_state, NHORIZWS);
7145 *flagp |= HASWIDTH|SIMPLE;
7146 goto finish_meta_pat;
7148 ret = reg_node(pRExC_state, VERTWS);
7149 *flagp |= HASWIDTH|SIMPLE;
7150 goto finish_meta_pat;
7152 ret = reg_node(pRExC_state, NVERTWS);
7153 *flagp |= HASWIDTH|SIMPLE;
7155 nextchar(pRExC_state);
7156 Set_Node_Length(ret, 2); /* MJD */
7161 char* const oldregxend = RExC_end;
7163 char* parse_start = RExC_parse - 2;
7166 if (RExC_parse[1] == '{') {
7167 /* a lovely hack--pretend we saw [\pX] instead */
7168 RExC_end = strchr(RExC_parse, '}');
7170 const U8 c = (U8)*RExC_parse;
7172 RExC_end = oldregxend;
7173 vFAIL2("Missing right brace on \\%c{}", c);
7178 RExC_end = RExC_parse + 2;
7179 if (RExC_end > oldregxend)
7180 RExC_end = oldregxend;
7184 ret = regclass(pRExC_state,depth+1);
7186 RExC_end = oldregxend;
7189 Set_Node_Offset(ret, parse_start + 2);
7190 Set_Node_Cur_Length(ret);
7191 nextchar(pRExC_state);
7192 *flagp |= HASWIDTH|SIMPLE;
7196 /* Handle \N and \N{NAME} here and not below because it can be
7197 multicharacter. join_exact() will join them up later on.
7198 Also this makes sure that things like /\N{BLAH}+/ and
7199 \N{BLAH} being multi char Just Happen. dmq*/
7201 ret= reg_namedseq(pRExC_state, NULL, flagp);
7203 case 'k': /* Handle \k<NAME> and \k'NAME' */
7206 char ch= RExC_parse[1];
7207 if (ch != '<' && ch != '\'' && ch != '{') {
7209 vFAIL2("Sequence %.2s... not terminated",parse_start);
7211 /* this pretty much dupes the code for (?P=...) in reg(), if
7212 you change this make sure you change that */
7213 char* name_start = (RExC_parse += 2);
7215 SV *sv_dat = reg_scan_name(pRExC_state,
7216 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7217 ch= (ch == '<') ? '>' : (ch == '{') ? '}' : '\'';
7218 if (RExC_parse == name_start || *RExC_parse != ch)
7219 vFAIL2("Sequence %.3s... not terminated",parse_start);
7222 num = add_data( pRExC_state, 1, "S" );
7223 RExC_rxi->data->data[num]=(void*)sv_dat;
7224 SvREFCNT_inc_simple_void(sv_dat);
7228 ret = reganode(pRExC_state,
7229 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
7233 /* override incorrect value set in reganode MJD */
7234 Set_Node_Offset(ret, parse_start+1);
7235 Set_Node_Cur_Length(ret); /* MJD */
7236 nextchar(pRExC_state);
7242 case '1': case '2': case '3': case '4':
7243 case '5': case '6': case '7': case '8': case '9':
7246 bool isg = *RExC_parse == 'g';
7251 if (*RExC_parse == '{') {
7255 if (*RExC_parse == '-') {
7259 if (hasbrace && !isDIGIT(*RExC_parse)) {
7260 if (isrel) RExC_parse--;
7262 goto parse_named_seq;
7264 num = atoi(RExC_parse);
7265 if (isg && num == 0)
7266 vFAIL("Reference to invalid group 0");
7268 num = RExC_npar - num;
7270 vFAIL("Reference to nonexistent or unclosed group");
7272 if (!isg && num > 9 && num >= RExC_npar)
7275 char * const parse_start = RExC_parse - 1; /* MJD */
7276 while (isDIGIT(*RExC_parse))
7278 if (parse_start == RExC_parse - 1)
7279 vFAIL("Unterminated \\g... pattern");
7281 if (*RExC_parse != '}')
7282 vFAIL("Unterminated \\g{...} pattern");
7286 if (num > (I32)RExC_rx->nparens)
7287 vFAIL("Reference to nonexistent group");
7290 ret = reganode(pRExC_state,
7291 (U8)(FOLD ? (LOC ? REFFL : REFF) : REF),
7295 /* override incorrect value set in reganode MJD */
7296 Set_Node_Offset(ret, parse_start+1);
7297 Set_Node_Cur_Length(ret); /* MJD */
7299 nextchar(pRExC_state);
7304 if (RExC_parse >= RExC_end)
7305 FAIL("Trailing \\");
7308 /* Do not generate "unrecognized" warnings here, we fall
7309 back into the quick-grab loop below */
7316 if (RExC_flags & RXf_PMf_EXTENDED) {
7317 if ( reg_skipcomment( pRExC_state ) )
7324 register STRLEN len;
7329 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
7331 parse_start = RExC_parse - 1;
7337 ret = reg_node(pRExC_state,
7338 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
7340 for (len = 0, p = RExC_parse - 1;
7341 len < 127 && p < RExC_end;
7344 char * const oldp = p;
7346 if (RExC_flags & RXf_PMf_EXTENDED)
7347 p = regwhite( pRExC_state, p );
7352 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7353 goto normal_default;
7363 /* Literal Escapes Switch
7365 This switch is meant to handle escape sequences that
7366 resolve to a literal character.
7368 Every escape sequence that represents something
7369 else, like an assertion or a char class, is handled
7370 in the switch marked 'Special Escapes' above in this
7371 routine, but also has an entry here as anything that
7372 isn't explicitly mentioned here will be treated as
7373 an unescaped equivalent literal.
7377 /* These are all the special escapes. */
7381 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7382 goto normal_default;
7383 case 'A': /* Start assertion */
7384 case 'b': case 'B': /* Word-boundary assertion*/
7385 case 'C': /* Single char !DANGEROUS! */
7386 case 'd': case 'D': /* digit class */
7387 case 'g': case 'G': /* generic-backref, pos assertion */
7388 case 'h': case 'H': /* HORIZWS */
7389 case 'k': case 'K': /* named backref, keep marker */
7390 case 'N': /* named char sequence */
7391 case 'p': case 'P': /* Unicode property */
7392 case 'R': /* LNBREAK */
7393 case 's': case 'S': /* space class */
7394 case 'v': case 'V': /* VERTWS */
7395 case 'w': case 'W': /* word class */
7396 case 'X': /* eXtended Unicode "combining character sequence" */
7397 case 'z': case 'Z': /* End of line/string assertion */
7401 /* Anything after here is an escape that resolves to a
7402 literal. (Except digits, which may or may not)
7421 ender = ASCII_TO_NATIVE('\033');
7425 ender = ASCII_TO_NATIVE('\007');
7430 char* const e = strchr(p, '}');
7434 vFAIL("Missing right brace on \\x{}");
7437 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
7438 | PERL_SCAN_DISALLOW_PREFIX;
7439 STRLEN numlen = e - p - 1;
7440 ender = grok_hex(p + 1, &numlen, &flags, NULL);
7447 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
7449 ender = grok_hex(p, &numlen, &flags, NULL);
7452 if (PL_encoding && ender < 0x100)
7453 goto recode_encoding;
7457 ender = UCHARAT(p++);
7458 ender = toCTRL(ender);
7460 case '0': case '1': case '2': case '3':case '4':
7461 case '5': case '6': case '7': case '8':case '9':
7463 (isDIGIT(p[1]) && atoi(p) >= RExC_npar) ) {
7466 ender = grok_oct(p, &numlen, &flags, NULL);
7468 /* An octal above 0xff is interpreted differently
7469 * depending on if the re is in utf8 or not. If it
7470 * is in utf8, the value will be itself, otherwise
7471 * it is interpreted as modulo 0x100. It has been
7472 * decided to discourage the use of octal above the
7473 * single-byte range. For now, warn only when
7474 * it ends up modulo */
7475 if (SIZE_ONLY && ender >= 0x100
7476 && ! UTF && ! PL_encoding
7477 && ckWARN2(WARN_DEPRECATED, WARN_REGEXP)) {
7478 vWARNdep(p, "Use of octal value above 377 is deprecated");
7486 if (PL_encoding && ender < 0x100)
7487 goto recode_encoding;
7491 SV* enc = PL_encoding;
7492 ender = reg_recode((const char)(U8)ender, &enc);
7493 if (!enc && SIZE_ONLY && ckWARN(WARN_REGEXP))
7494 vWARN(p, "Invalid escape in the specified encoding");
7500 FAIL("Trailing \\");
7503 if (!SIZE_ONLY&& isALPHA(*p) && ckWARN(WARN_REGEXP))
7504 vWARN2(p + 1, "Unrecognized escape \\%c passed through", UCHARAT(p));
7505 goto normal_default;
7510 if (UTF8_IS_START(*p) && UTF) {
7512 ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
7513 &numlen, UTF8_ALLOW_DEFAULT);
7520 if ( RExC_flags & RXf_PMf_EXTENDED)
7521 p = regwhite( pRExC_state, p );
7523 /* Prime the casefolded buffer. */
7524 ender = toFOLD_uni(ender, tmpbuf, &foldlen);
7526 if (p < RExC_end && ISMULT2(p)) { /* Back off on ?+*. */
7531 /* Emit all the Unicode characters. */
7533 for (foldbuf = tmpbuf;
7535 foldlen -= numlen) {
7536 ender = utf8_to_uvchr(foldbuf, &numlen);
7538 const STRLEN unilen = reguni(pRExC_state, ender, s);
7541 /* In EBCDIC the numlen
7542 * and unilen can differ. */
7544 if (numlen >= foldlen)
7548 break; /* "Can't happen." */
7552 const STRLEN unilen = reguni(pRExC_state, ender, s);
7561 REGC((char)ender, s++);
7567 /* Emit all the Unicode characters. */
7569 for (foldbuf = tmpbuf;
7571 foldlen -= numlen) {
7572 ender = utf8_to_uvchr(foldbuf, &numlen);
7574 const STRLEN unilen = reguni(pRExC_state, ender, s);
7577 /* In EBCDIC the numlen
7578 * and unilen can differ. */
7580 if (numlen >= foldlen)
7588 const STRLEN unilen = reguni(pRExC_state, ender, s);
7597 REGC((char)ender, s++);
7601 Set_Node_Cur_Length(ret); /* MJD */
7602 nextchar(pRExC_state);
7604 /* len is STRLEN which is unsigned, need to copy to signed */
7607 vFAIL("Internal disaster");
7611 if (len == 1 && UNI_IS_INVARIANT(ender))
7615 RExC_size += STR_SZ(len);
7618 RExC_emit += STR_SZ(len);
7628 S_regwhite( RExC_state_t *pRExC_state, char *p )
7630 const char *e = RExC_end;
7632 PERL_ARGS_ASSERT_REGWHITE;
7637 else if (*p == '#') {
7646 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
7654 /* Parse POSIX character classes: [[:foo:]], [[=foo=]], [[.foo.]].
7655 Character classes ([:foo:]) can also be negated ([:^foo:]).
7656 Returns a named class id (ANYOF_XXX) if successful, -1 otherwise.
7657 Equivalence classes ([=foo=]) and composites ([.foo.]) are parsed,
7658 but trigger failures because they are currently unimplemented. */
7660 #define POSIXCC_DONE(c) ((c) == ':')
7661 #define POSIXCC_NOTYET(c) ((c) == '=' || (c) == '.')
7662 #define POSIXCC(c) (POSIXCC_DONE(c) || POSIXCC_NOTYET(c))
7665 S_regpposixcc(pTHX_ RExC_state_t *pRExC_state, I32 value)
7668 I32 namedclass = OOB_NAMEDCLASS;
7670 PERL_ARGS_ASSERT_REGPPOSIXCC;
7672 if (value == '[' && RExC_parse + 1 < RExC_end &&
7673 /* I smell either [: or [= or [. -- POSIX has been here, right? */
7674 POSIXCC(UCHARAT(RExC_parse))) {
7675 const char c = UCHARAT(RExC_parse);
7676 char* const s = RExC_parse++;
7678 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != c)
7680 if (RExC_parse == RExC_end)
7681 /* Grandfather lone [:, [=, [. */
7684 const char* const t = RExC_parse++; /* skip over the c */
7687 if (UCHARAT(RExC_parse) == ']') {
7688 const char *posixcc = s + 1;
7689 RExC_parse++; /* skip over the ending ] */
7692 const I32 complement = *posixcc == '^' ? *posixcc++ : 0;
7693 const I32 skip = t - posixcc;
7695 /* Initially switch on the length of the name. */
7698 if (memEQ(posixcc, "word", 4)) /* this is not POSIX, this is the Perl \w */
7699 namedclass = complement ? ANYOF_NALNUM : ANYOF_ALNUM;
7702 /* Names all of length 5. */
7703 /* alnum alpha ascii blank cntrl digit graph lower
7704 print punct space upper */
7705 /* Offset 4 gives the best switch position. */
7706 switch (posixcc[4]) {
7708 if (memEQ(posixcc, "alph", 4)) /* alpha */
7709 namedclass = complement ? ANYOF_NALPHA : ANYOF_ALPHA;
7712 if (memEQ(posixcc, "spac", 4)) /* space */
7713 namedclass = complement ? ANYOF_NPSXSPC : ANYOF_PSXSPC;
7716 if (memEQ(posixcc, "grap", 4)) /* graph */
7717 namedclass = complement ? ANYOF_NGRAPH : ANYOF_GRAPH;
7720 if (memEQ(posixcc, "asci", 4)) /* ascii */
7721 namedclass = complement ? ANYOF_NASCII : ANYOF_ASCII;
7724 if (memEQ(posixcc, "blan", 4)) /* blank */
7725 namedclass = complement ? ANYOF_NBLANK : ANYOF_BLANK;
7728 if (memEQ(posixcc, "cntr", 4)) /* cntrl */
7729 namedclass = complement ? ANYOF_NCNTRL : ANYOF_CNTRL;
7732 if (memEQ(posixcc, "alnu", 4)) /* alnum */
7733 namedclass = complement ? ANYOF_NALNUMC : ANYOF_ALNUMC;
7736 if (memEQ(posixcc, "lowe", 4)) /* lower */
7737 namedclass = complement ? ANYOF_NLOWER : ANYOF_LOWER;
7738 else if (memEQ(posixcc, "uppe", 4)) /* upper */
7739 namedclass = complement ? ANYOF_NUPPER : ANYOF_UPPER;
7742 if (memEQ(posixcc, "digi", 4)) /* digit */
7743 namedclass = complement ? ANYOF_NDIGIT : ANYOF_DIGIT;
7744 else if (memEQ(posixcc, "prin", 4)) /* print */
7745 namedclass = complement ? ANYOF_NPRINT : ANYOF_PRINT;
7746 else if (memEQ(posixcc, "punc", 4)) /* punct */
7747 namedclass = complement ? ANYOF_NPUNCT : ANYOF_PUNCT;
7752 if (memEQ(posixcc, "xdigit", 6))
7753 namedclass = complement ? ANYOF_NXDIGIT : ANYOF_XDIGIT;
7757 if (namedclass == OOB_NAMEDCLASS)
7758 Simple_vFAIL3("POSIX class [:%.*s:] unknown",
7760 assert (posixcc[skip] == ':');
7761 assert (posixcc[skip+1] == ']');
7762 } else if (!SIZE_ONLY) {
7763 /* [[=foo=]] and [[.foo.]] are still future. */
7765 /* adjust RExC_parse so the warning shows after
7767 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse) != ']')
7769 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7772 /* Maternal grandfather:
7773 * "[:" ending in ":" but not in ":]" */
7783 S_checkposixcc(pTHX_ RExC_state_t *pRExC_state)
7787 PERL_ARGS_ASSERT_CHECKPOSIXCC;
7789 if (POSIXCC(UCHARAT(RExC_parse))) {
7790 const char *s = RExC_parse;
7791 const char c = *s++;
7795 if (*s && c == *s && s[1] == ']') {
7796 if (ckWARN(WARN_REGEXP))
7798 "POSIX syntax [%c %c] belongs inside character classes",
7801 /* [[=foo=]] and [[.foo.]] are still future. */
7802 if (POSIXCC_NOTYET(c)) {
7803 /* adjust RExC_parse so the error shows after
7805 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse++) != ']')
7807 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7814 #define _C_C_T_(NAME,TEST,WORD) \
7817 ANYOF_CLASS_SET(ret, ANYOF_##NAME); \
7819 for (value = 0; value < 256; value++) \
7821 ANYOF_BITMAP_SET(ret, value); \
7826 case ANYOF_N##NAME: \
7828 ANYOF_CLASS_SET(ret, ANYOF_N##NAME); \
7830 for (value = 0; value < 256; value++) \
7832 ANYOF_BITMAP_SET(ret, value); \
7838 #define _C_C_T_NOLOC_(NAME,TEST,WORD) \
7840 for (value = 0; value < 256; value++) \
7842 ANYOF_BITMAP_SET(ret, value); \
7846 case ANYOF_N##NAME: \
7847 for (value = 0; value < 256; value++) \
7849 ANYOF_BITMAP_SET(ret, value); \
7855 We dont use PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS as the direct test
7856 so that it is possible to override the option here without having to
7857 rebuild the entire core. as we are required to do if we change regcomp.h
7858 which is where PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS is defined.
7860 #if PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS
7861 #define BROKEN_UNICODE_CHARCLASS_MAPPINGS
7864 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
7865 #define POSIX_CC_UNI_NAME(CCNAME) CCNAME
7867 #define POSIX_CC_UNI_NAME(CCNAME) "Posix" CCNAME
7871 parse a class specification and produce either an ANYOF node that
7872 matches the pattern or if the pattern matches a single char only and
7873 that char is < 256 and we are case insensitive then we produce an
7878 S_regclass(pTHX_ RExC_state_t *pRExC_state, U32 depth)
7881 register UV nextvalue;
7882 register IV prevvalue = OOB_UNICODE;
7883 register IV range = 0;
7884 UV value = 0; /* XXX:dmq: needs to be referenceable (unfortunately) */
7885 register regnode *ret;
7888 char *rangebegin = NULL;
7889 bool need_class = 0;
7892 bool optimize_invert = TRUE;
7893 AV* unicode_alternate = NULL;
7895 UV literal_endpoint = 0;
7897 UV stored = 0; /* number of chars stored in the class */
7899 regnode * const orig_emit = RExC_emit; /* Save the original RExC_emit in
7900 case we need to change the emitted regop to an EXACT. */
7901 const char * orig_parse = RExC_parse;
7902 GET_RE_DEBUG_FLAGS_DECL;
7904 PERL_ARGS_ASSERT_REGCLASS;
7906 PERL_UNUSED_ARG(depth);
7909 DEBUG_PARSE("clas");
7911 /* Assume we are going to generate an ANYOF node. */
7912 ret = reganode(pRExC_state, ANYOF, 0);
7915 ANYOF_FLAGS(ret) = 0;
7917 if (UCHARAT(RExC_parse) == '^') { /* Complement of range. */
7921 ANYOF_FLAGS(ret) |= ANYOF_INVERT;
7925 RExC_size += ANYOF_SKIP;
7926 listsv = &PL_sv_undef; /* For code scanners: listsv always non-NULL. */
7929 RExC_emit += ANYOF_SKIP;
7931 ANYOF_FLAGS(ret) |= ANYOF_FOLD;
7933 ANYOF_FLAGS(ret) |= ANYOF_LOCALE;
7934 ANYOF_BITMAP_ZERO(ret);
7935 listsv = newSVpvs("# comment\n");
7938 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
7940 if (!SIZE_ONLY && POSIXCC(nextvalue))
7941 checkposixcc(pRExC_state);
7943 /* allow 1st char to be ] (allowing it to be - is dealt with later) */
7944 if (UCHARAT(RExC_parse) == ']')
7948 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != ']') {
7952 namedclass = OOB_NAMEDCLASS; /* initialize as illegal */
7955 rangebegin = RExC_parse;
7957 value = utf8n_to_uvchr((U8*)RExC_parse,
7958 RExC_end - RExC_parse,
7959 &numlen, UTF8_ALLOW_DEFAULT);
7960 RExC_parse += numlen;
7963 value = UCHARAT(RExC_parse++);
7965 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
7966 if (value == '[' && POSIXCC(nextvalue))
7967 namedclass = regpposixcc(pRExC_state, value);
7968 else if (value == '\\') {
7970 value = utf8n_to_uvchr((U8*)RExC_parse,
7971 RExC_end - RExC_parse,
7972 &numlen, UTF8_ALLOW_DEFAULT);
7973 RExC_parse += numlen;
7976 value = UCHARAT(RExC_parse++);
7977 /* Some compilers cannot handle switching on 64-bit integer
7978 * values, therefore value cannot be an UV. Yes, this will
7979 * be a problem later if we want switch on Unicode.
7980 * A similar issue a little bit later when switching on
7981 * namedclass. --jhi */
7982 switch ((I32)value) {
7983 case 'w': namedclass = ANYOF_ALNUM; break;
7984 case 'W': namedclass = ANYOF_NALNUM; break;
7985 case 's': namedclass = ANYOF_SPACE; break;
7986 case 'S': namedclass = ANYOF_NSPACE; break;
7987 case 'd': namedclass = ANYOF_DIGIT; break;
7988 case 'D': namedclass = ANYOF_NDIGIT; break;
7989 case 'v': namedclass = ANYOF_VERTWS; break;
7990 case 'V': namedclass = ANYOF_NVERTWS; break;
7991 case 'h': namedclass = ANYOF_HORIZWS; break;
7992 case 'H': namedclass = ANYOF_NHORIZWS; break;
7993 case 'N': /* Handle \N{NAME} in class */
7995 /* We only pay attention to the first char of
7996 multichar strings being returned. I kinda wonder
7997 if this makes sense as it does change the behaviour
7998 from earlier versions, OTOH that behaviour was broken
8000 UV v; /* value is register so we cant & it /grrr */
8001 if (reg_namedseq(pRExC_state, &v, NULL)) {
8011 if (RExC_parse >= RExC_end)
8012 vFAIL2("Empty \\%c{}", (U8)value);
8013 if (*RExC_parse == '{') {
8014 const U8 c = (U8)value;
8015 e = strchr(RExC_parse++, '}');
8017 vFAIL2("Missing right brace on \\%c{}", c);
8018 while (isSPACE(UCHARAT(RExC_parse)))
8020 if (e == RExC_parse)
8021 vFAIL2("Empty \\%c{}", c);
8023 while (isSPACE(UCHARAT(RExC_parse + n - 1)))
8031 if (UCHARAT(RExC_parse) == '^') {
8034 value = value == 'p' ? 'P' : 'p'; /* toggle */
8035 while (isSPACE(UCHARAT(RExC_parse))) {
8040 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::%.*s\n",
8041 (value=='p' ? '+' : '!'), (int)n, RExC_parse);
8044 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8045 namedclass = ANYOF_MAX; /* no official name, but it's named */
8048 case 'n': value = '\n'; break;
8049 case 'r': value = '\r'; break;
8050 case 't': value = '\t'; break;
8051 case 'f': value = '\f'; break;
8052 case 'b': value = '\b'; break;
8053 case 'e': value = ASCII_TO_NATIVE('\033');break;
8054 case 'a': value = ASCII_TO_NATIVE('\007');break;
8056 if (*RExC_parse == '{') {
8057 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
8058 | PERL_SCAN_DISALLOW_PREFIX;
8059 char * const e = strchr(RExC_parse++, '}');
8061 vFAIL("Missing right brace on \\x{}");
8063 numlen = e - RExC_parse;
8064 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8068 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
8070 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8071 RExC_parse += numlen;
8073 if (PL_encoding && value < 0x100)
8074 goto recode_encoding;
8077 value = UCHARAT(RExC_parse++);
8078 value = toCTRL(value);
8080 case '0': case '1': case '2': case '3': case '4':
8081 case '5': case '6': case '7': case '8': case '9':
8085 value = grok_oct(--RExC_parse, &numlen, &flags, NULL);
8086 RExC_parse += numlen;
8087 if (PL_encoding && value < 0x100)
8088 goto recode_encoding;
8093 SV* enc = PL_encoding;
8094 value = reg_recode((const char)(U8)value, &enc);
8095 if (!enc && SIZE_ONLY && ckWARN(WARN_REGEXP))
8097 "Invalid escape in the specified encoding");
8101 if (!SIZE_ONLY && isALPHA(value) && ckWARN(WARN_REGEXP))
8103 "Unrecognized escape \\%c in character class passed through",
8107 } /* end of \blah */
8113 if (namedclass > OOB_NAMEDCLASS) { /* this is a named class \blah */
8115 if (!SIZE_ONLY && !need_class)
8116 ANYOF_CLASS_ZERO(ret);
8120 /* a bad range like a-\d, a-[:digit:] ? */
8123 if (ckWARN(WARN_REGEXP)) {
8125 RExC_parse >= rangebegin ?
8126 RExC_parse - rangebegin : 0;
8128 "False [] range \"%*.*s\"",
8131 if (prevvalue < 256) {
8132 ANYOF_BITMAP_SET(ret, prevvalue);
8133 ANYOF_BITMAP_SET(ret, '-');
8136 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8137 Perl_sv_catpvf(aTHX_ listsv,
8138 "%04"UVxf"\n%04"UVxf"\n", (UV)prevvalue, (UV) '-');
8142 range = 0; /* this was not a true range */
8148 const char *what = NULL;
8151 if (namedclass > OOB_NAMEDCLASS)
8152 optimize_invert = FALSE;
8153 /* Possible truncation here but in some 64-bit environments
8154 * the compiler gets heartburn about switch on 64-bit values.
8155 * A similar issue a little earlier when switching on value.
8157 switch ((I32)namedclass) {
8159 case _C_C_T_(ALNUMC, isALNUMC(value), POSIX_CC_UNI_NAME("Alnum"));
8160 case _C_C_T_(ALPHA, isALPHA(value), POSIX_CC_UNI_NAME("Alpha"));
8161 case _C_C_T_(BLANK, isBLANK(value), POSIX_CC_UNI_NAME("Blank"));
8162 case _C_C_T_(CNTRL, isCNTRL(value), POSIX_CC_UNI_NAME("Cntrl"));
8163 case _C_C_T_(GRAPH, isGRAPH(value), POSIX_CC_UNI_NAME("Graph"));
8164 case _C_C_T_(LOWER, isLOWER(value), POSIX_CC_UNI_NAME("Lower"));
8165 case _C_C_T_(PRINT, isPRINT(value), POSIX_CC_UNI_NAME("Print"));
8166 case _C_C_T_(PSXSPC, isPSXSPC(value), POSIX_CC_UNI_NAME("Space"));
8167 case _C_C_T_(PUNCT, isPUNCT(value), POSIX_CC_UNI_NAME("Punct"));
8168 case _C_C_T_(UPPER, isUPPER(value), POSIX_CC_UNI_NAME("Upper"));
8169 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
8170 case _C_C_T_(ALNUM, isALNUM(value), "Word");
8171 case _C_C_T_(SPACE, isSPACE(value), "SpacePerl");
8173 case _C_C_T_(SPACE, isSPACE(value), "PerlSpace");
8174 case _C_C_T_(ALNUM, isALNUM(value), "PerlWord");
8176 case _C_C_T_(XDIGIT, isXDIGIT(value), "XDigit");
8177 case _C_C_T_NOLOC_(VERTWS, is_VERTWS_latin1(&value), "VertSpace");
8178 case _C_C_T_NOLOC_(HORIZWS, is_HORIZWS_latin1(&value), "HorizSpace");
8181 ANYOF_CLASS_SET(ret, ANYOF_ASCII);
8184 for (value = 0; value < 128; value++)
8185 ANYOF_BITMAP_SET(ret, value);
8187 for (value = 0; value < 256; value++) {
8189 ANYOF_BITMAP_SET(ret, value);
8198 ANYOF_CLASS_SET(ret, ANYOF_NASCII);
8201 for (value = 128; value < 256; value++)
8202 ANYOF_BITMAP_SET(ret, value);
8204 for (value = 0; value < 256; value++) {
8205 if (!isASCII(value))
8206 ANYOF_BITMAP_SET(ret, value);
8215 ANYOF_CLASS_SET(ret, ANYOF_DIGIT);
8217 /* consecutive digits assumed */
8218 for (value = '0'; value <= '9'; value++)
8219 ANYOF_BITMAP_SET(ret, value);
8222 what = POSIX_CC_UNI_NAME("Digit");
8226 ANYOF_CLASS_SET(ret, ANYOF_NDIGIT);
8228 /* consecutive digits assumed */
8229 for (value = 0; value < '0'; value++)
8230 ANYOF_BITMAP_SET(ret, value);
8231 for (value = '9' + 1; value < 256; value++)
8232 ANYOF_BITMAP_SET(ret, value);
8235 what = POSIX_CC_UNI_NAME("Digit");
8238 /* this is to handle \p and \P */
8241 vFAIL("Invalid [::] class");
8245 /* Strings such as "+utf8::isWord\n" */
8246 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::Is%s\n", yesno, what);
8249 ANYOF_FLAGS(ret) |= ANYOF_CLASS;
8252 } /* end of namedclass \blah */
8255 if (prevvalue > (IV)value) /* b-a */ {
8256 const int w = RExC_parse - rangebegin;
8257 Simple_vFAIL4("Invalid [] range \"%*.*s\"", w, w, rangebegin);
8258 range = 0; /* not a valid range */
8262 prevvalue = value; /* save the beginning of the range */
8263 if (*RExC_parse == '-' && RExC_parse+1 < RExC_end &&
8264 RExC_parse[1] != ']') {
8267 /* a bad range like \w-, [:word:]- ? */
8268 if (namedclass > OOB_NAMEDCLASS) {
8269 if (ckWARN(WARN_REGEXP)) {
8271 RExC_parse >= rangebegin ?
8272 RExC_parse - rangebegin : 0;
8274 "False [] range \"%*.*s\"",
8278 ANYOF_BITMAP_SET(ret, '-');
8280 range = 1; /* yeah, it's a range! */
8281 continue; /* but do it the next time */
8285 /* now is the next time */
8286 /*stored += (value - prevvalue + 1);*/
8288 if (prevvalue < 256) {
8289 const IV ceilvalue = value < 256 ? value : 255;
8292 /* In EBCDIC [\x89-\x91] should include
8293 * the \x8e but [i-j] should not. */
8294 if (literal_endpoint == 2 &&
8295 ((isLOWER(prevvalue) && isLOWER(ceilvalue)) ||
8296 (isUPPER(prevvalue) && isUPPER(ceilvalue))))
8298 if (isLOWER(prevvalue)) {
8299 for (i = prevvalue; i <= ceilvalue; i++)
8300 if (isLOWER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8302 ANYOF_BITMAP_SET(ret, i);
8305 for (i = prevvalue; i <= ceilvalue; i++)
8306 if (isUPPER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8308 ANYOF_BITMAP_SET(ret, i);
8314 for (i = prevvalue; i <= ceilvalue; i++) {
8315 if (!ANYOF_BITMAP_TEST(ret,i)) {
8317 ANYOF_BITMAP_SET(ret, i);
8321 if (value > 255 || UTF) {
8322 const UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
8323 const UV natvalue = NATIVE_TO_UNI(value);
8324 stored+=2; /* can't optimize this class */
8325 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8326 if (prevnatvalue < natvalue) { /* what about > ? */
8327 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\t%04"UVxf"\n",
8328 prevnatvalue, natvalue);
8330 else if (prevnatvalue == natvalue) {
8331 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n", natvalue);
8333 U8 foldbuf[UTF8_MAXBYTES_CASE+1];
8335 const UV f = to_uni_fold(natvalue, foldbuf, &foldlen);
8337 #ifdef EBCDIC /* RD t/uni/fold ff and 6b */
8338 if (RExC_precomp[0] == ':' &&
8339 RExC_precomp[1] == '[' &&
8340 (f == 0xDF || f == 0x92)) {
8341 f = NATIVE_TO_UNI(f);
8344 /* If folding and foldable and a single
8345 * character, insert also the folded version
8346 * to the charclass. */
8348 #ifdef EBCDIC /* RD tunifold ligatures s,t fb05, fb06 */
8349 if ((RExC_precomp[0] == ':' &&
8350 RExC_precomp[1] == '[' &&
8352 (value == 0xFB05 || value == 0xFB06))) ?
8353 foldlen == ((STRLEN)UNISKIP(f) - 1) :
8354 foldlen == (STRLEN)UNISKIP(f) )
8356 if (foldlen == (STRLEN)UNISKIP(f))
8358 Perl_sv_catpvf(aTHX_ listsv,
8361 /* Any multicharacter foldings
8362 * require the following transform:
8363 * [ABCDEF] -> (?:[ABCabcDEFd]|pq|rst)
8364 * where E folds into "pq" and F folds
8365 * into "rst", all other characters
8366 * fold to single characters. We save
8367 * away these multicharacter foldings,
8368 * to be later saved as part of the
8369 * additional "s" data. */
8372 if (!unicode_alternate)
8373 unicode_alternate = newAV();
8374 sv = newSVpvn_utf8((char*)foldbuf, foldlen,
8376 av_push(unicode_alternate, sv);
8380 /* If folding and the value is one of the Greek
8381 * sigmas insert a few more sigmas to make the
8382 * folding rules of the sigmas to work right.
8383 * Note that not all the possible combinations
8384 * are handled here: some of them are handled
8385 * by the standard folding rules, and some of
8386 * them (literal or EXACTF cases) are handled
8387 * during runtime in regexec.c:S_find_byclass(). */
8388 if (value == UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA) {
8389 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8390 (UV)UNICODE_GREEK_CAPITAL_LETTER_SIGMA);
8391 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8392 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8394 else if (value == UNICODE_GREEK_CAPITAL_LETTER_SIGMA)
8395 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8396 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8401 literal_endpoint = 0;
8405 range = 0; /* this range (if it was one) is done now */
8409 ANYOF_FLAGS(ret) |= ANYOF_LARGE;
8411 RExC_size += ANYOF_CLASS_ADD_SKIP;
8413 RExC_emit += ANYOF_CLASS_ADD_SKIP;
8419 /****** !SIZE_ONLY AFTER HERE *********/
8421 if( stored == 1 && (value < 128 || (value < 256 && !UTF))
8422 && !( ANYOF_FLAGS(ret) & ( ANYOF_FLAGS_ALL ^ ANYOF_FOLD ) )
8424 /* optimize single char class to an EXACT node
8425 but *only* when its not a UTF/high char */
8426 const char * cur_parse= RExC_parse;
8427 RExC_emit = (regnode *)orig_emit;
8428 RExC_parse = (char *)orig_parse;
8429 ret = reg_node(pRExC_state,
8430 (U8)((ANYOF_FLAGS(ret) & ANYOF_FOLD) ? EXACTF : EXACT));
8431 RExC_parse = (char *)cur_parse;
8432 *STRING(ret)= (char)value;
8434 RExC_emit += STR_SZ(1);
8436 SvREFCNT_dec(listsv);
8440 /* optimize case-insensitive simple patterns (e.g. /[a-z]/i) */
8441 if ( /* If the only flag is folding (plus possibly inversion). */
8442 ((ANYOF_FLAGS(ret) & (ANYOF_FLAGS_ALL ^ ANYOF_INVERT)) == ANYOF_FOLD)
8444 for (value = 0; value < 256; ++value) {
8445 if (ANYOF_BITMAP_TEST(ret, value)) {
8446 UV fold = PL_fold[value];
8449 ANYOF_BITMAP_SET(ret, fold);
8452 ANYOF_FLAGS(ret) &= ~ANYOF_FOLD;
8455 /* optimize inverted simple patterns (e.g. [^a-z]) */
8456 if (optimize_invert &&
8457 /* If the only flag is inversion. */
8458 (ANYOF_FLAGS(ret) & ANYOF_FLAGS_ALL) == ANYOF_INVERT) {
8459 for (value = 0; value < ANYOF_BITMAP_SIZE; ++value)
8460 ANYOF_BITMAP(ret)[value] ^= ANYOF_FLAGS_ALL;
8461 ANYOF_FLAGS(ret) = ANYOF_UNICODE_ALL;
8464 AV * const av = newAV();
8466 /* The 0th element stores the character class description
8467 * in its textual form: used later (regexec.c:Perl_regclass_swash())
8468 * to initialize the appropriate swash (which gets stored in
8469 * the 1st element), and also useful for dumping the regnode.
8470 * The 2nd element stores the multicharacter foldings,
8471 * used later (regexec.c:S_reginclass()). */
8472 av_store(av, 0, listsv);
8473 av_store(av, 1, NULL);
8474 av_store(av, 2, MUTABLE_SV(unicode_alternate));
8475 rv = newRV_noinc(MUTABLE_SV(av));
8476 n = add_data(pRExC_state, 1, "s");
8477 RExC_rxi->data->data[n] = (void*)rv;
8485 /* reg_skipcomment()
8487 Absorbs an /x style # comments from the input stream.
8488 Returns true if there is more text remaining in the stream.
8489 Will set the REG_SEEN_RUN_ON_COMMENT flag if the comment
8490 terminates the pattern without including a newline.
8492 Note its the callers responsibility to ensure that we are
8498 S_reg_skipcomment(pTHX_ RExC_state_t *pRExC_state)
8502 PERL_ARGS_ASSERT_REG_SKIPCOMMENT;
8504 while (RExC_parse < RExC_end)
8505 if (*RExC_parse++ == '\n') {
8510 /* we ran off the end of the pattern without ending
8511 the comment, so we have to add an \n when wrapping */
8512 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
8520 Advance that parse position, and optionall absorbs
8521 "whitespace" from the inputstream.
8523 Without /x "whitespace" means (?#...) style comments only,
8524 with /x this means (?#...) and # comments and whitespace proper.
8526 Returns the RExC_parse point from BEFORE the scan occurs.
8528 This is the /x friendly way of saying RExC_parse++.
8532 S_nextchar(pTHX_ RExC_state_t *pRExC_state)
8534 char* const retval = RExC_parse++;
8536 PERL_ARGS_ASSERT_NEXTCHAR;
8539 if (*RExC_parse == '(' && RExC_parse[1] == '?' &&
8540 RExC_parse[2] == '#') {
8541 while (*RExC_parse != ')') {
8542 if (RExC_parse == RExC_end)
8543 FAIL("Sequence (?#... not terminated");
8549 if (RExC_flags & RXf_PMf_EXTENDED) {
8550 if (isSPACE(*RExC_parse)) {
8554 else if (*RExC_parse == '#') {
8555 if ( reg_skipcomment( pRExC_state ) )
8564 - reg_node - emit a node
8566 STATIC regnode * /* Location. */
8567 S_reg_node(pTHX_ RExC_state_t *pRExC_state, U8 op)
8570 register regnode *ptr;
8571 regnode * const ret = RExC_emit;
8572 GET_RE_DEBUG_FLAGS_DECL;
8574 PERL_ARGS_ASSERT_REG_NODE;
8577 SIZE_ALIGN(RExC_size);
8581 if (RExC_emit >= RExC_emit_bound)
8582 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8584 NODE_ALIGN_FILL(ret);
8586 FILL_ADVANCE_NODE(ptr, op);
8587 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (ptr) - 1);
8588 #ifdef RE_TRACK_PATTERN_OFFSETS
8589 if (RExC_offsets) { /* MJD */
8590 MJD_OFFSET_DEBUG(("%s:%d: (op %s) %s %"UVuf" (len %"UVuf") (max %"UVuf").\n",
8591 "reg_node", __LINE__,
8593 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0]
8594 ? "Overwriting end of array!\n" : "OK",
8595 (UV)(RExC_emit - RExC_emit_start),
8596 (UV)(RExC_parse - RExC_start),
8597 (UV)RExC_offsets[0]));
8598 Set_Node_Offset(RExC_emit, RExC_parse + (op == END));
8606 - reganode - emit a node with an argument
8608 STATIC regnode * /* Location. */
8609 S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
8612 register regnode *ptr;
8613 regnode * const ret = RExC_emit;
8614 GET_RE_DEBUG_FLAGS_DECL;
8616 PERL_ARGS_ASSERT_REGANODE;
8619 SIZE_ALIGN(RExC_size);
8624 assert(2==regarglen[op]+1);
8626 Anything larger than this has to allocate the extra amount.
8627 If we changed this to be:
8629 RExC_size += (1 + regarglen[op]);
8631 then it wouldn't matter. Its not clear what side effect
8632 might come from that so its not done so far.
8637 if (RExC_emit >= RExC_emit_bound)
8638 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8640 NODE_ALIGN_FILL(ret);
8642 FILL_ADVANCE_NODE_ARG(ptr, op, arg);
8643 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (ptr) - 2);
8644 #ifdef RE_TRACK_PATTERN_OFFSETS
8645 if (RExC_offsets) { /* MJD */
8646 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8650 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0] ?
8651 "Overwriting end of array!\n" : "OK",
8652 (UV)(RExC_emit - RExC_emit_start),
8653 (UV)(RExC_parse - RExC_start),
8654 (UV)RExC_offsets[0]));
8655 Set_Cur_Node_Offset;
8663 - reguni - emit (if appropriate) a Unicode character
8666 S_reguni(pTHX_ const RExC_state_t *pRExC_state, UV uv, char* s)
8670 PERL_ARGS_ASSERT_REGUNI;
8672 return SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
8676 - reginsert - insert an operator in front of already-emitted operand
8678 * Means relocating the operand.
8681 S_reginsert(pTHX_ RExC_state_t *pRExC_state, U8 op, regnode *opnd, U32 depth)
8684 register regnode *src;
8685 register regnode *dst;
8686 register regnode *place;
8687 const int offset = regarglen[(U8)op];
8688 const int size = NODE_STEP_REGNODE + offset;
8689 GET_RE_DEBUG_FLAGS_DECL;
8691 PERL_ARGS_ASSERT_REGINSERT;
8692 PERL_UNUSED_ARG(depth);
8693 /* (PL_regkind[(U8)op] == CURLY ? EXTRA_STEP_2ARGS : 0); */
8694 DEBUG_PARSE_FMT("inst"," - %s",PL_reg_name[op]);
8703 if (RExC_open_parens) {
8705 /*DEBUG_PARSE_FMT("inst"," - %"IVdf, (IV)RExC_npar);*/
8706 for ( paren=0 ; paren < RExC_npar ; paren++ ) {
8707 if ( RExC_open_parens[paren] >= opnd ) {
8708 /*DEBUG_PARSE_FMT("open"," - %d",size);*/
8709 RExC_open_parens[paren] += size;
8711 /*DEBUG_PARSE_FMT("open"," - %s","ok");*/
8713 if ( RExC_close_parens[paren] >= opnd ) {
8714 /*DEBUG_PARSE_FMT("close"," - %d",size);*/
8715 RExC_close_parens[paren] += size;
8717 /*DEBUG_PARSE_FMT("close"," - %s","ok");*/
8722 while (src > opnd) {
8723 StructCopy(--src, --dst, regnode);
8724 #ifdef RE_TRACK_PATTERN_OFFSETS
8725 if (RExC_offsets) { /* MJD 20010112 */
8726 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s copy %"UVuf" -> %"UVuf" (max %"UVuf").\n",
8730 (UV)(dst - RExC_emit_start) > RExC_offsets[0]
8731 ? "Overwriting end of array!\n" : "OK",
8732 (UV)(src - RExC_emit_start),
8733 (UV)(dst - RExC_emit_start),
8734 (UV)RExC_offsets[0]));
8735 Set_Node_Offset_To_R(dst-RExC_emit_start, Node_Offset(src));
8736 Set_Node_Length_To_R(dst-RExC_emit_start, Node_Length(src));
8742 place = opnd; /* Op node, where operand used to be. */
8743 #ifdef RE_TRACK_PATTERN_OFFSETS
8744 if (RExC_offsets) { /* MJD */
8745 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8749 (UV)(place - RExC_emit_start) > RExC_offsets[0]
8750 ? "Overwriting end of array!\n" : "OK",
8751 (UV)(place - RExC_emit_start),
8752 (UV)(RExC_parse - RExC_start),
8753 (UV)RExC_offsets[0]));
8754 Set_Node_Offset(place, RExC_parse);
8755 Set_Node_Length(place, 1);
8758 src = NEXTOPER(place);
8759 FILL_ADVANCE_NODE(place, op);
8760 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (place) - 1);
8761 Zero(src, offset, regnode);
8765 - regtail - set the next-pointer at the end of a node chain of p to val.
8766 - SEE ALSO: regtail_study
8768 /* TODO: All three parms should be const */
8770 S_regtail(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8773 register regnode *scan;
8774 GET_RE_DEBUG_FLAGS_DECL;
8776 PERL_ARGS_ASSERT_REGTAIL;
8778 PERL_UNUSED_ARG(depth);
8784 /* Find last node. */
8787 regnode * const temp = regnext(scan);
8789 SV * const mysv=sv_newmortal();
8790 DEBUG_PARSE_MSG((scan==p ? "tail" : ""));
8791 regprop(RExC_rx, mysv, scan);
8792 PerlIO_printf(Perl_debug_log, "~ %s (%d) %s %s\n",
8793 SvPV_nolen_const(mysv), REG_NODE_NUM(scan),
8794 (temp == NULL ? "->" : ""),
8795 (temp == NULL ? PL_reg_name[OP(val)] : "")
8803 if (reg_off_by_arg[OP(scan)]) {
8804 ARG_SET(scan, val - scan);
8807 NEXT_OFF(scan) = val - scan;
8813 - regtail_study - set the next-pointer at the end of a node chain of p to val.
8814 - Look for optimizable sequences at the same time.
8815 - currently only looks for EXACT chains.
8817 This is expermental code. The idea is to use this routine to perform
8818 in place optimizations on branches and groups as they are constructed,
8819 with the long term intention of removing optimization from study_chunk so
8820 that it is purely analytical.
8822 Currently only used when in DEBUG mode. The macro REGTAIL_STUDY() is used
8823 to control which is which.
8826 /* TODO: All four parms should be const */
8829 S_regtail_study(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8832 register regnode *scan;
8834 #ifdef EXPERIMENTAL_INPLACESCAN
8837 GET_RE_DEBUG_FLAGS_DECL;
8839 PERL_ARGS_ASSERT_REGTAIL_STUDY;
8845 /* Find last node. */
8849 regnode * const temp = regnext(scan);
8850 #ifdef EXPERIMENTAL_INPLACESCAN
8851 if (PL_regkind[OP(scan)] == EXACT)
8852 if (join_exact(pRExC_state,scan,&min,1,val,depth+1))
8860 if( exact == PSEUDO )
8862 else if ( exact != OP(scan) )
8871 SV * const mysv=sv_newmortal();
8872 DEBUG_PARSE_MSG((scan==p ? "tsdy" : ""));
8873 regprop(RExC_rx, mysv, scan);
8874 PerlIO_printf(Perl_debug_log, "~ %s (%d) -> %s\n",
8875 SvPV_nolen_const(mysv),
8877 PL_reg_name[exact]);
8884 SV * const mysv_val=sv_newmortal();
8885 DEBUG_PARSE_MSG("");
8886 regprop(RExC_rx, mysv_val, val);
8887 PerlIO_printf(Perl_debug_log, "~ attach to %s (%"IVdf") offset to %"IVdf"\n",
8888 SvPV_nolen_const(mysv_val),
8889 (IV)REG_NODE_NUM(val),
8893 if (reg_off_by_arg[OP(scan)]) {
8894 ARG_SET(scan, val - scan);
8897 NEXT_OFF(scan) = val - scan;
8905 - regcurly - a little FSA that accepts {\d+,?\d*}
8908 S_regcurly(register const char *s)
8910 PERL_ARGS_ASSERT_REGCURLY;
8929 - regdump - dump a regexp onto Perl_debug_log in vaguely comprehensible form
8933 S_regdump_extflags(pTHX_ const char *lead, const U32 flags)
8938 for (bit=0; bit<32; bit++) {
8939 if (flags & (1<<bit)) {
8941 PerlIO_printf(Perl_debug_log, "%s",lead);
8942 PerlIO_printf(Perl_debug_log, "%s ",PL_reg_extflags_name[bit]);
8947 PerlIO_printf(Perl_debug_log, "\n");
8949 PerlIO_printf(Perl_debug_log, "%s[none-set]\n",lead);
8955 Perl_regdump(pTHX_ const regexp *r)
8959 SV * const sv = sv_newmortal();
8960 SV *dsv= sv_newmortal();
8962 GET_RE_DEBUG_FLAGS_DECL;
8964 PERL_ARGS_ASSERT_REGDUMP;
8966 (void)dumpuntil(r, ri->program, ri->program + 1, NULL, NULL, sv, 0, 0);
8968 /* Header fields of interest. */
8969 if (r->anchored_substr) {
8970 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->anchored_substr),
8971 RE_SV_DUMPLEN(r->anchored_substr), 30);
8972 PerlIO_printf(Perl_debug_log,
8973 "anchored %s%s at %"IVdf" ",
8974 s, RE_SV_TAIL(r->anchored_substr),
8975 (IV)r->anchored_offset);
8976 } else if (r->anchored_utf8) {
8977 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->anchored_utf8),
8978 RE_SV_DUMPLEN(r->anchored_utf8), 30);
8979 PerlIO_printf(Perl_debug_log,
8980 "anchored utf8 %s%s at %"IVdf" ",
8981 s, RE_SV_TAIL(r->anchored_utf8),
8982 (IV)r->anchored_offset);
8984 if (r->float_substr) {
8985 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->float_substr),
8986 RE_SV_DUMPLEN(r->float_substr), 30);
8987 PerlIO_printf(Perl_debug_log,
8988 "floating %s%s at %"IVdf"..%"UVuf" ",
8989 s, RE_SV_TAIL(r->float_substr),
8990 (IV)r->float_min_offset, (UV)r->float_max_offset);
8991 } else if (r->float_utf8) {
8992 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->float_utf8),
8993 RE_SV_DUMPLEN(r->float_utf8), 30);
8994 PerlIO_printf(Perl_debug_log,
8995 "floating utf8 %s%s at %"IVdf"..%"UVuf" ",
8996 s, RE_SV_TAIL(r->float_utf8),
8997 (IV)r->float_min_offset, (UV)r->float_max_offset);
8999 if (r->check_substr || r->check_utf8)
9000 PerlIO_printf(Perl_debug_log,
9002 (r->check_substr == r->float_substr
9003 && r->check_utf8 == r->float_utf8
9004 ? "(checking floating" : "(checking anchored"));
9005 if (r->extflags & RXf_NOSCAN)
9006 PerlIO_printf(Perl_debug_log, " noscan");
9007 if (r->extflags & RXf_CHECK_ALL)
9008 PerlIO_printf(Perl_debug_log, " isall");
9009 if (r->check_substr || r->check_utf8)
9010 PerlIO_printf(Perl_debug_log, ") ");
9012 if (ri->regstclass) {
9013 regprop(r, sv, ri->regstclass);
9014 PerlIO_printf(Perl_debug_log, "stclass %s ", SvPVX_const(sv));
9016 if (r->extflags & RXf_ANCH) {
9017 PerlIO_printf(Perl_debug_log, "anchored");
9018 if (r->extflags & RXf_ANCH_BOL)
9019 PerlIO_printf(Perl_debug_log, "(BOL)");
9020 if (r->extflags & RXf_ANCH_MBOL)
9021 PerlIO_printf(Perl_debug_log, "(MBOL)");
9022 if (r->extflags & RXf_ANCH_SBOL)
9023 PerlIO_printf(Perl_debug_log, "(SBOL)");
9024 if (r->extflags & RXf_ANCH_GPOS)
9025 PerlIO_printf(Perl_debug_log, "(GPOS)");
9026 PerlIO_putc(Perl_debug_log, ' ');
9028 if (r->extflags & RXf_GPOS_SEEN)
9029 PerlIO_printf(Perl_debug_log, "GPOS:%"UVuf" ", (UV)r->gofs);
9030 if (r->intflags & PREGf_SKIP)
9031 PerlIO_printf(Perl_debug_log, "plus ");
9032 if (r->intflags & PREGf_IMPLICIT)
9033 PerlIO_printf(Perl_debug_log, "implicit ");
9034 PerlIO_printf(Perl_debug_log, "minlen %"IVdf" ", (IV)r->minlen);
9035 if (r->extflags & RXf_EVAL_SEEN)
9036 PerlIO_printf(Perl_debug_log, "with eval ");
9037 PerlIO_printf(Perl_debug_log, "\n");
9038 DEBUG_FLAGS_r(regdump_extflags("r->extflags: ",r->extflags));
9040 PERL_ARGS_ASSERT_REGDUMP;
9041 PERL_UNUSED_CONTEXT;
9043 #endif /* DEBUGGING */
9047 - regprop - printable representation of opcode
9049 #define EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags) \
9052 Perl_sv_catpvf(aTHX_ sv,"%s][%s",PL_colors[1],PL_colors[0]); \
9053 if (flags & ANYOF_INVERT) \
9054 /*make sure the invert info is in each */ \
9055 sv_catpvs(sv, "^"); \
9061 Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
9066 RXi_GET_DECL(prog,progi);
9067 GET_RE_DEBUG_FLAGS_DECL;
9069 PERL_ARGS_ASSERT_REGPROP;
9073 if (OP(o) > REGNODE_MAX) /* regnode.type is unsigned */
9074 /* It would be nice to FAIL() here, but this may be called from
9075 regexec.c, and it would be hard to supply pRExC_state. */
9076 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(o), (int)REGNODE_MAX);
9077 sv_catpv(sv, PL_reg_name[OP(o)]); /* Take off const! */
9079 k = PL_regkind[OP(o)];
9083 /* Using is_utf8_string() (via PERL_PV_UNI_DETECT)
9084 * is a crude hack but it may be the best for now since
9085 * we have no flag "this EXACTish node was UTF-8"
9087 pv_pretty(sv, STRING(o), STR_LEN(o), 60, PL_colors[0], PL_colors[1],
9088 PERL_PV_ESCAPE_UNI_DETECT |
9089 PERL_PV_PRETTY_ELLIPSES |
9090 PERL_PV_PRETTY_LTGT |
9091 PERL_PV_PRETTY_NOCLEAR
9093 } else if (k == TRIE) {
9094 /* print the details of the trie in dumpuntil instead, as
9095 * progi->data isn't available here */
9096 const char op = OP(o);
9097 const U32 n = ARG(o);
9098 const reg_ac_data * const ac = IS_TRIE_AC(op) ?
9099 (reg_ac_data *)progi->data->data[n] :
9101 const reg_trie_data * const trie
9102 = (reg_trie_data*)progi->data->data[!IS_TRIE_AC(op) ? n : ac->trie];
9104 Perl_sv_catpvf(aTHX_ sv, "-%s",PL_reg_name[o->flags]);
9105 DEBUG_TRIE_COMPILE_r(
9106 Perl_sv_catpvf(aTHX_ sv,
9107 "<S:%"UVuf"/%"IVdf" W:%"UVuf" L:%"UVuf"/%"UVuf" C:%"UVuf"/%"UVuf">",
9108 (UV)trie->startstate,
9109 (IV)trie->statecount-1, /* -1 because of the unused 0 element */
9110 (UV)trie->wordcount,
9113 (UV)TRIE_CHARCOUNT(trie),
9114 (UV)trie->uniquecharcount
9117 if ( IS_ANYOF_TRIE(op) || trie->bitmap ) {
9119 int rangestart = -1;
9120 U8* bitmap = IS_ANYOF_TRIE(op) ? (U8*)ANYOF_BITMAP(o) : (U8*)TRIE_BITMAP(trie);
9122 for (i = 0; i <= 256; i++) {
9123 if (i < 256 && BITMAP_TEST(bitmap,i)) {
9124 if (rangestart == -1)
9126 } else if (rangestart != -1) {
9127 if (i <= rangestart + 3)
9128 for (; rangestart < i; rangestart++)
9129 put_byte(sv, rangestart);
9131 put_byte(sv, rangestart);
9133 put_byte(sv, i - 1);
9141 } else if (k == CURLY) {
9142 if (OP(o) == CURLYM || OP(o) == CURLYN || OP(o) == CURLYX)
9143 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* Parenth number */
9144 Perl_sv_catpvf(aTHX_ sv, " {%d,%d}", ARG1(o), ARG2(o));
9146 else if (k == WHILEM && o->flags) /* Ordinal/of */
9147 Perl_sv_catpvf(aTHX_ sv, "[%d/%d]", o->flags & 0xf, o->flags>>4);
9148 else if (k == REF || k == OPEN || k == CLOSE || k == GROUPP || OP(o)==ACCEPT) {
9149 Perl_sv_catpvf(aTHX_ sv, "%d", (int)ARG(o)); /* Parenth number */
9150 if ( RXp_PAREN_NAMES(prog) ) {
9151 if ( k != REF || OP(o) < NREF) {
9152 AV *list= MUTABLE_AV(progi->data->data[progi->name_list_idx]);
9153 SV **name= av_fetch(list, ARG(o), 0 );
9155 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9158 AV *list= MUTABLE_AV(progi->data->data[ progi->name_list_idx ]);
9159 SV *sv_dat= MUTABLE_SV(progi->data->data[ ARG( o ) ]);
9160 I32 *nums=(I32*)SvPVX(sv_dat);
9161 SV **name= av_fetch(list, nums[0], 0 );
9164 for ( n=0; n<SvIVX(sv_dat); n++ ) {
9165 Perl_sv_catpvf(aTHX_ sv, "%s%"IVdf,
9166 (n ? "," : ""), (IV)nums[n]);
9168 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9172 } else if (k == GOSUB)
9173 Perl_sv_catpvf(aTHX_ sv, "%d[%+d]", (int)ARG(o),(int)ARG2L(o)); /* Paren and offset */
9174 else if (k == VERB) {
9176 Perl_sv_catpvf(aTHX_ sv, ":%"SVf,
9177 SVfARG((MUTABLE_SV(progi->data->data[ ARG( o ) ]))));
9178 } else if (k == LOGICAL)
9179 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* 2: embedded, otherwise 1 */
9180 else if (k == FOLDCHAR)
9181 Perl_sv_catpvf(aTHX_ sv, "[0x%"UVXf"]", PTR2UV(ARG(o)) );
9182 else if (k == ANYOF) {
9183 int i, rangestart = -1;
9184 const U8 flags = ANYOF_FLAGS(o);
9187 /* Should be synchronized with * ANYOF_ #xdefines in regcomp.h */
9188 static const char * const anyofs[] = {
9221 if (flags & ANYOF_LOCALE)
9222 sv_catpvs(sv, "{loc}");
9223 if (flags & ANYOF_FOLD)
9224 sv_catpvs(sv, "{i}");
9225 Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
9226 if (flags & ANYOF_INVERT)
9229 /* output what the standard cp 0-255 bitmap matches */
9230 for (i = 0; i <= 256; i++) {
9231 if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
9232 if (rangestart == -1)
9234 } else if (rangestart != -1) {
9235 if (i <= rangestart + 3)
9236 for (; rangestart < i; rangestart++)
9237 put_byte(sv, rangestart);
9239 put_byte(sv, rangestart);
9241 put_byte(sv, i - 1);
9248 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9249 /* output any special charclass tests (used mostly under use locale) */
9250 if (o->flags & ANYOF_CLASS)
9251 for (i = 0; i < (int)(sizeof(anyofs)/sizeof(char*)); i++)
9252 if (ANYOF_CLASS_TEST(o,i)) {
9253 sv_catpv(sv, anyofs[i]);
9257 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9259 /* output information about the unicode matching */
9260 if (flags & ANYOF_UNICODE)
9261 sv_catpvs(sv, "{unicode}");
9262 else if (flags & ANYOF_UNICODE_ALL)
9263 sv_catpvs(sv, "{unicode_all}");
9267 SV * const sw = regclass_swash(prog, o, FALSE, &lv, 0);
9271 U8 s[UTF8_MAXBYTES_CASE+1];
9273 for (i = 0; i <= 256; i++) { /* just the first 256 */
9274 uvchr_to_utf8(s, i);
9276 if (i < 256 && swash_fetch(sw, s, TRUE)) {
9277 if (rangestart == -1)
9279 } else if (rangestart != -1) {
9280 if (i <= rangestart + 3)
9281 for (; rangestart < i; rangestart++) {
9282 const U8 * const e = uvchr_to_utf8(s,rangestart);
9284 for(p = s; p < e; p++)
9288 const U8 *e = uvchr_to_utf8(s,rangestart);
9290 for (p = s; p < e; p++)
9293 e = uvchr_to_utf8(s, i-1);
9294 for (p = s; p < e; p++)
9301 sv_catpvs(sv, "..."); /* et cetera */
9305 char *s = savesvpv(lv);
9306 char * const origs = s;
9308 while (*s && *s != '\n')
9312 const char * const t = ++s;
9330 Perl_sv_catpvf(aTHX_ sv, "%s]", PL_colors[1]);
9332 else if (k == BRANCHJ && (OP(o) == UNLESSM || OP(o) == IFMATCH))
9333 Perl_sv_catpvf(aTHX_ sv, "[%d]", -(o->flags));
9335 PERL_UNUSED_CONTEXT;
9336 PERL_UNUSED_ARG(sv);
9338 PERL_UNUSED_ARG(prog);
9339 #endif /* DEBUGGING */
9343 Perl_re_intuit_string(pTHX_ REGEXP * const r)
9344 { /* Assume that RE_INTUIT is set */
9346 struct regexp *const prog = (struct regexp *)SvANY(r);
9347 GET_RE_DEBUG_FLAGS_DECL;
9349 PERL_ARGS_ASSERT_RE_INTUIT_STRING;
9350 PERL_UNUSED_CONTEXT;
9354 const char * const s = SvPV_nolen_const(prog->check_substr
9355 ? prog->check_substr : prog->check_utf8);
9357 if (!PL_colorset) reginitcolors();
9358 PerlIO_printf(Perl_debug_log,
9359 "%sUsing REx %ssubstr:%s \"%s%.60s%s%s\"\n",
9361 prog->check_substr ? "" : "utf8 ",
9362 PL_colors[5],PL_colors[0],
9365 (strlen(s) > 60 ? "..." : ""));
9368 return prog->check_substr ? prog->check_substr : prog->check_utf8;
9374 handles refcounting and freeing the perl core regexp structure. When
9375 it is necessary to actually free the structure the first thing it
9376 does is call the 'free' method of the regexp_engine associated to to
9377 the regexp, allowing the handling of the void *pprivate; member
9378 first. (This routine is not overridable by extensions, which is why
9379 the extensions free is called first.)
9381 See regdupe and regdupe_internal if you change anything here.
9383 #ifndef PERL_IN_XSUB_RE
9385 Perl_pregfree(pTHX_ REGEXP *r)
9391 Perl_pregfree2(pTHX_ REGEXP *rx)
9394 struct regexp *const r = (struct regexp *)SvANY(rx);
9395 GET_RE_DEBUG_FLAGS_DECL;
9397 PERL_ARGS_ASSERT_PREGFREE2;
9400 ReREFCNT_dec(r->mother_re);
9402 CALLREGFREE_PVT(rx); /* free the private data */
9403 if (RXp_PAREN_NAMES(r))
9404 SvREFCNT_dec(RXp_PAREN_NAMES(r));
9407 if (r->anchored_substr)
9408 SvREFCNT_dec(r->anchored_substr);
9409 if (r->anchored_utf8)
9410 SvREFCNT_dec(r->anchored_utf8);
9411 if (r->float_substr)
9412 SvREFCNT_dec(r->float_substr);
9414 SvREFCNT_dec(r->float_utf8);
9415 Safefree(r->substrs);
9417 RX_MATCH_COPY_FREE(rx);
9418 #ifdef PERL_OLD_COPY_ON_WRITE
9420 SvREFCNT_dec(r->saved_copy);
9427 This is a hacky workaround to the structural issue of match results
9428 being stored in the regexp structure which is in turn stored in
9429 PL_curpm/PL_reg_curpm. The problem is that due to qr// the pattern
9430 could be PL_curpm in multiple contexts, and could require multiple
9431 result sets being associated with the pattern simultaneously, such
9432 as when doing a recursive match with (??{$qr})
9434 The solution is to make a lightweight copy of the regexp structure
9435 when a qr// is returned from the code executed by (??{$qr}) this
9436 lightweight copy doesnt actually own any of its data except for
9437 the starp/end and the actual regexp structure itself.
9443 Perl_reg_temp_copy (pTHX_ REGEXP *rx)
9445 REGEXP *ret_x = (REGEXP*) newSV_type(SVt_REGEXP);
9446 struct regexp *ret = (struct regexp *)SvANY(ret_x);
9447 struct regexp *const r = (struct regexp *)SvANY(rx);
9448 register const I32 npar = r->nparens+1;
9450 PERL_ARGS_ASSERT_REG_TEMP_COPY;
9452 (void)ReREFCNT_inc(rx);
9453 /* We can take advantage of the existing "copied buffer" mechanism in SVs
9454 by pointing directly at the buffer, but flagging that the allocated
9455 space in the copy is zero. As we've just done a struct copy, it's now
9456 a case of zero-ing that, rather than copying the current length. */
9457 SvPV_set(ret_x, RX_WRAPPED(rx));
9458 SvFLAGS(ret_x) |= SvFLAGS(rx) & (SVf_POK|SVp_POK|SVf_UTF8);
9459 memcpy(&(ret->xpv_cur), &(r->xpv_cur),
9460 sizeof(regexp) - STRUCT_OFFSET(regexp, xpv_cur));
9461 SvLEN_set(ret_x, 0);
9462 Newx(ret->offs, npar, regexp_paren_pair);
9463 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9465 Newx(ret->substrs, 1, struct reg_substr_data);
9466 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9468 SvREFCNT_inc_void(ret->anchored_substr);
9469 SvREFCNT_inc_void(ret->anchored_utf8);
9470 SvREFCNT_inc_void(ret->float_substr);
9471 SvREFCNT_inc_void(ret->float_utf8);
9473 /* check_substr and check_utf8, if non-NULL, point to either their
9474 anchored or float namesakes, and don't hold a second reference. */
9476 RX_MATCH_COPIED_off(ret_x);
9477 #ifdef PERL_OLD_COPY_ON_WRITE
9478 ret->saved_copy = NULL;
9480 ret->mother_re = rx;
9486 /* regfree_internal()
9488 Free the private data in a regexp. This is overloadable by
9489 extensions. Perl takes care of the regexp structure in pregfree(),
9490 this covers the *pprivate pointer which technically perldoesnt
9491 know about, however of course we have to handle the
9492 regexp_internal structure when no extension is in use.
9494 Note this is called before freeing anything in the regexp
9499 Perl_regfree_internal(pTHX_ REGEXP * const rx)
9502 struct regexp *const r = (struct regexp *)SvANY(rx);
9504 GET_RE_DEBUG_FLAGS_DECL;
9506 PERL_ARGS_ASSERT_REGFREE_INTERNAL;
9512 SV *dsv= sv_newmortal();
9513 RE_PV_QUOTED_DECL(s, RX_UTF8(rx),
9514 dsv, RX_PRECOMP(rx), RX_PRELEN(rx), 60);
9515 PerlIO_printf(Perl_debug_log,"%sFreeing REx:%s %s\n",
9516 PL_colors[4],PL_colors[5],s);
9519 #ifdef RE_TRACK_PATTERN_OFFSETS
9521 Safefree(ri->u.offsets); /* 20010421 MJD */
9524 int n = ri->data->count;
9525 PAD* new_comppad = NULL;
9530 /* If you add a ->what type here, update the comment in regcomp.h */
9531 switch (ri->data->what[n]) {
9535 SvREFCNT_dec(MUTABLE_SV(ri->data->data[n]));
9538 Safefree(ri->data->data[n]);
9541 new_comppad = MUTABLE_AV(ri->data->data[n]);
9544 if (new_comppad == NULL)
9545 Perl_croak(aTHX_ "panic: pregfree comppad");
9546 PAD_SAVE_LOCAL(old_comppad,
9547 /* Watch out for global destruction's random ordering. */
9548 (SvTYPE(new_comppad) == SVt_PVAV) ? new_comppad : NULL
9551 refcnt = OpREFCNT_dec((OP_4tree*)ri->data->data[n]);
9554 op_free((OP_4tree*)ri->data->data[n]);
9556 PAD_RESTORE_LOCAL(old_comppad);
9557 SvREFCNT_dec(MUTABLE_SV(new_comppad));
9563 { /* Aho Corasick add-on structure for a trie node.
9564 Used in stclass optimization only */
9566 reg_ac_data *aho=(reg_ac_data*)ri->data->data[n];
9568 refcount = --aho->refcount;
9571 PerlMemShared_free(aho->states);
9572 PerlMemShared_free(aho->fail);
9573 /* do this last!!!! */
9574 PerlMemShared_free(ri->data->data[n]);
9575 PerlMemShared_free(ri->regstclass);
9581 /* trie structure. */
9583 reg_trie_data *trie=(reg_trie_data*)ri->data->data[n];
9585 refcount = --trie->refcount;
9588 PerlMemShared_free(trie->charmap);
9589 PerlMemShared_free(trie->states);
9590 PerlMemShared_free(trie->trans);
9592 PerlMemShared_free(trie->bitmap);
9594 PerlMemShared_free(trie->wordlen);
9596 PerlMemShared_free(trie->jump);
9598 PerlMemShared_free(trie->nextword);
9599 /* do this last!!!! */
9600 PerlMemShared_free(ri->data->data[n]);
9605 Perl_croak(aTHX_ "panic: regfree data code '%c'", ri->data->what[n]);
9608 Safefree(ri->data->what);
9615 #define sv_dup_inc(s,t) SvREFCNT_inc(sv_dup(s,t))
9616 #define av_dup_inc(s,t) MUTABLE_AV(SvREFCNT_inc(sv_dup((const SV *)s,t)))
9617 #define hv_dup_inc(s,t) MUTABLE_HV(SvREFCNT_inc(sv_dup((const SV *)s,t)))
9618 #define SAVEPVN(p,n) ((p) ? savepvn(p,n) : NULL)
9621 re_dup - duplicate a regexp.
9623 This routine is expected to clone a given regexp structure. It is only
9624 compiled under USE_ITHREADS.
9626 After all of the core data stored in struct regexp is duplicated
9627 the regexp_engine.dupe method is used to copy any private data
9628 stored in the *pprivate pointer. This allows extensions to handle
9629 any duplication it needs to do.
9631 See pregfree() and regfree_internal() if you change anything here.
9633 #if defined(USE_ITHREADS)
9634 #ifndef PERL_IN_XSUB_RE
9636 Perl_re_dup_guts(pTHX_ const REGEXP *sstr, REGEXP *dstr, CLONE_PARAMS *param)
9640 const struct regexp *r = (const struct regexp *)SvANY(sstr);
9641 struct regexp *ret = (struct regexp *)SvANY(dstr);
9643 PERL_ARGS_ASSERT_RE_DUP_GUTS;
9645 npar = r->nparens+1;
9646 Newx(ret->offs, npar, regexp_paren_pair);
9647 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9649 /* no need to copy these */
9650 Newx(ret->swap, npar, regexp_paren_pair);
9654 /* Do it this way to avoid reading from *r after the StructCopy().
9655 That way, if any of the sv_dup_inc()s dislodge *r from the L1
9656 cache, it doesn't matter. */
9657 const bool anchored = r->check_substr
9658 ? r->check_substr == r->anchored_substr
9659 : r->check_utf8 == r->anchored_utf8;
9660 Newx(ret->substrs, 1, struct reg_substr_data);
9661 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9663 ret->anchored_substr = sv_dup_inc(ret->anchored_substr, param);
9664 ret->anchored_utf8 = sv_dup_inc(ret->anchored_utf8, param);
9665 ret->float_substr = sv_dup_inc(ret->float_substr, param);
9666 ret->float_utf8 = sv_dup_inc(ret->float_utf8, param);
9668 /* check_substr and check_utf8, if non-NULL, point to either their
9669 anchored or float namesakes, and don't hold a second reference. */
9671 if (ret->check_substr) {
9673 assert(r->check_utf8 == r->anchored_utf8);
9674 ret->check_substr = ret->anchored_substr;
9675 ret->check_utf8 = ret->anchored_utf8;
9677 assert(r->check_substr == r->float_substr);
9678 assert(r->check_utf8 == r->float_utf8);
9679 ret->check_substr = ret->float_substr;
9680 ret->check_utf8 = ret->float_utf8;
9682 } else if (ret->check_utf8) {
9684 ret->check_utf8 = ret->anchored_utf8;
9686 ret->check_utf8 = ret->float_utf8;
9691 RXp_PAREN_NAMES(ret) = hv_dup_inc(RXp_PAREN_NAMES(ret), param);
9694 RXi_SET(ret,CALLREGDUPE_PVT(dstr,param));
9696 if (RX_MATCH_COPIED(dstr))
9697 ret->subbeg = SAVEPVN(ret->subbeg, ret->sublen);
9700 #ifdef PERL_OLD_COPY_ON_WRITE
9701 ret->saved_copy = NULL;
9704 ret->mother_re = NULL;
9707 #endif /* PERL_IN_XSUB_RE */
9712 This is the internal complement to regdupe() which is used to copy
9713 the structure pointed to by the *pprivate pointer in the regexp.
9714 This is the core version of the extension overridable cloning hook.
9715 The regexp structure being duplicated will be copied by perl prior
9716 to this and will be provided as the regexp *r argument, however
9717 with the /old/ structures pprivate pointer value. Thus this routine
9718 may override any copying normally done by perl.
9720 It returns a pointer to the new regexp_internal structure.
9724 Perl_regdupe_internal(pTHX_ REGEXP * const rx, CLONE_PARAMS *param)
9727 struct regexp *const r = (struct regexp *)SvANY(rx);
9728 regexp_internal *reti;
9732 PERL_ARGS_ASSERT_REGDUPE_INTERNAL;
9734 npar = r->nparens+1;
9737 Newxc(reti, sizeof(regexp_internal) + len*sizeof(regnode), char, regexp_internal);
9738 Copy(ri->program, reti->program, len+1, regnode);
9741 reti->regstclass = NULL;
9745 const int count = ri->data->count;
9748 Newxc(d, sizeof(struct reg_data) + count*sizeof(void *),
9749 char, struct reg_data);
9750 Newx(d->what, count, U8);
9753 for (i = 0; i < count; i++) {
9754 d->what[i] = ri->data->what[i];
9755 switch (d->what[i]) {
9756 /* legal options are one of: sSfpontTu
9757 see also regcomp.h and pregfree() */
9760 case 'p': /* actually an AV, but the dup function is identical. */
9761 case 'u': /* actually an HV, but the dup function is identical. */
9762 d->data[i] = sv_dup_inc((const SV *)ri->data->data[i], param);
9765 /* This is cheating. */
9766 Newx(d->data[i], 1, struct regnode_charclass_class);
9767 StructCopy(ri->data->data[i], d->data[i],
9768 struct regnode_charclass_class);
9769 reti->regstclass = (regnode*)d->data[i];
9772 /* Compiled op trees are readonly and in shared memory,
9773 and can thus be shared without duplication. */
9775 d->data[i] = (void*)OpREFCNT_inc((OP*)ri->data->data[i]);
9779 /* Trie stclasses are readonly and can thus be shared
9780 * without duplication. We free the stclass in pregfree
9781 * when the corresponding reg_ac_data struct is freed.
9783 reti->regstclass= ri->regstclass;
9787 ((reg_trie_data*)ri->data->data[i])->refcount++;
9791 d->data[i] = ri->data->data[i];
9794 Perl_croak(aTHX_ "panic: re_dup unknown data code '%c'", ri->data->what[i]);
9803 reti->name_list_idx = ri->name_list_idx;
9805 #ifdef RE_TRACK_PATTERN_OFFSETS
9806 if (ri->u.offsets) {
9807 Newx(reti->u.offsets, 2*len+1, U32);
9808 Copy(ri->u.offsets, reti->u.offsets, 2*len+1, U32);
9811 SetProgLen(reti,len);
9817 #endif /* USE_ITHREADS */
9819 #ifndef PERL_IN_XSUB_RE
9822 - regnext - dig the "next" pointer out of a node
9825 Perl_regnext(pTHX_ register regnode *p)
9828 register I32 offset;
9833 offset = (reg_off_by_arg[OP(p)] ? ARG(p) : NEXT_OFF(p));
9842 S_re_croak2(pTHX_ const char* pat1,const char* pat2,...)
9845 STRLEN l1 = strlen(pat1);
9846 STRLEN l2 = strlen(pat2);
9849 const char *message;
9851 PERL_ARGS_ASSERT_RE_CROAK2;
9857 Copy(pat1, buf, l1 , char);
9858 Copy(pat2, buf + l1, l2 , char);
9859 buf[l1 + l2] = '\n';
9860 buf[l1 + l2 + 1] = '\0';
9862 /* ANSI variant takes additional second argument */
9863 va_start(args, pat2);
9867 msv = vmess(buf, &args);
9869 message = SvPV_const(msv,l1);
9872 Copy(message, buf, l1 , char);
9873 buf[l1-1] = '\0'; /* Overwrite \n */
9874 Perl_croak(aTHX_ "%s", buf);
9877 /* XXX Here's a total kludge. But we need to re-enter for swash routines. */
9879 #ifndef PERL_IN_XSUB_RE
9881 Perl_save_re_context(pTHX)
9885 struct re_save_state *state;
9887 SAVEVPTR(PL_curcop);
9888 SSGROW(SAVESTACK_ALLOC_FOR_RE_SAVE_STATE + 1);
9890 state = (struct re_save_state *)(PL_savestack + PL_savestack_ix);
9891 PL_savestack_ix += SAVESTACK_ALLOC_FOR_RE_SAVE_STATE;
9892 SSPUSHINT(SAVEt_RE_STATE);
9894 Copy(&PL_reg_state, state, 1, struct re_save_state);
9896 PL_reg_start_tmp = 0;
9897 PL_reg_start_tmpl = 0;
9898 PL_reg_oldsaved = NULL;
9899 PL_reg_oldsavedlen = 0;
9901 PL_reg_leftiter = 0;
9902 PL_reg_poscache = NULL;
9903 PL_reg_poscache_size = 0;
9904 #ifdef PERL_OLD_COPY_ON_WRITE
9908 /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
9910 const REGEXP * const rx = PM_GETRE(PL_curpm);
9913 for (i = 1; i <= RX_NPARENS(rx); i++) {
9914 char digits[TYPE_CHARS(long)];
9915 const STRLEN len = my_snprintf(digits, sizeof(digits), "%lu", (long)i);
9916 GV *const *const gvp
9917 = (GV**)hv_fetch(PL_defstash, digits, len, 0);
9920 GV * const gv = *gvp;
9921 if (SvTYPE(gv) == SVt_PVGV && GvSV(gv))
9931 clear_re(pTHX_ void *r)
9934 ReREFCNT_dec((REGEXP *)r);
9940 S_put_byte(pTHX_ SV *sv, int c)
9942 PERL_ARGS_ASSERT_PUT_BYTE;
9944 /* Our definition of isPRINT() ignores locales, so only bytes that are
9945 not part of UTF-8 are considered printable. I assume that the same
9946 holds for UTF-EBCDIC.
9947 Also, code point 255 is not printable in either (it's E0 in EBCDIC,
9948 which Wikipedia says:
9950 EO, or Eight Ones, is an 8-bit EBCDIC character code represented as all
9951 ones (binary 1111 1111, hexadecimal FF). It is similar, but not
9952 identical, to the ASCII delete (DEL) or rubout control character.
9953 ) So the old condition can be simplified to !isPRINT(c) */
9955 Perl_sv_catpvf(aTHX_ sv, "\\%o", c);
9957 const char string = c;
9958 if (c == '-' || c == ']' || c == '\\' || c == '^')
9959 sv_catpvs(sv, "\\");
9960 sv_catpvn(sv, &string, 1);
9965 #define CLEAR_OPTSTART \
9966 if (optstart) STMT_START { \
9967 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log, " (%"IVdf" nodes)\n", (IV)(node - optstart))); \
9971 #define DUMPUNTIL(b,e) CLEAR_OPTSTART; node=dumpuntil(r,start,(b),(e),last,sv,indent+1,depth+1);
9973 STATIC const regnode *
9974 S_dumpuntil(pTHX_ const regexp *r, const regnode *start, const regnode *node,
9975 const regnode *last, const regnode *plast,
9976 SV* sv, I32 indent, U32 depth)
9979 register U8 op = PSEUDO; /* Arbitrary non-END op. */
9980 register const regnode *next;
9981 const regnode *optstart= NULL;
9984 GET_RE_DEBUG_FLAGS_DECL;
9986 PERL_ARGS_ASSERT_DUMPUNTIL;
9988 #ifdef DEBUG_DUMPUNTIL
9989 PerlIO_printf(Perl_debug_log, "--- %d : %d - %d - %d\n",indent,node-start,
9990 last ? last-start : 0,plast ? plast-start : 0);
9993 if (plast && plast < last)
9996 while (PL_regkind[op] != END && (!last || node < last)) {
9997 /* While that wasn't END last time... */
10000 if (op == CLOSE || op == WHILEM)
10002 next = regnext((regnode *)node);
10005 if (OP(node) == OPTIMIZED) {
10006 if (!optstart && RE_DEBUG_FLAG(RE_DEBUG_COMPILE_OPTIMISE))
10013 regprop(r, sv, node);
10014 PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
10015 (int)(2*indent + 1), "", SvPVX_const(sv));
10017 if (OP(node) != OPTIMIZED) {
10018 if (next == NULL) /* Next ptr. */
10019 PerlIO_printf(Perl_debug_log, " (0)");
10020 else if (PL_regkind[(U8)op] == BRANCH && PL_regkind[OP(next)] != BRANCH )
10021 PerlIO_printf(Perl_debug_log, " (FAIL)");
10023 PerlIO_printf(Perl_debug_log, " (%"IVdf")", (IV)(next - start));
10024 (void)PerlIO_putc(Perl_debug_log, '\n');
10028 if (PL_regkind[(U8)op] == BRANCHJ) {
10031 register const regnode *nnode = (OP(next) == LONGJMP
10032 ? regnext((regnode *)next)
10034 if (last && nnode > last)
10036 DUMPUNTIL(NEXTOPER(NEXTOPER(node)), nnode);
10039 else if (PL_regkind[(U8)op] == BRANCH) {
10041 DUMPUNTIL(NEXTOPER(node), next);
10043 else if ( PL_regkind[(U8)op] == TRIE ) {
10044 const regnode *this_trie = node;
10045 const char op = OP(node);
10046 const U32 n = ARG(node);
10047 const reg_ac_data * const ac = op>=AHOCORASICK ?
10048 (reg_ac_data *)ri->data->data[n] :
10050 const reg_trie_data * const trie =
10051 (reg_trie_data*)ri->data->data[op<AHOCORASICK ? n : ac->trie];
10053 AV *const trie_words = MUTABLE_AV(ri->data->data[n + TRIE_WORDS_OFFSET]);
10055 const regnode *nextbranch= NULL;
10058 for (word_idx= 0; word_idx < (I32)trie->wordcount; word_idx++) {
10059 SV ** const elem_ptr = av_fetch(trie_words,word_idx,0);
10061 PerlIO_printf(Perl_debug_log, "%*s%s ",
10062 (int)(2*(indent+3)), "",
10063 elem_ptr ? pv_pretty(sv, SvPV_nolen_const(*elem_ptr), SvCUR(*elem_ptr), 60,
10064 PL_colors[0], PL_colors[1],
10065 (SvUTF8(*elem_ptr) ? PERL_PV_ESCAPE_UNI : 0) |
10066 PERL_PV_PRETTY_ELLIPSES |
10067 PERL_PV_PRETTY_LTGT
10072 U16 dist= trie->jump[word_idx+1];
10073 PerlIO_printf(Perl_debug_log, "(%"UVuf")\n",
10074 (UV)((dist ? this_trie + dist : next) - start));
10077 nextbranch= this_trie + trie->jump[0];
10078 DUMPUNTIL(this_trie + dist, nextbranch);
10080 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
10081 nextbranch= regnext((regnode *)nextbranch);
10083 PerlIO_printf(Perl_debug_log, "\n");
10086 if (last && next > last)
10091 else if ( op == CURLY ) { /* "next" might be very big: optimizer */
10092 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS,
10093 NEXTOPER(node) + EXTRA_STEP_2ARGS + 1);
10095 else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
10097 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS, next);
10099 else if ( op == PLUS || op == STAR) {
10100 DUMPUNTIL(NEXTOPER(node), NEXTOPER(node) + 1);
10102 else if (op == ANYOF) {
10103 /* arglen 1 + class block */
10104 node += 1 + ((ANYOF_FLAGS(node) & ANYOF_LARGE)
10105 ? ANYOF_CLASS_SKIP : ANYOF_SKIP);
10106 node = NEXTOPER(node);
10108 else if (PL_regkind[(U8)op] == EXACT) {
10109 /* Literal string, where present. */
10110 node += NODE_SZ_STR(node) - 1;
10111 node = NEXTOPER(node);
10114 node = NEXTOPER(node);
10115 node += regarglen[(U8)op];
10117 if (op == CURLYX || op == OPEN)
10121 #ifdef DEBUG_DUMPUNTIL
10122 PerlIO_printf(Perl_debug_log, "--- %d\n", (int)indent);
10127 #endif /* DEBUGGING */
10131 * c-indentation-style: bsd
10132 * c-basic-offset: 4
10133 * indent-tabs-mode: t
10136 * ex: set ts=8 sts=4 sw=4 noet:
projects / perl / modules / re-engine-Hooks.git / blob