5 * "A fair jaw-cracker dwarf-language must be." --Samwise Gamgee
8 /* This file contains functions for compiling a regular expression. See
9 * also regexec.c which funnily enough, contains functions for executing
10 * a regular expression.
12 * This file is also copied at build time to ext/re/re_comp.c, where
13 * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
14 * This causes the main functions to be compiled under new names and with
15 * debugging support added, which makes "use re 'debug'" work.
18 /* NOTE: this is derived from Henry Spencer's regexp code, and should not
19 * confused with the original package (see point 3 below). Thanks, Henry!
22 /* Additional note: this code is very heavily munged from Henry's version
23 * in places. In some spots I've traded clarity for efficiency, so don't
24 * blame Henry for some of the lack of readability.
27 /* The names of the functions have been changed from regcomp and
28 * regexec to pregcomp and pregexec in order to avoid conflicts
29 * with the POSIX routines of the same names.
32 #ifdef PERL_EXT_RE_BUILD
37 * pregcomp and pregexec -- regsub and regerror are not used in perl
39 * Copyright (c) 1986 by University of Toronto.
40 * Written by Henry Spencer. Not derived from licensed software.
42 * Permission is granted to anyone to use this software for any
43 * purpose on any computer system, and to redistribute it freely,
44 * subject to the following restrictions:
46 * 1. The author is not responsible for the consequences of use of
47 * this software, no matter how awful, even if they arise
50 * 2. The origin of this software must not be misrepresented, either
51 * by explicit claim or by omission.
53 * 3. Altered versions must be plainly marked as such, and must not
54 * be misrepresented as being the original software.
57 **** Alterations to Henry's code are...
59 **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
60 **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007 by Larry Wall and others
62 **** You may distribute under the terms of either the GNU General Public
63 **** License or the Artistic License, as specified in the README file.
66 * Beware that some of this code is subtly aware of the way operator
67 * precedence is structured in regular expressions. Serious changes in
68 * regular-expression syntax might require a total rethink.
71 #define PERL_IN_REGCOMP_C
74 #ifndef PERL_IN_XSUB_RE
79 #ifdef PERL_IN_XSUB_RE
90 # if defined(BUGGY_MSC6)
91 /* MSC 6.00A breaks on op/regexp.t test 85 unless we turn this off */
92 # pragma optimize("a",off)
93 /* But MSC 6.00A is happy with 'w', for aliases only across function calls*/
94 # pragma optimize("w",on )
95 # endif /* BUGGY_MSC6 */
102 typedef struct RExC_state_t {
103 U32 flags; /* are we folding, multilining? */
104 char *precomp; /* uncompiled string. */
105 regexp *rx; /* perl core regexp structure */
106 regexp_internal *rxi; /* internal data for regexp object pprivate field */
107 char *start; /* Start of input for compile */
108 char *end; /* End of input for compile */
109 char *parse; /* Input-scan pointer. */
110 I32 whilem_seen; /* number of WHILEM in this expr */
111 regnode *emit_start; /* Start of emitted-code area */
112 regnode *emit_bound; /* First regnode outside of the allocated space */
113 regnode *emit; /* Code-emit pointer; ®dummy = don't = compiling */
114 I32 naughty; /* How bad is this pattern? */
115 I32 sawback; /* Did we see \1, ...? */
117 I32 size; /* Code size. */
118 I32 npar; /* Capture buffer count, (OPEN). */
119 I32 cpar; /* Capture buffer count, (CLOSE). */
120 I32 nestroot; /* root parens we are in - used by accept */
124 regnode **open_parens; /* pointers to open parens */
125 regnode **close_parens; /* pointers to close parens */
126 regnode *opend; /* END node in program */
127 I32 utf8; /* whether the pattern is utf8 or not */
128 I32 orig_utf8; /* whether the pattern was originally in utf8 */
129 /* XXX use this for future optimisation of case
130 * where pattern must be upgraded to utf8. */
131 HV *charnames; /* cache of named sequences */
132 HV *paren_names; /* Paren names */
134 regnode **recurse; /* Recurse regops */
135 I32 recurse_count; /* Number of recurse regops */
137 char *starttry; /* -Dr: where regtry was called. */
138 #define RExC_starttry (pRExC_state->starttry)
141 const char *lastparse;
143 AV *paren_name_list; /* idx -> name */
144 #define RExC_lastparse (pRExC_state->lastparse)
145 #define RExC_lastnum (pRExC_state->lastnum)
146 #define RExC_paren_name_list (pRExC_state->paren_name_list)
150 #define RExC_flags (pRExC_state->flags)
151 #define RExC_precomp (pRExC_state->precomp)
152 #define RExC_rx (pRExC_state->rx)
153 #define RExC_rxi (pRExC_state->rxi)
154 #define RExC_start (pRExC_state->start)
155 #define RExC_end (pRExC_state->end)
156 #define RExC_parse (pRExC_state->parse)
157 #define RExC_whilem_seen (pRExC_state->whilem_seen)
158 #ifdef RE_TRACK_PATTERN_OFFSETS
159 #define RExC_offsets (pRExC_state->rxi->u.offsets) /* I am not like the others */
161 #define RExC_emit (pRExC_state->emit)
162 #define RExC_emit_start (pRExC_state->emit_start)
163 #define RExC_emit_bound (pRExC_state->emit_bound)
164 #define RExC_naughty (pRExC_state->naughty)
165 #define RExC_sawback (pRExC_state->sawback)
166 #define RExC_seen (pRExC_state->seen)
167 #define RExC_size (pRExC_state->size)
168 #define RExC_npar (pRExC_state->npar)
169 #define RExC_nestroot (pRExC_state->nestroot)
170 #define RExC_extralen (pRExC_state->extralen)
171 #define RExC_seen_zerolen (pRExC_state->seen_zerolen)
172 #define RExC_seen_evals (pRExC_state->seen_evals)
173 #define RExC_utf8 (pRExC_state->utf8)
174 #define RExC_orig_utf8 (pRExC_state->orig_utf8)
175 #define RExC_charnames (pRExC_state->charnames)
176 #define RExC_open_parens (pRExC_state->open_parens)
177 #define RExC_close_parens (pRExC_state->close_parens)
178 #define RExC_opend (pRExC_state->opend)
179 #define RExC_paren_names (pRExC_state->paren_names)
180 #define RExC_recurse (pRExC_state->recurse)
181 #define RExC_recurse_count (pRExC_state->recurse_count)
184 #define ISMULT1(c) ((c) == '*' || (c) == '+' || (c) == '?')
185 #define ISMULT2(s) ((*s) == '*' || (*s) == '+' || (*s) == '?' || \
186 ((*s) == '{' && regcurly(s)))
189 #undef SPSTART /* dratted cpp namespace... */
192 * Flags to be passed up and down.
194 #define WORST 0 /* Worst case. */
195 #define HASWIDTH 0x01 /* Known to match non-null strings. */
196 #define SIMPLE 0x02 /* Simple enough to be STAR/PLUS operand. */
197 #define SPSTART 0x04 /* Starts with * or +. */
198 #define TRYAGAIN 0x08 /* Weeded out a declaration. */
199 #define POSTPONED 0x10 /* (?1),(?&name), (??{...}) or similar */
201 #define REG_NODE_NUM(x) ((x) ? (int)((x)-RExC_emit_start) : -1)
203 /* whether trie related optimizations are enabled */
204 #if PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION
205 #define TRIE_STUDY_OPT
206 #define FULL_TRIE_STUDY
212 #define PBYTE(u8str,paren) ((U8*)(u8str))[(paren) >> 3]
213 #define PBITVAL(paren) (1 << ((paren) & 7))
214 #define PAREN_TEST(u8str,paren) ( PBYTE(u8str,paren) & PBITVAL(paren))
215 #define PAREN_SET(u8str,paren) PBYTE(u8str,paren) |= PBITVAL(paren)
216 #define PAREN_UNSET(u8str,paren) PBYTE(u8str,paren) &= (~PBITVAL(paren))
219 /* About scan_data_t.
221 During optimisation we recurse through the regexp program performing
222 various inplace (keyhole style) optimisations. In addition study_chunk
223 and scan_commit populate this data structure with information about
224 what strings MUST appear in the pattern. We look for the longest
225 string that must appear for at a fixed location, and we look for the
226 longest string that may appear at a floating location. So for instance
231 Both 'FOO' and 'A' are fixed strings. Both 'B' and 'BAR' are floating
232 strings (because they follow a .* construct). study_chunk will identify
233 both FOO and BAR as being the longest fixed and floating strings respectively.
235 The strings can be composites, for instance
239 will result in a composite fixed substring 'foo'.
241 For each string some basic information is maintained:
243 - offset or min_offset
244 This is the position the string must appear at, or not before.
245 It also implicitly (when combined with minlenp) tells us how many
246 character must match before the string we are searching.
247 Likewise when combined with minlenp and the length of the string
248 tells us how many characters must appear after the string we have
252 Only used for floating strings. This is the rightmost point that
253 the string can appear at. Ifset to I32 max it indicates that the
254 string can occur infinitely far to the right.
257 A pointer to the minimum length of the pattern that the string
258 was found inside. This is important as in the case of positive
259 lookahead or positive lookbehind we can have multiple patterns
264 The minimum length of the pattern overall is 3, the minimum length
265 of the lookahead part is 3, but the minimum length of the part that
266 will actually match is 1. So 'FOO's minimum length is 3, but the
267 minimum length for the F is 1. This is important as the minimum length
268 is used to determine offsets in front of and behind the string being
269 looked for. Since strings can be composites this is the length of the
270 pattern at the time it was commited with a scan_commit. Note that
271 the length is calculated by study_chunk, so that the minimum lengths
272 are not known until the full pattern has been compiled, thus the
273 pointer to the value.
277 In the case of lookbehind the string being searched for can be
278 offset past the start point of the final matching string.
279 If this value was just blithely removed from the min_offset it would
280 invalidate some of the calculations for how many chars must match
281 before or after (as they are derived from min_offset and minlen and
282 the length of the string being searched for).
283 When the final pattern is compiled and the data is moved from the
284 scan_data_t structure into the regexp structure the information
285 about lookbehind is factored in, with the information that would
286 have been lost precalculated in the end_shift field for the
289 The fields pos_min and pos_delta are used to store the minimum offset
290 and the delta to the maximum offset at the current point in the pattern.
294 typedef struct scan_data_t {
295 /*I32 len_min; unused */
296 /*I32 len_delta; unused */
300 I32 last_end; /* min value, <0 unless valid. */
303 SV **longest; /* Either &l_fixed, or &l_float. */
304 SV *longest_fixed; /* longest fixed string found in pattern */
305 I32 offset_fixed; /* offset where it starts */
306 I32 *minlen_fixed; /* pointer to the minlen relevent to the string */
307 I32 lookbehind_fixed; /* is the position of the string modfied by LB */
308 SV *longest_float; /* longest floating string found in pattern */
309 I32 offset_float_min; /* earliest point in string it can appear */
310 I32 offset_float_max; /* latest point in string it can appear */
311 I32 *minlen_float; /* pointer to the minlen relevent to the string */
312 I32 lookbehind_float; /* is the position of the string modified by LB */
316 struct regnode_charclass_class *start_class;
320 * Forward declarations for pregcomp()'s friends.
323 static const scan_data_t zero_scan_data =
324 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ,0};
326 #define SF_BEFORE_EOL (SF_BEFORE_SEOL|SF_BEFORE_MEOL)
327 #define SF_BEFORE_SEOL 0x0001
328 #define SF_BEFORE_MEOL 0x0002
329 #define SF_FIX_BEFORE_EOL (SF_FIX_BEFORE_SEOL|SF_FIX_BEFORE_MEOL)
330 #define SF_FL_BEFORE_EOL (SF_FL_BEFORE_SEOL|SF_FL_BEFORE_MEOL)
333 # define SF_FIX_SHIFT_EOL (0+2)
334 # define SF_FL_SHIFT_EOL (0+4)
336 # define SF_FIX_SHIFT_EOL (+2)
337 # define SF_FL_SHIFT_EOL (+4)
340 #define SF_FIX_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FIX_SHIFT_EOL)
341 #define SF_FIX_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FIX_SHIFT_EOL)
343 #define SF_FL_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FL_SHIFT_EOL)
344 #define SF_FL_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FL_SHIFT_EOL) /* 0x20 */
345 #define SF_IS_INF 0x0040
346 #define SF_HAS_PAR 0x0080
347 #define SF_IN_PAR 0x0100
348 #define SF_HAS_EVAL 0x0200
349 #define SCF_DO_SUBSTR 0x0400
350 #define SCF_DO_STCLASS_AND 0x0800
351 #define SCF_DO_STCLASS_OR 0x1000
352 #define SCF_DO_STCLASS (SCF_DO_STCLASS_AND|SCF_DO_STCLASS_OR)
353 #define SCF_WHILEM_VISITED_POS 0x2000
355 #define SCF_TRIE_RESTUDY 0x4000 /* Do restudy? */
356 #define SCF_SEEN_ACCEPT 0x8000
358 #define UTF (RExC_utf8 != 0)
359 #define LOC ((RExC_flags & RXf_PMf_LOCALE) != 0)
360 #define FOLD ((RExC_flags & RXf_PMf_FOLD) != 0)
362 #define OOB_UNICODE 12345678
363 #define OOB_NAMEDCLASS -1
365 #define CHR_SVLEN(sv) (UTF ? sv_len_utf8(sv) : SvCUR(sv))
366 #define CHR_DIST(a,b) (UTF ? utf8_distance(a,b) : a - b)
369 /* length of regex to show in messages that don't mark a position within */
370 #define RegexLengthToShowInErrorMessages 127
373 * If MARKER[12] are adjusted, be sure to adjust the constants at the top
374 * of t/op/regmesg.t, the tests in t/op/re_tests, and those in
375 * op/pragma/warn/regcomp.
377 #define MARKER1 "<-- HERE" /* marker as it appears in the description */
378 #define MARKER2 " <-- HERE " /* marker as it appears within the regex */
380 #define REPORT_LOCATION " in regex; marked by " MARKER1 " in m/%.*s" MARKER2 "%s/"
383 * Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
384 * arg. Show regex, up to a maximum length. If it's too long, chop and add
387 #define _FAIL(code) STMT_START { \
388 const char *ellipses = ""; \
389 IV len = RExC_end - RExC_precomp; \
392 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx); \
393 if (len > RegexLengthToShowInErrorMessages) { \
394 /* chop 10 shorter than the max, to ensure meaning of "..." */ \
395 len = RegexLengthToShowInErrorMessages - 10; \
401 #define FAIL(msg) _FAIL( \
402 Perl_croak(aTHX_ "%s in regex m/%.*s%s/", \
403 msg, (int)len, RExC_precomp, ellipses))
405 #define FAIL2(msg,arg) _FAIL( \
406 Perl_croak(aTHX_ msg " in regex m/%.*s%s/", \
407 arg, (int)len, RExC_precomp, ellipses))
410 * Simple_vFAIL -- like FAIL, but marks the current location in the scan
412 #define Simple_vFAIL(m) STMT_START { \
413 const IV offset = RExC_parse - RExC_precomp; \
414 Perl_croak(aTHX_ "%s" REPORT_LOCATION, \
415 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
419 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL()
421 #define vFAIL(m) STMT_START { \
423 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx); \
428 * Like Simple_vFAIL(), but accepts two arguments.
430 #define Simple_vFAIL2(m,a1) STMT_START { \
431 const IV offset = RExC_parse - RExC_precomp; \
432 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, \
433 (int)offset, RExC_precomp, RExC_precomp + offset); \
437 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL2().
439 #define vFAIL2(m,a1) STMT_START { \
441 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx); \
442 Simple_vFAIL2(m, a1); \
447 * Like Simple_vFAIL(), but accepts three arguments.
449 #define Simple_vFAIL3(m, a1, a2) STMT_START { \
450 const IV offset = RExC_parse - RExC_precomp; \
451 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, \
452 (int)offset, RExC_precomp, RExC_precomp + offset); \
456 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL3().
458 #define vFAIL3(m,a1,a2) STMT_START { \
460 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx); \
461 Simple_vFAIL3(m, a1, a2); \
465 * Like Simple_vFAIL(), but accepts four arguments.
467 #define Simple_vFAIL4(m, a1, a2, a3) STMT_START { \
468 const IV offset = RExC_parse - RExC_precomp; \
469 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, \
470 (int)offset, RExC_precomp, RExC_precomp + offset); \
473 #define vWARN(loc,m) STMT_START { \
474 const IV offset = loc - RExC_precomp; \
475 Perl_warner(aTHX_ packWARN(WARN_REGEXP), "%s" REPORT_LOCATION, \
476 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
479 #define vWARNdep(loc,m) STMT_START { \
480 const IV offset = loc - RExC_precomp; \
481 Perl_warner(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
482 "%s" REPORT_LOCATION, \
483 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
487 #define vWARN2(loc, m, a1) STMT_START { \
488 const IV offset = loc - RExC_precomp; \
489 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
490 a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
493 #define vWARN3(loc, m, a1, a2) STMT_START { \
494 const IV offset = loc - RExC_precomp; \
495 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
496 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
499 #define vWARN4(loc, m, a1, a2, a3) STMT_START { \
500 const IV offset = loc - RExC_precomp; \
501 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
502 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
505 #define vWARN5(loc, m, a1, a2, a3, a4) STMT_START { \
506 const IV offset = loc - RExC_precomp; \
507 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
508 a1, a2, a3, a4, (int)offset, RExC_precomp, RExC_precomp + offset); \
512 /* Allow for side effects in s */
513 #define REGC(c,s) STMT_START { \
514 if (!SIZE_ONLY) *(s) = (c); else (void)(s); \
517 /* Macros for recording node offsets. 20001227 mjd@plover.com
518 * Nodes are numbered 1, 2, 3, 4. Node #n's position is recorded in
519 * element 2*n-1 of the array. Element #2n holds the byte length node #n.
520 * Element 0 holds the number n.
521 * Position is 1 indexed.
523 #ifndef RE_TRACK_PATTERN_OFFSETS
524 #define Set_Node_Offset_To_R(node,byte)
525 #define Set_Node_Offset(node,byte)
526 #define Set_Cur_Node_Offset
527 #define Set_Node_Length_To_R(node,len)
528 #define Set_Node_Length(node,len)
529 #define Set_Node_Cur_Length(node)
530 #define Node_Offset(n)
531 #define Node_Length(n)
532 #define Set_Node_Offset_Length(node,offset,len)
533 #define ProgLen(ri) ri->u.proglen
534 #define SetProgLen(ri,x) ri->u.proglen = x
536 #define ProgLen(ri) ri->u.offsets[0]
537 #define SetProgLen(ri,x) ri->u.offsets[0] = x
538 #define Set_Node_Offset_To_R(node,byte) STMT_START { \
540 MJD_OFFSET_DEBUG(("** (%d) offset of node %d is %d.\n", \
541 __LINE__, (int)(node), (int)(byte))); \
543 Perl_croak(aTHX_ "value of node is %d in Offset macro", (int)(node)); \
545 RExC_offsets[2*(node)-1] = (byte); \
550 #define Set_Node_Offset(node,byte) \
551 Set_Node_Offset_To_R((node)-RExC_emit_start, (byte)-RExC_start)
552 #define Set_Cur_Node_Offset Set_Node_Offset(RExC_emit, RExC_parse)
554 #define Set_Node_Length_To_R(node,len) STMT_START { \
556 MJD_OFFSET_DEBUG(("** (%d) size of node %d is %d.\n", \
557 __LINE__, (int)(node), (int)(len))); \
559 Perl_croak(aTHX_ "value of node is %d in Length macro", (int)(node)); \
561 RExC_offsets[2*(node)] = (len); \
566 #define Set_Node_Length(node,len) \
567 Set_Node_Length_To_R((node)-RExC_emit_start, len)
568 #define Set_Cur_Node_Length(len) Set_Node_Length(RExC_emit, len)
569 #define Set_Node_Cur_Length(node) \
570 Set_Node_Length(node, RExC_parse - parse_start)
572 /* Get offsets and lengths */
573 #define Node_Offset(n) (RExC_offsets[2*((n)-RExC_emit_start)-1])
574 #define Node_Length(n) (RExC_offsets[2*((n)-RExC_emit_start)])
576 #define Set_Node_Offset_Length(node,offset,len) STMT_START { \
577 Set_Node_Offset_To_R((node)-RExC_emit_start, (offset)); \
578 Set_Node_Length_To_R((node)-RExC_emit_start, (len)); \
582 #if PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS
583 #define EXPERIMENTAL_INPLACESCAN
584 #endif /*RE_TRACK_PATTERN_OFFSETS*/
586 #define DEBUG_STUDYDATA(str,data,depth) \
587 DEBUG_OPTIMISE_MORE_r(if(data){ \
588 PerlIO_printf(Perl_debug_log, \
589 "%*s" str "Pos:%"IVdf"/%"IVdf \
590 " Flags: 0x%"UVXf" Whilem_c: %"IVdf" Lcp: %"IVdf" %s", \
591 (int)(depth)*2, "", \
592 (IV)((data)->pos_min), \
593 (IV)((data)->pos_delta), \
594 (UV)((data)->flags), \
595 (IV)((data)->whilem_c), \
596 (IV)((data)->last_closep ? *((data)->last_closep) : -1), \
597 is_inf ? "INF " : "" \
599 if ((data)->last_found) \
600 PerlIO_printf(Perl_debug_log, \
601 "Last:'%s' %"IVdf":%"IVdf"/%"IVdf" %sFixed:'%s' @ %"IVdf \
602 " %sFloat: '%s' @ %"IVdf"/%"IVdf"", \
603 SvPVX_const((data)->last_found), \
604 (IV)((data)->last_end), \
605 (IV)((data)->last_start_min), \
606 (IV)((data)->last_start_max), \
607 ((data)->longest && \
608 (data)->longest==&((data)->longest_fixed)) ? "*" : "", \
609 SvPVX_const((data)->longest_fixed), \
610 (IV)((data)->offset_fixed), \
611 ((data)->longest && \
612 (data)->longest==&((data)->longest_float)) ? "*" : "", \
613 SvPVX_const((data)->longest_float), \
614 (IV)((data)->offset_float_min), \
615 (IV)((data)->offset_float_max) \
617 PerlIO_printf(Perl_debug_log,"\n"); \
620 static void clear_re(pTHX_ void *r);
622 /* Mark that we cannot extend a found fixed substring at this point.
623 Update the longest found anchored substring and the longest found
624 floating substrings if needed. */
627 S_scan_commit(pTHX_ const RExC_state_t *pRExC_state, scan_data_t *data, I32 *minlenp, int is_inf)
629 const STRLEN l = CHR_SVLEN(data->last_found);
630 const STRLEN old_l = CHR_SVLEN(*data->longest);
631 GET_RE_DEBUG_FLAGS_DECL;
633 if ((l >= old_l) && ((l > old_l) || (data->flags & SF_BEFORE_EOL))) {
634 SvSetMagicSV(*data->longest, data->last_found);
635 if (*data->longest == data->longest_fixed) {
636 data->offset_fixed = l ? data->last_start_min : data->pos_min;
637 if (data->flags & SF_BEFORE_EOL)
639 |= ((data->flags & SF_BEFORE_EOL) << SF_FIX_SHIFT_EOL);
641 data->flags &= ~SF_FIX_BEFORE_EOL;
642 data->minlen_fixed=minlenp;
643 data->lookbehind_fixed=0;
645 else { /* *data->longest == data->longest_float */
646 data->offset_float_min = l ? data->last_start_min : data->pos_min;
647 data->offset_float_max = (l
648 ? data->last_start_max
649 : data->pos_min + data->pos_delta);
650 if (is_inf || (U32)data->offset_float_max > (U32)I32_MAX)
651 data->offset_float_max = I32_MAX;
652 if (data->flags & SF_BEFORE_EOL)
654 |= ((data->flags & SF_BEFORE_EOL) << SF_FL_SHIFT_EOL);
656 data->flags &= ~SF_FL_BEFORE_EOL;
657 data->minlen_float=minlenp;
658 data->lookbehind_float=0;
661 SvCUR_set(data->last_found, 0);
663 SV * const sv = data->last_found;
664 if (SvUTF8(sv) && SvMAGICAL(sv)) {
665 MAGIC * const mg = mg_find(sv, PERL_MAGIC_utf8);
671 data->flags &= ~SF_BEFORE_EOL;
672 DEBUG_STUDYDATA("commit: ",data,0);
675 /* Can match anything (initialization) */
677 S_cl_anything(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
679 ANYOF_CLASS_ZERO(cl);
680 ANYOF_BITMAP_SETALL(cl);
681 cl->flags = ANYOF_EOS|ANYOF_UNICODE_ALL;
683 cl->flags |= ANYOF_LOCALE;
686 /* Can match anything (initialization) */
688 S_cl_is_anything(const struct regnode_charclass_class *cl)
692 for (value = 0; value <= ANYOF_MAX; value += 2)
693 if (ANYOF_CLASS_TEST(cl, value) && ANYOF_CLASS_TEST(cl, value + 1))
695 if (!(cl->flags & ANYOF_UNICODE_ALL))
697 if (!ANYOF_BITMAP_TESTALLSET((const void*)cl))
702 /* Can match anything (initialization) */
704 S_cl_init(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
706 Zero(cl, 1, struct regnode_charclass_class);
708 cl_anything(pRExC_state, cl);
712 S_cl_init_zero(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
714 Zero(cl, 1, struct regnode_charclass_class);
716 cl_anything(pRExC_state, cl);
718 cl->flags |= ANYOF_LOCALE;
721 /* 'And' a given class with another one. Can create false positives */
722 /* We assume that cl is not inverted */
724 S_cl_and(struct regnode_charclass_class *cl,
725 const struct regnode_charclass_class *and_with)
728 assert(and_with->type == ANYOF);
729 if (!(and_with->flags & ANYOF_CLASS)
730 && !(cl->flags & ANYOF_CLASS)
731 && (and_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
732 && !(and_with->flags & ANYOF_FOLD)
733 && !(cl->flags & ANYOF_FOLD)) {
736 if (and_with->flags & ANYOF_INVERT)
737 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
738 cl->bitmap[i] &= ~and_with->bitmap[i];
740 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
741 cl->bitmap[i] &= and_with->bitmap[i];
742 } /* XXXX: logic is complicated otherwise, leave it along for a moment. */
743 if (!(and_with->flags & ANYOF_EOS))
744 cl->flags &= ~ANYOF_EOS;
746 if (cl->flags & ANYOF_UNICODE_ALL && and_with->flags & ANYOF_UNICODE &&
747 !(and_with->flags & ANYOF_INVERT)) {
748 cl->flags &= ~ANYOF_UNICODE_ALL;
749 cl->flags |= ANYOF_UNICODE;
750 ARG_SET(cl, ARG(and_with));
752 if (!(and_with->flags & ANYOF_UNICODE_ALL) &&
753 !(and_with->flags & ANYOF_INVERT))
754 cl->flags &= ~ANYOF_UNICODE_ALL;
755 if (!(and_with->flags & (ANYOF_UNICODE|ANYOF_UNICODE_ALL)) &&
756 !(and_with->flags & ANYOF_INVERT))
757 cl->flags &= ~ANYOF_UNICODE;
760 /* 'OR' a given class with another one. Can create false positives */
761 /* We assume that cl is not inverted */
763 S_cl_or(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, const struct regnode_charclass_class *or_with)
765 if (or_with->flags & ANYOF_INVERT) {
767 * (B1 | CL1) | (!B2 & !CL2) = (B1 | !B2 & !CL2) | (CL1 | (!B2 & !CL2))
768 * <= (B1 | !B2) | (CL1 | !CL2)
769 * which is wasteful if CL2 is small, but we ignore CL2:
770 * (B1 | CL1) | (!B2 & !CL2) <= (B1 | CL1) | !B2 = (B1 | !B2) | CL1
771 * XXXX Can we handle case-fold? Unclear:
772 * (OK1(i) | OK1(i')) | !(OK1(i) | OK1(i')) =
773 * (OK1(i) | OK1(i')) | (!OK1(i) & !OK1(i'))
775 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
776 && !(or_with->flags & ANYOF_FOLD)
777 && !(cl->flags & ANYOF_FOLD) ) {
780 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
781 cl->bitmap[i] |= ~or_with->bitmap[i];
782 } /* XXXX: logic is complicated otherwise */
784 cl_anything(pRExC_state, cl);
787 /* (B1 | CL1) | (B2 | CL2) = (B1 | B2) | (CL1 | CL2)) */
788 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
789 && (!(or_with->flags & ANYOF_FOLD)
790 || (cl->flags & ANYOF_FOLD)) ) {
793 /* OR char bitmap and class bitmap separately */
794 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
795 cl->bitmap[i] |= or_with->bitmap[i];
796 if (or_with->flags & ANYOF_CLASS) {
797 for (i = 0; i < ANYOF_CLASSBITMAP_SIZE; i++)
798 cl->classflags[i] |= or_with->classflags[i];
799 cl->flags |= ANYOF_CLASS;
802 else { /* XXXX: logic is complicated, leave it along for a moment. */
803 cl_anything(pRExC_state, cl);
806 if (or_with->flags & ANYOF_EOS)
807 cl->flags |= ANYOF_EOS;
809 if (cl->flags & ANYOF_UNICODE && or_with->flags & ANYOF_UNICODE &&
810 ARG(cl) != ARG(or_with)) {
811 cl->flags |= ANYOF_UNICODE_ALL;
812 cl->flags &= ~ANYOF_UNICODE;
814 if (or_with->flags & ANYOF_UNICODE_ALL) {
815 cl->flags |= ANYOF_UNICODE_ALL;
816 cl->flags &= ~ANYOF_UNICODE;
820 #define TRIE_LIST_ITEM(state,idx) (trie->states[state].trans.list)[ idx ]
821 #define TRIE_LIST_CUR(state) ( TRIE_LIST_ITEM( state, 0 ).forid )
822 #define TRIE_LIST_LEN(state) ( TRIE_LIST_ITEM( state, 0 ).newstate )
823 #define TRIE_LIST_USED(idx) ( trie->states[state].trans.list ? (TRIE_LIST_CUR( idx ) - 1) : 0 )
828 dump_trie(trie,widecharmap,revcharmap)
829 dump_trie_interim_list(trie,widecharmap,revcharmap,next_alloc)
830 dump_trie_interim_table(trie,widecharmap,revcharmap,next_alloc)
832 These routines dump out a trie in a somewhat readable format.
833 The _interim_ variants are used for debugging the interim
834 tables that are used to generate the final compressed
835 representation which is what dump_trie expects.
837 Part of the reason for their existance is to provide a form
838 of documentation as to how the different representations function.
843 Dumps the final compressed table form of the trie to Perl_debug_log.
844 Used for debugging make_trie().
848 S_dump_trie(pTHX_ const struct _reg_trie_data *trie, HV *widecharmap,
849 AV *revcharmap, U32 depth)
852 SV *sv=sv_newmortal();
853 int colwidth= widecharmap ? 6 : 4;
854 GET_RE_DEBUG_FLAGS_DECL;
857 PerlIO_printf( Perl_debug_log, "%*sChar : %-6s%-6s%-4s ",
858 (int)depth * 2 + 2,"",
859 "Match","Base","Ofs" );
861 for( state = 0 ; state < trie->uniquecharcount ; state++ ) {
862 SV ** const tmp = av_fetch( revcharmap, state, 0);
864 PerlIO_printf( Perl_debug_log, "%*s",
866 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
867 PL_colors[0], PL_colors[1],
868 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
869 PERL_PV_ESCAPE_FIRSTCHAR
874 PerlIO_printf( Perl_debug_log, "\n%*sState|-----------------------",
875 (int)depth * 2 + 2,"");
877 for( state = 0 ; state < trie->uniquecharcount ; state++ )
878 PerlIO_printf( Perl_debug_log, "%.*s", colwidth, "--------");
879 PerlIO_printf( Perl_debug_log, "\n");
881 for( state = 1 ; state < trie->statecount ; state++ ) {
882 const U32 base = trie->states[ state ].trans.base;
884 PerlIO_printf( Perl_debug_log, "%*s#%4"UVXf"|", (int)depth * 2 + 2,"", (UV)state);
886 if ( trie->states[ state ].wordnum ) {
887 PerlIO_printf( Perl_debug_log, " W%4X", trie->states[ state ].wordnum );
889 PerlIO_printf( Perl_debug_log, "%6s", "" );
892 PerlIO_printf( Perl_debug_log, " @%4"UVXf" ", (UV)base );
897 while( ( base + ofs < trie->uniquecharcount ) ||
898 ( base + ofs - trie->uniquecharcount < trie->lasttrans
899 && trie->trans[ base + ofs - trie->uniquecharcount ].check != state))
902 PerlIO_printf( Perl_debug_log, "+%2"UVXf"[ ", (UV)ofs);
904 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
905 if ( ( base + ofs >= trie->uniquecharcount ) &&
906 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
907 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
909 PerlIO_printf( Perl_debug_log, "%*"UVXf,
911 (UV)trie->trans[ base + ofs - trie->uniquecharcount ].next );
913 PerlIO_printf( Perl_debug_log, "%*s",colwidth," ." );
917 PerlIO_printf( Perl_debug_log, "]");
920 PerlIO_printf( Perl_debug_log, "\n" );
924 Dumps a fully constructed but uncompressed trie in list form.
925 List tries normally only are used for construction when the number of
926 possible chars (trie->uniquecharcount) is very high.
927 Used for debugging make_trie().
930 S_dump_trie_interim_list(pTHX_ const struct _reg_trie_data *trie,
931 HV *widecharmap, AV *revcharmap, U32 next_alloc,
935 SV *sv=sv_newmortal();
936 int colwidth= widecharmap ? 6 : 4;
937 GET_RE_DEBUG_FLAGS_DECL;
938 /* print out the table precompression. */
939 PerlIO_printf( Perl_debug_log, "%*sState :Word | Transition Data\n%*s%s",
940 (int)depth * 2 + 2,"", (int)depth * 2 + 2,"",
941 "------:-----+-----------------\n" );
943 for( state=1 ; state < next_alloc ; state ++ ) {
946 PerlIO_printf( Perl_debug_log, "%*s %4"UVXf" :",
947 (int)depth * 2 + 2,"", (UV)state );
948 if ( ! trie->states[ state ].wordnum ) {
949 PerlIO_printf( Perl_debug_log, "%5s| ","");
951 PerlIO_printf( Perl_debug_log, "W%4x| ",
952 trie->states[ state ].wordnum
955 for( charid = 1 ; charid <= TRIE_LIST_USED( state ) ; charid++ ) {
956 SV ** const tmp = av_fetch( revcharmap, TRIE_LIST_ITEM(state,charid).forid, 0);
958 PerlIO_printf( Perl_debug_log, "%*s:%3X=%4"UVXf" | ",
960 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
961 PL_colors[0], PL_colors[1],
962 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
963 PERL_PV_ESCAPE_FIRSTCHAR
965 TRIE_LIST_ITEM(state,charid).forid,
966 (UV)TRIE_LIST_ITEM(state,charid).newstate
969 PerlIO_printf(Perl_debug_log, "\n%*s| ",
970 (int)((depth * 2) + 14), "");
973 PerlIO_printf( Perl_debug_log, "\n");
978 Dumps a fully constructed but uncompressed trie in table form.
979 This is the normal DFA style state transition table, with a few
980 twists to facilitate compression later.
981 Used for debugging make_trie().
984 S_dump_trie_interim_table(pTHX_ const struct _reg_trie_data *trie,
985 HV *widecharmap, AV *revcharmap, U32 next_alloc,
990 SV *sv=sv_newmortal();
991 int colwidth= widecharmap ? 6 : 4;
992 GET_RE_DEBUG_FLAGS_DECL;
995 print out the table precompression so that we can do a visual check
996 that they are identical.
999 PerlIO_printf( Perl_debug_log, "%*sChar : ",(int)depth * 2 + 2,"" );
1001 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1002 SV ** const tmp = av_fetch( revcharmap, charid, 0);
1004 PerlIO_printf( Perl_debug_log, "%*s",
1006 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1007 PL_colors[0], PL_colors[1],
1008 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1009 PERL_PV_ESCAPE_FIRSTCHAR
1015 PerlIO_printf( Perl_debug_log, "\n%*sState+-",(int)depth * 2 + 2,"" );
1017 for( charid=0 ; charid < trie->uniquecharcount ; charid++ ) {
1018 PerlIO_printf( Perl_debug_log, "%.*s", colwidth,"--------");
1021 PerlIO_printf( Perl_debug_log, "\n" );
1023 for( state=1 ; state < next_alloc ; state += trie->uniquecharcount ) {
1025 PerlIO_printf( Perl_debug_log, "%*s%4"UVXf" : ",
1026 (int)depth * 2 + 2,"",
1027 (UV)TRIE_NODENUM( state ) );
1029 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1030 UV v=(UV)SAFE_TRIE_NODENUM( trie->trans[ state + charid ].next );
1032 PerlIO_printf( Perl_debug_log, "%*"UVXf, colwidth, v );
1034 PerlIO_printf( Perl_debug_log, "%*s", colwidth, "." );
1036 if ( ! trie->states[ TRIE_NODENUM( state ) ].wordnum ) {
1037 PerlIO_printf( Perl_debug_log, " (%4"UVXf")\n", (UV)trie->trans[ state ].check );
1039 PerlIO_printf( Perl_debug_log, " (%4"UVXf") W%4X\n", (UV)trie->trans[ state ].check,
1040 trie->states[ TRIE_NODENUM( state ) ].wordnum );
1047 /* make_trie(startbranch,first,last,tail,word_count,flags,depth)
1048 startbranch: the first branch in the whole branch sequence
1049 first : start branch of sequence of branch-exact nodes.
1050 May be the same as startbranch
1051 last : Thing following the last branch.
1052 May be the same as tail.
1053 tail : item following the branch sequence
1054 count : words in the sequence
1055 flags : currently the OP() type we will be building one of /EXACT(|F|Fl)/
1056 depth : indent depth
1058 Inplace optimizes a sequence of 2 or more Branch-Exact nodes into a TRIE node.
1060 A trie is an N'ary tree where the branches are determined by digital
1061 decomposition of the key. IE, at the root node you look up the 1st character and
1062 follow that branch repeat until you find the end of the branches. Nodes can be
1063 marked as "accepting" meaning they represent a complete word. Eg:
1067 would convert into the following structure. Numbers represent states, letters
1068 following numbers represent valid transitions on the letter from that state, if
1069 the number is in square brackets it represents an accepting state, otherwise it
1070 will be in parenthesis.
1072 +-h->+-e->[3]-+-r->(8)-+-s->[9]
1076 (1) +-i->(6)-+-s->[7]
1078 +-s->(3)-+-h->(4)-+-e->[5]
1080 Accept Word Mapping: 3=>1 (he),5=>2 (she), 7=>3 (his), 9=>4 (hers)
1082 This shows that when matching against the string 'hers' we will begin at state 1
1083 read 'h' and move to state 2, read 'e' and move to state 3 which is accepting,
1084 then read 'r' and go to state 8 followed by 's' which takes us to state 9 which
1085 is also accepting. Thus we know that we can match both 'he' and 'hers' with a
1086 single traverse. We store a mapping from accepting to state to which word was
1087 matched, and then when we have multiple possibilities we try to complete the
1088 rest of the regex in the order in which they occured in the alternation.
1090 The only prior NFA like behaviour that would be changed by the TRIE support is
1091 the silent ignoring of duplicate alternations which are of the form:
1093 / (DUPE|DUPE) X? (?{ ... }) Y /x
1095 Thus EVAL blocks follwing a trie may be called a different number of times with
1096 and without the optimisation. With the optimisations dupes will be silently
1097 ignored. This inconsistant behaviour of EVAL type nodes is well established as
1098 the following demonstrates:
1100 'words'=~/(word|word|word)(?{ print $1 })[xyz]/
1102 which prints out 'word' three times, but
1104 'words'=~/(word|word|word)(?{ print $1 })S/
1106 which doesnt print it out at all. This is due to other optimisations kicking in.
1108 Example of what happens on a structural level:
1110 The regexp /(ac|ad|ab)+/ will produce the folowing debug output:
1112 1: CURLYM[1] {1,32767}(18)
1123 This would be optimizable with startbranch=5, first=5, last=16, tail=16
1124 and should turn into:
1126 1: CURLYM[1] {1,32767}(18)
1128 [Words:3 Chars Stored:6 Unique Chars:4 States:5 NCP:1]
1136 Cases where tail != last would be like /(?foo|bar)baz/:
1146 which would be optimizable with startbranch=1, first=1, last=7, tail=8
1147 and would end up looking like:
1150 [Words:2 Chars Stored:6 Unique Chars:5 States:7 NCP:1]
1157 d = uvuni_to_utf8_flags(d, uv, 0);
1159 is the recommended Unicode-aware way of saying
1164 #define TRIE_STORE_REVCHAR \
1166 SV *tmp = newSVpvs(""); \
1167 if (UTF) SvUTF8_on(tmp); \
1168 Perl_sv_catpvf( aTHX_ tmp, "%c", (int)uvc ); \
1169 av_push( revcharmap, tmp ); \
1172 #define TRIE_READ_CHAR STMT_START { \
1176 if ( foldlen > 0 ) { \
1177 uvc = utf8n_to_uvuni( scan, UTF8_MAXLEN, &len, uniflags ); \
1182 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1183 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1184 foldlen -= UNISKIP( uvc ); \
1185 scan = foldbuf + UNISKIP( uvc ); \
1188 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1198 #define TRIE_LIST_PUSH(state,fid,ns) STMT_START { \
1199 if ( TRIE_LIST_CUR( state ) >=TRIE_LIST_LEN( state ) ) { \
1200 U32 ging = TRIE_LIST_LEN( state ) *= 2; \
1201 Renew( trie->states[ state ].trans.list, ging, reg_trie_trans_le ); \
1203 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).forid = fid; \
1204 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).newstate = ns; \
1205 TRIE_LIST_CUR( state )++; \
1208 #define TRIE_LIST_NEW(state) STMT_START { \
1209 Newxz( trie->states[ state ].trans.list, \
1210 4, reg_trie_trans_le ); \
1211 TRIE_LIST_CUR( state ) = 1; \
1212 TRIE_LIST_LEN( state ) = 4; \
1215 #define TRIE_HANDLE_WORD(state) STMT_START { \
1216 U16 dupe= trie->states[ state ].wordnum; \
1217 regnode * const noper_next = regnext( noper ); \
1219 if (trie->wordlen) \
1220 trie->wordlen[ curword ] = wordlen; \
1222 /* store the word for dumping */ \
1224 if (OP(noper) != NOTHING) \
1225 tmp = newSVpvn(STRING(noper), STR_LEN(noper)); \
1227 tmp = newSVpvn( "", 0 ); \
1228 if ( UTF ) SvUTF8_on( tmp ); \
1229 av_push( trie_words, tmp ); \
1234 if ( noper_next < tail ) { \
1236 trie->jump = (U16 *) PerlMemShared_calloc( word_count + 1, sizeof(U16) ); \
1237 trie->jump[curword] = (U16)(noper_next - convert); \
1239 jumper = noper_next; \
1241 nextbranch= regnext(cur); \
1245 /* So it's a dupe. This means we need to maintain a */\
1246 /* linked-list from the first to the next. */\
1247 /* we only allocate the nextword buffer when there */\
1248 /* a dupe, so first time we have to do the allocation */\
1249 if (!trie->nextword) \
1250 trie->nextword = (U16 *) \
1251 PerlMemShared_calloc( word_count + 1, sizeof(U16)); \
1252 while ( trie->nextword[dupe] ) \
1253 dupe= trie->nextword[dupe]; \
1254 trie->nextword[dupe]= curword; \
1256 /* we haven't inserted this word yet. */ \
1257 trie->states[ state ].wordnum = curword; \
1262 #define TRIE_TRANS_STATE(state,base,ucharcount,charid,special) \
1263 ( ( base + charid >= ucharcount \
1264 && base + charid < ubound \
1265 && state == trie->trans[ base - ucharcount + charid ].check \
1266 && trie->trans[ base - ucharcount + charid ].next ) \
1267 ? trie->trans[ base - ucharcount + charid ].next \
1268 : ( state==1 ? special : 0 ) \
1272 #define MADE_JUMP_TRIE 2
1273 #define MADE_EXACT_TRIE 4
1276 S_make_trie(pTHX_ RExC_state_t *pRExC_state, regnode *startbranch, regnode *first, regnode *last, regnode *tail, U32 word_count, U32 flags, U32 depth)
1279 /* first pass, loop through and scan words */
1280 reg_trie_data *trie;
1281 HV *widecharmap = NULL;
1282 AV *revcharmap = newAV();
1284 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1289 regnode *jumper = NULL;
1290 regnode *nextbranch = NULL;
1291 regnode *convert = NULL;
1292 /* we just use folder as a flag in utf8 */
1293 const U8 * const folder = ( flags == EXACTF
1295 : ( flags == EXACTFL
1302 const U32 data_slot = add_data( pRExC_state, 4, "tuuu" );
1303 AV *trie_words = NULL;
1304 /* along with revcharmap, this only used during construction but both are
1305 * useful during debugging so we store them in the struct when debugging.
1308 const U32 data_slot = add_data( pRExC_state, 2, "tu" );
1309 STRLEN trie_charcount=0;
1311 SV *re_trie_maxbuff;
1312 GET_RE_DEBUG_FLAGS_DECL;
1314 PERL_UNUSED_ARG(depth);
1317 trie = (reg_trie_data *) PerlMemShared_calloc( 1, sizeof(reg_trie_data) );
1319 trie->startstate = 1;
1320 trie->wordcount = word_count;
1321 RExC_rxi->data->data[ data_slot ] = (void*)trie;
1322 trie->charmap = (U16 *) PerlMemShared_calloc( 256, sizeof(U16) );
1323 if (!(UTF && folder))
1324 trie->bitmap = (char *) PerlMemShared_calloc( ANYOF_BITMAP_SIZE, 1 );
1326 trie_words = newAV();
1329 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
1330 if (!SvIOK(re_trie_maxbuff)) {
1331 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
1334 PerlIO_printf( Perl_debug_log,
1335 "%*smake_trie start==%d, first==%d, last==%d, tail==%d depth=%d\n",
1336 (int)depth * 2 + 2, "",
1337 REG_NODE_NUM(startbranch),REG_NODE_NUM(first),
1338 REG_NODE_NUM(last), REG_NODE_NUM(tail),
1342 /* Find the node we are going to overwrite */
1343 if ( first == startbranch && OP( last ) != BRANCH ) {
1344 /* whole branch chain */
1347 /* branch sub-chain */
1348 convert = NEXTOPER( first );
1351 /* -- First loop and Setup --
1353 We first traverse the branches and scan each word to determine if it
1354 contains widechars, and how many unique chars there are, this is
1355 important as we have to build a table with at least as many columns as we
1358 We use an array of integers to represent the character codes 0..255
1359 (trie->charmap) and we use a an HV* to store unicode characters. We use the
1360 native representation of the character value as the key and IV's for the
1363 *TODO* If we keep track of how many times each character is used we can
1364 remap the columns so that the table compression later on is more
1365 efficient in terms of memory by ensuring most common value is in the
1366 middle and the least common are on the outside. IMO this would be better
1367 than a most to least common mapping as theres a decent chance the most
1368 common letter will share a node with the least common, meaning the node
1369 will not be compressable. With a middle is most common approach the worst
1370 case is when we have the least common nodes twice.
1374 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1375 regnode * const noper = NEXTOPER( cur );
1376 const U8 *uc = (U8*)STRING( noper );
1377 const U8 * const e = uc + STR_LEN( noper );
1379 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1380 const U8 *scan = (U8*)NULL;
1381 U32 wordlen = 0; /* required init */
1383 bool set_bit = trie->bitmap ? 1 : 0; /*store the first char in the bitmap?*/
1385 if (OP(noper) == NOTHING) {
1389 if ( set_bit ) /* bitmap only alloced when !(UTF&&Folding) */
1390 TRIE_BITMAP_SET(trie,*uc); /* store the raw first byte
1391 regardless of encoding */
1393 for ( ; uc < e ; uc += len ) {
1394 TRIE_CHARCOUNT(trie)++;
1398 if ( !trie->charmap[ uvc ] ) {
1399 trie->charmap[ uvc ]=( ++trie->uniquecharcount );
1401 trie->charmap[ folder[ uvc ] ] = trie->charmap[ uvc ];
1405 /* store the codepoint in the bitmap, and if its ascii
1406 also store its folded equivelent. */
1407 TRIE_BITMAP_SET(trie,uvc);
1408 if ( folder ) TRIE_BITMAP_SET(trie,folder[ uvc ]);
1409 set_bit = 0; /* We've done our bit :-) */
1414 widecharmap = newHV();
1416 svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 1 );
1419 Perl_croak( aTHX_ "error creating/fetching widecharmap entry for 0x%"UVXf, uvc );
1421 if ( !SvTRUE( *svpp ) ) {
1422 sv_setiv( *svpp, ++trie->uniquecharcount );
1427 if( cur == first ) {
1430 } else if (chars < trie->minlen) {
1432 } else if (chars > trie->maxlen) {
1436 } /* end first pass */
1437 DEBUG_TRIE_COMPILE_r(
1438 PerlIO_printf( Perl_debug_log, "%*sTRIE(%s): W:%d C:%d Uq:%d Min:%d Max:%d\n",
1439 (int)depth * 2 + 2,"",
1440 ( widecharmap ? "UTF8" : "NATIVE" ), (int)word_count,
1441 (int)TRIE_CHARCOUNT(trie), trie->uniquecharcount,
1442 (int)trie->minlen, (int)trie->maxlen )
1444 trie->wordlen = (U32 *) PerlMemShared_calloc( word_count, sizeof(U32) );
1447 We now know what we are dealing with in terms of unique chars and
1448 string sizes so we can calculate how much memory a naive
1449 representation using a flat table will take. If it's over a reasonable
1450 limit (as specified by ${^RE_TRIE_MAXBUF}) we use a more memory
1451 conservative but potentially much slower representation using an array
1454 At the end we convert both representations into the same compressed
1455 form that will be used in regexec.c for matching with. The latter
1456 is a form that cannot be used to construct with but has memory
1457 properties similar to the list form and access properties similar
1458 to the table form making it both suitable for fast searches and
1459 small enough that its feasable to store for the duration of a program.
1461 See the comment in the code where the compressed table is produced
1462 inplace from the flat tabe representation for an explanation of how
1463 the compression works.
1468 if ( (IV)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1) > SvIV(re_trie_maxbuff) ) {
1470 Second Pass -- Array Of Lists Representation
1472 Each state will be represented by a list of charid:state records
1473 (reg_trie_trans_le) the first such element holds the CUR and LEN
1474 points of the allocated array. (See defines above).
1476 We build the initial structure using the lists, and then convert
1477 it into the compressed table form which allows faster lookups
1478 (but cant be modified once converted).
1481 STRLEN transcount = 1;
1483 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1484 "%*sCompiling trie using list compiler\n",
1485 (int)depth * 2 + 2, ""));
1487 trie->states = (reg_trie_state *)
1488 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1489 sizeof(reg_trie_state) );
1493 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1495 regnode * const noper = NEXTOPER( cur );
1496 U8 *uc = (U8*)STRING( noper );
1497 const U8 * const e = uc + STR_LEN( noper );
1498 U32 state = 1; /* required init */
1499 U16 charid = 0; /* sanity init */
1500 U8 *scan = (U8*)NULL; /* sanity init */
1501 STRLEN foldlen = 0; /* required init */
1502 U32 wordlen = 0; /* required init */
1503 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1505 if (OP(noper) != NOTHING) {
1506 for ( ; uc < e ; uc += len ) {
1511 charid = trie->charmap[ uvc ];
1513 SV** const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1517 charid=(U16)SvIV( *svpp );
1520 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1527 if ( !trie->states[ state ].trans.list ) {
1528 TRIE_LIST_NEW( state );
1530 for ( check = 1; check <= TRIE_LIST_USED( state ); check++ ) {
1531 if ( TRIE_LIST_ITEM( state, check ).forid == charid ) {
1532 newstate = TRIE_LIST_ITEM( state, check ).newstate;
1537 newstate = next_alloc++;
1538 TRIE_LIST_PUSH( state, charid, newstate );
1543 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1547 TRIE_HANDLE_WORD(state);
1549 } /* end second pass */
1551 /* next alloc is the NEXT state to be allocated */
1552 trie->statecount = next_alloc;
1553 trie->states = (reg_trie_state *)
1554 PerlMemShared_realloc( trie->states,
1556 * sizeof(reg_trie_state) );
1558 /* and now dump it out before we compress it */
1559 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_list(trie, widecharmap,
1560 revcharmap, next_alloc,
1564 trie->trans = (reg_trie_trans *)
1565 PerlMemShared_calloc( transcount, sizeof(reg_trie_trans) );
1572 for( state=1 ; state < next_alloc ; state ++ ) {
1576 DEBUG_TRIE_COMPILE_MORE_r(
1577 PerlIO_printf( Perl_debug_log, "tp: %d zp: %d ",tp,zp)
1581 if (trie->states[state].trans.list) {
1582 U16 minid=TRIE_LIST_ITEM( state, 1).forid;
1586 for( idx = 2 ; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1587 const U16 forid = TRIE_LIST_ITEM( state, idx).forid;
1588 if ( forid < minid ) {
1590 } else if ( forid > maxid ) {
1594 if ( transcount < tp + maxid - minid + 1) {
1596 trie->trans = (reg_trie_trans *)
1597 PerlMemShared_realloc( trie->trans,
1599 * sizeof(reg_trie_trans) );
1600 Zero( trie->trans + (transcount / 2), transcount / 2 , reg_trie_trans );
1602 base = trie->uniquecharcount + tp - minid;
1603 if ( maxid == minid ) {
1605 for ( ; zp < tp ; zp++ ) {
1606 if ( ! trie->trans[ zp ].next ) {
1607 base = trie->uniquecharcount + zp - minid;
1608 trie->trans[ zp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1609 trie->trans[ zp ].check = state;
1615 trie->trans[ tp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1616 trie->trans[ tp ].check = state;
1621 for ( idx=1; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1622 const U32 tid = base - trie->uniquecharcount + TRIE_LIST_ITEM( state, idx ).forid;
1623 trie->trans[ tid ].next = TRIE_LIST_ITEM( state, idx ).newstate;
1624 trie->trans[ tid ].check = state;
1626 tp += ( maxid - minid + 1 );
1628 Safefree(trie->states[ state ].trans.list);
1631 DEBUG_TRIE_COMPILE_MORE_r(
1632 PerlIO_printf( Perl_debug_log, " base: %d\n",base);
1635 trie->states[ state ].trans.base=base;
1637 trie->lasttrans = tp + 1;
1641 Second Pass -- Flat Table Representation.
1643 we dont use the 0 slot of either trans[] or states[] so we add 1 to each.
1644 We know that we will need Charcount+1 trans at most to store the data
1645 (one row per char at worst case) So we preallocate both structures
1646 assuming worst case.
1648 We then construct the trie using only the .next slots of the entry
1651 We use the .check field of the first entry of the node temporarily to
1652 make compression both faster and easier by keeping track of how many non
1653 zero fields are in the node.
1655 Since trans are numbered from 1 any 0 pointer in the table is a FAIL
1658 There are two terms at use here: state as a TRIE_NODEIDX() which is a
1659 number representing the first entry of the node, and state as a
1660 TRIE_NODENUM() which is the trans number. state 1 is TRIE_NODEIDX(1) and
1661 TRIE_NODENUM(1), state 2 is TRIE_NODEIDX(2) and TRIE_NODENUM(3) if there
1662 are 2 entrys per node. eg:
1670 The table is internally in the right hand, idx form. However as we also
1671 have to deal with the states array which is indexed by nodenum we have to
1672 use TRIE_NODENUM() to convert.
1675 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1676 "%*sCompiling trie using table compiler\n",
1677 (int)depth * 2 + 2, ""));
1679 trie->trans = (reg_trie_trans *)
1680 PerlMemShared_calloc( ( TRIE_CHARCOUNT(trie) + 1 )
1681 * trie->uniquecharcount + 1,
1682 sizeof(reg_trie_trans) );
1683 trie->states = (reg_trie_state *)
1684 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1685 sizeof(reg_trie_state) );
1686 next_alloc = trie->uniquecharcount + 1;
1689 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1691 regnode * const noper = NEXTOPER( cur );
1692 const U8 *uc = (U8*)STRING( noper );
1693 const U8 * const e = uc + STR_LEN( noper );
1695 U32 state = 1; /* required init */
1697 U16 charid = 0; /* sanity init */
1698 U32 accept_state = 0; /* sanity init */
1699 U8 *scan = (U8*)NULL; /* sanity init */
1701 STRLEN foldlen = 0; /* required init */
1702 U32 wordlen = 0; /* required init */
1703 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1705 if ( OP(noper) != NOTHING ) {
1706 for ( ; uc < e ; uc += len ) {
1711 charid = trie->charmap[ uvc ];
1713 SV* const * const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1714 charid = svpp ? (U16)SvIV(*svpp) : 0;
1718 if ( !trie->trans[ state + charid ].next ) {
1719 trie->trans[ state + charid ].next = next_alloc;
1720 trie->trans[ state ].check++;
1721 next_alloc += trie->uniquecharcount;
1723 state = trie->trans[ state + charid ].next;
1725 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1727 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1730 accept_state = TRIE_NODENUM( state );
1731 TRIE_HANDLE_WORD(accept_state);
1733 } /* end second pass */
1735 /* and now dump it out before we compress it */
1736 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_table(trie, widecharmap,
1738 next_alloc, depth+1));
1742 * Inplace compress the table.*
1744 For sparse data sets the table constructed by the trie algorithm will
1745 be mostly 0/FAIL transitions or to put it another way mostly empty.
1746 (Note that leaf nodes will not contain any transitions.)
1748 This algorithm compresses the tables by eliminating most such
1749 transitions, at the cost of a modest bit of extra work during lookup:
1751 - Each states[] entry contains a .base field which indicates the
1752 index in the state[] array wheres its transition data is stored.
1754 - If .base is 0 there are no valid transitions from that node.
1756 - If .base is nonzero then charid is added to it to find an entry in
1759 -If trans[states[state].base+charid].check!=state then the
1760 transition is taken to be a 0/Fail transition. Thus if there are fail
1761 transitions at the front of the node then the .base offset will point
1762 somewhere inside the previous nodes data (or maybe even into a node
1763 even earlier), but the .check field determines if the transition is
1767 The following process inplace converts the table to the compressed
1768 table: We first do not compress the root node 1,and mark its all its
1769 .check pointers as 1 and set its .base pointer as 1 as well. This
1770 allows to do a DFA construction from the compressed table later, and
1771 ensures that any .base pointers we calculate later are greater than
1774 - We set 'pos' to indicate the first entry of the second node.
1776 - We then iterate over the columns of the node, finding the first and
1777 last used entry at l and m. We then copy l..m into pos..(pos+m-l),
1778 and set the .check pointers accordingly, and advance pos
1779 appropriately and repreat for the next node. Note that when we copy
1780 the next pointers we have to convert them from the original
1781 NODEIDX form to NODENUM form as the former is not valid post
1784 - If a node has no transitions used we mark its base as 0 and do not
1785 advance the pos pointer.
1787 - If a node only has one transition we use a second pointer into the
1788 structure to fill in allocated fail transitions from other states.
1789 This pointer is independent of the main pointer and scans forward
1790 looking for null transitions that are allocated to a state. When it
1791 finds one it writes the single transition into the "hole". If the
1792 pointer doesnt find one the single transition is appended as normal.
1794 - Once compressed we can Renew/realloc the structures to release the
1797 See "Table-Compression Methods" in sec 3.9 of the Red Dragon,
1798 specifically Fig 3.47 and the associated pseudocode.
1802 const U32 laststate = TRIE_NODENUM( next_alloc );
1805 trie->statecount = laststate;
1807 for ( state = 1 ; state < laststate ; state++ ) {
1809 const U32 stateidx = TRIE_NODEIDX( state );
1810 const U32 o_used = trie->trans[ stateidx ].check;
1811 U32 used = trie->trans[ stateidx ].check;
1812 trie->trans[ stateidx ].check = 0;
1814 for ( charid = 0 ; used && charid < trie->uniquecharcount ; charid++ ) {
1815 if ( flag || trie->trans[ stateidx + charid ].next ) {
1816 if ( trie->trans[ stateidx + charid ].next ) {
1818 for ( ; zp < pos ; zp++ ) {
1819 if ( ! trie->trans[ zp ].next ) {
1823 trie->states[ state ].trans.base = zp + trie->uniquecharcount - charid ;
1824 trie->trans[ zp ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1825 trie->trans[ zp ].check = state;
1826 if ( ++zp > pos ) pos = zp;
1833 trie->states[ state ].trans.base = pos + trie->uniquecharcount - charid ;
1835 trie->trans[ pos ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1836 trie->trans[ pos ].check = state;
1841 trie->lasttrans = pos + 1;
1842 trie->states = (reg_trie_state *)
1843 PerlMemShared_realloc( trie->states, laststate
1844 * sizeof(reg_trie_state) );
1845 DEBUG_TRIE_COMPILE_MORE_r(
1846 PerlIO_printf( Perl_debug_log,
1847 "%*sAlloc: %d Orig: %"IVdf" elements, Final:%"IVdf". Savings of %%%5.2f\n",
1848 (int)depth * 2 + 2,"",
1849 (int)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1 ),
1852 ( ( next_alloc - pos ) * 100 ) / (double)next_alloc );
1855 } /* end table compress */
1857 DEBUG_TRIE_COMPILE_MORE_r(
1858 PerlIO_printf(Perl_debug_log, "%*sStatecount:%"UVxf" Lasttrans:%"UVxf"\n",
1859 (int)depth * 2 + 2, "",
1860 (UV)trie->statecount,
1861 (UV)trie->lasttrans)
1863 /* resize the trans array to remove unused space */
1864 trie->trans = (reg_trie_trans *)
1865 PerlMemShared_realloc( trie->trans, trie->lasttrans
1866 * sizeof(reg_trie_trans) );
1868 /* and now dump out the compressed format */
1869 DEBUG_TRIE_COMPILE_r(dump_trie(trie, widecharmap, revcharmap, depth+1));
1871 { /* Modify the program and insert the new TRIE node*/
1872 U8 nodetype =(U8)(flags & 0xFF);
1876 regnode *optimize = NULL;
1877 #ifdef RE_TRACK_PATTERN_OFFSETS
1880 U32 mjd_nodelen = 0;
1881 #endif /* RE_TRACK_PATTERN_OFFSETS */
1882 #endif /* DEBUGGING */
1884 This means we convert either the first branch or the first Exact,
1885 depending on whether the thing following (in 'last') is a branch
1886 or not and whther first is the startbranch (ie is it a sub part of
1887 the alternation or is it the whole thing.)
1888 Assuming its a sub part we conver the EXACT otherwise we convert
1889 the whole branch sequence, including the first.
1891 /* Find the node we are going to overwrite */
1892 if ( first != startbranch || OP( last ) == BRANCH ) {
1893 /* branch sub-chain */
1894 NEXT_OFF( first ) = (U16)(last - first);
1895 #ifdef RE_TRACK_PATTERN_OFFSETS
1897 mjd_offset= Node_Offset((convert));
1898 mjd_nodelen= Node_Length((convert));
1901 /* whole branch chain */
1903 #ifdef RE_TRACK_PATTERN_OFFSETS
1906 const regnode *nop = NEXTOPER( convert );
1907 mjd_offset= Node_Offset((nop));
1908 mjd_nodelen= Node_Length((nop));
1912 PerlIO_printf(Perl_debug_log, "%*sMJD offset:%"UVuf" MJD length:%"UVuf"\n",
1913 (int)depth * 2 + 2, "",
1914 (UV)mjd_offset, (UV)mjd_nodelen)
1917 /* But first we check to see if there is a common prefix we can
1918 split out as an EXACT and put in front of the TRIE node. */
1919 trie->startstate= 1;
1920 if ( trie->bitmap && !widecharmap && !trie->jump ) {
1922 for ( state = 1 ; state < trie->statecount-1 ; state++ ) {
1926 const U32 base = trie->states[ state ].trans.base;
1928 if ( trie->states[state].wordnum )
1931 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
1932 if ( ( base + ofs >= trie->uniquecharcount ) &&
1933 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
1934 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
1936 if ( ++count > 1 ) {
1937 SV **tmp = av_fetch( revcharmap, ofs, 0);
1938 const U8 *ch = (U8*)SvPV_nolen_const( *tmp );
1939 if ( state == 1 ) break;
1941 Zero(trie->bitmap, ANYOF_BITMAP_SIZE, char);
1943 PerlIO_printf(Perl_debug_log,
1944 "%*sNew Start State=%"UVuf" Class: [",
1945 (int)depth * 2 + 2, "",
1948 SV ** const tmp = av_fetch( revcharmap, idx, 0);
1949 const U8 * const ch = (U8*)SvPV_nolen_const( *tmp );
1951 TRIE_BITMAP_SET(trie,*ch);
1953 TRIE_BITMAP_SET(trie, folder[ *ch ]);
1955 PerlIO_printf(Perl_debug_log, (char*)ch)
1959 TRIE_BITMAP_SET(trie,*ch);
1961 TRIE_BITMAP_SET(trie,folder[ *ch ]);
1962 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"%s", ch));
1968 SV **tmp = av_fetch( revcharmap, idx, 0);
1969 char *ch = SvPV_nolen( *tmp );
1971 SV *sv=sv_newmortal();
1972 PerlIO_printf( Perl_debug_log,
1973 "%*sPrefix State: %"UVuf" Idx:%"UVuf" Char='%s'\n",
1974 (int)depth * 2 + 2, "",
1976 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 6,
1977 PL_colors[0], PL_colors[1],
1978 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1979 PERL_PV_ESCAPE_FIRSTCHAR
1984 OP( convert ) = nodetype;
1985 str=STRING(convert);
1996 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"]\n"));
2002 regnode *n = convert+NODE_SZ_STR(convert);
2003 NEXT_OFF(convert) = NODE_SZ_STR(convert);
2004 trie->startstate = state;
2005 trie->minlen -= (state - 1);
2006 trie->maxlen -= (state - 1);
2008 regnode *fix = convert;
2009 U32 word = trie->wordcount;
2011 Set_Node_Offset_Length(convert, mjd_offset, state - 1);
2012 while( ++fix < n ) {
2013 Set_Node_Offset_Length(fix, 0, 0);
2016 SV ** const tmp = av_fetch( trie_words, word, 0 );
2018 if ( STR_LEN(convert) <= SvCUR(*tmp) )
2019 sv_chop(*tmp, SvPV_nolen(*tmp) + STR_LEN(convert));
2021 sv_chop(*tmp, SvPV_nolen(*tmp) + SvCUR(*tmp));
2028 NEXT_OFF(convert) = (U16)(tail - convert);
2029 DEBUG_r(optimize= n);
2035 if ( trie->maxlen ) {
2036 NEXT_OFF( convert ) = (U16)(tail - convert);
2037 ARG_SET( convert, data_slot );
2038 /* Store the offset to the first unabsorbed branch in
2039 jump[0], which is otherwise unused by the jump logic.
2040 We use this when dumping a trie and during optimisation. */
2042 trie->jump[0] = (U16)(nextbranch - convert);
2045 if ( !trie->states[trie->startstate].wordnum && trie->bitmap &&
2046 ( (char *)jumper - (char *)convert) >= (int)sizeof(struct regnode_charclass) )
2048 OP( convert ) = TRIEC;
2049 Copy(trie->bitmap, ((struct regnode_charclass *)convert)->bitmap, ANYOF_BITMAP_SIZE, char);
2050 PerlMemShared_free(trie->bitmap);
2053 OP( convert ) = TRIE;
2055 /* store the type in the flags */
2056 convert->flags = nodetype;
2060 + regarglen[ OP( convert ) ];
2062 /* XXX We really should free up the resource in trie now,
2063 as we won't use them - (which resources?) dmq */
2065 /* needed for dumping*/
2066 DEBUG_r(if (optimize) {
2067 regnode *opt = convert;
2069 while ( ++opt < optimize) {
2070 Set_Node_Offset_Length(opt,0,0);
2073 Try to clean up some of the debris left after the
2076 while( optimize < jumper ) {
2077 mjd_nodelen += Node_Length((optimize));
2078 OP( optimize ) = OPTIMIZED;
2079 Set_Node_Offset_Length(optimize,0,0);
2082 Set_Node_Offset_Length(convert,mjd_offset,mjd_nodelen);
2084 } /* end node insert */
2085 RExC_rxi->data->data[ data_slot + 1 ] = (void*)widecharmap;
2087 RExC_rxi->data->data[ data_slot + TRIE_WORDS_OFFSET ] = (void*)trie_words;
2088 RExC_rxi->data->data[ data_slot + 3 ] = (void*)revcharmap;
2090 SvREFCNT_dec(revcharmap);
2094 : trie->startstate>1
2100 S_make_trie_failtable(pTHX_ RExC_state_t *pRExC_state, regnode *source, regnode *stclass, U32 depth)
2102 /* The Trie is constructed and compressed now so we can build a fail array now if its needed
2104 This is basically the Aho-Corasick algorithm. Its from exercise 3.31 and 3.32 in the
2105 "Red Dragon" -- Compilers, principles, techniques, and tools. Aho, Sethi, Ullman 1985/88
2108 We find the fail state for each state in the trie, this state is the longest proper
2109 suffix of the current states 'word' that is also a proper prefix of another word in our
2110 trie. State 1 represents the word '' and is the thus the default fail state. This allows
2111 the DFA not to have to restart after its tried and failed a word at a given point, it
2112 simply continues as though it had been matching the other word in the first place.
2114 'abcdgu'=~/abcdefg|cdgu/
2115 When we get to 'd' we are still matching the first word, we would encounter 'g' which would
2116 fail, which would bring use to the state representing 'd' in the second word where we would
2117 try 'g' and succeed, prodceding to match 'cdgu'.
2119 /* add a fail transition */
2120 const U32 trie_offset = ARG(source);
2121 reg_trie_data *trie=(reg_trie_data *)RExC_rxi->data->data[trie_offset];
2123 const U32 ucharcount = trie->uniquecharcount;
2124 const U32 numstates = trie->statecount;
2125 const U32 ubound = trie->lasttrans + ucharcount;
2129 U32 base = trie->states[ 1 ].trans.base;
2132 const U32 data_slot = add_data( pRExC_state, 1, "T" );
2133 GET_RE_DEBUG_FLAGS_DECL;
2135 PERL_UNUSED_ARG(depth);
2139 ARG_SET( stclass, data_slot );
2140 aho = (reg_ac_data *) PerlMemShared_calloc( 1, sizeof(reg_ac_data) );
2141 RExC_rxi->data->data[ data_slot ] = (void*)aho;
2142 aho->trie=trie_offset;
2143 aho->states=(reg_trie_state *)PerlMemShared_malloc( numstates * sizeof(reg_trie_state) );
2144 Copy( trie->states, aho->states, numstates, reg_trie_state );
2145 Newxz( q, numstates, U32);
2146 aho->fail = (U32 *) PerlMemShared_calloc( numstates, sizeof(U32) );
2149 /* initialize fail[0..1] to be 1 so that we always have
2150 a valid final fail state */
2151 fail[ 0 ] = fail[ 1 ] = 1;
2153 for ( charid = 0; charid < ucharcount ; charid++ ) {
2154 const U32 newstate = TRIE_TRANS_STATE( 1, base, ucharcount, charid, 0 );
2156 q[ q_write ] = newstate;
2157 /* set to point at the root */
2158 fail[ q[ q_write++ ] ]=1;
2161 while ( q_read < q_write) {
2162 const U32 cur = q[ q_read++ % numstates ];
2163 base = trie->states[ cur ].trans.base;
2165 for ( charid = 0 ; charid < ucharcount ; charid++ ) {
2166 const U32 ch_state = TRIE_TRANS_STATE( cur, base, ucharcount, charid, 1 );
2168 U32 fail_state = cur;
2171 fail_state = fail[ fail_state ];
2172 fail_base = aho->states[ fail_state ].trans.base;
2173 } while ( !TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 ) );
2175 fail_state = TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 );
2176 fail[ ch_state ] = fail_state;
2177 if ( !aho->states[ ch_state ].wordnum && aho->states[ fail_state ].wordnum )
2179 aho->states[ ch_state ].wordnum = aho->states[ fail_state ].wordnum;
2181 q[ q_write++ % numstates] = ch_state;
2185 /* restore fail[0..1] to 0 so that we "fall out" of the AC loop
2186 when we fail in state 1, this allows us to use the
2187 charclass scan to find a valid start char. This is based on the principle
2188 that theres a good chance the string being searched contains lots of stuff
2189 that cant be a start char.
2191 fail[ 0 ] = fail[ 1 ] = 0;
2192 DEBUG_TRIE_COMPILE_r({
2193 PerlIO_printf(Perl_debug_log,
2194 "%*sStclass Failtable (%"UVuf" states): 0",
2195 (int)(depth * 2), "", (UV)numstates
2197 for( q_read=1; q_read<numstates; q_read++ ) {
2198 PerlIO_printf(Perl_debug_log, ", %"UVuf, (UV)fail[q_read]);
2200 PerlIO_printf(Perl_debug_log, "\n");
2203 /*RExC_seen |= REG_SEEN_TRIEDFA;*/
2208 * There are strange code-generation bugs caused on sparc64 by gcc-2.95.2.
2209 * These need to be revisited when a newer toolchain becomes available.
2211 #if defined(__sparc64__) && defined(__GNUC__)
2212 # if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
2213 # undef SPARC64_GCC_WORKAROUND
2214 # define SPARC64_GCC_WORKAROUND 1
2218 #define DEBUG_PEEP(str,scan,depth) \
2219 DEBUG_OPTIMISE_r({if (scan){ \
2220 SV * const mysv=sv_newmortal(); \
2221 regnode *Next = regnext(scan); \
2222 regprop(RExC_rx, mysv, scan); \
2223 PerlIO_printf(Perl_debug_log, "%*s" str ">%3d: %s (%d)\n", \
2224 (int)depth*2, "", REG_NODE_NUM(scan), SvPV_nolen_const(mysv),\
2225 Next ? (REG_NODE_NUM(Next)) : 0 ); \
2232 #define JOIN_EXACT(scan,min,flags) \
2233 if (PL_regkind[OP(scan)] == EXACT) \
2234 join_exact(pRExC_state,(scan),(min),(flags),NULL,depth+1)
2237 S_join_exact(pTHX_ RExC_state_t *pRExC_state, regnode *scan, I32 *min, U32 flags,regnode *val, U32 depth) {
2238 /* Merge several consecutive EXACTish nodes into one. */
2239 regnode *n = regnext(scan);
2241 regnode *next = scan + NODE_SZ_STR(scan);
2245 regnode *stop = scan;
2246 GET_RE_DEBUG_FLAGS_DECL;
2248 PERL_UNUSED_ARG(depth);
2250 #ifndef EXPERIMENTAL_INPLACESCAN
2251 PERL_UNUSED_ARG(flags);
2252 PERL_UNUSED_ARG(val);
2254 DEBUG_PEEP("join",scan,depth);
2256 /* Skip NOTHING, merge EXACT*. */
2258 ( PL_regkind[OP(n)] == NOTHING ||
2259 (stringok && (OP(n) == OP(scan))))
2261 && NEXT_OFF(scan) + NEXT_OFF(n) < I16_MAX) {
2263 if (OP(n) == TAIL || n > next)
2265 if (PL_regkind[OP(n)] == NOTHING) {
2266 DEBUG_PEEP("skip:",n,depth);
2267 NEXT_OFF(scan) += NEXT_OFF(n);
2268 next = n + NODE_STEP_REGNODE;
2275 else if (stringok) {
2276 const unsigned int oldl = STR_LEN(scan);
2277 regnode * const nnext = regnext(n);
2279 DEBUG_PEEP("merg",n,depth);
2282 if (oldl + STR_LEN(n) > U8_MAX)
2284 NEXT_OFF(scan) += NEXT_OFF(n);
2285 STR_LEN(scan) += STR_LEN(n);
2286 next = n + NODE_SZ_STR(n);
2287 /* Now we can overwrite *n : */
2288 Move(STRING(n), STRING(scan) + oldl, STR_LEN(n), char);
2296 #ifdef EXPERIMENTAL_INPLACESCAN
2297 if (flags && !NEXT_OFF(n)) {
2298 DEBUG_PEEP("atch", val, depth);
2299 if (reg_off_by_arg[OP(n)]) {
2300 ARG_SET(n, val - n);
2303 NEXT_OFF(n) = val - n;
2310 if (UTF && ( OP(scan) == EXACTF ) && ( STR_LEN(scan) >= 6 ) ) {
2312 Two problematic code points in Unicode casefolding of EXACT nodes:
2314 U+0390 - GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
2315 U+03B0 - GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
2321 U+03B9 U+0308 U+0301 0xCE 0xB9 0xCC 0x88 0xCC 0x81
2322 U+03C5 U+0308 U+0301 0xCF 0x85 0xCC 0x88 0xCC 0x81
2324 This means that in case-insensitive matching (or "loose matching",
2325 as Unicode calls it), an EXACTF of length six (the UTF-8 encoded byte
2326 length of the above casefolded versions) can match a target string
2327 of length two (the byte length of UTF-8 encoded U+0390 or U+03B0).
2328 This would rather mess up the minimum length computation.
2330 What we'll do is to look for the tail four bytes, and then peek
2331 at the preceding two bytes to see whether we need to decrease
2332 the minimum length by four (six minus two).
2334 Thanks to the design of UTF-8, there cannot be false matches:
2335 A sequence of valid UTF-8 bytes cannot be a subsequence of
2336 another valid sequence of UTF-8 bytes.
2339 char * const s0 = STRING(scan), *s, *t;
2340 char * const s1 = s0 + STR_LEN(scan) - 1;
2341 char * const s2 = s1 - 4;
2342 #ifdef EBCDIC /* RD tunifold greek 0390 and 03B0 */
2343 const char t0[] = "\xaf\x49\xaf\x42";
2345 const char t0[] = "\xcc\x88\xcc\x81";
2347 const char * const t1 = t0 + 3;
2350 s < s2 && (t = ninstr(s, s1, t0, t1));
2353 if (((U8)t[-1] == 0x68 && (U8)t[-2] == 0xB4) ||
2354 ((U8)t[-1] == 0x46 && (U8)t[-2] == 0xB5))
2356 if (((U8)t[-1] == 0xB9 && (U8)t[-2] == 0xCE) ||
2357 ((U8)t[-1] == 0x85 && (U8)t[-2] == 0xCF))
2365 n = scan + NODE_SZ_STR(scan);
2367 if (PL_regkind[OP(n)] != NOTHING || OP(n) == NOTHING) {
2374 DEBUG_OPTIMISE_r(if (merged){DEBUG_PEEP("finl",scan,depth)});
2378 /* REx optimizer. Converts nodes into quickier variants "in place".
2379 Finds fixed substrings. */
2381 /* Stops at toplevel WHILEM as well as at "last". At end *scanp is set
2382 to the position after last scanned or to NULL. */
2384 #define INIT_AND_WITHP \
2385 assert(!and_withp); \
2386 Newx(and_withp,1,struct regnode_charclass_class); \
2387 SAVEFREEPV(and_withp)
2389 /* this is a chain of data about sub patterns we are processing that
2390 need to be handled seperately/specially in study_chunk. Its so
2391 we can simulate recursion without losing state. */
2393 typedef struct scan_frame {
2394 regnode *last; /* last node to process in this frame */
2395 regnode *next; /* next node to process when last is reached */
2396 struct scan_frame *prev; /*previous frame*/
2397 I32 stop; /* what stopparen do we use */
2401 #define SCAN_COMMIT(s, data, m) scan_commit(s, data, m, is_inf)
2403 #define CASE_SYNST_FNC(nAmE) \
2405 if (flags & SCF_DO_STCLASS_AND) { \
2406 for (value = 0; value < 256; value++) \
2407 if (!is_ ## nAmE ## _cp(value)) \
2408 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2411 for (value = 0; value < 256; value++) \
2412 if (is_ ## nAmE ## _cp(value)) \
2413 ANYOF_BITMAP_SET(data->start_class, value); \
2417 if (flags & SCF_DO_STCLASS_AND) { \
2418 for (value = 0; value < 256; value++) \
2419 if (is_ ## nAmE ## _cp(value)) \
2420 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2423 for (value = 0; value < 256; value++) \
2424 if (!is_ ## nAmE ## _cp(value)) \
2425 ANYOF_BITMAP_SET(data->start_class, value); \
2432 S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
2433 I32 *minlenp, I32 *deltap,
2438 struct regnode_charclass_class *and_withp,
2439 U32 flags, U32 depth)
2440 /* scanp: Start here (read-write). */
2441 /* deltap: Write maxlen-minlen here. */
2442 /* last: Stop before this one. */
2443 /* data: string data about the pattern */
2444 /* stopparen: treat close N as END */
2445 /* recursed: which subroutines have we recursed into */
2446 /* and_withp: Valid if flags & SCF_DO_STCLASS_OR */
2449 I32 min = 0, pars = 0, code;
2450 regnode *scan = *scanp, *next;
2452 int is_inf = (flags & SCF_DO_SUBSTR) && (data->flags & SF_IS_INF);
2453 int is_inf_internal = 0; /* The studied chunk is infinite */
2454 I32 is_par = OP(scan) == OPEN ? ARG(scan) : 0;
2455 scan_data_t data_fake;
2456 SV *re_trie_maxbuff = NULL;
2457 regnode *first_non_open = scan;
2458 I32 stopmin = I32_MAX;
2459 scan_frame *frame = NULL;
2461 GET_RE_DEBUG_FLAGS_DECL;
2464 StructCopy(&zero_scan_data, &data_fake, scan_data_t);
2468 while (first_non_open && OP(first_non_open) == OPEN)
2469 first_non_open=regnext(first_non_open);
2474 while ( scan && OP(scan) != END && scan < last ){
2475 /* Peephole optimizer: */
2476 DEBUG_STUDYDATA("Peep:", data,depth);
2477 DEBUG_PEEP("Peep",scan,depth);
2478 JOIN_EXACT(scan,&min,0);
2480 /* Follow the next-chain of the current node and optimize
2481 away all the NOTHINGs from it. */
2482 if (OP(scan) != CURLYX) {
2483 const int max = (reg_off_by_arg[OP(scan)]
2485 /* I32 may be smaller than U16 on CRAYs! */
2486 : (I32_MAX < U16_MAX ? I32_MAX : U16_MAX));
2487 int off = (reg_off_by_arg[OP(scan)] ? ARG(scan) : NEXT_OFF(scan));
2491 /* Skip NOTHING and LONGJMP. */
2492 while ((n = regnext(n))
2493 && ((PL_regkind[OP(n)] == NOTHING && (noff = NEXT_OFF(n)))
2494 || ((OP(n) == LONGJMP) && (noff = ARG(n))))
2495 && off + noff < max)
2497 if (reg_off_by_arg[OP(scan)])
2500 NEXT_OFF(scan) = off;
2505 /* The principal pseudo-switch. Cannot be a switch, since we
2506 look into several different things. */
2507 if (OP(scan) == BRANCH || OP(scan) == BRANCHJ
2508 || OP(scan) == IFTHEN) {
2509 next = regnext(scan);
2511 /* demq: the op(next)==code check is to see if we have "branch-branch" AFAICT */
2513 if (OP(next) == code || code == IFTHEN) {
2514 /* NOTE - There is similar code to this block below for handling
2515 TRIE nodes on a re-study. If you change stuff here check there
2517 I32 max1 = 0, min1 = I32_MAX, num = 0;
2518 struct regnode_charclass_class accum;
2519 regnode * const startbranch=scan;
2521 if (flags & SCF_DO_SUBSTR)
2522 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot merge strings after this. */
2523 if (flags & SCF_DO_STCLASS)
2524 cl_init_zero(pRExC_state, &accum);
2526 while (OP(scan) == code) {
2527 I32 deltanext, minnext, f = 0, fake;
2528 struct regnode_charclass_class this_class;
2531 data_fake.flags = 0;
2533 data_fake.whilem_c = data->whilem_c;
2534 data_fake.last_closep = data->last_closep;
2537 data_fake.last_closep = &fake;
2539 data_fake.pos_delta = delta;
2540 next = regnext(scan);
2541 scan = NEXTOPER(scan);
2543 scan = NEXTOPER(scan);
2544 if (flags & SCF_DO_STCLASS) {
2545 cl_init(pRExC_state, &this_class);
2546 data_fake.start_class = &this_class;
2547 f = SCF_DO_STCLASS_AND;
2549 if (flags & SCF_WHILEM_VISITED_POS)
2550 f |= SCF_WHILEM_VISITED_POS;
2552 /* we suppose the run is continuous, last=next...*/
2553 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
2555 stopparen, recursed, NULL, f,depth+1);
2558 if (max1 < minnext + deltanext)
2559 max1 = minnext + deltanext;
2560 if (deltanext == I32_MAX)
2561 is_inf = is_inf_internal = 1;
2563 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
2565 if (data_fake.flags & SCF_SEEN_ACCEPT) {
2566 if ( stopmin > minnext)
2567 stopmin = min + min1;
2568 flags &= ~SCF_DO_SUBSTR;
2570 data->flags |= SCF_SEEN_ACCEPT;
2573 if (data_fake.flags & SF_HAS_EVAL)
2574 data->flags |= SF_HAS_EVAL;
2575 data->whilem_c = data_fake.whilem_c;
2577 if (flags & SCF_DO_STCLASS)
2578 cl_or(pRExC_state, &accum, &this_class);
2580 if (code == IFTHEN && num < 2) /* Empty ELSE branch */
2582 if (flags & SCF_DO_SUBSTR) {
2583 data->pos_min += min1;
2584 data->pos_delta += max1 - min1;
2585 if (max1 != min1 || is_inf)
2586 data->longest = &(data->longest_float);
2589 delta += max1 - min1;
2590 if (flags & SCF_DO_STCLASS_OR) {
2591 cl_or(pRExC_state, data->start_class, &accum);
2593 cl_and(data->start_class, and_withp);
2594 flags &= ~SCF_DO_STCLASS;
2597 else if (flags & SCF_DO_STCLASS_AND) {
2599 cl_and(data->start_class, &accum);
2600 flags &= ~SCF_DO_STCLASS;
2603 /* Switch to OR mode: cache the old value of
2604 * data->start_class */
2606 StructCopy(data->start_class, and_withp,
2607 struct regnode_charclass_class);
2608 flags &= ~SCF_DO_STCLASS_AND;
2609 StructCopy(&accum, data->start_class,
2610 struct regnode_charclass_class);
2611 flags |= SCF_DO_STCLASS_OR;
2612 data->start_class->flags |= ANYOF_EOS;
2616 if (PERL_ENABLE_TRIE_OPTIMISATION && OP( startbranch ) == BRANCH ) {
2619 Assuming this was/is a branch we are dealing with: 'scan' now
2620 points at the item that follows the branch sequence, whatever
2621 it is. We now start at the beginning of the sequence and look
2628 which would be constructed from a pattern like /A|LIST|OF|WORDS/
2630 If we can find such a subseqence we need to turn the first
2631 element into a trie and then add the subsequent branch exact
2632 strings to the trie.
2636 1. patterns where the whole set of branch can be converted.
2638 2. patterns where only a subset can be converted.
2640 In case 1 we can replace the whole set with a single regop
2641 for the trie. In case 2 we need to keep the start and end
2644 'BRANCH EXACT; BRANCH EXACT; BRANCH X'
2645 becomes BRANCH TRIE; BRANCH X;
2647 There is an additional case, that being where there is a
2648 common prefix, which gets split out into an EXACT like node
2649 preceding the TRIE node.
2651 If x(1..n)==tail then we can do a simple trie, if not we make
2652 a "jump" trie, such that when we match the appropriate word
2653 we "jump" to the appopriate tail node. Essentailly we turn
2654 a nested if into a case structure of sorts.
2659 if (!re_trie_maxbuff) {
2660 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
2661 if (!SvIOK(re_trie_maxbuff))
2662 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
2664 if ( SvIV(re_trie_maxbuff)>=0 ) {
2666 regnode *first = (regnode *)NULL;
2667 regnode *last = (regnode *)NULL;
2668 regnode *tail = scan;
2673 SV * const mysv = sv_newmortal(); /* for dumping */
2675 /* var tail is used because there may be a TAIL
2676 regop in the way. Ie, the exacts will point to the
2677 thing following the TAIL, but the last branch will
2678 point at the TAIL. So we advance tail. If we
2679 have nested (?:) we may have to move through several
2683 while ( OP( tail ) == TAIL ) {
2684 /* this is the TAIL generated by (?:) */
2685 tail = regnext( tail );
2690 regprop(RExC_rx, mysv, tail );
2691 PerlIO_printf( Perl_debug_log, "%*s%s%s\n",
2692 (int)depth * 2 + 2, "",
2693 "Looking for TRIE'able sequences. Tail node is: ",
2694 SvPV_nolen_const( mysv )
2700 step through the branches, cur represents each
2701 branch, noper is the first thing to be matched
2702 as part of that branch and noper_next is the
2703 regnext() of that node. if noper is an EXACT
2704 and noper_next is the same as scan (our current
2705 position in the regex) then the EXACT branch is
2706 a possible optimization target. Once we have
2707 two or more consequetive such branches we can
2708 create a trie of the EXACT's contents and stich
2709 it in place. If the sequence represents all of
2710 the branches we eliminate the whole thing and
2711 replace it with a single TRIE. If it is a
2712 subsequence then we need to stitch it in. This
2713 means the first branch has to remain, and needs
2714 to be repointed at the item on the branch chain
2715 following the last branch optimized. This could
2716 be either a BRANCH, in which case the
2717 subsequence is internal, or it could be the
2718 item following the branch sequence in which
2719 case the subsequence is at the end.
2723 /* dont use tail as the end marker for this traverse */
2724 for ( cur = startbranch ; cur != scan ; cur = regnext( cur ) ) {
2725 regnode * const noper = NEXTOPER( cur );
2726 #if defined(DEBUGGING) || defined(NOJUMPTRIE)
2727 regnode * const noper_next = regnext( noper );
2731 regprop(RExC_rx, mysv, cur);
2732 PerlIO_printf( Perl_debug_log, "%*s- %s (%d)",
2733 (int)depth * 2 + 2,"", SvPV_nolen_const( mysv ), REG_NODE_NUM(cur) );
2735 regprop(RExC_rx, mysv, noper);
2736 PerlIO_printf( Perl_debug_log, " -> %s",
2737 SvPV_nolen_const(mysv));
2740 regprop(RExC_rx, mysv, noper_next );
2741 PerlIO_printf( Perl_debug_log,"\t=> %s\t",
2742 SvPV_nolen_const(mysv));
2744 PerlIO_printf( Perl_debug_log, "(First==%d,Last==%d,Cur==%d)\n",
2745 REG_NODE_NUM(first), REG_NODE_NUM(last), REG_NODE_NUM(cur) );
2747 if ( (((first && optype!=NOTHING) ? OP( noper ) == optype
2748 : PL_regkind[ OP( noper ) ] == EXACT )
2749 || OP(noper) == NOTHING )
2751 && noper_next == tail
2756 if ( !first || optype == NOTHING ) {
2757 if (!first) first = cur;
2758 optype = OP( noper );
2764 make_trie( pRExC_state,
2765 startbranch, first, cur, tail, count,
2768 if ( PL_regkind[ OP( noper ) ] == EXACT
2770 && noper_next == tail
2775 optype = OP( noper );
2785 regprop(RExC_rx, mysv, cur);
2786 PerlIO_printf( Perl_debug_log,
2787 "%*s- %s (%d) <SCAN FINISHED>\n", (int)depth * 2 + 2,
2788 "", SvPV_nolen_const( mysv ),REG_NODE_NUM(cur));
2792 made= make_trie( pRExC_state, startbranch, first, scan, tail, count, optype, depth+1 );
2793 #ifdef TRIE_STUDY_OPT
2794 if ( ((made == MADE_EXACT_TRIE &&
2795 startbranch == first)
2796 || ( first_non_open == first )) &&
2798 flags |= SCF_TRIE_RESTUDY;
2799 if ( startbranch == first
2802 RExC_seen &=~REG_TOP_LEVEL_BRANCHES;
2812 else if ( code == BRANCHJ ) { /* single branch is optimized. */
2813 scan = NEXTOPER(NEXTOPER(scan));
2814 } else /* single branch is optimized. */
2815 scan = NEXTOPER(scan);
2817 } else if (OP(scan) == SUSPEND || OP(scan) == GOSUB || OP(scan) == GOSTART) {
2818 scan_frame *newframe = NULL;
2823 if (OP(scan) != SUSPEND) {
2824 /* set the pointer */
2825 if (OP(scan) == GOSUB) {
2827 RExC_recurse[ARG2L(scan)] = scan;
2828 start = RExC_open_parens[paren-1];
2829 end = RExC_close_parens[paren-1];
2832 start = RExC_rxi->program + 1;
2836 Newxz(recursed, (((RExC_npar)>>3) +1), U8);
2837 SAVEFREEPV(recursed);
2839 if (!PAREN_TEST(recursed,paren+1)) {
2840 PAREN_SET(recursed,paren+1);
2841 Newx(newframe,1,scan_frame);
2843 if (flags & SCF_DO_SUBSTR) {
2844 SCAN_COMMIT(pRExC_state,data,minlenp);
2845 data->longest = &(data->longest_float);
2847 is_inf = is_inf_internal = 1;
2848 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
2849 cl_anything(pRExC_state, data->start_class);
2850 flags &= ~SCF_DO_STCLASS;
2853 Newx(newframe,1,scan_frame);
2856 end = regnext(scan);
2861 SAVEFREEPV(newframe);
2862 newframe->next = regnext(scan);
2863 newframe->last = last;
2864 newframe->stop = stopparen;
2865 newframe->prev = frame;
2875 else if (OP(scan) == EXACT) {
2876 I32 l = STR_LEN(scan);
2879 const U8 * const s = (U8*)STRING(scan);
2880 l = utf8_length(s, s + l);
2881 uc = utf8_to_uvchr(s, NULL);
2883 uc = *((U8*)STRING(scan));
2886 if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
2887 /* The code below prefers earlier match for fixed
2888 offset, later match for variable offset. */
2889 if (data->last_end == -1) { /* Update the start info. */
2890 data->last_start_min = data->pos_min;
2891 data->last_start_max = is_inf
2892 ? I32_MAX : data->pos_min + data->pos_delta;
2894 sv_catpvn(data->last_found, STRING(scan), STR_LEN(scan));
2896 SvUTF8_on(data->last_found);
2898 SV * const sv = data->last_found;
2899 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
2900 mg_find(sv, PERL_MAGIC_utf8) : NULL;
2901 if (mg && mg->mg_len >= 0)
2902 mg->mg_len += utf8_length((U8*)STRING(scan),
2903 (U8*)STRING(scan)+STR_LEN(scan));
2905 data->last_end = data->pos_min + l;
2906 data->pos_min += l; /* As in the first entry. */
2907 data->flags &= ~SF_BEFORE_EOL;
2909 if (flags & SCF_DO_STCLASS_AND) {
2910 /* Check whether it is compatible with what we know already! */
2914 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
2915 && !ANYOF_BITMAP_TEST(data->start_class, uc)
2916 && (!(data->start_class->flags & ANYOF_FOLD)
2917 || !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
2920 ANYOF_CLASS_ZERO(data->start_class);
2921 ANYOF_BITMAP_ZERO(data->start_class);
2923 ANYOF_BITMAP_SET(data->start_class, uc);
2924 data->start_class->flags &= ~ANYOF_EOS;
2926 data->start_class->flags &= ~ANYOF_UNICODE_ALL;
2928 else if (flags & SCF_DO_STCLASS_OR) {
2929 /* false positive possible if the class is case-folded */
2931 ANYOF_BITMAP_SET(data->start_class, uc);
2933 data->start_class->flags |= ANYOF_UNICODE_ALL;
2934 data->start_class->flags &= ~ANYOF_EOS;
2935 cl_and(data->start_class, and_withp);
2937 flags &= ~SCF_DO_STCLASS;
2939 else if (PL_regkind[OP(scan)] == EXACT) { /* But OP != EXACT! */
2940 I32 l = STR_LEN(scan);
2941 UV uc = *((U8*)STRING(scan));
2943 /* Search for fixed substrings supports EXACT only. */
2944 if (flags & SCF_DO_SUBSTR) {
2946 SCAN_COMMIT(pRExC_state, data, minlenp);
2949 const U8 * const s = (U8 *)STRING(scan);
2950 l = utf8_length(s, s + l);
2951 uc = utf8_to_uvchr(s, NULL);
2954 if (flags & SCF_DO_SUBSTR)
2956 if (flags & SCF_DO_STCLASS_AND) {
2957 /* Check whether it is compatible with what we know already! */
2961 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
2962 && !ANYOF_BITMAP_TEST(data->start_class, uc)
2963 && !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
2965 ANYOF_CLASS_ZERO(data->start_class);
2966 ANYOF_BITMAP_ZERO(data->start_class);
2968 ANYOF_BITMAP_SET(data->start_class, uc);
2969 data->start_class->flags &= ~ANYOF_EOS;
2970 data->start_class->flags |= ANYOF_FOLD;
2971 if (OP(scan) == EXACTFL)
2972 data->start_class->flags |= ANYOF_LOCALE;
2975 else if (flags & SCF_DO_STCLASS_OR) {
2976 if (data->start_class->flags & ANYOF_FOLD) {
2977 /* false positive possible if the class is case-folded.
2978 Assume that the locale settings are the same... */
2980 ANYOF_BITMAP_SET(data->start_class, uc);
2981 data->start_class->flags &= ~ANYOF_EOS;
2983 cl_and(data->start_class, and_withp);
2985 flags &= ~SCF_DO_STCLASS;
2987 else if (strchr((const char*)PL_varies,OP(scan))) {
2988 I32 mincount, maxcount, minnext, deltanext, fl = 0;
2989 I32 f = flags, pos_before = 0;
2990 regnode * const oscan = scan;
2991 struct regnode_charclass_class this_class;
2992 struct regnode_charclass_class *oclass = NULL;
2993 I32 next_is_eval = 0;
2995 switch (PL_regkind[OP(scan)]) {
2996 case WHILEM: /* End of (?:...)* . */
2997 scan = NEXTOPER(scan);
3000 if (flags & (SCF_DO_SUBSTR | SCF_DO_STCLASS)) {
3001 next = NEXTOPER(scan);
3002 if (OP(next) == EXACT || (flags & SCF_DO_STCLASS)) {
3004 maxcount = REG_INFTY;
3005 next = regnext(scan);
3006 scan = NEXTOPER(scan);
3010 if (flags & SCF_DO_SUBSTR)
3015 if (flags & SCF_DO_STCLASS) {
3017 maxcount = REG_INFTY;
3018 next = regnext(scan);
3019 scan = NEXTOPER(scan);
3022 is_inf = is_inf_internal = 1;
3023 scan = regnext(scan);
3024 if (flags & SCF_DO_SUBSTR) {
3025 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot extend fixed substrings */
3026 data->longest = &(data->longest_float);
3028 goto optimize_curly_tail;
3030 if (stopparen>0 && (OP(scan)==CURLYN || OP(scan)==CURLYM)
3031 && (scan->flags == stopparen))
3036 mincount = ARG1(scan);
3037 maxcount = ARG2(scan);
3039 next = regnext(scan);
3040 if (OP(scan) == CURLYX) {
3041 I32 lp = (data ? *(data->last_closep) : 0);
3042 scan->flags = ((lp <= (I32)U8_MAX) ? (U8)lp : U8_MAX);
3044 scan = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
3045 next_is_eval = (OP(scan) == EVAL);
3047 if (flags & SCF_DO_SUBSTR) {
3048 if (mincount == 0) SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot extend fixed substrings */
3049 pos_before = data->pos_min;
3053 data->flags &= ~(SF_HAS_PAR|SF_IN_PAR|SF_HAS_EVAL);
3055 data->flags |= SF_IS_INF;
3057 if (flags & SCF_DO_STCLASS) {
3058 cl_init(pRExC_state, &this_class);
3059 oclass = data->start_class;
3060 data->start_class = &this_class;
3061 f |= SCF_DO_STCLASS_AND;
3062 f &= ~SCF_DO_STCLASS_OR;
3064 /* These are the cases when once a subexpression
3065 fails at a particular position, it cannot succeed
3066 even after backtracking at the enclosing scope.
3068 XXXX what if minimal match and we are at the
3069 initial run of {n,m}? */
3070 if ((mincount != maxcount - 1) && (maxcount != REG_INFTY))
3071 f &= ~SCF_WHILEM_VISITED_POS;
3073 /* This will finish on WHILEM, setting scan, or on NULL: */
3074 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
3075 last, data, stopparen, recursed, NULL,
3077 ? (f & ~SCF_DO_SUBSTR) : f),depth+1);
3079 if (flags & SCF_DO_STCLASS)
3080 data->start_class = oclass;
3081 if (mincount == 0 || minnext == 0) {
3082 if (flags & SCF_DO_STCLASS_OR) {
3083 cl_or(pRExC_state, data->start_class, &this_class);
3085 else if (flags & SCF_DO_STCLASS_AND) {
3086 /* Switch to OR mode: cache the old value of
3087 * data->start_class */
3089 StructCopy(data->start_class, and_withp,
3090 struct regnode_charclass_class);
3091 flags &= ~SCF_DO_STCLASS_AND;
3092 StructCopy(&this_class, data->start_class,
3093 struct regnode_charclass_class);
3094 flags |= SCF_DO_STCLASS_OR;
3095 data->start_class->flags |= ANYOF_EOS;
3097 } else { /* Non-zero len */
3098 if (flags & SCF_DO_STCLASS_OR) {
3099 cl_or(pRExC_state, data->start_class, &this_class);
3100 cl_and(data->start_class, and_withp);
3102 else if (flags & SCF_DO_STCLASS_AND)
3103 cl_and(data->start_class, &this_class);
3104 flags &= ~SCF_DO_STCLASS;
3106 if (!scan) /* It was not CURLYX, but CURLY. */
3108 if ( /* ? quantifier ok, except for (?{ ... }) */
3109 (next_is_eval || !(mincount == 0 && maxcount == 1))
3110 && (minnext == 0) && (deltanext == 0)
3111 && data && !(data->flags & (SF_HAS_PAR|SF_IN_PAR))
3112 && maxcount <= REG_INFTY/3 /* Complement check for big count */
3113 && ckWARN(WARN_REGEXP))
3116 "Quantifier unexpected on zero-length expression");
3119 min += minnext * mincount;
3120 is_inf_internal |= ((maxcount == REG_INFTY
3121 && (minnext + deltanext) > 0)
3122 || deltanext == I32_MAX);
3123 is_inf |= is_inf_internal;
3124 delta += (minnext + deltanext) * maxcount - minnext * mincount;
3126 /* Try powerful optimization CURLYX => CURLYN. */
3127 if ( OP(oscan) == CURLYX && data
3128 && data->flags & SF_IN_PAR
3129 && !(data->flags & SF_HAS_EVAL)
3130 && !deltanext && minnext == 1 ) {
3131 /* Try to optimize to CURLYN. */
3132 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS;
3133 regnode * const nxt1 = nxt;
3140 if (!strchr((const char*)PL_simple,OP(nxt))
3141 && !(PL_regkind[OP(nxt)] == EXACT
3142 && STR_LEN(nxt) == 1))
3148 if (OP(nxt) != CLOSE)
3150 if (RExC_open_parens) {
3151 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3152 RExC_close_parens[ARG(nxt1)-1]=nxt+2; /*close->while*/
3154 /* Now we know that nxt2 is the only contents: */
3155 oscan->flags = (U8)ARG(nxt);
3157 OP(nxt1) = NOTHING; /* was OPEN. */
3160 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3161 NEXT_OFF(nxt1+ 1) = 0; /* just for consistancy. */
3162 NEXT_OFF(nxt2) = 0; /* just for consistancy with CURLY. */
3163 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3164 OP(nxt + 1) = OPTIMIZED; /* was count. */
3165 NEXT_OFF(nxt+ 1) = 0; /* just for consistancy. */
3170 /* Try optimization CURLYX => CURLYM. */
3171 if ( OP(oscan) == CURLYX && data
3172 && !(data->flags & SF_HAS_PAR)
3173 && !(data->flags & SF_HAS_EVAL)
3174 && !deltanext /* atom is fixed width */
3175 && minnext != 0 /* CURLYM can't handle zero width */
3177 /* XXXX How to optimize if data == 0? */
3178 /* Optimize to a simpler form. */
3179 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN */
3183 while ( (nxt2 = regnext(nxt)) /* skip over embedded stuff*/
3184 && (OP(nxt2) != WHILEM))
3186 OP(nxt2) = SUCCEED; /* Whas WHILEM */
3187 /* Need to optimize away parenths. */
3188 if (data->flags & SF_IN_PAR) {
3189 /* Set the parenth number. */
3190 regnode *nxt1 = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN*/
3192 if (OP(nxt) != CLOSE)
3193 FAIL("Panic opt close");
3194 oscan->flags = (U8)ARG(nxt);
3195 if (RExC_open_parens) {
3196 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3197 RExC_close_parens[ARG(nxt1)-1]=nxt2+1; /*close->NOTHING*/
3199 OP(nxt1) = OPTIMIZED; /* was OPEN. */
3200 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3203 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3204 OP(nxt + 1) = OPTIMIZED; /* was count. */
3205 NEXT_OFF(nxt1 + 1) = 0; /* just for consistancy. */
3206 NEXT_OFF(nxt + 1) = 0; /* just for consistancy. */
3209 while ( nxt1 && (OP(nxt1) != WHILEM)) {
3210 regnode *nnxt = regnext(nxt1);
3213 if (reg_off_by_arg[OP(nxt1)])
3214 ARG_SET(nxt1, nxt2 - nxt1);
3215 else if (nxt2 - nxt1 < U16_MAX)
3216 NEXT_OFF(nxt1) = nxt2 - nxt1;
3218 OP(nxt) = NOTHING; /* Cannot beautify */
3223 /* Optimize again: */
3224 study_chunk(pRExC_state, &nxt1, minlenp, &deltanext, nxt,
3225 NULL, stopparen, recursed, NULL, 0,depth+1);
3230 else if ((OP(oscan) == CURLYX)
3231 && (flags & SCF_WHILEM_VISITED_POS)
3232 /* See the comment on a similar expression above.
3233 However, this time it not a subexpression
3234 we care about, but the expression itself. */
3235 && (maxcount == REG_INFTY)
3236 && data && ++data->whilem_c < 16) {
3237 /* This stays as CURLYX, we can put the count/of pair. */
3238 /* Find WHILEM (as in regexec.c) */
3239 regnode *nxt = oscan + NEXT_OFF(oscan);
3241 if (OP(PREVOPER(nxt)) == NOTHING) /* LONGJMP */
3243 PREVOPER(nxt)->flags = (U8)(data->whilem_c
3244 | (RExC_whilem_seen << 4)); /* On WHILEM */
3246 if (data && fl & (SF_HAS_PAR|SF_IN_PAR))
3248 if (flags & SCF_DO_SUBSTR) {
3249 SV *last_str = NULL;
3250 int counted = mincount != 0;
3252 if (data->last_end > 0 && mincount != 0) { /* Ends with a string. */
3253 #if defined(SPARC64_GCC_WORKAROUND)
3256 const char *s = NULL;
3259 if (pos_before >= data->last_start_min)
3262 b = data->last_start_min;
3265 s = SvPV_const(data->last_found, l);
3266 old = b - data->last_start_min;
3269 I32 b = pos_before >= data->last_start_min
3270 ? pos_before : data->last_start_min;
3272 const char * const s = SvPV_const(data->last_found, l);
3273 I32 old = b - data->last_start_min;
3277 old = utf8_hop((U8*)s, old) - (U8*)s;
3280 /* Get the added string: */
3281 last_str = newSVpvn(s + old, l);
3283 SvUTF8_on(last_str);
3284 if (deltanext == 0 && pos_before == b) {
3285 /* What was added is a constant string */
3287 SvGROW(last_str, (mincount * l) + 1);
3288 repeatcpy(SvPVX(last_str) + l,
3289 SvPVX_const(last_str), l, mincount - 1);
3290 SvCUR_set(last_str, SvCUR(last_str) * mincount);
3291 /* Add additional parts. */
3292 SvCUR_set(data->last_found,
3293 SvCUR(data->last_found) - l);
3294 sv_catsv(data->last_found, last_str);
3296 SV * sv = data->last_found;
3298 SvUTF8(sv) && SvMAGICAL(sv) ?
3299 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3300 if (mg && mg->mg_len >= 0)
3301 mg->mg_len += CHR_SVLEN(last_str);
3303 data->last_end += l * (mincount - 1);
3306 /* start offset must point into the last copy */
3307 data->last_start_min += minnext * (mincount - 1);
3308 data->last_start_max += is_inf ? I32_MAX
3309 : (maxcount - 1) * (minnext + data->pos_delta);
3312 /* It is counted once already... */
3313 data->pos_min += minnext * (mincount - counted);
3314 data->pos_delta += - counted * deltanext +
3315 (minnext + deltanext) * maxcount - minnext * mincount;
3316 if (mincount != maxcount) {
3317 /* Cannot extend fixed substrings found inside
3319 SCAN_COMMIT(pRExC_state,data,minlenp);
3320 if (mincount && last_str) {
3321 SV * const sv = data->last_found;
3322 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3323 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3327 sv_setsv(sv, last_str);
3328 data->last_end = data->pos_min;
3329 data->last_start_min =
3330 data->pos_min - CHR_SVLEN(last_str);
3331 data->last_start_max = is_inf
3333 : data->pos_min + data->pos_delta
3334 - CHR_SVLEN(last_str);
3336 data->longest = &(data->longest_float);
3338 SvREFCNT_dec(last_str);
3340 if (data && (fl & SF_HAS_EVAL))
3341 data->flags |= SF_HAS_EVAL;
3342 optimize_curly_tail:
3343 if (OP(oscan) != CURLYX) {
3344 while (PL_regkind[OP(next = regnext(oscan))] == NOTHING
3346 NEXT_OFF(oscan) += NEXT_OFF(next);
3349 default: /* REF and CLUMP only? */
3350 if (flags & SCF_DO_SUBSTR) {
3351 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3352 data->longest = &(data->longest_float);
3354 is_inf = is_inf_internal = 1;
3355 if (flags & SCF_DO_STCLASS_OR)
3356 cl_anything(pRExC_state, data->start_class);
3357 flags &= ~SCF_DO_STCLASS;
3361 else if (OP(scan) == LNBREAK) {
3362 if (flags & SCF_DO_STCLASS) {
3364 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3365 if (flags & SCF_DO_STCLASS_AND) {
3366 for (value = 0; value < 256; value++)
3367 if (!is_VERTWS_cp(value))
3368 ANYOF_BITMAP_CLEAR(data->start_class, value);
3371 for (value = 0; value < 256; value++)
3372 if (is_VERTWS_cp(value))
3373 ANYOF_BITMAP_SET(data->start_class, value);
3375 if (flags & SCF_DO_STCLASS_OR)
3376 cl_and(data->start_class, and_withp);
3377 flags &= ~SCF_DO_STCLASS;
3381 if (flags & SCF_DO_SUBSTR) {
3382 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3384 data->pos_delta += 1;
3385 data->longest = &(data->longest_float);
3389 else if (OP(scan) == FOLDCHAR) {
3390 int d = ARG(scan)==0xDF ? 1 : 2;
3391 flags &= ~SCF_DO_STCLASS;
3394 if (flags & SCF_DO_SUBSTR) {
3395 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3397 data->pos_delta += d;
3398 data->longest = &(data->longest_float);
3401 else if (strchr((const char*)PL_simple,OP(scan))) {
3404 if (flags & SCF_DO_SUBSTR) {
3405 SCAN_COMMIT(pRExC_state,data,minlenp);
3409 if (flags & SCF_DO_STCLASS) {
3410 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3412 /* Some of the logic below assumes that switching
3413 locale on will only add false positives. */
3414 switch (PL_regkind[OP(scan)]) {
3418 /* Perl_croak(aTHX_ "panic: unexpected simple REx opcode %d", OP(scan)); */
3419 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3420 cl_anything(pRExC_state, data->start_class);
3423 if (OP(scan) == SANY)
3425 if (flags & SCF_DO_STCLASS_OR) { /* Everything but \n */
3426 value = (ANYOF_BITMAP_TEST(data->start_class,'\n')
3427 || (data->start_class->flags & ANYOF_CLASS));
3428 cl_anything(pRExC_state, data->start_class);
3430 if (flags & SCF_DO_STCLASS_AND || !value)
3431 ANYOF_BITMAP_CLEAR(data->start_class,'\n');
3434 if (flags & SCF_DO_STCLASS_AND)
3435 cl_and(data->start_class,
3436 (struct regnode_charclass_class*)scan);
3438 cl_or(pRExC_state, data->start_class,
3439 (struct regnode_charclass_class*)scan);
3442 if (flags & SCF_DO_STCLASS_AND) {
3443 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3444 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3445 for (value = 0; value < 256; value++)
3446 if (!isALNUM(value))
3447 ANYOF_BITMAP_CLEAR(data->start_class, value);
3451 if (data->start_class->flags & ANYOF_LOCALE)
3452 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3454 for (value = 0; value < 256; value++)
3456 ANYOF_BITMAP_SET(data->start_class, value);
3461 if (flags & SCF_DO_STCLASS_AND) {
3462 if (data->start_class->flags & ANYOF_LOCALE)
3463 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3466 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3467 data->start_class->flags |= ANYOF_LOCALE;
3471 if (flags & SCF_DO_STCLASS_AND) {
3472 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3473 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3474 for (value = 0; value < 256; value++)
3476 ANYOF_BITMAP_CLEAR(data->start_class, value);
3480 if (data->start_class->flags & ANYOF_LOCALE)
3481 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3483 for (value = 0; value < 256; value++)
3484 if (!isALNUM(value))
3485 ANYOF_BITMAP_SET(data->start_class, value);
3490 if (flags & SCF_DO_STCLASS_AND) {
3491 if (data->start_class->flags & ANYOF_LOCALE)
3492 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3495 data->start_class->flags |= ANYOF_LOCALE;
3496 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3500 if (flags & SCF_DO_STCLASS_AND) {
3501 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3502 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3503 for (value = 0; value < 256; value++)
3504 if (!isSPACE(value))
3505 ANYOF_BITMAP_CLEAR(data->start_class, value);
3509 if (data->start_class->flags & ANYOF_LOCALE)
3510 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3512 for (value = 0; value < 256; value++)
3514 ANYOF_BITMAP_SET(data->start_class, value);
3519 if (flags & SCF_DO_STCLASS_AND) {
3520 if (data->start_class->flags & ANYOF_LOCALE)
3521 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3524 data->start_class->flags |= ANYOF_LOCALE;
3525 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3529 if (flags & SCF_DO_STCLASS_AND) {
3530 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3531 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3532 for (value = 0; value < 256; value++)
3534 ANYOF_BITMAP_CLEAR(data->start_class, value);
3538 if (data->start_class->flags & ANYOF_LOCALE)
3539 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3541 for (value = 0; value < 256; value++)
3542 if (!isSPACE(value))
3543 ANYOF_BITMAP_SET(data->start_class, value);
3548 if (flags & SCF_DO_STCLASS_AND) {
3549 if (data->start_class->flags & ANYOF_LOCALE) {
3550 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3551 for (value = 0; value < 256; value++)
3552 if (!isSPACE(value))
3553 ANYOF_BITMAP_CLEAR(data->start_class, value);
3557 data->start_class->flags |= ANYOF_LOCALE;
3558 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3562 if (flags & SCF_DO_STCLASS_AND) {
3563 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NDIGIT);
3564 for (value = 0; value < 256; value++)
3565 if (!isDIGIT(value))
3566 ANYOF_BITMAP_CLEAR(data->start_class, value);
3569 if (data->start_class->flags & ANYOF_LOCALE)
3570 ANYOF_CLASS_SET(data->start_class,ANYOF_DIGIT);
3572 for (value = 0; value < 256; value++)
3574 ANYOF_BITMAP_SET(data->start_class, value);
3579 if (flags & SCF_DO_STCLASS_AND) {
3580 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_DIGIT);
3581 for (value = 0; value < 256; value++)
3583 ANYOF_BITMAP_CLEAR(data->start_class, value);
3586 if (data->start_class->flags & ANYOF_LOCALE)
3587 ANYOF_CLASS_SET(data->start_class,ANYOF_NDIGIT);
3589 for (value = 0; value < 256; value++)
3590 if (!isDIGIT(value))
3591 ANYOF_BITMAP_SET(data->start_class, value);
3595 CASE_SYNST_FNC(VERTWS);
3596 CASE_SYNST_FNC(HORIZWS);
3599 if (flags & SCF_DO_STCLASS_OR)
3600 cl_and(data->start_class, and_withp);
3601 flags &= ~SCF_DO_STCLASS;
3604 else if (PL_regkind[OP(scan)] == EOL && flags & SCF_DO_SUBSTR) {
3605 data->flags |= (OP(scan) == MEOL
3609 else if ( PL_regkind[OP(scan)] == BRANCHJ
3610 /* Lookbehind, or need to calculate parens/evals/stclass: */
3611 && (scan->flags || data || (flags & SCF_DO_STCLASS))
3612 && (OP(scan) == IFMATCH || OP(scan) == UNLESSM)) {
3613 if ( !PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3614 || OP(scan) == UNLESSM )
3616 /* Negative Lookahead/lookbehind
3617 In this case we can't do fixed string optimisation.
3620 I32 deltanext, minnext, fake = 0;
3622 struct regnode_charclass_class intrnl;
3625 data_fake.flags = 0;
3627 data_fake.whilem_c = data->whilem_c;
3628 data_fake.last_closep = data->last_closep;
3631 data_fake.last_closep = &fake;
3632 data_fake.pos_delta = delta;
3633 if ( flags & SCF_DO_STCLASS && !scan->flags
3634 && OP(scan) == IFMATCH ) { /* Lookahead */
3635 cl_init(pRExC_state, &intrnl);
3636 data_fake.start_class = &intrnl;
3637 f |= SCF_DO_STCLASS_AND;
3639 if (flags & SCF_WHILEM_VISITED_POS)
3640 f |= SCF_WHILEM_VISITED_POS;
3641 next = regnext(scan);
3642 nscan = NEXTOPER(NEXTOPER(scan));
3643 minnext = study_chunk(pRExC_state, &nscan, minlenp, &deltanext,
3644 last, &data_fake, stopparen, recursed, NULL, f, depth+1);
3647 FAIL("Variable length lookbehind not implemented");
3649 else if (minnext > (I32)U8_MAX) {
3650 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3652 scan->flags = (U8)minnext;
3655 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3657 if (data_fake.flags & SF_HAS_EVAL)
3658 data->flags |= SF_HAS_EVAL;
3659 data->whilem_c = data_fake.whilem_c;
3661 if (f & SCF_DO_STCLASS_AND) {
3662 const int was = (data->start_class->flags & ANYOF_EOS);
3664 cl_and(data->start_class, &intrnl);
3666 data->start_class->flags |= ANYOF_EOS;
3669 #if PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3671 /* Positive Lookahead/lookbehind
3672 In this case we can do fixed string optimisation,
3673 but we must be careful about it. Note in the case of
3674 lookbehind the positions will be offset by the minimum
3675 length of the pattern, something we won't know about
3676 until after the recurse.
3678 I32 deltanext, fake = 0;
3680 struct regnode_charclass_class intrnl;
3682 /* We use SAVEFREEPV so that when the full compile
3683 is finished perl will clean up the allocated
3684 minlens when its all done. This was we don't
3685 have to worry about freeing them when we know
3686 they wont be used, which would be a pain.
3689 Newx( minnextp, 1, I32 );
3690 SAVEFREEPV(minnextp);
3693 StructCopy(data, &data_fake, scan_data_t);
3694 if ((flags & SCF_DO_SUBSTR) && data->last_found) {
3697 SCAN_COMMIT(pRExC_state, &data_fake,minlenp);
3698 data_fake.last_found=newSVsv(data->last_found);
3702 data_fake.last_closep = &fake;
3703 data_fake.flags = 0;
3704 data_fake.pos_delta = delta;
3706 data_fake.flags |= SF_IS_INF;
3707 if ( flags & SCF_DO_STCLASS && !scan->flags
3708 && OP(scan) == IFMATCH ) { /* Lookahead */
3709 cl_init(pRExC_state, &intrnl);
3710 data_fake.start_class = &intrnl;
3711 f |= SCF_DO_STCLASS_AND;
3713 if (flags & SCF_WHILEM_VISITED_POS)
3714 f |= SCF_WHILEM_VISITED_POS;
3715 next = regnext(scan);
3716 nscan = NEXTOPER(NEXTOPER(scan));
3718 *minnextp = study_chunk(pRExC_state, &nscan, minnextp, &deltanext,
3719 last, &data_fake, stopparen, recursed, NULL, f,depth+1);
3722 FAIL("Variable length lookbehind not implemented");
3724 else if (*minnextp > (I32)U8_MAX) {
3725 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3727 scan->flags = (U8)*minnextp;
3732 if (f & SCF_DO_STCLASS_AND) {
3733 const int was = (data->start_class->flags & ANYOF_EOS);
3735 cl_and(data->start_class, &intrnl);
3737 data->start_class->flags |= ANYOF_EOS;
3740 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3742 if (data_fake.flags & SF_HAS_EVAL)
3743 data->flags |= SF_HAS_EVAL;
3744 data->whilem_c = data_fake.whilem_c;
3745 if ((flags & SCF_DO_SUBSTR) && data_fake.last_found) {
3746 if (RExC_rx->minlen<*minnextp)
3747 RExC_rx->minlen=*minnextp;
3748 SCAN_COMMIT(pRExC_state, &data_fake, minnextp);
3749 SvREFCNT_dec(data_fake.last_found);
3751 if ( data_fake.minlen_fixed != minlenp )
3753 data->offset_fixed= data_fake.offset_fixed;
3754 data->minlen_fixed= data_fake.minlen_fixed;
3755 data->lookbehind_fixed+= scan->flags;
3757 if ( data_fake.minlen_float != minlenp )
3759 data->minlen_float= data_fake.minlen_float;
3760 data->offset_float_min=data_fake.offset_float_min;
3761 data->offset_float_max=data_fake.offset_float_max;
3762 data->lookbehind_float+= scan->flags;
3771 else if (OP(scan) == OPEN) {
3772 if (stopparen != (I32)ARG(scan))
3775 else if (OP(scan) == CLOSE) {
3776 if (stopparen == (I32)ARG(scan)) {
3779 if ((I32)ARG(scan) == is_par) {
3780 next = regnext(scan);
3782 if ( next && (OP(next) != WHILEM) && next < last)
3783 is_par = 0; /* Disable optimization */
3786 *(data->last_closep) = ARG(scan);
3788 else if (OP(scan) == EVAL) {
3790 data->flags |= SF_HAS_EVAL;
3792 else if ( PL_regkind[OP(scan)] == ENDLIKE ) {
3793 if (flags & SCF_DO_SUBSTR) {
3794 SCAN_COMMIT(pRExC_state,data,minlenp);
3795 flags &= ~SCF_DO_SUBSTR;
3797 if (data && OP(scan)==ACCEPT) {
3798 data->flags |= SCF_SEEN_ACCEPT;
3803 else if (OP(scan) == LOGICAL && scan->flags == 2) /* Embedded follows */
3805 if (flags & SCF_DO_SUBSTR) {
3806 SCAN_COMMIT(pRExC_state,data,minlenp);
3807 data->longest = &(data->longest_float);
3809 is_inf = is_inf_internal = 1;
3810 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3811 cl_anything(pRExC_state, data->start_class);
3812 flags &= ~SCF_DO_STCLASS;
3814 else if (OP(scan) == GPOS) {
3815 if (!(RExC_rx->extflags & RXf_GPOS_FLOAT) &&
3816 !(delta || is_inf || (data && data->pos_delta)))
3818 if (!(RExC_rx->extflags & RXf_ANCH) && (flags & SCF_DO_SUBSTR))
3819 RExC_rx->extflags |= RXf_ANCH_GPOS;
3820 if (RExC_rx->gofs < (U32)min)
3821 RExC_rx->gofs = min;
3823 RExC_rx->extflags |= RXf_GPOS_FLOAT;
3827 #ifdef TRIE_STUDY_OPT
3828 #ifdef FULL_TRIE_STUDY
3829 else if (PL_regkind[OP(scan)] == TRIE) {
3830 /* NOTE - There is similar code to this block above for handling
3831 BRANCH nodes on the initial study. If you change stuff here
3833 regnode *trie_node= scan;
3834 regnode *tail= regnext(scan);
3835 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
3836 I32 max1 = 0, min1 = I32_MAX;
3837 struct regnode_charclass_class accum;
3839 if (flags & SCF_DO_SUBSTR) /* XXXX Add !SUSPEND? */
3840 SCAN_COMMIT(pRExC_state, data,minlenp); /* Cannot merge strings after this. */
3841 if (flags & SCF_DO_STCLASS)
3842 cl_init_zero(pRExC_state, &accum);
3848 const regnode *nextbranch= NULL;
3851 for ( word=1 ; word <= trie->wordcount ; word++)
3853 I32 deltanext=0, minnext=0, f = 0, fake;
3854 struct regnode_charclass_class this_class;
3856 data_fake.flags = 0;
3858 data_fake.whilem_c = data->whilem_c;
3859 data_fake.last_closep = data->last_closep;
3862 data_fake.last_closep = &fake;
3863 data_fake.pos_delta = delta;
3864 if (flags & SCF_DO_STCLASS) {
3865 cl_init(pRExC_state, &this_class);
3866 data_fake.start_class = &this_class;
3867 f = SCF_DO_STCLASS_AND;
3869 if (flags & SCF_WHILEM_VISITED_POS)
3870 f |= SCF_WHILEM_VISITED_POS;
3872 if (trie->jump[word]) {
3874 nextbranch = trie_node + trie->jump[0];
3875 scan= trie_node + trie->jump[word];
3876 /* We go from the jump point to the branch that follows
3877 it. Note this means we need the vestigal unused branches
3878 even though they arent otherwise used.
3880 minnext = study_chunk(pRExC_state, &scan, minlenp,
3881 &deltanext, (regnode *)nextbranch, &data_fake,
3882 stopparen, recursed, NULL, f,depth+1);
3884 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
3885 nextbranch= regnext((regnode*)nextbranch);
3887 if (min1 > (I32)(minnext + trie->minlen))
3888 min1 = minnext + trie->minlen;
3889 if (max1 < (I32)(minnext + deltanext + trie->maxlen))
3890 max1 = minnext + deltanext + trie->maxlen;
3891 if (deltanext == I32_MAX)
3892 is_inf = is_inf_internal = 1;
3894 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3896 if (data_fake.flags & SCF_SEEN_ACCEPT) {
3897 if ( stopmin > min + min1)
3898 stopmin = min + min1;
3899 flags &= ~SCF_DO_SUBSTR;
3901 data->flags |= SCF_SEEN_ACCEPT;
3904 if (data_fake.flags & SF_HAS_EVAL)
3905 data->flags |= SF_HAS_EVAL;
3906 data->whilem_c = data_fake.whilem_c;
3908 if (flags & SCF_DO_STCLASS)
3909 cl_or(pRExC_state, &accum, &this_class);
3912 if (flags & SCF_DO_SUBSTR) {
3913 data->pos_min += min1;
3914 data->pos_delta += max1 - min1;
3915 if (max1 != min1 || is_inf)
3916 data->longest = &(data->longest_float);
3919 delta += max1 - min1;
3920 if (flags & SCF_DO_STCLASS_OR) {
3921 cl_or(pRExC_state, data->start_class, &accum);
3923 cl_and(data->start_class, and_withp);
3924 flags &= ~SCF_DO_STCLASS;
3927 else if (flags & SCF_DO_STCLASS_AND) {
3929 cl_and(data->start_class, &accum);
3930 flags &= ~SCF_DO_STCLASS;
3933 /* Switch to OR mode: cache the old value of
3934 * data->start_class */
3936 StructCopy(data->start_class, and_withp,
3937 struct regnode_charclass_class);
3938 flags &= ~SCF_DO_STCLASS_AND;
3939 StructCopy(&accum, data->start_class,
3940 struct regnode_charclass_class);
3941 flags |= SCF_DO_STCLASS_OR;
3942 data->start_class->flags |= ANYOF_EOS;
3949 else if (PL_regkind[OP(scan)] == TRIE) {
3950 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
3953 min += trie->minlen;
3954 delta += (trie->maxlen - trie->minlen);
3955 flags &= ~SCF_DO_STCLASS; /* xxx */
3956 if (flags & SCF_DO_SUBSTR) {
3957 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3958 data->pos_min += trie->minlen;
3959 data->pos_delta += (trie->maxlen - trie->minlen);
3960 if (trie->maxlen != trie->minlen)
3961 data->longest = &(data->longest_float);
3963 if (trie->jump) /* no more substrings -- for now /grr*/
3964 flags &= ~SCF_DO_SUBSTR;
3966 #endif /* old or new */
3967 #endif /* TRIE_STUDY_OPT */
3969 /* Else: zero-length, ignore. */
3970 scan = regnext(scan);
3975 stopparen = frame->stop;
3976 frame = frame->prev;
3977 goto fake_study_recurse;
3982 DEBUG_STUDYDATA("pre-fin:",data,depth);
3985 *deltap = is_inf_internal ? I32_MAX : delta;
3986 if (flags & SCF_DO_SUBSTR && is_inf)
3987 data->pos_delta = I32_MAX - data->pos_min;
3988 if (is_par > (I32)U8_MAX)
3990 if (is_par && pars==1 && data) {
3991 data->flags |= SF_IN_PAR;
3992 data->flags &= ~SF_HAS_PAR;
3994 else if (pars && data) {
3995 data->flags |= SF_HAS_PAR;
3996 data->flags &= ~SF_IN_PAR;
3998 if (flags & SCF_DO_STCLASS_OR)
3999 cl_and(data->start_class, and_withp);
4000 if (flags & SCF_TRIE_RESTUDY)
4001 data->flags |= SCF_TRIE_RESTUDY;
4003 DEBUG_STUDYDATA("post-fin:",data,depth);
4005 return min < stopmin ? min : stopmin;
4009 S_add_data(RExC_state_t *pRExC_state, U32 n, const char *s)
4011 U32 count = RExC_rxi->data ? RExC_rxi->data->count : 0;
4013 Renewc(RExC_rxi->data,
4014 sizeof(*RExC_rxi->data) + sizeof(void*) * (count + n - 1),
4015 char, struct reg_data);
4017 Renew(RExC_rxi->data->what, count + n, U8);
4019 Newx(RExC_rxi->data->what, n, U8);
4020 RExC_rxi->data->count = count + n;
4021 Copy(s, RExC_rxi->data->what + count, n, U8);
4025 /*XXX: todo make this not included in a non debugging perl */
4026 #ifndef PERL_IN_XSUB_RE
4028 Perl_reginitcolors(pTHX)
4031 const char * const s = PerlEnv_getenv("PERL_RE_COLORS");
4033 char *t = savepv(s);
4037 t = strchr(t, '\t');
4043 PL_colors[i] = t = (char *)"";
4048 PL_colors[i++] = (char *)"";
4055 #ifdef TRIE_STUDY_OPT
4056 #define CHECK_RESTUDY_GOTO \
4058 (data.flags & SCF_TRIE_RESTUDY) \
4062 #define CHECK_RESTUDY_GOTO
4066 - pregcomp - compile a regular expression into internal code
4068 * We can't allocate space until we know how big the compiled form will be,
4069 * but we can't compile it (and thus know how big it is) until we've got a
4070 * place to put the code. So we cheat: we compile it twice, once with code
4071 * generation turned off and size counting turned on, and once "for real".
4072 * This also means that we don't allocate space until we are sure that the
4073 * thing really will compile successfully, and we never have to move the
4074 * code and thus invalidate pointers into it. (Note that it has to be in
4075 * one piece because free() must be able to free it all.) [NB: not true in perl]
4077 * Beware that the optimization-preparation code in here knows about some
4078 * of the structure of the compiled regexp. [I'll say.]
4083 #ifndef PERL_IN_XSUB_RE
4084 #define RE_ENGINE_PTR &PL_core_reg_engine
4086 extern const struct regexp_engine my_reg_engine;
4087 #define RE_ENGINE_PTR &my_reg_engine
4090 #ifndef PERL_IN_XSUB_RE
4092 Perl_pregcomp(pTHX_ const SV * const pattern, const U32 flags)
4095 HV * const table = GvHV(PL_hintgv);
4096 /* Dispatch a request to compile a regexp to correct
4099 SV **ptr= hv_fetchs(table, "regcomp", FALSE);
4100 GET_RE_DEBUG_FLAGS_DECL;
4101 if (ptr && SvIOK(*ptr) && SvIV(*ptr)) {
4102 const regexp_engine *eng=INT2PTR(regexp_engine*,SvIV(*ptr));
4104 PerlIO_printf(Perl_debug_log, "Using engine %"UVxf"\n",
4107 return CALLREGCOMP_ENG(eng, pattern, flags);
4110 return Perl_re_compile(aTHX_ pattern, flags);
4115 Perl_re_compile(pTHX_ const SV * const pattern, const U32 pm_flags)
4119 register regexp_internal *ri;
4121 char* exp = SvPV((SV*)pattern, plen);
4122 char* xend = exp + plen;
4129 RExC_state_t RExC_state;
4130 RExC_state_t * const pRExC_state = &RExC_state;
4131 #ifdef TRIE_STUDY_OPT
4133 RExC_state_t copyRExC_state;
4135 GET_RE_DEBUG_FLAGS_DECL;
4136 DEBUG_r(if (!PL_colorset) reginitcolors());
4138 RExC_utf8 = RExC_orig_utf8 = pm_flags & RXf_UTF8;
4141 SV *dsv= sv_newmortal();
4142 RE_PV_QUOTED_DECL(s, RExC_utf8,
4143 dsv, exp, plen, 60);
4144 PerlIO_printf(Perl_debug_log, "%sCompiling REx%s %s\n",
4145 PL_colors[4],PL_colors[5],s);
4150 RExC_flags = pm_flags;
4154 RExC_seen_zerolen = *exp == '^' ? -1 : 0;
4155 RExC_seen_evals = 0;
4158 /* First pass: determine size, legality. */
4166 RExC_emit = &PL_regdummy;
4167 RExC_whilem_seen = 0;
4168 RExC_charnames = NULL;
4169 RExC_open_parens = NULL;
4170 RExC_close_parens = NULL;
4172 RExC_paren_names = NULL;
4174 RExC_paren_name_list = NULL;
4176 RExC_recurse = NULL;
4177 RExC_recurse_count = 0;
4179 #if 0 /* REGC() is (currently) a NOP at the first pass.
4180 * Clever compilers notice this and complain. --jhi */
4181 REGC((U8)REG_MAGIC, (char*)RExC_emit);
4183 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "Starting first pass (sizing)\n"));
4184 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4185 RExC_precomp = NULL;
4188 if (RExC_utf8 && !RExC_orig_utf8) {
4189 /* It's possible to write a regexp in ascii that represents unicode
4190 codepoints outside of the byte range, such as via \x{100}. If we
4191 detect such a sequence we have to convert the entire pattern to utf8
4192 and then recompile, as our sizing calculation will have been based
4193 on 1 byte == 1 character, but we will need to use utf8 to encode
4194 at least some part of the pattern, and therefore must convert the whole
4196 XXX: somehow figure out how to make this less expensive...
4199 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log,
4200 "UTF8 mismatch! Converting to utf8 for resizing and compile\n"));
4201 exp = (char*)Perl_bytes_to_utf8(aTHX_ (U8*)exp, &len);
4203 RExC_orig_utf8 = RExC_utf8;
4205 goto redo_first_pass;
4208 PerlIO_printf(Perl_debug_log,
4209 "Required size %"IVdf" nodes\n"
4210 "Starting second pass (creation)\n",
4213 RExC_lastparse=NULL;
4215 /* Small enough for pointer-storage convention?
4216 If extralen==0, this means that we will not need long jumps. */
4217 if (RExC_size >= 0x10000L && RExC_extralen)
4218 RExC_size += RExC_extralen;
4221 if (RExC_whilem_seen > 15)
4222 RExC_whilem_seen = 15;
4224 /* Allocate space and zero-initialize. Note, the two step process
4225 of zeroing when in debug mode, thus anything assigned has to
4226 happen after that */
4227 Newxz(r, 1, regexp);
4228 Newxc(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode),
4229 char, regexp_internal);
4230 if ( r == NULL || ri == NULL )
4231 FAIL("Regexp out of space");
4233 /* avoid reading uninitialized memory in DEBUGGING code in study_chunk() */
4234 Zero(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode), char);
4236 /* bulk initialize base fields with 0. */
4237 Zero(ri, sizeof(regexp_internal), char);
4240 /* non-zero initialization begins here */
4242 r->engine= RE_ENGINE_PTR;
4245 r->extflags = pm_flags;
4247 bool has_k = ((r->extflags & RXf_PMf_KEEPCOPY) == RXf_PMf_KEEPCOPY);
4248 bool has_minus = ((r->extflags & RXf_PMf_STD_PMMOD) != RXf_PMf_STD_PMMOD);
4249 bool has_runon = ((RExC_seen & REG_SEEN_RUN_ON_COMMENT)==REG_SEEN_RUN_ON_COMMENT);
4250 U16 reganch = (U16)((r->extflags & RXf_PMf_STD_PMMOD) >> 12);
4251 const char *fptr = STD_PAT_MODS; /*"msix"*/
4253 r->wraplen = r->prelen + has_minus + has_k + has_runon
4254 + (sizeof(STD_PAT_MODS) - 1)
4255 + (sizeof("(?:)") - 1);
4257 Newx(r->wrapped, r->wraplen + 1, char );
4261 *p++ = KEEPCOPY_PAT_MOD; /*'k'*/
4263 char *r = p + (sizeof(STD_PAT_MODS) - 1) + has_minus - 1;
4264 char *colon = r + 1;
4267 while((ch = *fptr++)) {
4281 Copy(RExC_precomp, p, r->prelen, char);
4291 r->nparens = RExC_npar - 1; /* set early to validate backrefs */
4293 if (RExC_seen & REG_SEEN_RECURSE) {
4294 Newxz(RExC_open_parens, RExC_npar,regnode *);
4295 SAVEFREEPV(RExC_open_parens);
4296 Newxz(RExC_close_parens,RExC_npar,regnode *);
4297 SAVEFREEPV(RExC_close_parens);
4300 /* Useful during FAIL. */
4301 #ifdef RE_TRACK_PATTERN_OFFSETS
4302 Newxz(ri->u.offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
4303 DEBUG_OFFSETS_r(PerlIO_printf(Perl_debug_log,
4304 "%s %"UVuf" bytes for offset annotations.\n",
4305 ri->u.offsets ? "Got" : "Couldn't get",
4306 (UV)((2*RExC_size+1) * sizeof(U32))));
4308 SetProgLen(ri,RExC_size);
4312 /* Second pass: emit code. */
4313 RExC_flags = pm_flags; /* don't let top level (?i) bleed */
4318 RExC_emit_start = ri->program;
4319 RExC_emit = ri->program;
4320 RExC_emit_bound = ri->program + RExC_size + 1;
4322 /* Store the count of eval-groups for security checks: */
4323 RExC_rx->seen_evals = RExC_seen_evals;
4324 REGC((U8)REG_MAGIC, (char*) RExC_emit++);
4325 if (reg(pRExC_state, 0, &flags,1) == NULL)
4328 /* XXXX To minimize changes to RE engine we always allocate
4329 3-units-long substrs field. */
4330 Newx(r->substrs, 1, struct reg_substr_data);
4331 if (RExC_recurse_count) {
4332 Newxz(RExC_recurse,RExC_recurse_count,regnode *);
4333 SAVEFREEPV(RExC_recurse);
4337 r->minlen = minlen = sawplus = sawopen = 0;
4338 Zero(r->substrs, 1, struct reg_substr_data);
4340 #ifdef TRIE_STUDY_OPT
4343 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log,"Restudying\n"));
4345 RExC_state = copyRExC_state;
4346 if (seen & REG_TOP_LEVEL_BRANCHES)
4347 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
4349 RExC_seen &= ~REG_TOP_LEVEL_BRANCHES;
4350 if (data.last_found) {
4351 SvREFCNT_dec(data.longest_fixed);
4352 SvREFCNT_dec(data.longest_float);
4353 SvREFCNT_dec(data.last_found);
4355 StructCopy(&zero_scan_data, &data, scan_data_t);
4357 StructCopy(&zero_scan_data, &data, scan_data_t);
4358 copyRExC_state = RExC_state;
4361 StructCopy(&zero_scan_data, &data, scan_data_t);
4364 /* Dig out information for optimizations. */
4365 r->extflags = pm_flags; /* Again? */
4366 /*dmq: removed as part of de-PMOP: pm->op_pmflags = RExC_flags; */
4369 r->extflags |= RXf_UTF8; /* Unicode in it? */
4370 ri->regstclass = NULL;
4371 if (RExC_naughty >= 10) /* Probably an expensive pattern. */
4372 r->intflags |= PREGf_NAUGHTY;
4373 scan = ri->program + 1; /* First BRANCH. */
4375 /* testing for BRANCH here tells us whether there is "must appear"
4376 data in the pattern. If there is then we can use it for optimisations */
4377 if (!(RExC_seen & REG_TOP_LEVEL_BRANCHES)) { /* Only one top-level choice. */
4379 STRLEN longest_float_length, longest_fixed_length;
4380 struct regnode_charclass_class ch_class; /* pointed to by data */
4382 I32 last_close = 0; /* pointed to by data */
4383 regnode *first= scan;
4384 regnode *first_next= regnext(first);
4386 /* Skip introductions and multiplicators >= 1. */
4387 while ((OP(first) == OPEN && (sawopen = 1)) ||
4388 /* An OR of *one* alternative - should not happen now. */
4389 (OP(first) == BRANCH && OP(first_next) != BRANCH) ||
4390 /* for now we can't handle lookbehind IFMATCH*/
4391 (OP(first) == IFMATCH && !first->flags) ||
4392 (OP(first) == PLUS) ||
4393 (OP(first) == MINMOD) ||
4394 /* An {n,m} with n>0 */
4395 (PL_regkind[OP(first)] == CURLY && ARG1(first) > 0) ||
4396 (OP(first) == NOTHING && PL_regkind[OP(first_next)] != END ))
4399 if (OP(first) == PLUS)
4402 first += regarglen[OP(first)];
4403 if (OP(first) == IFMATCH) {
4404 first = NEXTOPER(first);
4405 first += EXTRA_STEP_2ARGS;
4406 } else /* XXX possible optimisation for /(?=)/ */
4407 first = NEXTOPER(first);
4408 first_next= regnext(first);
4411 /* Starting-point info. */
4413 DEBUG_PEEP("first:",first,0);
4414 /* Ignore EXACT as we deal with it later. */
4415 if (PL_regkind[OP(first)] == EXACT) {
4416 if (OP(first) == EXACT)
4417 NOOP; /* Empty, get anchored substr later. */
4418 else if ((OP(first) == EXACTF || OP(first) == EXACTFL))
4419 ri->regstclass = first;
4422 else if (PL_regkind[OP(first)] == TRIE &&
4423 ((reg_trie_data *)ri->data->data[ ARG(first) ])->minlen>0)
4426 /* this can happen only on restudy */
4427 if ( OP(first) == TRIE ) {
4428 struct regnode_1 *trieop = (struct regnode_1 *)
4429 PerlMemShared_calloc(1, sizeof(struct regnode_1));
4430 StructCopy(first,trieop,struct regnode_1);
4431 trie_op=(regnode *)trieop;
4433 struct regnode_charclass *trieop = (struct regnode_charclass *)
4434 PerlMemShared_calloc(1, sizeof(struct regnode_charclass));
4435 StructCopy(first,trieop,struct regnode_charclass);
4436 trie_op=(regnode *)trieop;
4439 make_trie_failtable(pRExC_state, (regnode *)first, trie_op, 0);
4440 ri->regstclass = trie_op;
4443 else if (strchr((const char*)PL_simple,OP(first)))
4444 ri->regstclass = first;
4445 else if (PL_regkind[OP(first)] == BOUND ||
4446 PL_regkind[OP(first)] == NBOUND)
4447 ri->regstclass = first;
4448 else if (PL_regkind[OP(first)] == BOL) {
4449 r->extflags |= (OP(first) == MBOL
4451 : (OP(first) == SBOL
4454 first = NEXTOPER(first);
4457 else if (OP(first) == GPOS) {
4458 r->extflags |= RXf_ANCH_GPOS;
4459 first = NEXTOPER(first);
4462 else if ((!sawopen || !RExC_sawback) &&
4463 (OP(first) == STAR &&
4464 PL_regkind[OP(NEXTOPER(first))] == REG_ANY) &&
4465 !(r->extflags & RXf_ANCH) && !(RExC_seen & REG_SEEN_EVAL))
4467 /* turn .* into ^.* with an implied $*=1 */
4469 (OP(NEXTOPER(first)) == REG_ANY)
4472 r->extflags |= type;
4473 r->intflags |= PREGf_IMPLICIT;
4474 first = NEXTOPER(first);
4477 if (sawplus && (!sawopen || !RExC_sawback)
4478 && !(RExC_seen & REG_SEEN_EVAL)) /* May examine pos and $& */
4479 /* x+ must match at the 1st pos of run of x's */
4480 r->intflags |= PREGf_SKIP;
4482 /* Scan is after the zeroth branch, first is atomic matcher. */
4483 #ifdef TRIE_STUDY_OPT
4486 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4487 (IV)(first - scan + 1))
4491 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4492 (IV)(first - scan + 1))
4498 * If there's something expensive in the r.e., find the
4499 * longest literal string that must appear and make it the
4500 * regmust. Resolve ties in favor of later strings, since
4501 * the regstart check works with the beginning of the r.e.
4502 * and avoiding duplication strengthens checking. Not a
4503 * strong reason, but sufficient in the absence of others.
4504 * [Now we resolve ties in favor of the earlier string if
4505 * it happens that c_offset_min has been invalidated, since the
4506 * earlier string may buy us something the later one won't.]
4509 data.longest_fixed = newSVpvs("");
4510 data.longest_float = newSVpvs("");
4511 data.last_found = newSVpvs("");
4512 data.longest = &(data.longest_fixed);
4514 if (!ri->regstclass) {
4515 cl_init(pRExC_state, &ch_class);
4516 data.start_class = &ch_class;
4517 stclass_flag = SCF_DO_STCLASS_AND;
4518 } else /* XXXX Check for BOUND? */
4520 data.last_closep = &last_close;
4522 minlen = study_chunk(pRExC_state, &first, &minlen, &fake, scan + RExC_size, /* Up to end */
4523 &data, -1, NULL, NULL,
4524 SCF_DO_SUBSTR | SCF_WHILEM_VISITED_POS | stclass_flag,0);
4530 if ( RExC_npar == 1 && data.longest == &(data.longest_fixed)
4531 && data.last_start_min == 0 && data.last_end > 0
4532 && !RExC_seen_zerolen
4533 && !(RExC_seen & REG_SEEN_VERBARG)
4534 && (!(RExC_seen & REG_SEEN_GPOS) || (r->extflags & RXf_ANCH_GPOS)))
4535 r->extflags |= RXf_CHECK_ALL;
4536 scan_commit(pRExC_state, &data,&minlen,0);
4537 SvREFCNT_dec(data.last_found);
4539 /* Note that code very similar to this but for anchored string
4540 follows immediately below, changes may need to be made to both.
4543 longest_float_length = CHR_SVLEN(data.longest_float);
4544 if (longest_float_length
4545 || (data.flags & SF_FL_BEFORE_EOL
4546 && (!(data.flags & SF_FL_BEFORE_MEOL)
4547 || (RExC_flags & RXf_PMf_MULTILINE))))
4551 if (SvCUR(data.longest_fixed) /* ok to leave SvCUR */
4552 && data.offset_fixed == data.offset_float_min
4553 && SvCUR(data.longest_fixed) == SvCUR(data.longest_float))
4554 goto remove_float; /* As in (a)+. */
4556 /* copy the information about the longest float from the reg_scan_data
4557 over to the program. */
4558 if (SvUTF8(data.longest_float)) {
4559 r->float_utf8 = data.longest_float;
4560 r->float_substr = NULL;
4562 r->float_substr = data.longest_float;
4563 r->float_utf8 = NULL;
4565 /* float_end_shift is how many chars that must be matched that
4566 follow this item. We calculate it ahead of time as once the
4567 lookbehind offset is added in we lose the ability to correctly
4569 ml = data.minlen_float ? *(data.minlen_float)
4570 : (I32)longest_float_length;
4571 r->float_end_shift = ml - data.offset_float_min
4572 - longest_float_length + (SvTAIL(data.longest_float) != 0)
4573 + data.lookbehind_float;
4574 r->float_min_offset = data.offset_float_min - data.lookbehind_float;
4575 r->float_max_offset = data.offset_float_max;
4576 if (data.offset_float_max < I32_MAX) /* Don't offset infinity */
4577 r->float_max_offset -= data.lookbehind_float;
4579 t = (data.flags & SF_FL_BEFORE_EOL /* Can't have SEOL and MULTI */
4580 && (!(data.flags & SF_FL_BEFORE_MEOL)
4581 || (RExC_flags & RXf_PMf_MULTILINE)));
4582 fbm_compile(data.longest_float, t ? FBMcf_TAIL : 0);
4586 r->float_substr = r->float_utf8 = NULL;
4587 SvREFCNT_dec(data.longest_float);
4588 longest_float_length = 0;
4591 /* Note that code very similar to this but for floating string
4592 is immediately above, changes may need to be made to both.
4595 longest_fixed_length = CHR_SVLEN(data.longest_fixed);
4596 if (longest_fixed_length
4597 || (data.flags & SF_FIX_BEFORE_EOL /* Cannot have SEOL and MULTI */
4598 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4599 || (RExC_flags & RXf_PMf_MULTILINE))))
4603 /* copy the information about the longest fixed
4604 from the reg_scan_data over to the program. */
4605 if (SvUTF8(data.longest_fixed)) {
4606 r->anchored_utf8 = data.longest_fixed;
4607 r->anchored_substr = NULL;
4609 r->anchored_substr = data.longest_fixed;
4610 r->anchored_utf8 = NULL;
4612 /* fixed_end_shift is how many chars that must be matched that
4613 follow this item. We calculate it ahead of time as once the
4614 lookbehind offset is added in we lose the ability to correctly
4616 ml = data.minlen_fixed ? *(data.minlen_fixed)
4617 : (I32)longest_fixed_length;
4618 r->anchored_end_shift = ml - data.offset_fixed
4619 - longest_fixed_length + (SvTAIL(data.longest_fixed) != 0)
4620 + data.lookbehind_fixed;
4621 r->anchored_offset = data.offset_fixed - data.lookbehind_fixed;
4623 t = (data.flags & SF_FIX_BEFORE_EOL /* Can't have SEOL and MULTI */
4624 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4625 || (RExC_flags & RXf_PMf_MULTILINE)));
4626 fbm_compile(data.longest_fixed, t ? FBMcf_TAIL : 0);
4629 r->anchored_substr = r->anchored_utf8 = NULL;
4630 SvREFCNT_dec(data.longest_fixed);
4631 longest_fixed_length = 0;
4634 && (OP(ri->regstclass) == REG_ANY || OP(ri->regstclass) == SANY))
4635 ri->regstclass = NULL;
4636 if ((!(r->anchored_substr || r->anchored_utf8) || r->anchored_offset)
4638 && !(data.start_class->flags & ANYOF_EOS)
4639 && !cl_is_anything(data.start_class))
4641 const U32 n = add_data(pRExC_state, 1, "f");
4643 Newx(RExC_rxi->data->data[n], 1,
4644 struct regnode_charclass_class);
4645 StructCopy(data.start_class,
4646 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4647 struct regnode_charclass_class);
4648 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4649 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4650 DEBUG_COMPILE_r({ SV *sv = sv_newmortal();
4651 regprop(r, sv, (regnode*)data.start_class);
4652 PerlIO_printf(Perl_debug_log,
4653 "synthetic stclass \"%s\".\n",
4654 SvPVX_const(sv));});
4657 /* A temporary algorithm prefers floated substr to fixed one to dig more info. */
4658 if (longest_fixed_length > longest_float_length) {
4659 r->check_end_shift = r->anchored_end_shift;
4660 r->check_substr = r->anchored_substr;
4661 r->check_utf8 = r->anchored_utf8;
4662 r->check_offset_min = r->check_offset_max = r->anchored_offset;
4663 if (r->extflags & RXf_ANCH_SINGLE)
4664 r->extflags |= RXf_NOSCAN;
4667 r->check_end_shift = r->float_end_shift;
4668 r->check_substr = r->float_substr;
4669 r->check_utf8 = r->float_utf8;
4670 r->check_offset_min = r->float_min_offset;
4671 r->check_offset_max = r->float_max_offset;
4673 /* XXXX Currently intuiting is not compatible with ANCH_GPOS.
4674 This should be changed ASAP! */
4675 if ((r->check_substr || r->check_utf8) && !(r->extflags & RXf_ANCH_GPOS)) {
4676 r->extflags |= RXf_USE_INTUIT;
4677 if (SvTAIL(r->check_substr ? r->check_substr : r->check_utf8))
4678 r->extflags |= RXf_INTUIT_TAIL;
4680 /* XXX Unneeded? dmq (shouldn't as this is handled elsewhere)
4681 if ( (STRLEN)minlen < longest_float_length )
4682 minlen= longest_float_length;
4683 if ( (STRLEN)minlen < longest_fixed_length )
4684 minlen= longest_fixed_length;
4688 /* Several toplevels. Best we can is to set minlen. */
4690 struct regnode_charclass_class ch_class;
4693 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "\nMulti Top Level\n"));
4695 scan = ri->program + 1;
4696 cl_init(pRExC_state, &ch_class);
4697 data.start_class = &ch_class;
4698 data.last_closep = &last_close;
4701 minlen = study_chunk(pRExC_state, &scan, &minlen, &fake, scan + RExC_size,
4702 &data, -1, NULL, NULL, SCF_DO_STCLASS_AND|SCF_WHILEM_VISITED_POS,0);
4706 r->check_substr = r->check_utf8 = r->anchored_substr = r->anchored_utf8
4707 = r->float_substr = r->float_utf8 = NULL;
4708 if (!(data.start_class->flags & ANYOF_EOS)
4709 && !cl_is_anything(data.start_class))
4711 const U32 n = add_data(pRExC_state, 1, "f");
4713 Newx(RExC_rxi->data->data[n], 1,
4714 struct regnode_charclass_class);
4715 StructCopy(data.start_class,
4716 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4717 struct regnode_charclass_class);
4718 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4719 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4720 DEBUG_COMPILE_r({ SV* sv = sv_newmortal();
4721 regprop(r, sv, (regnode*)data.start_class);
4722 PerlIO_printf(Perl_debug_log,
4723 "synthetic stclass \"%s\".\n",
4724 SvPVX_const(sv));});
4728 /* Guard against an embedded (?=) or (?<=) with a longer minlen than
4729 the "real" pattern. */
4731 PerlIO_printf(Perl_debug_log,"minlen: %"IVdf" r->minlen:%"IVdf"\n",
4732 (IV)minlen, (IV)r->minlen);
4734 r->minlenret = minlen;
4735 if (r->minlen < minlen)
4738 if (RExC_seen & REG_SEEN_GPOS)
4739 r->extflags |= RXf_GPOS_SEEN;
4740 if (RExC_seen & REG_SEEN_LOOKBEHIND)
4741 r->extflags |= RXf_LOOKBEHIND_SEEN;
4742 if (RExC_seen & REG_SEEN_EVAL)
4743 r->extflags |= RXf_EVAL_SEEN;
4744 if (RExC_seen & REG_SEEN_CANY)
4745 r->extflags |= RXf_CANY_SEEN;
4746 if (RExC_seen & REG_SEEN_VERBARG)
4747 r->intflags |= PREGf_VERBARG_SEEN;
4748 if (RExC_seen & REG_SEEN_CUTGROUP)
4749 r->intflags |= PREGf_CUTGROUP_SEEN;
4750 if (RExC_paren_names)
4751 r->paren_names = (HV*)SvREFCNT_inc(RExC_paren_names);
4753 r->paren_names = NULL;
4754 if (r->prelen == 3 && strnEQ("\\s+", r->precomp, 3)) /* precomp = "\\s+)" */
4755 r->extflags |= RXf_WHITE;
4756 else if (r->prelen == 1 && r->precomp[0] == '^')
4757 r->extflags |= RXf_START_ONLY;
4760 if (RExC_paren_names) {
4761 ri->name_list_idx = add_data( pRExC_state, 1, "p" );
4762 ri->data->data[ri->name_list_idx] = (void*)SvREFCNT_inc(RExC_paren_name_list);
4765 ri->name_list_idx = 0;
4767 if (RExC_recurse_count) {
4768 for ( ; RExC_recurse_count ; RExC_recurse_count-- ) {
4769 const regnode *scan = RExC_recurse[RExC_recurse_count-1];
4770 ARG2L_SET( scan, RExC_open_parens[ARG(scan)-1] - scan );
4773 Newxz(r->offs, RExC_npar, regexp_paren_pair);
4774 /* assume we don't need to swap parens around before we match */
4777 PerlIO_printf(Perl_debug_log,"Final program:\n");
4780 #ifdef RE_TRACK_PATTERN_OFFSETS
4781 DEBUG_OFFSETS_r(if (ri->u.offsets) {
4782 const U32 len = ri->u.offsets[0];
4784 GET_RE_DEBUG_FLAGS_DECL;
4785 PerlIO_printf(Perl_debug_log, "Offsets: [%"UVuf"]\n\t", (UV)ri->u.offsets[0]);
4786 for (i = 1; i <= len; i++) {
4787 if (ri->u.offsets[i*2-1] || ri->u.offsets[i*2])
4788 PerlIO_printf(Perl_debug_log, "%"UVuf":%"UVuf"[%"UVuf"] ",
4789 (UV)i, (UV)ri->u.offsets[i*2-1], (UV)ri->u.offsets[i*2]);
4791 PerlIO_printf(Perl_debug_log, "\n");
4797 #undef RE_ENGINE_PTR
4801 Perl_reg_named_buff_fetch(pTHX_ REGEXP * const rx, SV * const namesv, const U32 flags)
4803 AV *retarray = NULL;
4808 if (rx && rx->paren_names) {
4809 HE *he_str = hv_fetch_ent( rx->paren_names, namesv, 0, 0 );
4812 SV* sv_dat=HeVAL(he_str);
4813 I32 *nums=(I32*)SvPVX(sv_dat);
4814 for ( i=0; i<SvIVX(sv_dat); i++ ) {
4815 if ((I32)(rx->nparens) >= nums[i]
4816 && rx->offs[nums[i]].start != -1
4817 && rx->offs[nums[i]].end != -1)
4820 CALLREG_NUMBUF_FETCH(rx,nums[i],ret);
4824 ret = newSVsv(&PL_sv_undef);
4827 SvREFCNT_inc_simple_void(ret);
4828 av_push(retarray, ret);
4832 return (SV*)retarray;
4839 Perl_reg_numbered_buff_fetch(pTHX_ REGEXP * const rx, const I32 paren, SV * const sv)
4846 sv_setsv(sv,&PL_sv_undef);
4850 if (paren == -2 && rx->offs[0].start != -1) {
4852 i = rx->offs[0].start;
4856 if (paren == -1 && rx->offs[0].end != -1) {
4858 s = rx->subbeg + rx->offs[0].end;
4859 i = rx->sublen - rx->offs[0].end;
4862 if ( 0 <= paren && paren <= (I32)rx->nparens &&
4863 (s1 = rx->offs[paren].start) != -1 &&
4864 (t1 = rx->offs[paren].end) != -1)
4868 s = rx->subbeg + s1;
4870 sv_setsv(sv,&PL_sv_undef);
4873 assert(rx->sublen >= (s - rx->subbeg) + i );
4875 const int oldtainted = PL_tainted;
4877 sv_setpvn(sv, s, i);
4878 PL_tainted = oldtainted;
4879 if ( (rx->extflags & RXf_CANY_SEEN)
4880 ? (RX_MATCH_UTF8(rx)
4881 && (!i || is_utf8_string((U8*)s, i)))
4882 : (RX_MATCH_UTF8(rx)) )
4889 if (RX_MATCH_TAINTED(rx)) {
4890 if (SvTYPE(sv) >= SVt_PVMG) {
4891 MAGIC* const mg = SvMAGIC(sv);
4894 SvMAGIC_set(sv, mg->mg_moremagic);
4896 if ((mgt = SvMAGIC(sv))) {
4897 mg->mg_moremagic = mgt;
4898 SvMAGIC_set(sv, mg);
4908 sv_setsv(sv,&PL_sv_undef);
4914 Perl_reg_numbered_buff_store(pTHX_ REGEXP * const rx, const I32 paren,
4915 SV const * const value)
4917 PERL_UNUSED_ARG(rx);
4918 PERL_UNUSED_ARG(paren);
4919 PERL_UNUSED_ARG(value);
4922 Perl_croak(aTHX_ PL_no_modify);
4926 Perl_reg_numbered_buff_length(pTHX_ REGEXP * const rx, const SV * const sv,
4932 /* Some of this code was originally in C<Perl_magic_len> in F<mg.c> */
4935 if (rx->offs[0].start != -1) {
4936 i = rx->offs[0].start;
4945 if (rx->offs[0].end != -1) {
4946 i = rx->sublen - rx->offs[0].end;
4948 s1 = rx->offs[0].end;
4954 default: /* $&, $1, $2, ... */
4955 if (paren <= (I32)rx->nparens &&
4956 (s1 = rx->offs[paren].start) != -1 &&
4957 (t1 = rx->offs[paren].end) != -1)
4962 if (ckWARN(WARN_UNINITIALIZED))
4963 report_uninit((SV*)sv);
4968 if (i > 0 && RX_MATCH_UTF8(rx)) {
4969 const char * const s = rx->subbeg + s1;
4974 if (is_utf8_string_loclen((U8*)s, i, &ep, &el))
4981 Perl_reg_qr_package(pTHX_ REGEXP * const rx)
4983 PERL_UNUSED_ARG(rx);
4984 return newSVpvs("Regexp");
4987 /* Scans the name of a named buffer from the pattern.
4988 * If flags is REG_RSN_RETURN_NULL returns null.
4989 * If flags is REG_RSN_RETURN_NAME returns an SV* containing the name
4990 * If flags is REG_RSN_RETURN_DATA returns the data SV* corresponding
4991 * to the parsed name as looked up in the RExC_paren_names hash.
4992 * If there is an error throws a vFAIL().. type exception.
4995 #define REG_RSN_RETURN_NULL 0
4996 #define REG_RSN_RETURN_NAME 1
4997 #define REG_RSN_RETURN_DATA 2
5000 S_reg_scan_name(pTHX_ RExC_state_t *pRExC_state, U32 flags) {
5001 char *name_start = RExC_parse;
5003 if (isIDFIRST_lazy_if(RExC_parse, UTF)) {
5004 /* skip IDFIRST by using do...while */
5007 RExC_parse += UTF8SKIP(RExC_parse);
5008 } while (isALNUM_utf8((U8*)RExC_parse));
5012 } while (isALNUM(*RExC_parse));
5016 SV* sv_name = sv_2mortal(Perl_newSVpvn(aTHX_ name_start,
5017 (int)(RExC_parse - name_start)));
5020 if ( flags == REG_RSN_RETURN_NAME)
5022 else if (flags==REG_RSN_RETURN_DATA) {
5025 if ( ! sv_name ) /* should not happen*/
5026 Perl_croak(aTHX_ "panic: no svname in reg_scan_name");
5027 if (RExC_paren_names)
5028 he_str = hv_fetch_ent( RExC_paren_names, sv_name, 0, 0 );
5030 sv_dat = HeVAL(he_str);
5032 vFAIL("Reference to nonexistent named group");
5036 Perl_croak(aTHX_ "panic: bad flag in reg_scan_name");
5043 #define DEBUG_PARSE_MSG(funcname) DEBUG_PARSE_r({ \
5044 int rem=(int)(RExC_end - RExC_parse); \
5053 if (RExC_lastparse!=RExC_parse) \
5054 PerlIO_printf(Perl_debug_log," >%.*s%-*s", \
5057 iscut ? "..." : "<" \
5060 PerlIO_printf(Perl_debug_log,"%16s",""); \
5063 num = RExC_size + 1; \
5065 num=REG_NODE_NUM(RExC_emit); \
5066 if (RExC_lastnum!=num) \
5067 PerlIO_printf(Perl_debug_log,"|%4d",num); \
5069 PerlIO_printf(Perl_debug_log,"|%4s",""); \
5070 PerlIO_printf(Perl_debug_log,"|%*s%-4s", \
5071 (int)((depth*2)), "", \
5075 RExC_lastparse=RExC_parse; \
5080 #define DEBUG_PARSE(funcname) DEBUG_PARSE_r({ \
5081 DEBUG_PARSE_MSG((funcname)); \
5082 PerlIO_printf(Perl_debug_log,"%4s","\n"); \
5084 #define DEBUG_PARSE_FMT(funcname,fmt,args) DEBUG_PARSE_r({ \
5085 DEBUG_PARSE_MSG((funcname)); \
5086 PerlIO_printf(Perl_debug_log,fmt "\n",args); \
5089 - reg - regular expression, i.e. main body or parenthesized thing
5091 * Caller must absorb opening parenthesis.
5093 * Combining parenthesis handling with the base level of regular expression
5094 * is a trifle forced, but the need to tie the tails of the branches to what
5095 * follows makes it hard to avoid.
5097 #define REGTAIL(x,y,z) regtail((x),(y),(z),depth+1)
5099 #define REGTAIL_STUDY(x,y,z) regtail_study((x),(y),(z),depth+1)
5101 #define REGTAIL_STUDY(x,y,z) regtail((x),(y),(z),depth+1)
5105 S_reg(pTHX_ RExC_state_t *pRExC_state, I32 paren, I32 *flagp,U32 depth)
5106 /* paren: Parenthesized? 0=top, 1=(, inside: changed to letter. */
5109 register regnode *ret; /* Will be the head of the group. */
5110 register regnode *br;
5111 register regnode *lastbr;
5112 register regnode *ender = NULL;
5113 register I32 parno = 0;
5115 const I32 oregflags = RExC_flags;
5116 bool have_branch = 0;
5118 I32 freeze_paren = 0;
5119 I32 after_freeze = 0;
5121 /* for (?g), (?gc), and (?o) warnings; warning
5122 about (?c) will warn about (?g) -- japhy */
5124 #define WASTED_O 0x01
5125 #define WASTED_G 0x02
5126 #define WASTED_C 0x04
5127 #define WASTED_GC (0x02|0x04)
5128 I32 wastedflags = 0x00;
5130 char * parse_start = RExC_parse; /* MJD */
5131 char * const oregcomp_parse = RExC_parse;
5133 GET_RE_DEBUG_FLAGS_DECL;
5134 DEBUG_PARSE("reg ");
5136 *flagp = 0; /* Tentatively. */
5139 /* Make an OPEN node, if parenthesized. */
5141 if ( *RExC_parse == '*') { /* (*VERB:ARG) */
5142 char *start_verb = RExC_parse;
5143 STRLEN verb_len = 0;
5144 char *start_arg = NULL;
5145 unsigned char op = 0;
5147 int internal_argval = 0; /* internal_argval is only useful if !argok */
5148 while ( *RExC_parse && *RExC_parse != ')' ) {
5149 if ( *RExC_parse == ':' ) {
5150 start_arg = RExC_parse + 1;
5156 verb_len = RExC_parse - start_verb;
5159 while ( *RExC_parse && *RExC_parse != ')' )
5161 if ( *RExC_parse != ')' )
5162 vFAIL("Unterminated verb pattern argument");
5163 if ( RExC_parse == start_arg )
5166 if ( *RExC_parse != ')' )
5167 vFAIL("Unterminated verb pattern");
5170 switch ( *start_verb ) {
5171 case 'A': /* (*ACCEPT) */
5172 if ( memEQs(start_verb,verb_len,"ACCEPT") ) {
5174 internal_argval = RExC_nestroot;
5177 case 'C': /* (*COMMIT) */
5178 if ( memEQs(start_verb,verb_len,"COMMIT") )
5181 case 'F': /* (*FAIL) */
5182 if ( verb_len==1 || memEQs(start_verb,verb_len,"FAIL") ) {
5187 case ':': /* (*:NAME) */
5188 case 'M': /* (*MARK:NAME) */
5189 if ( verb_len==0 || memEQs(start_verb,verb_len,"MARK") ) {
5194 case 'P': /* (*PRUNE) */
5195 if ( memEQs(start_verb,verb_len,"PRUNE") )
5198 case 'S': /* (*SKIP) */
5199 if ( memEQs(start_verb,verb_len,"SKIP") )
5202 case 'T': /* (*THEN) */
5203 /* [19:06] <TimToady> :: is then */
5204 if ( memEQs(start_verb,verb_len,"THEN") ) {
5206 RExC_seen |= REG_SEEN_CUTGROUP;
5212 vFAIL3("Unknown verb pattern '%.*s'",
5213 verb_len, start_verb);
5216 if ( start_arg && internal_argval ) {
5217 vFAIL3("Verb pattern '%.*s' may not have an argument",
5218 verb_len, start_verb);
5219 } else if ( argok < 0 && !start_arg ) {
5220 vFAIL3("Verb pattern '%.*s' has a mandatory argument",
5221 verb_len, start_verb);
5223 ret = reganode(pRExC_state, op, internal_argval);
5224 if ( ! internal_argval && ! SIZE_ONLY ) {
5226 SV *sv = newSVpvn( start_arg, RExC_parse - start_arg);
5227 ARG(ret) = add_data( pRExC_state, 1, "S" );
5228 RExC_rxi->data->data[ARG(ret)]=(void*)sv;
5235 if (!internal_argval)
5236 RExC_seen |= REG_SEEN_VERBARG;
5237 } else if ( start_arg ) {
5238 vFAIL3("Verb pattern '%.*s' may not have an argument",
5239 verb_len, start_verb);
5241 ret = reg_node(pRExC_state, op);
5243 nextchar(pRExC_state);
5246 if (*RExC_parse == '?') { /* (?...) */
5247 bool is_logical = 0;
5248 const char * const seqstart = RExC_parse;
5251 paren = *RExC_parse++;
5252 ret = NULL; /* For look-ahead/behind. */
5255 case 'P': /* (?P...) variants for those used to PCRE/Python */
5256 paren = *RExC_parse++;
5257 if ( paren == '<') /* (?P<...>) named capture */
5259 else if (paren == '>') { /* (?P>name) named recursion */
5260 goto named_recursion;
5262 else if (paren == '=') { /* (?P=...) named backref */
5263 /* this pretty much dupes the code for \k<NAME> in regatom(), if
5264 you change this make sure you change that */
5265 char* name_start = RExC_parse;
5267 SV *sv_dat = reg_scan_name(pRExC_state,
5268 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5269 if (RExC_parse == name_start || *RExC_parse != ')')
5270 vFAIL2("Sequence %.3s... not terminated",parse_start);
5273 num = add_data( pRExC_state, 1, "S" );
5274 RExC_rxi->data->data[num]=(void*)sv_dat;
5275 SvREFCNT_inc_simple_void(sv_dat);
5278 ret = reganode(pRExC_state,
5279 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
5283 Set_Node_Offset(ret, parse_start+1);
5284 Set_Node_Cur_Length(ret); /* MJD */
5286 nextchar(pRExC_state);
5290 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5292 case '<': /* (?<...) */
5293 if (*RExC_parse == '!')
5295 else if (*RExC_parse != '=')
5301 case '\'': /* (?'...') */
5302 name_start= RExC_parse;
5303 svname = reg_scan_name(pRExC_state,
5304 SIZE_ONLY ? /* reverse test from the others */
5305 REG_RSN_RETURN_NAME :
5306 REG_RSN_RETURN_NULL);
5307 if (RExC_parse == name_start) {
5309 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5312 if (*RExC_parse != paren)
5313 vFAIL2("Sequence (?%c... not terminated",
5314 paren=='>' ? '<' : paren);
5318 if (!svname) /* shouldnt happen */
5320 "panic: reg_scan_name returned NULL");
5321 if (!RExC_paren_names) {
5322 RExC_paren_names= newHV();
5323 sv_2mortal((SV*)RExC_paren_names);
5325 RExC_paren_name_list= newAV();
5326 sv_2mortal((SV*)RExC_paren_name_list);
5329 he_str = hv_fetch_ent( RExC_paren_names, svname, 1, 0 );
5331 sv_dat = HeVAL(he_str);
5333 /* croak baby croak */
5335 "panic: paren_name hash element allocation failed");
5336 } else if ( SvPOK(sv_dat) ) {
5337 /* (?|...) can mean we have dupes so scan to check
5338 its already been stored. Maybe a flag indicating
5339 we are inside such a construct would be useful,
5340 but the arrays are likely to be quite small, so
5341 for now we punt -- dmq */
5342 IV count = SvIV(sv_dat);
5343 I32 *pv = (I32*)SvPVX(sv_dat);
5345 for ( i = 0 ; i < count ; i++ ) {
5346 if ( pv[i] == RExC_npar ) {
5352 pv = (I32*)SvGROW(sv_dat, SvCUR(sv_dat) + sizeof(I32)+1);
5353 SvCUR_set(sv_dat, SvCUR(sv_dat) + sizeof(I32));
5354 pv[count] = RExC_npar;
5358 (void)SvUPGRADE(sv_dat,SVt_PVNV);
5359 sv_setpvn(sv_dat, (char *)&(RExC_npar), sizeof(I32));
5364 if (!av_store(RExC_paren_name_list, RExC_npar, SvREFCNT_inc(svname)))
5365 SvREFCNT_dec(svname);
5368 /*sv_dump(sv_dat);*/
5370 nextchar(pRExC_state);
5372 goto capturing_parens;
5374 RExC_seen |= REG_SEEN_LOOKBEHIND;
5376 case '=': /* (?=...) */
5377 case '!': /* (?!...) */
5378 RExC_seen_zerolen++;
5379 if (*RExC_parse == ')') {
5380 ret=reg_node(pRExC_state, OPFAIL);
5381 nextchar(pRExC_state);
5385 case '|': /* (?|...) */
5386 /* branch reset, behave like a (?:...) except that
5387 buffers in alternations share the same numbers */
5389 after_freeze = freeze_paren = RExC_npar;
5391 case ':': /* (?:...) */
5392 case '>': /* (?>...) */
5394 case '$': /* (?$...) */
5395 case '@': /* (?@...) */
5396 vFAIL2("Sequence (?%c...) not implemented", (int)paren);
5398 case '#': /* (?#...) */
5399 while (*RExC_parse && *RExC_parse != ')')
5401 if (*RExC_parse != ')')
5402 FAIL("Sequence (?#... not terminated");
5403 nextchar(pRExC_state);
5406 case '0' : /* (?0) */
5407 case 'R' : /* (?R) */
5408 if (*RExC_parse != ')')
5409 FAIL("Sequence (?R) not terminated");
5410 ret = reg_node(pRExC_state, GOSTART);
5411 *flagp |= POSTPONED;
5412 nextchar(pRExC_state);
5415 { /* named and numeric backreferences */
5417 case '&': /* (?&NAME) */
5418 parse_start = RExC_parse - 1;
5421 SV *sv_dat = reg_scan_name(pRExC_state,
5422 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5423 num = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5425 goto gen_recurse_regop;
5428 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5430 vFAIL("Illegal pattern");
5432 goto parse_recursion;
5434 case '-': /* (?-1) */
5435 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5436 RExC_parse--; /* rewind to let it be handled later */
5440 case '1': case '2': case '3': case '4': /* (?1) */
5441 case '5': case '6': case '7': case '8': case '9':
5444 num = atoi(RExC_parse);
5445 parse_start = RExC_parse - 1; /* MJD */
5446 if (*RExC_parse == '-')
5448 while (isDIGIT(*RExC_parse))
5450 if (*RExC_parse!=')')
5451 vFAIL("Expecting close bracket");
5454 if ( paren == '-' ) {
5456 Diagram of capture buffer numbering.
5457 Top line is the normal capture buffer numbers
5458 Botton line is the negative indexing as from
5462 /(a(x)y)(a(b(c(?-2)d)e)f)(g(h))/
5466 num = RExC_npar + num;
5469 vFAIL("Reference to nonexistent group");
5471 } else if ( paren == '+' ) {
5472 num = RExC_npar + num - 1;
5475 ret = reganode(pRExC_state, GOSUB, num);
5477 if (num > (I32)RExC_rx->nparens) {
5479 vFAIL("Reference to nonexistent group");
5481 ARG2L_SET( ret, RExC_recurse_count++);
5483 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
5484 "Recurse #%"UVuf" to %"IVdf"\n", (UV)ARG(ret), (IV)ARG2L(ret)));
5488 RExC_seen |= REG_SEEN_RECURSE;
5489 Set_Node_Length(ret, 1 + regarglen[OP(ret)]); /* MJD */
5490 Set_Node_Offset(ret, parse_start); /* MJD */
5492 *flagp |= POSTPONED;
5493 nextchar(pRExC_state);
5495 } /* named and numeric backreferences */
5498 case '?': /* (??...) */
5500 if (*RExC_parse != '{') {
5502 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5505 *flagp |= POSTPONED;
5506 paren = *RExC_parse++;
5508 case '{': /* (?{...}) */
5513 char *s = RExC_parse;
5515 RExC_seen_zerolen++;
5516 RExC_seen |= REG_SEEN_EVAL;
5517 while (count && (c = *RExC_parse)) {
5528 if (*RExC_parse != ')') {
5530 vFAIL("Sequence (?{...}) not terminated or not {}-balanced");
5534 OP_4tree *sop, *rop;
5535 SV * const sv = newSVpvn(s, RExC_parse - 1 - s);
5538 Perl_save_re_context(aTHX);
5539 rop = sv_compile_2op(sv, &sop, "re", &pad);
5540 sop->op_private |= OPpREFCOUNTED;
5541 /* re_dup will OpREFCNT_inc */
5542 OpREFCNT_set(sop, 1);
5545 n = add_data(pRExC_state, 3, "nop");
5546 RExC_rxi->data->data[n] = (void*)rop;
5547 RExC_rxi->data->data[n+1] = (void*)sop;
5548 RExC_rxi->data->data[n+2] = (void*)pad;
5551 else { /* First pass */
5552 if (PL_reginterp_cnt < ++RExC_seen_evals
5554 /* No compiled RE interpolated, has runtime
5555 components ===> unsafe. */
5556 FAIL("Eval-group not allowed at runtime, use re 'eval'");
5557 if (PL_tainting && PL_tainted)
5558 FAIL("Eval-group in insecure regular expression");
5559 #if PERL_VERSION > 8
5560 if (IN_PERL_COMPILETIME)
5565 nextchar(pRExC_state);
5567 ret = reg_node(pRExC_state, LOGICAL);
5570 REGTAIL(pRExC_state, ret, reganode(pRExC_state, EVAL, n));
5571 /* deal with the length of this later - MJD */
5574 ret = reganode(pRExC_state, EVAL, n);
5575 Set_Node_Length(ret, RExC_parse - parse_start + 1);
5576 Set_Node_Offset(ret, parse_start);
5579 case '(': /* (?(?{...})...) and (?(?=...)...) */
5582 if (RExC_parse[0] == '?') { /* (?(?...)) */
5583 if (RExC_parse[1] == '=' || RExC_parse[1] == '!'
5584 || RExC_parse[1] == '<'
5585 || RExC_parse[1] == '{') { /* Lookahead or eval. */
5588 ret = reg_node(pRExC_state, LOGICAL);
5591 REGTAIL(pRExC_state, ret, reg(pRExC_state, 1, &flag,depth+1));
5595 else if ( RExC_parse[0] == '<' /* (?(<NAME>)...) */
5596 || RExC_parse[0] == '\'' ) /* (?('NAME')...) */
5598 char ch = RExC_parse[0] == '<' ? '>' : '\'';
5599 char *name_start= RExC_parse++;
5601 SV *sv_dat=reg_scan_name(pRExC_state,
5602 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5603 if (RExC_parse == name_start || *RExC_parse != ch)
5604 vFAIL2("Sequence (?(%c... not terminated",
5605 (ch == '>' ? '<' : ch));
5608 num = add_data( pRExC_state, 1, "S" );
5609 RExC_rxi->data->data[num]=(void*)sv_dat;
5610 SvREFCNT_inc_simple_void(sv_dat);
5612 ret = reganode(pRExC_state,NGROUPP,num);
5613 goto insert_if_check_paren;
5615 else if (RExC_parse[0] == 'D' &&
5616 RExC_parse[1] == 'E' &&
5617 RExC_parse[2] == 'F' &&
5618 RExC_parse[3] == 'I' &&
5619 RExC_parse[4] == 'N' &&
5620 RExC_parse[5] == 'E')
5622 ret = reganode(pRExC_state,DEFINEP,0);
5625 goto insert_if_check_paren;
5627 else if (RExC_parse[0] == 'R') {
5630 if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
5631 parno = atoi(RExC_parse++);
5632 while (isDIGIT(*RExC_parse))
5634 } else if (RExC_parse[0] == '&') {
5637 sv_dat = reg_scan_name(pRExC_state,
5638 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5639 parno = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5641 ret = reganode(pRExC_state,INSUBP,parno);
5642 goto insert_if_check_paren;
5644 else if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
5647 parno = atoi(RExC_parse++);
5649 while (isDIGIT(*RExC_parse))
5651 ret = reganode(pRExC_state, GROUPP, parno);
5653 insert_if_check_paren:
5654 if ((c = *nextchar(pRExC_state)) != ')')
5655 vFAIL("Switch condition not recognized");
5657 REGTAIL(pRExC_state, ret, reganode(pRExC_state, IFTHEN, 0));
5658 br = regbranch(pRExC_state, &flags, 1,depth+1);
5660 br = reganode(pRExC_state, LONGJMP, 0);
5662 REGTAIL(pRExC_state, br, reganode(pRExC_state, LONGJMP, 0));
5663 c = *nextchar(pRExC_state);
5668 vFAIL("(?(DEFINE)....) does not allow branches");
5669 lastbr = reganode(pRExC_state, IFTHEN, 0); /* Fake one for optimizer. */
5670 regbranch(pRExC_state, &flags, 1,depth+1);
5671 REGTAIL(pRExC_state, ret, lastbr);
5674 c = *nextchar(pRExC_state);
5679 vFAIL("Switch (?(condition)... contains too many branches");
5680 ender = reg_node(pRExC_state, TAIL);
5681 REGTAIL(pRExC_state, br, ender);
5683 REGTAIL(pRExC_state, lastbr, ender);
5684 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender);
5687 REGTAIL(pRExC_state, ret, ender);
5688 RExC_size++; /* XXX WHY do we need this?!!
5689 For large programs it seems to be required
5690 but I can't figure out why. -- dmq*/
5694 vFAIL2("Unknown switch condition (?(%.2s", RExC_parse);
5698 RExC_parse--; /* for vFAIL to print correctly */
5699 vFAIL("Sequence (? incomplete");
5703 parse_flags: /* (?i) */
5705 U32 posflags = 0, negflags = 0;
5706 U32 *flagsp = &posflags;
5708 while (*RExC_parse) {
5709 /* && strchr("iogcmsx", *RExC_parse) */
5710 /* (?g), (?gc) and (?o) are useless here
5711 and must be globally applied -- japhy */
5712 switch (*RExC_parse) {
5713 CASE_STD_PMMOD_FLAGS_PARSE_SET(flagsp);
5716 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
5717 const I32 wflagbit = *RExC_parse == 'o' ? WASTED_O : WASTED_G;
5718 if (! (wastedflags & wflagbit) ) {
5719 wastedflags |= wflagbit;
5722 "Useless (%s%c) - %suse /%c modifier",
5723 flagsp == &negflags ? "?-" : "?",
5725 flagsp == &negflags ? "don't " : "",
5733 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
5734 if (! (wastedflags & WASTED_C) ) {
5735 wastedflags |= WASTED_GC;
5738 "Useless (%sc) - %suse /gc modifier",
5739 flagsp == &negflags ? "?-" : "?",
5740 flagsp == &negflags ? "don't " : ""
5746 if (flagsp == &negflags) {
5747 if (SIZE_ONLY && ckWARN(WARN_REGEXP))
5748 vWARN(RExC_parse + 1,"Useless use of (?-k)");
5750 *flagsp |= RXf_PMf_KEEPCOPY;
5754 if (flagsp == &negflags) {
5756 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5760 wastedflags = 0; /* reset so (?g-c) warns twice */
5766 RExC_flags |= posflags;
5767 RExC_flags &= ~negflags;
5768 nextchar(pRExC_state);
5779 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5784 }} /* one for the default block, one for the switch */
5791 ret = reganode(pRExC_state, OPEN, parno);
5794 RExC_nestroot = parno;
5795 if (RExC_seen & REG_SEEN_RECURSE
5796 && !RExC_open_parens[parno-1])
5798 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
5799 "Setting open paren #%"IVdf" to %d\n",
5800 (IV)parno, REG_NODE_NUM(ret)));
5801 RExC_open_parens[parno-1]= ret;
5804 Set_Node_Length(ret, 1); /* MJD */
5805 Set_Node_Offset(ret, RExC_parse); /* MJD */
5813 /* Pick up the branches, linking them together. */
5814 parse_start = RExC_parse; /* MJD */
5815 br = regbranch(pRExC_state, &flags, 1,depth+1);
5816 /* branch_len = (paren != 0); */
5820 if (*RExC_parse == '|') {
5821 if (!SIZE_ONLY && RExC_extralen) {
5822 reginsert(pRExC_state, BRANCHJ, br, depth+1);
5825 reginsert(pRExC_state, BRANCH, br, depth+1);
5826 Set_Node_Length(br, paren != 0);
5827 Set_Node_Offset_To_R(br-RExC_emit_start, parse_start-RExC_start);
5831 RExC_extralen += 1; /* For BRANCHJ-BRANCH. */
5833 else if (paren == ':') {
5834 *flagp |= flags&SIMPLE;
5836 if (is_open) { /* Starts with OPEN. */
5837 REGTAIL(pRExC_state, ret, br); /* OPEN -> first. */
5839 else if (paren != '?') /* Not Conditional */
5841 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
5843 while (*RExC_parse == '|') {
5844 if (!SIZE_ONLY && RExC_extralen) {
5845 ender = reganode(pRExC_state, LONGJMP,0);
5846 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender); /* Append to the previous. */
5849 RExC_extralen += 2; /* Account for LONGJMP. */
5850 nextchar(pRExC_state);
5852 if (RExC_npar > after_freeze)
5853 after_freeze = RExC_npar;
5854 RExC_npar = freeze_paren;
5856 br = regbranch(pRExC_state, &flags, 0, depth+1);
5860 REGTAIL(pRExC_state, lastbr, br); /* BRANCH -> BRANCH. */
5862 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
5865 if (have_branch || paren != ':') {
5866 /* Make a closing node, and hook it on the end. */
5869 ender = reg_node(pRExC_state, TAIL);
5872 ender = reganode(pRExC_state, CLOSE, parno);
5873 if (!SIZE_ONLY && RExC_seen & REG_SEEN_RECURSE) {
5874 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
5875 "Setting close paren #%"IVdf" to %d\n",
5876 (IV)parno, REG_NODE_NUM(ender)));
5877 RExC_close_parens[parno-1]= ender;
5878 if (RExC_nestroot == parno)
5881 Set_Node_Offset(ender,RExC_parse+1); /* MJD */
5882 Set_Node_Length(ender,1); /* MJD */
5888 *flagp &= ~HASWIDTH;
5891 ender = reg_node(pRExC_state, SUCCEED);
5894 ender = reg_node(pRExC_state, END);
5896 assert(!RExC_opend); /* there can only be one! */
5901 REGTAIL(pRExC_state, lastbr, ender);
5903 if (have_branch && !SIZE_ONLY) {
5905 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
5907 /* Hook the tails of the branches to the closing node. */
5908 for (br = ret; br; br = regnext(br)) {
5909 const U8 op = PL_regkind[OP(br)];
5911 REGTAIL_STUDY(pRExC_state, NEXTOPER(br), ender);
5913 else if (op == BRANCHJ) {
5914 REGTAIL_STUDY(pRExC_state, NEXTOPER(NEXTOPER(br)), ender);
5922 static const char parens[] = "=!<,>";
5924 if (paren && (p = strchr(parens, paren))) {
5925 U8 node = ((p - parens) % 2) ? UNLESSM : IFMATCH;
5926 int flag = (p - parens) > 1;
5929 node = SUSPEND, flag = 0;
5930 reginsert(pRExC_state, node,ret, depth+1);
5931 Set_Node_Cur_Length(ret);
5932 Set_Node_Offset(ret, parse_start + 1);
5934 REGTAIL_STUDY(pRExC_state, ret, reg_node(pRExC_state, TAIL));
5938 /* Check for proper termination. */
5940 RExC_flags = oregflags;
5941 if (RExC_parse >= RExC_end || *nextchar(pRExC_state) != ')') {
5942 RExC_parse = oregcomp_parse;
5943 vFAIL("Unmatched (");
5946 else if (!paren && RExC_parse < RExC_end) {
5947 if (*RExC_parse == ')') {
5949 vFAIL("Unmatched )");
5952 FAIL("Junk on end of regexp"); /* "Can't happen". */
5956 RExC_npar = after_freeze;
5961 - regbranch - one alternative of an | operator
5963 * Implements the concatenation operator.
5966 S_regbranch(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, I32 first, U32 depth)
5969 register regnode *ret;
5970 register regnode *chain = NULL;
5971 register regnode *latest;
5972 I32 flags = 0, c = 0;
5973 GET_RE_DEBUG_FLAGS_DECL;
5974 DEBUG_PARSE("brnc");
5979 if (!SIZE_ONLY && RExC_extralen)
5980 ret = reganode(pRExC_state, BRANCHJ,0);
5982 ret = reg_node(pRExC_state, BRANCH);
5983 Set_Node_Length(ret, 1);
5987 if (!first && SIZE_ONLY)
5988 RExC_extralen += 1; /* BRANCHJ */
5990 *flagp = WORST; /* Tentatively. */
5993 nextchar(pRExC_state);
5994 while (RExC_parse < RExC_end && *RExC_parse != '|' && *RExC_parse != ')') {
5996 latest = regpiece(pRExC_state, &flags,depth+1);
5997 if (latest == NULL) {
5998 if (flags & TRYAGAIN)
6002 else if (ret == NULL)
6004 *flagp |= flags&(HASWIDTH|POSTPONED);
6005 if (chain == NULL) /* First piece. */
6006 *flagp |= flags&SPSTART;
6009 REGTAIL(pRExC_state, chain, latest);
6014 if (chain == NULL) { /* Loop ran zero times. */
6015 chain = reg_node(pRExC_state, NOTHING);
6020 *flagp |= flags&SIMPLE;
6027 - regpiece - something followed by possible [*+?]
6029 * Note that the branching code sequences used for ? and the general cases
6030 * of * and + are somewhat optimized: they use the same NOTHING node as
6031 * both the endmarker for their branch list and the body of the last branch.
6032 * It might seem that this node could be dispensed with entirely, but the
6033 * endmarker role is not redundant.
6036 S_regpiece(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6039 register regnode *ret;
6041 register char *next;
6043 const char * const origparse = RExC_parse;
6045 I32 max = REG_INFTY;
6047 const char *maxpos = NULL;
6048 GET_RE_DEBUG_FLAGS_DECL;
6049 DEBUG_PARSE("piec");
6051 ret = regatom(pRExC_state, &flags,depth+1);
6053 if (flags & TRYAGAIN)
6060 if (op == '{' && regcurly(RExC_parse)) {
6062 parse_start = RExC_parse; /* MJD */
6063 next = RExC_parse + 1;
6064 while (isDIGIT(*next) || *next == ',') {
6073 if (*next == '}') { /* got one */
6077 min = atoi(RExC_parse);
6081 maxpos = RExC_parse;
6083 if (!max && *maxpos != '0')
6084 max = REG_INFTY; /* meaning "infinity" */
6085 else if (max >= REG_INFTY)
6086 vFAIL2("Quantifier in {,} bigger than %d", REG_INFTY - 1);
6088 nextchar(pRExC_state);
6091 if ((flags&SIMPLE)) {
6092 RExC_naughty += 2 + RExC_naughty / 2;
6093 reginsert(pRExC_state, CURLY, ret, depth+1);
6094 Set_Node_Offset(ret, parse_start+1); /* MJD */
6095 Set_Node_Cur_Length(ret);
6098 regnode * const w = reg_node(pRExC_state, WHILEM);
6101 REGTAIL(pRExC_state, ret, w);
6102 if (!SIZE_ONLY && RExC_extralen) {
6103 reginsert(pRExC_state, LONGJMP,ret, depth+1);
6104 reginsert(pRExC_state, NOTHING,ret, depth+1);
6105 NEXT_OFF(ret) = 3; /* Go over LONGJMP. */
6107 reginsert(pRExC_state, CURLYX,ret, depth+1);
6109 Set_Node_Offset(ret, parse_start+1);
6110 Set_Node_Length(ret,
6111 op == '{' ? (RExC_parse - parse_start) : 1);
6113 if (!SIZE_ONLY && RExC_extralen)
6114 NEXT_OFF(ret) = 3; /* Go over NOTHING to LONGJMP. */
6115 REGTAIL(pRExC_state, ret, reg_node(pRExC_state, NOTHING));
6117 RExC_whilem_seen++, RExC_extralen += 3;
6118 RExC_naughty += 4 + RExC_naughty; /* compound interest */
6126 if (max && max < min)
6127 vFAIL("Can't do {n,m} with n > m");
6129 ARG1_SET(ret, (U16)min);
6130 ARG2_SET(ret, (U16)max);
6142 #if 0 /* Now runtime fix should be reliable. */
6144 /* if this is reinstated, don't forget to put this back into perldiag:
6146 =item Regexp *+ operand could be empty at {#} in regex m/%s/
6148 (F) The part of the regexp subject to either the * or + quantifier
6149 could match an empty string. The {#} shows in the regular
6150 expression about where the problem was discovered.
6154 if (!(flags&HASWIDTH) && op != '?')
6155 vFAIL("Regexp *+ operand could be empty");
6158 parse_start = RExC_parse;
6159 nextchar(pRExC_state);
6161 *flagp = (op != '+') ? (WORST|SPSTART|HASWIDTH) : (WORST|HASWIDTH);
6163 if (op == '*' && (flags&SIMPLE)) {
6164 reginsert(pRExC_state, STAR, ret, depth+1);
6168 else if (op == '*') {
6172 else if (op == '+' && (flags&SIMPLE)) {
6173 reginsert(pRExC_state, PLUS, ret, depth+1);
6177 else if (op == '+') {
6181 else if (op == '?') {
6186 if (!SIZE_ONLY && !(flags&(HASWIDTH|POSTPONED)) && max > REG_INFTY/3 && ckWARN(WARN_REGEXP)) {
6188 "%.*s matches null string many times",
6189 (int)(RExC_parse >= origparse ? RExC_parse - origparse : 0),
6193 if (RExC_parse < RExC_end && *RExC_parse == '?') {
6194 nextchar(pRExC_state);
6195 reginsert(pRExC_state, MINMOD, ret, depth+1);
6196 REGTAIL(pRExC_state, ret, ret + NODE_STEP_REGNODE);
6198 #ifndef REG_ALLOW_MINMOD_SUSPEND
6201 if (RExC_parse < RExC_end && *RExC_parse == '+') {
6203 nextchar(pRExC_state);
6204 ender = reg_node(pRExC_state, SUCCEED);
6205 REGTAIL(pRExC_state, ret, ender);
6206 reginsert(pRExC_state, SUSPEND, ret, depth+1);
6208 ender = reg_node(pRExC_state, TAIL);
6209 REGTAIL(pRExC_state, ret, ender);
6213 if (RExC_parse < RExC_end && ISMULT2(RExC_parse)) {
6215 vFAIL("Nested quantifiers");
6222 /* reg_namedseq(pRExC_state,UVp)
6224 This is expected to be called by a parser routine that has
6225 recognized'\N' and needs to handle the rest. RExC_parse is
6226 expected to point at the first char following the N at the time
6229 If valuep is non-null then it is assumed that we are parsing inside
6230 of a charclass definition and the first codepoint in the resolved
6231 string is returned via *valuep and the routine will return NULL.
6232 In this mode if a multichar string is returned from the charnames
6233 handler a warning will be issued, and only the first char in the
6234 sequence will be examined. If the string returned is zero length
6235 then the value of *valuep is undefined and NON-NULL will
6236 be returned to indicate failure. (This will NOT be a valid pointer
6239 If value is null then it is assumed that we are parsing normal text
6240 and inserts a new EXACT node into the program containing the resolved
6241 string and returns a pointer to the new node. If the string is
6242 zerolength a NOTHING node is emitted.
6244 On success RExC_parse is set to the char following the endbrace.
6245 Parsing failures will generate a fatal errorvia vFAIL(...)
6247 NOTE: We cache all results from the charnames handler locally in
6248 the RExC_charnames hash (created on first use) to prevent a charnames
6249 handler from playing silly-buggers and returning a short string and
6250 then a long string for a given pattern. Since the regexp program
6251 size is calculated during an initial parse this would result
6252 in a buffer overrun so we cache to prevent the charname result from
6253 changing during the course of the parse.
6257 S_reg_namedseq(pTHX_ RExC_state_t *pRExC_state, UV *valuep)
6259 char * name; /* start of the content of the name */
6260 char * endbrace; /* endbrace following the name */
6263 STRLEN len; /* this has various purposes throughout the code */
6264 bool cached = 0; /* if this is true then we shouldn't refcount dev sv_str */
6265 regnode *ret = NULL;
6267 if (*RExC_parse != '{') {
6268 vFAIL("Missing braces on \\N{}");
6270 name = RExC_parse+1;
6271 endbrace = strchr(RExC_parse, '}');
6274 vFAIL("Missing right brace on \\N{}");
6276 RExC_parse = endbrace + 1;
6279 /* RExC_parse points at the beginning brace,
6280 endbrace points at the last */
6281 if ( name[0]=='U' && name[1]=='+' ) {
6282 /* its a "unicode hex" notation {U+89AB} */
6283 I32 fl = PERL_SCAN_ALLOW_UNDERSCORES
6284 | PERL_SCAN_DISALLOW_PREFIX
6285 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
6287 len = (STRLEN)(endbrace - name - 2);
6288 cp = grok_hex(name + 2, &len, &fl, NULL);
6289 if ( len != (STRLEN)(endbrace - name - 2) ) {
6298 sv_str= Perl_newSVpvf_nocontext("%c",(int)cp);
6300 /* fetch the charnames handler for this scope */
6301 HV * const table = GvHV(PL_hintgv);
6303 hv_fetchs(table, "charnames", FALSE) :
6305 SV *cv= cvp ? *cvp : NULL;
6308 /* create an SV with the name as argument */
6309 sv_name = newSVpvn(name, endbrace - name);
6311 if (!table || !(PL_hints & HINT_LOCALIZE_HH)) {
6312 vFAIL2("Constant(\\N{%s}) unknown: "
6313 "(possibly a missing \"use charnames ...\")",
6316 if (!cvp || !SvOK(*cvp)) { /* when $^H{charnames} = undef; */
6317 vFAIL2("Constant(\\N{%s}): "
6318 "$^H{charnames} is not defined",SvPVX(sv_name));
6323 if (!RExC_charnames) {
6324 /* make sure our cache is allocated */
6325 RExC_charnames = newHV();
6326 sv_2mortal((SV*)RExC_charnames);
6328 /* see if we have looked this one up before */
6329 he_str = hv_fetch_ent( RExC_charnames, sv_name, 0, 0 );
6331 sv_str = HeVAL(he_str);
6344 count= call_sv(cv, G_SCALAR);
6346 if (count == 1) { /* XXXX is this right? dmq */
6348 SvREFCNT_inc_simple_void(sv_str);
6356 if ( !sv_str || !SvOK(sv_str) ) {
6357 vFAIL2("Constant(\\N{%s}): Call to &{$^H{charnames}} "
6358 "did not return a defined value",SvPVX(sv_name));
6360 if (hv_store_ent( RExC_charnames, sv_name, sv_str, 0))
6365 char *p = SvPV(sv_str, len);
6368 if ( SvUTF8(sv_str) ) {
6369 *valuep = utf8_to_uvchr((U8*)p, &numlen);
6373 We have to turn on utf8 for high bit chars otherwise
6374 we get failures with
6376 "ss" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i
6377 "SS" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i
6379 This is different from what \x{} would do with the same
6380 codepoint, where the condition is > 0xFF.
6387 /* warn if we havent used the whole string? */
6389 if (numlen<len && SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6391 "Ignoring excess chars from \\N{%s} in character class",
6395 } else if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6397 "Ignoring zero length \\N{%s} in character class",
6402 SvREFCNT_dec(sv_name);
6404 SvREFCNT_dec(sv_str);
6405 return len ? NULL : (regnode *)&len;
6406 } else if(SvCUR(sv_str)) {
6412 char * parse_start = name-3; /* needed for the offsets */
6414 GET_RE_DEBUG_FLAGS_DECL; /* needed for the offsets */
6416 ret = reg_node(pRExC_state,
6417 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
6420 if ( RExC_utf8 && !SvUTF8(sv_str) ) {
6421 sv_utf8_upgrade(sv_str);
6422 } else if ( !RExC_utf8 && SvUTF8(sv_str) ) {
6426 p = SvPV(sv_str, len);
6428 /* len is the length written, charlen is the size the char read */
6429 for ( len = 0; p < pend; p += charlen ) {
6431 UV uvc = utf8_to_uvchr((U8*)p, &charlen);
6433 STRLEN foldlen,numlen;
6434 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
6435 uvc = toFOLD_uni(uvc, tmpbuf, &foldlen);
6436 /* Emit all the Unicode characters. */
6438 for (foldbuf = tmpbuf;
6442 uvc = utf8_to_uvchr(foldbuf, &numlen);
6444 const STRLEN unilen = reguni(pRExC_state, uvc, s);
6447 /* In EBCDIC the numlen
6448 * and unilen can differ. */
6450 if (numlen >= foldlen)
6454 break; /* "Can't happen." */
6457 const STRLEN unilen = reguni(pRExC_state, uvc, s);
6469 RExC_size += STR_SZ(len);
6472 RExC_emit += STR_SZ(len);
6474 Set_Node_Cur_Length(ret); /* MJD */
6476 nextchar(pRExC_state);
6478 ret = reg_node(pRExC_state,NOTHING);
6481 SvREFCNT_dec(sv_str);
6484 SvREFCNT_dec(sv_name);
6494 * It returns the code point in utf8 for the value in *encp.
6495 * value: a code value in the source encoding
6496 * encp: a pointer to an Encode object
6498 * If the result from Encode is not a single character,
6499 * it returns U+FFFD (Replacement character) and sets *encp to NULL.
6502 S_reg_recode(pTHX_ const char value, SV **encp)
6505 SV * const sv = sv_2mortal(newSVpvn(&value, numlen));
6506 const char * const s = *encp ? sv_recode_to_utf8(sv, *encp) : SvPVX(sv);
6507 const STRLEN newlen = SvCUR(sv);
6508 UV uv = UNICODE_REPLACEMENT;
6512 ? utf8n_to_uvchr((U8*)s, newlen, &numlen, UTF8_ALLOW_DEFAULT)
6515 if (!newlen || numlen != newlen) {
6516 uv = UNICODE_REPLACEMENT;
6524 - regatom - the lowest level
6526 Try to identify anything special at the start of the pattern. If there
6527 is, then handle it as required. This may involve generating a single regop,
6528 such as for an assertion; or it may involve recursing, such as to
6529 handle a () structure.
6531 If the string doesn't start with something special then we gobble up
6532 as much literal text as we can.
6534 Once we have been able to handle whatever type of thing started the
6535 sequence, we return.
6537 Note: we have to be careful with escapes, as they can be both literal
6538 and special, and in the case of \10 and friends can either, depending
6539 on context. Specifically there are two seperate switches for handling
6540 escape sequences, with the one for handling literal escapes requiring
6541 a dummy entry for all of the special escapes that are actually handled
6546 S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6549 register regnode *ret = NULL;
6551 char *parse_start = RExC_parse;
6552 GET_RE_DEBUG_FLAGS_DECL;
6553 DEBUG_PARSE("atom");
6554 *flagp = WORST; /* Tentatively. */
6558 switch ((U8)*RExC_parse) {
6560 RExC_seen_zerolen++;
6561 nextchar(pRExC_state);
6562 if (RExC_flags & RXf_PMf_MULTILINE)
6563 ret = reg_node(pRExC_state, MBOL);
6564 else if (RExC_flags & RXf_PMf_SINGLELINE)
6565 ret = reg_node(pRExC_state, SBOL);
6567 ret = reg_node(pRExC_state, BOL);
6568 Set_Node_Length(ret, 1); /* MJD */
6571 nextchar(pRExC_state);
6573 RExC_seen_zerolen++;
6574 if (RExC_flags & RXf_PMf_MULTILINE)
6575 ret = reg_node(pRExC_state, MEOL);
6576 else if (RExC_flags & RXf_PMf_SINGLELINE)
6577 ret = reg_node(pRExC_state, SEOL);
6579 ret = reg_node(pRExC_state, EOL);
6580 Set_Node_Length(ret, 1); /* MJD */
6583 nextchar(pRExC_state);
6584 if (RExC_flags & RXf_PMf_SINGLELINE)
6585 ret = reg_node(pRExC_state, SANY);
6587 ret = reg_node(pRExC_state, REG_ANY);
6588 *flagp |= HASWIDTH|SIMPLE;
6590 Set_Node_Length(ret, 1); /* MJD */
6594 char * const oregcomp_parse = ++RExC_parse;
6595 ret = regclass(pRExC_state,depth+1);
6596 if (*RExC_parse != ']') {
6597 RExC_parse = oregcomp_parse;
6598 vFAIL("Unmatched [");
6600 nextchar(pRExC_state);
6601 *flagp |= HASWIDTH|SIMPLE;
6602 Set_Node_Length(ret, RExC_parse - oregcomp_parse + 1); /* MJD */
6606 nextchar(pRExC_state);
6607 ret = reg(pRExC_state, 1, &flags,depth+1);
6609 if (flags & TRYAGAIN) {
6610 if (RExC_parse == RExC_end) {
6611 /* Make parent create an empty node if needed. */
6619 *flagp |= flags&(HASWIDTH|SPSTART|SIMPLE|POSTPONED);
6623 if (flags & TRYAGAIN) {
6627 vFAIL("Internal urp");
6628 /* Supposed to be caught earlier. */
6631 if (!regcurly(RExC_parse)) {
6640 vFAIL("Quantifier follows nothing");
6647 if ((cp = what_len_TRICKYFOLD_safe(RExC_parse,RExC_end,UTF,len))) {
6648 *flagp |= HASWIDTH; /* could be SIMPLE too, but needs a handler in regexec.regrepeat */
6649 RExC_parse+=len-1; /* we get one from nextchar() as well. :-( */
6650 ret = reganode(pRExC_state, FOLDCHAR, cp);
6651 Set_Node_Length(ret, 1); /* MJD */
6652 nextchar(pRExC_state); /* kill whitespace under /x */
6660 This switch handles escape sequences that resolve to some kind
6661 of special regop and not to literal text. Escape sequnces that
6662 resolve to literal text are handled below in the switch marked
6665 Every entry in this switch *must* have a corresponding entry
6666 in the literal escape switch. However, the opposite is not
6667 required, as the default for this switch is to jump to the
6668 literal text handling code.
6670 switch (*++RExC_parse) {
6671 /* Special Escapes */
6673 RExC_seen_zerolen++;
6674 ret = reg_node(pRExC_state, SBOL);
6676 goto finish_meta_pat;
6678 ret = reg_node(pRExC_state, GPOS);
6679 RExC_seen |= REG_SEEN_GPOS;
6681 goto finish_meta_pat;
6683 RExC_seen_zerolen++;
6684 ret = reg_node(pRExC_state, KEEPS);
6686 goto finish_meta_pat;
6688 ret = reg_node(pRExC_state, SEOL);
6690 RExC_seen_zerolen++; /* Do not optimize RE away */
6691 goto finish_meta_pat;
6693 ret = reg_node(pRExC_state, EOS);
6695 RExC_seen_zerolen++; /* Do not optimize RE away */
6696 goto finish_meta_pat;
6698 ret = reg_node(pRExC_state, CANY);
6699 RExC_seen |= REG_SEEN_CANY;
6700 *flagp |= HASWIDTH|SIMPLE;
6701 goto finish_meta_pat;
6703 ret = reg_node(pRExC_state, CLUMP);
6705 goto finish_meta_pat;
6707 ret = reg_node(pRExC_state, (U8)(LOC ? ALNUML : ALNUM));
6708 *flagp |= HASWIDTH|SIMPLE;
6709 goto finish_meta_pat;
6711 ret = reg_node(pRExC_state, (U8)(LOC ? NALNUML : NALNUM));
6712 *flagp |= HASWIDTH|SIMPLE;
6713 goto finish_meta_pat;
6715 RExC_seen_zerolen++;
6716 RExC_seen |= REG_SEEN_LOOKBEHIND;
6717 ret = reg_node(pRExC_state, (U8)(LOC ? BOUNDL : BOUND));
6719 goto finish_meta_pat;
6721 RExC_seen_zerolen++;
6722 RExC_seen |= REG_SEEN_LOOKBEHIND;
6723 ret = reg_node(pRExC_state, (U8)(LOC ? NBOUNDL : NBOUND));
6725 goto finish_meta_pat;
6727 ret = reg_node(pRExC_state, (U8)(LOC ? SPACEL : SPACE));
6728 *flagp |= HASWIDTH|SIMPLE;
6729 goto finish_meta_pat;
6731 ret = reg_node(pRExC_state, (U8)(LOC ? NSPACEL : NSPACE));
6732 *flagp |= HASWIDTH|SIMPLE;
6733 goto finish_meta_pat;
6735 ret = reg_node(pRExC_state, DIGIT);
6736 *flagp |= HASWIDTH|SIMPLE;
6737 goto finish_meta_pat;
6739 ret = reg_node(pRExC_state, NDIGIT);
6740 *flagp |= HASWIDTH|SIMPLE;
6741 goto finish_meta_pat;
6743 ret = reg_node(pRExC_state, LNBREAK);
6744 *flagp |= HASWIDTH|SIMPLE;
6745 goto finish_meta_pat;
6747 ret = reg_node(pRExC_state, HORIZWS);
6748 *flagp |= HASWIDTH|SIMPLE;
6749 goto finish_meta_pat;
6751 ret = reg_node(pRExC_state, NHORIZWS);
6752 *flagp |= HASWIDTH|SIMPLE;
6753 goto finish_meta_pat;
6755 ret = reg_node(pRExC_state, VERTWS);
6756 *flagp |= HASWIDTH|SIMPLE;
6757 goto finish_meta_pat;
6759 ret = reg_node(pRExC_state, NVERTWS);
6760 *flagp |= HASWIDTH|SIMPLE;
6762 nextchar(pRExC_state);
6763 Set_Node_Length(ret, 2); /* MJD */
6768 char* const oldregxend = RExC_end;
6770 char* parse_start = RExC_parse - 2;
6773 if (RExC_parse[1] == '{') {
6774 /* a lovely hack--pretend we saw [\pX] instead */
6775 RExC_end = strchr(RExC_parse, '}');
6777 const U8 c = (U8)*RExC_parse;
6779 RExC_end = oldregxend;
6780 vFAIL2("Missing right brace on \\%c{}", c);
6785 RExC_end = RExC_parse + 2;
6786 if (RExC_end > oldregxend)
6787 RExC_end = oldregxend;
6791 ret = regclass(pRExC_state,depth+1);
6793 RExC_end = oldregxend;
6796 Set_Node_Offset(ret, parse_start + 2);
6797 Set_Node_Cur_Length(ret);
6798 nextchar(pRExC_state);
6799 *flagp |= HASWIDTH|SIMPLE;
6803 /* Handle \N{NAME} here and not below because it can be
6804 multicharacter. join_exact() will join them up later on.
6805 Also this makes sure that things like /\N{BLAH}+/ and
6806 \N{BLAH} being multi char Just Happen. dmq*/
6808 ret= reg_namedseq(pRExC_state, NULL);
6810 case 'k': /* Handle \k<NAME> and \k'NAME' */
6813 char ch= RExC_parse[1];
6814 if (ch != '<' && ch != '\'' && ch != '{') {
6816 vFAIL2("Sequence %.2s... not terminated",parse_start);
6818 /* this pretty much dupes the code for (?P=...) in reg(), if
6819 you change this make sure you change that */
6820 char* name_start = (RExC_parse += 2);
6822 SV *sv_dat = reg_scan_name(pRExC_state,
6823 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
6824 ch= (ch == '<') ? '>' : (ch == '{') ? '}' : '\'';
6825 if (RExC_parse == name_start || *RExC_parse != ch)
6826 vFAIL2("Sequence %.3s... not terminated",parse_start);
6829 num = add_data( pRExC_state, 1, "S" );
6830 RExC_rxi->data->data[num]=(void*)sv_dat;
6831 SvREFCNT_inc_simple_void(sv_dat);
6835 ret = reganode(pRExC_state,
6836 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
6840 /* override incorrect value set in reganode MJD */
6841 Set_Node_Offset(ret, parse_start+1);
6842 Set_Node_Cur_Length(ret); /* MJD */
6843 nextchar(pRExC_state);
6849 case '1': case '2': case '3': case '4':
6850 case '5': case '6': case '7': case '8': case '9':
6853 bool isg = *RExC_parse == 'g';
6858 if (*RExC_parse == '{') {
6862 if (*RExC_parse == '-') {
6866 if (hasbrace && !isDIGIT(*RExC_parse)) {
6867 if (isrel) RExC_parse--;
6869 goto parse_named_seq;
6871 num = atoi(RExC_parse);
6873 num = RExC_npar - num;
6875 vFAIL("Reference to nonexistent or unclosed group");
6877 if (!isg && num > 9 && num >= RExC_npar)
6880 char * const parse_start = RExC_parse - 1; /* MJD */
6881 while (isDIGIT(*RExC_parse))
6883 if (parse_start == RExC_parse - 1)
6884 vFAIL("Unterminated \\g... pattern");
6886 if (*RExC_parse != '}')
6887 vFAIL("Unterminated \\g{...} pattern");
6891 if (num > (I32)RExC_rx->nparens)
6892 vFAIL("Reference to nonexistent group");
6895 ret = reganode(pRExC_state,
6896 (U8)(FOLD ? (LOC ? REFFL : REFF) : REF),
6900 /* override incorrect value set in reganode MJD */
6901 Set_Node_Offset(ret, parse_start+1);
6902 Set_Node_Cur_Length(ret); /* MJD */
6904 nextchar(pRExC_state);
6909 if (RExC_parse >= RExC_end)
6910 FAIL("Trailing \\");
6913 /* Do not generate "unrecognized" warnings here, we fall
6914 back into the quick-grab loop below */
6921 if (RExC_flags & RXf_PMf_EXTENDED) {
6922 if ( reg_skipcomment( pRExC_state ) )
6929 register STRLEN len;
6934 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
6936 parse_start = RExC_parse - 1;
6942 ret = reg_node(pRExC_state,
6943 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
6945 for (len = 0, p = RExC_parse - 1;
6946 len < 127 && p < RExC_end;
6949 char * const oldp = p;
6951 if (RExC_flags & RXf_PMf_EXTENDED)
6952 p = regwhite( pRExC_state, p );
6957 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
6958 goto normal_default;
6968 /* Literal Escapes Switch
6970 This switch is meant to handle escape sequences that
6971 resolve to a literal character.
6973 Every escape sequence that represents something
6974 else, like an assertion or a char class, is handled
6975 in the switch marked 'Special Escapes' above in this
6976 routine, but also has an entry here as anything that
6977 isn't explicitly mentioned here will be treated as
6978 an unescaped equivalent literal.
6982 /* These are all the special escapes. */
6983 case 'A': /* Start assertion */
6984 case 'b': case 'B': /* Word-boundary assertion*/
6985 case 'C': /* Single char !DANGEROUS! */
6986 case 'd': case 'D': /* digit class */
6987 case 'g': case 'G': /* generic-backref, pos assertion */
6988 case 'h': case 'H': /* HORIZWS */
6989 case 'k': case 'K': /* named backref, keep marker */
6990 case 'N': /* named char sequence */
6991 case 'p': case 'P': /* unicode property */
6992 case 'R': /* LNBREAK */
6993 case 's': case 'S': /* space class */
6994 case 'v': case 'V': /* VERTWS */
6995 case 'w': case 'W': /* word class */
6996 case 'X': /* eXtended Unicode "combining character sequence" */
6997 case 'z': case 'Z': /* End of line/string assertion */
7001 /* Anything after here is an escape that resolves to a
7002 literal. (Except digits, which may or may not)
7021 ender = ASCII_TO_NATIVE('\033');
7025 ender = ASCII_TO_NATIVE('\007');
7030 char* const e = strchr(p, '}');
7034 vFAIL("Missing right brace on \\x{}");
7037 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
7038 | PERL_SCAN_DISALLOW_PREFIX;
7039 STRLEN numlen = e - p - 1;
7040 ender = grok_hex(p + 1, &numlen, &flags, NULL);
7047 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
7049 ender = grok_hex(p, &numlen, &flags, NULL);
7052 if (PL_encoding && ender < 0x100)
7053 goto recode_encoding;
7057 ender = UCHARAT(p++);
7058 ender = toCTRL(ender);
7060 case '0': case '1': case '2': case '3':case '4':
7061 case '5': case '6': case '7': case '8':case '9':
7063 (isDIGIT(p[1]) && atoi(p) >= RExC_npar) ) {
7066 ender = grok_oct(p, &numlen, &flags, NULL);
7073 if (PL_encoding && ender < 0x100)
7074 goto recode_encoding;
7078 SV* enc = PL_encoding;
7079 ender = reg_recode((const char)(U8)ender, &enc);
7080 if (!enc && SIZE_ONLY && ckWARN(WARN_REGEXP))
7081 vWARN(p, "Invalid escape in the specified encoding");
7087 FAIL("Trailing \\");
7090 if (!SIZE_ONLY&& isALPHA(*p) && ckWARN(WARN_REGEXP))
7091 vWARN2(p + 1, "Unrecognized escape \\%c passed through", UCHARAT(p));
7092 goto normal_default;
7097 if (UTF8_IS_START(*p) && UTF) {
7099 ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
7100 &numlen, UTF8_ALLOW_DEFAULT);
7107 if ( RExC_flags & RXf_PMf_EXTENDED)
7108 p = regwhite( pRExC_state, p );
7110 /* Prime the casefolded buffer. */
7111 ender = toFOLD_uni(ender, tmpbuf, &foldlen);
7113 if (p < RExC_end && ISMULT2(p)) { /* Back off on ?+*. */
7118 /* Emit all the Unicode characters. */
7120 for (foldbuf = tmpbuf;
7122 foldlen -= numlen) {
7123 ender = utf8_to_uvchr(foldbuf, &numlen);
7125 const STRLEN unilen = reguni(pRExC_state, ender, s);
7128 /* In EBCDIC the numlen
7129 * and unilen can differ. */
7131 if (numlen >= foldlen)
7135 break; /* "Can't happen." */
7139 const STRLEN unilen = reguni(pRExC_state, ender, s);
7148 REGC((char)ender, s++);
7154 /* Emit all the Unicode characters. */
7156 for (foldbuf = tmpbuf;
7158 foldlen -= numlen) {
7159 ender = utf8_to_uvchr(foldbuf, &numlen);
7161 const STRLEN unilen = reguni(pRExC_state, ender, s);
7164 /* In EBCDIC the numlen
7165 * and unilen can differ. */
7167 if (numlen >= foldlen)
7175 const STRLEN unilen = reguni(pRExC_state, ender, s);
7184 REGC((char)ender, s++);
7188 Set_Node_Cur_Length(ret); /* MJD */
7189 nextchar(pRExC_state);
7191 /* len is STRLEN which is unsigned, need to copy to signed */
7194 vFAIL("Internal disaster");
7198 if (len == 1 && UNI_IS_INVARIANT(ender))
7202 RExC_size += STR_SZ(len);
7205 RExC_emit += STR_SZ(len);
7215 S_regwhite( RExC_state_t *pRExC_state, char *p )
7217 const char *e = RExC_end;
7221 else if (*p == '#') {
7230 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
7238 /* Parse POSIX character classes: [[:foo:]], [[=foo=]], [[.foo.]].
7239 Character classes ([:foo:]) can also be negated ([:^foo:]).
7240 Returns a named class id (ANYOF_XXX) if successful, -1 otherwise.
7241 Equivalence classes ([=foo=]) and composites ([.foo.]) are parsed,
7242 but trigger failures because they are currently unimplemented. */
7244 #define POSIXCC_DONE(c) ((c) == ':')
7245 #define POSIXCC_NOTYET(c) ((c) == '=' || (c) == '.')
7246 #define POSIXCC(c) (POSIXCC_DONE(c) || POSIXCC_NOTYET(c))
7249 S_regpposixcc(pTHX_ RExC_state_t *pRExC_state, I32 value)
7252 I32 namedclass = OOB_NAMEDCLASS;
7254 if (value == '[' && RExC_parse + 1 < RExC_end &&
7255 /* I smell either [: or [= or [. -- POSIX has been here, right? */
7256 POSIXCC(UCHARAT(RExC_parse))) {
7257 const char c = UCHARAT(RExC_parse);
7258 char* const s = RExC_parse++;
7260 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != c)
7262 if (RExC_parse == RExC_end)
7263 /* Grandfather lone [:, [=, [. */
7266 const char* const t = RExC_parse++; /* skip over the c */
7269 if (UCHARAT(RExC_parse) == ']') {
7270 const char *posixcc = s + 1;
7271 RExC_parse++; /* skip over the ending ] */
7274 const I32 complement = *posixcc == '^' ? *posixcc++ : 0;
7275 const I32 skip = t - posixcc;
7277 /* Initially switch on the length of the name. */
7280 if (memEQ(posixcc, "word", 4)) /* this is not POSIX, this is the Perl \w */
7281 namedclass = complement ? ANYOF_NALNUM : ANYOF_ALNUM;
7284 /* Names all of length 5. */
7285 /* alnum alpha ascii blank cntrl digit graph lower
7286 print punct space upper */
7287 /* Offset 4 gives the best switch position. */
7288 switch (posixcc[4]) {
7290 if (memEQ(posixcc, "alph", 4)) /* alpha */
7291 namedclass = complement ? ANYOF_NALPHA : ANYOF_ALPHA;
7294 if (memEQ(posixcc, "spac", 4)) /* space */
7295 namedclass = complement ? ANYOF_NPSXSPC : ANYOF_PSXSPC;
7298 if (memEQ(posixcc, "grap", 4)) /* graph */
7299 namedclass = complement ? ANYOF_NGRAPH : ANYOF_GRAPH;
7302 if (memEQ(posixcc, "asci", 4)) /* ascii */
7303 namedclass = complement ? ANYOF_NASCII : ANYOF_ASCII;
7306 if (memEQ(posixcc, "blan", 4)) /* blank */
7307 namedclass = complement ? ANYOF_NBLANK : ANYOF_BLANK;
7310 if (memEQ(posixcc, "cntr", 4)) /* cntrl */
7311 namedclass = complement ? ANYOF_NCNTRL : ANYOF_CNTRL;
7314 if (memEQ(posixcc, "alnu", 4)) /* alnum */
7315 namedclass = complement ? ANYOF_NALNUMC : ANYOF_ALNUMC;
7318 if (memEQ(posixcc, "lowe", 4)) /* lower */
7319 namedclass = complement ? ANYOF_NLOWER : ANYOF_LOWER;
7320 else if (memEQ(posixcc, "uppe", 4)) /* upper */
7321 namedclass = complement ? ANYOF_NUPPER : ANYOF_UPPER;
7324 if (memEQ(posixcc, "digi", 4)) /* digit */
7325 namedclass = complement ? ANYOF_NDIGIT : ANYOF_DIGIT;
7326 else if (memEQ(posixcc, "prin", 4)) /* print */
7327 namedclass = complement ? ANYOF_NPRINT : ANYOF_PRINT;
7328 else if (memEQ(posixcc, "punc", 4)) /* punct */
7329 namedclass = complement ? ANYOF_NPUNCT : ANYOF_PUNCT;
7334 if (memEQ(posixcc, "xdigit", 6))
7335 namedclass = complement ? ANYOF_NXDIGIT : ANYOF_XDIGIT;
7339 if (namedclass == OOB_NAMEDCLASS)
7340 Simple_vFAIL3("POSIX class [:%.*s:] unknown",
7342 assert (posixcc[skip] == ':');
7343 assert (posixcc[skip+1] == ']');
7344 } else if (!SIZE_ONLY) {
7345 /* [[=foo=]] and [[.foo.]] are still future. */
7347 /* adjust RExC_parse so the warning shows after
7349 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse) != ']')
7351 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7354 /* Maternal grandfather:
7355 * "[:" ending in ":" but not in ":]" */
7365 S_checkposixcc(pTHX_ RExC_state_t *pRExC_state)
7368 if (POSIXCC(UCHARAT(RExC_parse))) {
7369 const char *s = RExC_parse;
7370 const char c = *s++;
7374 if (*s && c == *s && s[1] == ']') {
7375 if (ckWARN(WARN_REGEXP))
7377 "POSIX syntax [%c %c] belongs inside character classes",
7380 /* [[=foo=]] and [[.foo.]] are still future. */
7381 if (POSIXCC_NOTYET(c)) {
7382 /* adjust RExC_parse so the error shows after
7384 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse++) != ']')
7386 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7393 #define _C_C_T_(NAME,TEST,WORD) \
7396 ANYOF_CLASS_SET(ret, ANYOF_##NAME); \
7398 for (value = 0; value < 256; value++) \
7400 ANYOF_BITMAP_SET(ret, value); \
7405 case ANYOF_N##NAME: \
7407 ANYOF_CLASS_SET(ret, ANYOF_N##NAME); \
7409 for (value = 0; value < 256; value++) \
7411 ANYOF_BITMAP_SET(ret, value); \
7417 #define _C_C_T_NOLOC_(NAME,TEST,WORD) \
7419 for (value = 0; value < 256; value++) \
7421 ANYOF_BITMAP_SET(ret, value); \
7425 case ANYOF_N##NAME: \
7426 for (value = 0; value < 256; value++) \
7428 ANYOF_BITMAP_SET(ret, value); \
7434 parse a class specification and produce either an ANYOF node that
7435 matches the pattern or if the pattern matches a single char only and
7436 that char is < 256 and we are case insensitive then we produce an
7441 S_regclass(pTHX_ RExC_state_t *pRExC_state, U32 depth)
7444 register UV nextvalue;
7445 register IV prevvalue = OOB_UNICODE;
7446 register IV range = 0;
7447 UV value = 0; /* XXX:dmq: needs to be referenceable (unfortunately) */
7448 register regnode *ret;
7451 char *rangebegin = NULL;
7452 bool need_class = 0;
7455 bool optimize_invert = TRUE;
7456 AV* unicode_alternate = NULL;
7458 UV literal_endpoint = 0;
7460 UV stored = 0; /* number of chars stored in the class */
7462 regnode * const orig_emit = RExC_emit; /* Save the original RExC_emit in
7463 case we need to change the emitted regop to an EXACT. */
7464 const char * orig_parse = RExC_parse;
7465 GET_RE_DEBUG_FLAGS_DECL;
7467 PERL_UNUSED_ARG(depth);
7470 DEBUG_PARSE("clas");
7472 /* Assume we are going to generate an ANYOF node. */
7473 ret = reganode(pRExC_state, ANYOF, 0);
7476 ANYOF_FLAGS(ret) = 0;
7478 if (UCHARAT(RExC_parse) == '^') { /* Complement of range. */
7482 ANYOF_FLAGS(ret) |= ANYOF_INVERT;
7486 RExC_size += ANYOF_SKIP;
7487 listsv = &PL_sv_undef; /* For code scanners: listsv always non-NULL. */
7490 RExC_emit += ANYOF_SKIP;
7492 ANYOF_FLAGS(ret) |= ANYOF_FOLD;
7494 ANYOF_FLAGS(ret) |= ANYOF_LOCALE;
7495 ANYOF_BITMAP_ZERO(ret);
7496 listsv = newSVpvs("# comment\n");
7499 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
7501 if (!SIZE_ONLY && POSIXCC(nextvalue))
7502 checkposixcc(pRExC_state);
7504 /* allow 1st char to be ] (allowing it to be - is dealt with later) */
7505 if (UCHARAT(RExC_parse) == ']')
7509 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != ']') {
7513 namedclass = OOB_NAMEDCLASS; /* initialize as illegal */
7516 rangebegin = RExC_parse;
7518 value = utf8n_to_uvchr((U8*)RExC_parse,
7519 RExC_end - RExC_parse,
7520 &numlen, UTF8_ALLOW_DEFAULT);
7521 RExC_parse += numlen;
7524 value = UCHARAT(RExC_parse++);
7526 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
7527 if (value == '[' && POSIXCC(nextvalue))
7528 namedclass = regpposixcc(pRExC_state, value);
7529 else if (value == '\\') {
7531 value = utf8n_to_uvchr((U8*)RExC_parse,
7532 RExC_end - RExC_parse,
7533 &numlen, UTF8_ALLOW_DEFAULT);
7534 RExC_parse += numlen;
7537 value = UCHARAT(RExC_parse++);
7538 /* Some compilers cannot handle switching on 64-bit integer
7539 * values, therefore value cannot be an UV. Yes, this will
7540 * be a problem later if we want switch on Unicode.
7541 * A similar issue a little bit later when switching on
7542 * namedclass. --jhi */
7543 switch ((I32)value) {
7544 case 'w': namedclass = ANYOF_ALNUM; break;
7545 case 'W': namedclass = ANYOF_NALNUM; break;
7546 case 's': namedclass = ANYOF_SPACE; break;
7547 case 'S': namedclass = ANYOF_NSPACE; break;
7548 case 'd': namedclass = ANYOF_DIGIT; break;
7549 case 'D': namedclass = ANYOF_NDIGIT; break;
7550 case 'v': namedclass = ANYOF_VERTWS; break;
7551 case 'V': namedclass = ANYOF_NVERTWS; break;
7552 case 'h': namedclass = ANYOF_HORIZWS; break;
7553 case 'H': namedclass = ANYOF_NHORIZWS; break;
7554 case 'N': /* Handle \N{NAME} in class */
7556 /* We only pay attention to the first char of
7557 multichar strings being returned. I kinda wonder
7558 if this makes sense as it does change the behaviour
7559 from earlier versions, OTOH that behaviour was broken
7561 UV v; /* value is register so we cant & it /grrr */
7562 if (reg_namedseq(pRExC_state, &v)) {
7572 if (RExC_parse >= RExC_end)
7573 vFAIL2("Empty \\%c{}", (U8)value);
7574 if (*RExC_parse == '{') {
7575 const U8 c = (U8)value;
7576 e = strchr(RExC_parse++, '}');
7578 vFAIL2("Missing right brace on \\%c{}", c);
7579 while (isSPACE(UCHARAT(RExC_parse)))
7581 if (e == RExC_parse)
7582 vFAIL2("Empty \\%c{}", c);
7584 while (isSPACE(UCHARAT(RExC_parse + n - 1)))
7592 if (UCHARAT(RExC_parse) == '^') {
7595 value = value == 'p' ? 'P' : 'p'; /* toggle */
7596 while (isSPACE(UCHARAT(RExC_parse))) {
7601 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::%.*s\n",
7602 (value=='p' ? '+' : '!'), (int)n, RExC_parse);
7605 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
7606 namedclass = ANYOF_MAX; /* no official name, but it's named */
7609 case 'n': value = '\n'; break;
7610 case 'r': value = '\r'; break;
7611 case 't': value = '\t'; break;
7612 case 'f': value = '\f'; break;
7613 case 'b': value = '\b'; break;
7614 case 'e': value = ASCII_TO_NATIVE('\033');break;
7615 case 'a': value = ASCII_TO_NATIVE('\007');break;
7617 if (*RExC_parse == '{') {
7618 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
7619 | PERL_SCAN_DISALLOW_PREFIX;
7620 char * const e = strchr(RExC_parse++, '}');
7622 vFAIL("Missing right brace on \\x{}");
7624 numlen = e - RExC_parse;
7625 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
7629 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
7631 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
7632 RExC_parse += numlen;
7634 if (PL_encoding && value < 0x100)
7635 goto recode_encoding;
7638 value = UCHARAT(RExC_parse++);
7639 value = toCTRL(value);
7641 case '0': case '1': case '2': case '3': case '4':
7642 case '5': case '6': case '7': case '8': case '9':
7646 value = grok_oct(--RExC_parse, &numlen, &flags, NULL);
7647 RExC_parse += numlen;
7648 if (PL_encoding && value < 0x100)
7649 goto recode_encoding;
7654 SV* enc = PL_encoding;
7655 value = reg_recode((const char)(U8)value, &enc);
7656 if (!enc && SIZE_ONLY && ckWARN(WARN_REGEXP))
7658 "Invalid escape in the specified encoding");
7662 if (!SIZE_ONLY && isALPHA(value) && ckWARN(WARN_REGEXP))
7664 "Unrecognized escape \\%c in character class passed through",
7668 } /* end of \blah */
7674 if (namedclass > OOB_NAMEDCLASS) { /* this is a named class \blah */
7676 if (!SIZE_ONLY && !need_class)
7677 ANYOF_CLASS_ZERO(ret);
7681 /* a bad range like a-\d, a-[:digit:] ? */
7684 if (ckWARN(WARN_REGEXP)) {
7686 RExC_parse >= rangebegin ?
7687 RExC_parse - rangebegin : 0;
7689 "False [] range \"%*.*s\"",
7692 if (prevvalue < 256) {
7693 ANYOF_BITMAP_SET(ret, prevvalue);
7694 ANYOF_BITMAP_SET(ret, '-');
7697 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
7698 Perl_sv_catpvf(aTHX_ listsv,
7699 "%04"UVxf"\n%04"UVxf"\n", (UV)prevvalue, (UV) '-');
7703 range = 0; /* this was not a true range */
7709 const char *what = NULL;
7712 if (namedclass > OOB_NAMEDCLASS)
7713 optimize_invert = FALSE;
7714 /* Possible truncation here but in some 64-bit environments
7715 * the compiler gets heartburn about switch on 64-bit values.
7716 * A similar issue a little earlier when switching on value.
7718 switch ((I32)namedclass) {
7719 case _C_C_T_(ALNUM, isALNUM(value), "Word");
7720 case _C_C_T_(ALNUMC, isALNUMC(value), "Alnum");
7721 case _C_C_T_(ALPHA, isALPHA(value), "Alpha");
7722 case _C_C_T_(BLANK, isBLANK(value), "Blank");
7723 case _C_C_T_(CNTRL, isCNTRL(value), "Cntrl");
7724 case _C_C_T_(GRAPH, isGRAPH(value), "Graph");
7725 case _C_C_T_(LOWER, isLOWER(value), "Lower");
7726 case _C_C_T_(PRINT, isPRINT(value), "Print");
7727 case _C_C_T_(PSXSPC, isPSXSPC(value), "Space");
7728 case _C_C_T_(PUNCT, isPUNCT(value), "Punct");
7729 case _C_C_T_(SPACE, isSPACE(value), "SpacePerl");
7730 case _C_C_T_(UPPER, isUPPER(value), "Upper");
7731 case _C_C_T_(XDIGIT, isXDIGIT(value), "XDigit");
7732 case _C_C_T_NOLOC_(VERTWS, is_VERTWS_latin1(&value), "VertSpace");
7733 case _C_C_T_NOLOC_(HORIZWS, is_HORIZWS_latin1(&value), "HorizSpace");
7736 ANYOF_CLASS_SET(ret, ANYOF_ASCII);
7739 for (value = 0; value < 128; value++)
7740 ANYOF_BITMAP_SET(ret, value);
7742 for (value = 0; value < 256; value++) {
7744 ANYOF_BITMAP_SET(ret, value);
7753 ANYOF_CLASS_SET(ret, ANYOF_NASCII);
7756 for (value = 128; value < 256; value++)
7757 ANYOF_BITMAP_SET(ret, value);
7759 for (value = 0; value < 256; value++) {
7760 if (!isASCII(value))
7761 ANYOF_BITMAP_SET(ret, value);
7770 ANYOF_CLASS_SET(ret, ANYOF_DIGIT);
7772 /* consecutive digits assumed */
7773 for (value = '0'; value <= '9'; value++)
7774 ANYOF_BITMAP_SET(ret, value);
7781 ANYOF_CLASS_SET(ret, ANYOF_NDIGIT);
7783 /* consecutive digits assumed */
7784 for (value = 0; value < '0'; value++)
7785 ANYOF_BITMAP_SET(ret, value);
7786 for (value = '9' + 1; value < 256; value++)
7787 ANYOF_BITMAP_SET(ret, value);
7793 /* this is to handle \p and \P */
7796 vFAIL("Invalid [::] class");
7800 /* Strings such as "+utf8::isWord\n" */
7801 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::Is%s\n", yesno, what);
7804 ANYOF_FLAGS(ret) |= ANYOF_CLASS;
7807 } /* end of namedclass \blah */
7810 if (prevvalue > (IV)value) /* b-a */ {
7811 const int w = RExC_parse - rangebegin;
7812 Simple_vFAIL4("Invalid [] range \"%*.*s\"", w, w, rangebegin);
7813 range = 0; /* not a valid range */
7817 prevvalue = value; /* save the beginning of the range */
7818 if (*RExC_parse == '-' && RExC_parse+1 < RExC_end &&
7819 RExC_parse[1] != ']') {
7822 /* a bad range like \w-, [:word:]- ? */
7823 if (namedclass > OOB_NAMEDCLASS) {
7824 if (ckWARN(WARN_REGEXP)) {
7826 RExC_parse >= rangebegin ?
7827 RExC_parse - rangebegin : 0;
7829 "False [] range \"%*.*s\"",
7833 ANYOF_BITMAP_SET(ret, '-');
7835 range = 1; /* yeah, it's a range! */
7836 continue; /* but do it the next time */
7840 /* now is the next time */
7841 /*stored += (value - prevvalue + 1);*/
7843 if (prevvalue < 256) {
7844 const IV ceilvalue = value < 256 ? value : 255;
7847 /* In EBCDIC [\x89-\x91] should include
7848 * the \x8e but [i-j] should not. */
7849 if (literal_endpoint == 2 &&
7850 ((isLOWER(prevvalue) && isLOWER(ceilvalue)) ||
7851 (isUPPER(prevvalue) && isUPPER(ceilvalue))))
7853 if (isLOWER(prevvalue)) {
7854 for (i = prevvalue; i <= ceilvalue; i++)
7856 ANYOF_BITMAP_SET(ret, i);
7858 for (i = prevvalue; i <= ceilvalue; i++)
7860 ANYOF_BITMAP_SET(ret, i);
7865 for (i = prevvalue; i <= ceilvalue; i++) {
7866 if (!ANYOF_BITMAP_TEST(ret,i)) {
7868 ANYOF_BITMAP_SET(ret, i);
7872 if (value > 255 || UTF) {
7873 const UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
7874 const UV natvalue = NATIVE_TO_UNI(value);
7875 stored+=2; /* can't optimize this class */
7876 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
7877 if (prevnatvalue < natvalue) { /* what about > ? */
7878 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\t%04"UVxf"\n",
7879 prevnatvalue, natvalue);
7881 else if (prevnatvalue == natvalue) {
7882 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n", natvalue);
7884 U8 foldbuf[UTF8_MAXBYTES_CASE+1];
7886 const UV f = to_uni_fold(natvalue, foldbuf, &foldlen);
7888 #ifdef EBCDIC /* RD t/uni/fold ff and 6b */
7889 if (RExC_precomp[0] == ':' &&
7890 RExC_precomp[1] == '[' &&
7891 (f == 0xDF || f == 0x92)) {
7892 f = NATIVE_TO_UNI(f);
7895 /* If folding and foldable and a single
7896 * character, insert also the folded version
7897 * to the charclass. */
7899 #ifdef EBCDIC /* RD tunifold ligatures s,t fb05, fb06 */
7900 if ((RExC_precomp[0] == ':' &&
7901 RExC_precomp[1] == '[' &&
7903 (value == 0xFB05 || value == 0xFB06))) ?
7904 foldlen == ((STRLEN)UNISKIP(f) - 1) :
7905 foldlen == (STRLEN)UNISKIP(f) )
7907 if (foldlen == (STRLEN)UNISKIP(f))
7909 Perl_sv_catpvf(aTHX_ listsv,
7912 /* Any multicharacter foldings
7913 * require the following transform:
7914 * [ABCDEF] -> (?:[ABCabcDEFd]|pq|rst)
7915 * where E folds into "pq" and F folds
7916 * into "rst", all other characters
7917 * fold to single characters. We save
7918 * away these multicharacter foldings,
7919 * to be later saved as part of the
7920 * additional "s" data. */
7923 if (!unicode_alternate)
7924 unicode_alternate = newAV();
7925 sv = newSVpvn((char*)foldbuf, foldlen);
7927 av_push(unicode_alternate, sv);
7931 /* If folding and the value is one of the Greek
7932 * sigmas insert a few more sigmas to make the
7933 * folding rules of the sigmas to work right.
7934 * Note that not all the possible combinations
7935 * are handled here: some of them are handled
7936 * by the standard folding rules, and some of
7937 * them (literal or EXACTF cases) are handled
7938 * during runtime in regexec.c:S_find_byclass(). */
7939 if (value == UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA) {
7940 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
7941 (UV)UNICODE_GREEK_CAPITAL_LETTER_SIGMA);
7942 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
7943 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
7945 else if (value == UNICODE_GREEK_CAPITAL_LETTER_SIGMA)
7946 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
7947 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
7952 literal_endpoint = 0;
7956 range = 0; /* this range (if it was one) is done now */
7960 ANYOF_FLAGS(ret) |= ANYOF_LARGE;
7962 RExC_size += ANYOF_CLASS_ADD_SKIP;
7964 RExC_emit += ANYOF_CLASS_ADD_SKIP;
7970 /****** !SIZE_ONLY AFTER HERE *********/
7972 if( stored == 1 && (value < 128 || (value < 256 && !UTF))
7973 && !( ANYOF_FLAGS(ret) & ( ANYOF_FLAGS_ALL ^ ANYOF_FOLD ) )
7975 /* optimize single char class to an EXACT node
7976 but *only* when its not a UTF/high char */
7977 const char * cur_parse= RExC_parse;
7978 RExC_emit = (regnode *)orig_emit;
7979 RExC_parse = (char *)orig_parse;
7980 ret = reg_node(pRExC_state,
7981 (U8)((ANYOF_FLAGS(ret) & ANYOF_FOLD) ? EXACTF : EXACT));
7982 RExC_parse = (char *)cur_parse;
7983 *STRING(ret)= (char)value;
7985 RExC_emit += STR_SZ(1);
7988 /* optimize case-insensitive simple patterns (e.g. /[a-z]/i) */
7989 if ( /* If the only flag is folding (plus possibly inversion). */
7990 ((ANYOF_FLAGS(ret) & (ANYOF_FLAGS_ALL ^ ANYOF_INVERT)) == ANYOF_FOLD)
7992 for (value = 0; value < 256; ++value) {
7993 if (ANYOF_BITMAP_TEST(ret, value)) {
7994 UV fold = PL_fold[value];
7997 ANYOF_BITMAP_SET(ret, fold);
8000 ANYOF_FLAGS(ret) &= ~ANYOF_FOLD;
8003 /* optimize inverted simple patterns (e.g. [^a-z]) */
8004 if (optimize_invert &&
8005 /* If the only flag is inversion. */
8006 (ANYOF_FLAGS(ret) & ANYOF_FLAGS_ALL) == ANYOF_INVERT) {
8007 for (value = 0; value < ANYOF_BITMAP_SIZE; ++value)
8008 ANYOF_BITMAP(ret)[value] ^= ANYOF_FLAGS_ALL;
8009 ANYOF_FLAGS(ret) = ANYOF_UNICODE_ALL;
8012 AV * const av = newAV();
8014 /* The 0th element stores the character class description
8015 * in its textual form: used later (regexec.c:Perl_regclass_swash())
8016 * to initialize the appropriate swash (which gets stored in
8017 * the 1st element), and also useful for dumping the regnode.
8018 * The 2nd element stores the multicharacter foldings,
8019 * used later (regexec.c:S_reginclass()). */
8020 av_store(av, 0, listsv);
8021 av_store(av, 1, NULL);
8022 av_store(av, 2, (SV*)unicode_alternate);
8023 rv = newRV_noinc((SV*)av);
8024 n = add_data(pRExC_state, 1, "s");
8025 RExC_rxi->data->data[n] = (void*)rv;
8033 /* reg_skipcomment()
8035 Absorbs an /x style # comments from the input stream.
8036 Returns true if there is more text remaining in the stream.
8037 Will set the REG_SEEN_RUN_ON_COMMENT flag if the comment
8038 terminates the pattern without including a newline.
8040 Note its the callers responsibility to ensure that we are
8046 S_reg_skipcomment(pTHX_ RExC_state_t *pRExC_state)
8049 while (RExC_parse < RExC_end)
8050 if (*RExC_parse++ == '\n') {
8055 /* we ran off the end of the pattern without ending
8056 the comment, so we have to add an \n when wrapping */
8057 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
8065 Advance that parse position, and optionall absorbs
8066 "whitespace" from the inputstream.
8068 Without /x "whitespace" means (?#...) style comments only,
8069 with /x this means (?#...) and # comments and whitespace proper.
8071 Returns the RExC_parse point from BEFORE the scan occurs.
8073 This is the /x friendly way of saying RExC_parse++.
8077 S_nextchar(pTHX_ RExC_state_t *pRExC_state)
8079 char* const retval = RExC_parse++;
8082 if (*RExC_parse == '(' && RExC_parse[1] == '?' &&
8083 RExC_parse[2] == '#') {
8084 while (*RExC_parse != ')') {
8085 if (RExC_parse == RExC_end)
8086 FAIL("Sequence (?#... not terminated");
8092 if (RExC_flags & RXf_PMf_EXTENDED) {
8093 if (isSPACE(*RExC_parse)) {
8097 else if (*RExC_parse == '#') {
8098 if ( reg_skipcomment( pRExC_state ) )
8107 - reg_node - emit a node
8109 STATIC regnode * /* Location. */
8110 S_reg_node(pTHX_ RExC_state_t *pRExC_state, U8 op)
8113 register regnode *ptr;
8114 regnode * const ret = RExC_emit;
8115 GET_RE_DEBUG_FLAGS_DECL;
8118 SIZE_ALIGN(RExC_size);
8122 if (RExC_emit >= RExC_emit_bound)
8123 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8125 NODE_ALIGN_FILL(ret);
8127 FILL_ADVANCE_NODE(ptr, op);
8128 #ifdef RE_TRACK_PATTERN_OFFSETS
8129 if (RExC_offsets) { /* MJD */
8130 MJD_OFFSET_DEBUG(("%s:%d: (op %s) %s %"UVuf" (len %"UVuf") (max %"UVuf").\n",
8131 "reg_node", __LINE__,
8133 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0]
8134 ? "Overwriting end of array!\n" : "OK",
8135 (UV)(RExC_emit - RExC_emit_start),
8136 (UV)(RExC_parse - RExC_start),
8137 (UV)RExC_offsets[0]));
8138 Set_Node_Offset(RExC_emit, RExC_parse + (op == END));
8146 - reganode - emit a node with an argument
8148 STATIC regnode * /* Location. */
8149 S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
8152 register regnode *ptr;
8153 regnode * const ret = RExC_emit;
8154 GET_RE_DEBUG_FLAGS_DECL;
8157 SIZE_ALIGN(RExC_size);
8162 assert(2==regarglen[op]+1);
8164 Anything larger than this has to allocate the extra amount.
8165 If we changed this to be:
8167 RExC_size += (1 + regarglen[op]);
8169 then it wouldn't matter. Its not clear what side effect
8170 might come from that so its not done so far.
8175 if (RExC_emit >= RExC_emit_bound)
8176 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8178 NODE_ALIGN_FILL(ret);
8180 FILL_ADVANCE_NODE_ARG(ptr, op, arg);
8181 #ifdef RE_TRACK_PATTERN_OFFSETS
8182 if (RExC_offsets) { /* MJD */
8183 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8187 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0] ?
8188 "Overwriting end of array!\n" : "OK",
8189 (UV)(RExC_emit - RExC_emit_start),
8190 (UV)(RExC_parse - RExC_start),
8191 (UV)RExC_offsets[0]));
8192 Set_Cur_Node_Offset;
8200 - reguni - emit (if appropriate) a Unicode character
8203 S_reguni(pTHX_ const RExC_state_t *pRExC_state, UV uv, char* s)
8206 return SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
8210 - reginsert - insert an operator in front of already-emitted operand
8212 * Means relocating the operand.
8215 S_reginsert(pTHX_ RExC_state_t *pRExC_state, U8 op, regnode *opnd, U32 depth)
8218 register regnode *src;
8219 register regnode *dst;
8220 register regnode *place;
8221 const int offset = regarglen[(U8)op];
8222 const int size = NODE_STEP_REGNODE + offset;
8223 GET_RE_DEBUG_FLAGS_DECL;
8224 PERL_UNUSED_ARG(depth);
8225 /* (PL_regkind[(U8)op] == CURLY ? EXTRA_STEP_2ARGS : 0); */
8226 DEBUG_PARSE_FMT("inst"," - %s",PL_reg_name[op]);
8235 if (RExC_open_parens) {
8237 /*DEBUG_PARSE_FMT("inst"," - %"IVdf, (IV)RExC_npar);*/
8238 for ( paren=0 ; paren < RExC_npar ; paren++ ) {
8239 if ( RExC_open_parens[paren] >= opnd ) {
8240 /*DEBUG_PARSE_FMT("open"," - %d",size);*/
8241 RExC_open_parens[paren] += size;
8243 /*DEBUG_PARSE_FMT("open"," - %s","ok");*/
8245 if ( RExC_close_parens[paren] >= opnd ) {
8246 /*DEBUG_PARSE_FMT("close"," - %d",size);*/
8247 RExC_close_parens[paren] += size;
8249 /*DEBUG_PARSE_FMT("close"," - %s","ok");*/
8254 while (src > opnd) {
8255 StructCopy(--src, --dst, regnode);
8256 #ifdef RE_TRACK_PATTERN_OFFSETS
8257 if (RExC_offsets) { /* MJD 20010112 */
8258 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s copy %"UVuf" -> %"UVuf" (max %"UVuf").\n",
8262 (UV)(dst - RExC_emit_start) > RExC_offsets[0]
8263 ? "Overwriting end of array!\n" : "OK",
8264 (UV)(src - RExC_emit_start),
8265 (UV)(dst - RExC_emit_start),
8266 (UV)RExC_offsets[0]));
8267 Set_Node_Offset_To_R(dst-RExC_emit_start, Node_Offset(src));
8268 Set_Node_Length_To_R(dst-RExC_emit_start, Node_Length(src));
8274 place = opnd; /* Op node, where operand used to be. */
8275 #ifdef RE_TRACK_PATTERN_OFFSETS
8276 if (RExC_offsets) { /* MJD */
8277 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8281 (UV)(place - RExC_emit_start) > RExC_offsets[0]
8282 ? "Overwriting end of array!\n" : "OK",
8283 (UV)(place - RExC_emit_start),
8284 (UV)(RExC_parse - RExC_start),
8285 (UV)RExC_offsets[0]));
8286 Set_Node_Offset(place, RExC_parse);
8287 Set_Node_Length(place, 1);
8290 src = NEXTOPER(place);
8291 FILL_ADVANCE_NODE(place, op);
8292 Zero(src, offset, regnode);
8296 - regtail - set the next-pointer at the end of a node chain of p to val.
8297 - SEE ALSO: regtail_study
8299 /* TODO: All three parms should be const */
8301 S_regtail(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8304 register regnode *scan;
8305 GET_RE_DEBUG_FLAGS_DECL;
8307 PERL_UNUSED_ARG(depth);
8313 /* Find last node. */
8316 regnode * const temp = regnext(scan);
8318 SV * const mysv=sv_newmortal();
8319 DEBUG_PARSE_MSG((scan==p ? "tail" : ""));
8320 regprop(RExC_rx, mysv, scan);
8321 PerlIO_printf(Perl_debug_log, "~ %s (%d) %s %s\n",
8322 SvPV_nolen_const(mysv), REG_NODE_NUM(scan),
8323 (temp == NULL ? "->" : ""),
8324 (temp == NULL ? PL_reg_name[OP(val)] : "")
8332 if (reg_off_by_arg[OP(scan)]) {
8333 ARG_SET(scan, val - scan);
8336 NEXT_OFF(scan) = val - scan;
8342 - regtail_study - set the next-pointer at the end of a node chain of p to val.
8343 - Look for optimizable sequences at the same time.
8344 - currently only looks for EXACT chains.
8346 This is expermental code. The idea is to use this routine to perform
8347 in place optimizations on branches and groups as they are constructed,
8348 with the long term intention of removing optimization from study_chunk so
8349 that it is purely analytical.
8351 Currently only used when in DEBUG mode. The macro REGTAIL_STUDY() is used
8352 to control which is which.
8355 /* TODO: All four parms should be const */
8358 S_regtail_study(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8361 register regnode *scan;
8363 #ifdef EXPERIMENTAL_INPLACESCAN
8367 GET_RE_DEBUG_FLAGS_DECL;
8373 /* Find last node. */
8377 regnode * const temp = regnext(scan);
8378 #ifdef EXPERIMENTAL_INPLACESCAN
8379 if (PL_regkind[OP(scan)] == EXACT)
8380 if (join_exact(pRExC_state,scan,&min,1,val,depth+1))
8388 if( exact == PSEUDO )
8390 else if ( exact != OP(scan) )
8399 SV * const mysv=sv_newmortal();
8400 DEBUG_PARSE_MSG((scan==p ? "tsdy" : ""));
8401 regprop(RExC_rx, mysv, scan);
8402 PerlIO_printf(Perl_debug_log, "~ %s (%d) -> %s\n",
8403 SvPV_nolen_const(mysv),
8405 PL_reg_name[exact]);
8412 SV * const mysv_val=sv_newmortal();
8413 DEBUG_PARSE_MSG("");
8414 regprop(RExC_rx, mysv_val, val);
8415 PerlIO_printf(Perl_debug_log, "~ attach to %s (%"IVdf") offset to %"IVdf"\n",
8416 SvPV_nolen_const(mysv_val),
8417 (IV)REG_NODE_NUM(val),
8421 if (reg_off_by_arg[OP(scan)]) {
8422 ARG_SET(scan, val - scan);
8425 NEXT_OFF(scan) = val - scan;
8433 - regcurly - a little FSA that accepts {\d+,?\d*}
8436 S_regcurly(register const char *s)
8455 - regdump - dump a regexp onto Perl_debug_log in vaguely comprehensible form
8458 Perl_regdump(pTHX_ const regexp *r)
8462 SV * const sv = sv_newmortal();
8463 SV *dsv= sv_newmortal();
8466 (void)dumpuntil(r, ri->program, ri->program + 1, NULL, NULL, sv, 0, 0);
8468 /* Header fields of interest. */
8469 if (r->anchored_substr) {
8470 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->anchored_substr),
8471 RE_SV_DUMPLEN(r->anchored_substr), 30);
8472 PerlIO_printf(Perl_debug_log,
8473 "anchored %s%s at %"IVdf" ",
8474 s, RE_SV_TAIL(r->anchored_substr),
8475 (IV)r->anchored_offset);
8476 } else if (r->anchored_utf8) {
8477 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->anchored_utf8),
8478 RE_SV_DUMPLEN(r->anchored_utf8), 30);
8479 PerlIO_printf(Perl_debug_log,
8480 "anchored utf8 %s%s at %"IVdf" ",
8481 s, RE_SV_TAIL(r->anchored_utf8),
8482 (IV)r->anchored_offset);
8484 if (r->float_substr) {
8485 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->float_substr),
8486 RE_SV_DUMPLEN(r->float_substr), 30);
8487 PerlIO_printf(Perl_debug_log,
8488 "floating %s%s at %"IVdf"..%"UVuf" ",
8489 s, RE_SV_TAIL(r->float_substr),
8490 (IV)r->float_min_offset, (UV)r->float_max_offset);
8491 } else if (r->float_utf8) {
8492 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->float_utf8),
8493 RE_SV_DUMPLEN(r->float_utf8), 30);
8494 PerlIO_printf(Perl_debug_log,
8495 "floating utf8 %s%s at %"IVdf"..%"UVuf" ",
8496 s, RE_SV_TAIL(r->float_utf8),
8497 (IV)r->float_min_offset, (UV)r->float_max_offset);
8499 if (r->check_substr || r->check_utf8)
8500 PerlIO_printf(Perl_debug_log,
8502 (r->check_substr == r->float_substr
8503 && r->check_utf8 == r->float_utf8
8504 ? "(checking floating" : "(checking anchored"));
8505 if (r->extflags & RXf_NOSCAN)
8506 PerlIO_printf(Perl_debug_log, " noscan");
8507 if (r->extflags & RXf_CHECK_ALL)
8508 PerlIO_printf(Perl_debug_log, " isall");
8509 if (r->check_substr || r->check_utf8)
8510 PerlIO_printf(Perl_debug_log, ") ");
8512 if (ri->regstclass) {
8513 regprop(r, sv, ri->regstclass);
8514 PerlIO_printf(Perl_debug_log, "stclass %s ", SvPVX_const(sv));
8516 if (r->extflags & RXf_ANCH) {
8517 PerlIO_printf(Perl_debug_log, "anchored");
8518 if (r->extflags & RXf_ANCH_BOL)
8519 PerlIO_printf(Perl_debug_log, "(BOL)");
8520 if (r->extflags & RXf_ANCH_MBOL)
8521 PerlIO_printf(Perl_debug_log, "(MBOL)");
8522 if (r->extflags & RXf_ANCH_SBOL)
8523 PerlIO_printf(Perl_debug_log, "(SBOL)");
8524 if (r->extflags & RXf_ANCH_GPOS)
8525 PerlIO_printf(Perl_debug_log, "(GPOS)");
8526 PerlIO_putc(Perl_debug_log, ' ');
8528 if (r->extflags & RXf_GPOS_SEEN)
8529 PerlIO_printf(Perl_debug_log, "GPOS:%"UVuf" ", (UV)r->gofs);
8530 if (r->intflags & PREGf_SKIP)
8531 PerlIO_printf(Perl_debug_log, "plus ");
8532 if (r->intflags & PREGf_IMPLICIT)
8533 PerlIO_printf(Perl_debug_log, "implicit ");
8534 PerlIO_printf(Perl_debug_log, "minlen %"IVdf" ", (IV)r->minlen);
8535 if (r->extflags & RXf_EVAL_SEEN)
8536 PerlIO_printf(Perl_debug_log, "with eval ");
8537 PerlIO_printf(Perl_debug_log, "\n");
8539 PERL_UNUSED_CONTEXT;
8541 #endif /* DEBUGGING */
8545 - regprop - printable representation of opcode
8548 Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
8553 RXi_GET_DECL(prog,progi);
8554 GET_RE_DEBUG_FLAGS_DECL;
8557 sv_setpvn(sv, "", 0);
8559 if (OP(o) > REGNODE_MAX) /* regnode.type is unsigned */
8560 /* It would be nice to FAIL() here, but this may be called from
8561 regexec.c, and it would be hard to supply pRExC_state. */
8562 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(o), (int)REGNODE_MAX);
8563 sv_catpv(sv, PL_reg_name[OP(o)]); /* Take off const! */
8565 k = PL_regkind[OP(o)];
8568 SV * const dsv = sv_2mortal(newSVpvs(""));
8569 /* Using is_utf8_string() (via PERL_PV_UNI_DETECT)
8570 * is a crude hack but it may be the best for now since
8571 * we have no flag "this EXACTish node was UTF-8"
8573 const char * const s =
8574 pv_pretty(dsv, STRING(o), STR_LEN(o), 60,
8575 PL_colors[0], PL_colors[1],
8576 PERL_PV_ESCAPE_UNI_DETECT |
8577 PERL_PV_PRETTY_ELIPSES |
8580 Perl_sv_catpvf(aTHX_ sv, " %s", s );
8581 } else if (k == TRIE) {
8582 /* print the details of the trie in dumpuntil instead, as
8583 * progi->data isn't available here */
8584 const char op = OP(o);
8585 const U32 n = ARG(o);
8586 const reg_ac_data * const ac = IS_TRIE_AC(op) ?
8587 (reg_ac_data *)progi->data->data[n] :
8589 const reg_trie_data * const trie
8590 = (reg_trie_data*)progi->data->data[!IS_TRIE_AC(op) ? n : ac->trie];
8592 Perl_sv_catpvf(aTHX_ sv, "-%s",PL_reg_name[o->flags]);
8593 DEBUG_TRIE_COMPILE_r(
8594 Perl_sv_catpvf(aTHX_ sv,
8595 "<S:%"UVuf"/%"IVdf" W:%"UVuf" L:%"UVuf"/%"UVuf" C:%"UVuf"/%"UVuf">",
8596 (UV)trie->startstate,
8597 (IV)trie->statecount-1, /* -1 because of the unused 0 element */
8598 (UV)trie->wordcount,
8601 (UV)TRIE_CHARCOUNT(trie),
8602 (UV)trie->uniquecharcount
8605 if ( IS_ANYOF_TRIE(op) || trie->bitmap ) {
8607 int rangestart = -1;
8608 U8* bitmap = IS_ANYOF_TRIE(op) ? (U8*)ANYOF_BITMAP(o) : (U8*)TRIE_BITMAP(trie);
8609 Perl_sv_catpvf(aTHX_ sv, "[");
8610 for (i = 0; i <= 256; i++) {
8611 if (i < 256 && BITMAP_TEST(bitmap,i)) {
8612 if (rangestart == -1)
8614 } else if (rangestart != -1) {
8615 if (i <= rangestart + 3)
8616 for (; rangestart < i; rangestart++)
8617 put_byte(sv, rangestart);
8619 put_byte(sv, rangestart);
8621 put_byte(sv, i - 1);
8626 Perl_sv_catpvf(aTHX_ sv, "]");
8629 } else if (k == CURLY) {
8630 if (OP(o) == CURLYM || OP(o) == CURLYN || OP(o) == CURLYX)
8631 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* Parenth number */
8632 Perl_sv_catpvf(aTHX_ sv, " {%d,%d}", ARG1(o), ARG2(o));
8634 else if (k == WHILEM && o->flags) /* Ordinal/of */
8635 Perl_sv_catpvf(aTHX_ sv, "[%d/%d]", o->flags & 0xf, o->flags>>4);
8636 else if (k == REF || k == OPEN || k == CLOSE || k == GROUPP || OP(o)==ACCEPT) {
8637 Perl_sv_catpvf(aTHX_ sv, "%d", (int)ARG(o)); /* Parenth number */
8638 if ( prog->paren_names ) {
8639 if ( k != REF || OP(o) < NREF) {
8640 AV *list= (AV *)progi->data->data[progi->name_list_idx];
8641 SV **name= av_fetch(list, ARG(o), 0 );
8643 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
8646 AV *list= (AV *)progi->data->data[ progi->name_list_idx ];
8647 SV *sv_dat=(SV*)progi->data->data[ ARG( o ) ];
8648 I32 *nums=(I32*)SvPVX(sv_dat);
8649 SV **name= av_fetch(list, nums[0], 0 );
8652 for ( n=0; n<SvIVX(sv_dat); n++ ) {
8653 Perl_sv_catpvf(aTHX_ sv, "%s%"IVdf,
8654 (n ? "," : ""), (IV)nums[n]);
8656 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
8660 } else if (k == GOSUB)
8661 Perl_sv_catpvf(aTHX_ sv, "%d[%+d]", (int)ARG(o),(int)ARG2L(o)); /* Paren and offset */
8662 else if (k == VERB) {
8664 Perl_sv_catpvf(aTHX_ sv, ":%"SVf,
8665 SVfARG((SV*)progi->data->data[ ARG( o ) ]));
8666 } else if (k == LOGICAL)
8667 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* 2: embedded, otherwise 1 */
8668 else if (k == FOLDCHAR)
8669 Perl_sv_catpvf(aTHX_ sv, "[0x%"UVXf"]",ARG(o) );
8670 else if (k == ANYOF) {
8671 int i, rangestart = -1;
8672 const U8 flags = ANYOF_FLAGS(o);
8674 /* Should be synchronized with * ANYOF_ #xdefines in regcomp.h */
8675 static const char * const anyofs[] = {
8708 if (flags & ANYOF_LOCALE)
8709 sv_catpvs(sv, "{loc}");
8710 if (flags & ANYOF_FOLD)
8711 sv_catpvs(sv, "{i}");
8712 Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
8713 if (flags & ANYOF_INVERT)
8715 for (i = 0; i <= 256; i++) {
8716 if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
8717 if (rangestart == -1)
8719 } else if (rangestart != -1) {
8720 if (i <= rangestart + 3)
8721 for (; rangestart < i; rangestart++)
8722 put_byte(sv, rangestart);
8724 put_byte(sv, rangestart);
8726 put_byte(sv, i - 1);
8732 if (o->flags & ANYOF_CLASS)
8733 for (i = 0; i < (int)(sizeof(anyofs)/sizeof(char*)); i++)
8734 if (ANYOF_CLASS_TEST(o,i))
8735 sv_catpv(sv, anyofs[i]);
8737 if (flags & ANYOF_UNICODE)
8738 sv_catpvs(sv, "{unicode}");
8739 else if (flags & ANYOF_UNICODE_ALL)
8740 sv_catpvs(sv, "{unicode_all}");
8744 SV * const sw = regclass_swash(prog, o, FALSE, &lv, 0);
8748 U8 s[UTF8_MAXBYTES_CASE+1];
8750 for (i = 0; i <= 256; i++) { /* just the first 256 */
8751 uvchr_to_utf8(s, i);
8753 if (i < 256 && swash_fetch(sw, s, TRUE)) {
8754 if (rangestart == -1)
8756 } else if (rangestart != -1) {
8757 if (i <= rangestart + 3)
8758 for (; rangestart < i; rangestart++) {
8759 const U8 * const e = uvchr_to_utf8(s,rangestart);
8761 for(p = s; p < e; p++)
8765 const U8 *e = uvchr_to_utf8(s,rangestart);
8767 for (p = s; p < e; p++)
8770 e = uvchr_to_utf8(s, i-1);
8771 for (p = s; p < e; p++)
8778 sv_catpvs(sv, "..."); /* et cetera */
8782 char *s = savesvpv(lv);
8783 char * const origs = s;
8785 while (*s && *s != '\n')
8789 const char * const t = ++s;
8807 Perl_sv_catpvf(aTHX_ sv, "%s]", PL_colors[1]);
8809 else if (k == BRANCHJ && (OP(o) == UNLESSM || OP(o) == IFMATCH))
8810 Perl_sv_catpvf(aTHX_ sv, "[%d]", -(o->flags));
8812 PERL_UNUSED_CONTEXT;
8813 PERL_UNUSED_ARG(sv);
8815 PERL_UNUSED_ARG(prog);
8816 #endif /* DEBUGGING */
8820 Perl_re_intuit_string(pTHX_ REGEXP * const prog)
8821 { /* Assume that RE_INTUIT is set */
8823 GET_RE_DEBUG_FLAGS_DECL;
8824 PERL_UNUSED_CONTEXT;
8828 const char * const s = SvPV_nolen_const(prog->check_substr
8829 ? prog->check_substr : prog->check_utf8);
8831 if (!PL_colorset) reginitcolors();
8832 PerlIO_printf(Perl_debug_log,
8833 "%sUsing REx %ssubstr:%s \"%s%.60s%s%s\"\n",
8835 prog->check_substr ? "" : "utf8 ",
8836 PL_colors[5],PL_colors[0],
8839 (strlen(s) > 60 ? "..." : ""));
8842 return prog->check_substr ? prog->check_substr : prog->check_utf8;
8848 handles refcounting and freeing the perl core regexp structure. When
8849 it is necessary to actually free the structure the first thing it
8850 does is call the 'free' method of the regexp_engine associated to to
8851 the regexp, allowing the handling of the void *pprivate; member
8852 first. (This routine is not overridable by extensions, which is why
8853 the extensions free is called first.)
8855 See regdupe and regdupe_internal if you change anything here.
8857 #ifndef PERL_IN_XSUB_RE
8859 Perl_pregfree(pTHX_ struct regexp *r)
8862 GET_RE_DEBUG_FLAGS_DECL;
8864 if (!r || (--r->refcnt > 0))
8867 ReREFCNT_dec(r->mother_re);
8869 CALLREGFREE_PVT(r); /* free the private data */
8871 SvREFCNT_dec(r->paren_names);
8872 Safefree(r->wrapped);
8875 if (r->anchored_substr)
8876 SvREFCNT_dec(r->anchored_substr);
8877 if (r->anchored_utf8)
8878 SvREFCNT_dec(r->anchored_utf8);
8879 if (r->float_substr)
8880 SvREFCNT_dec(r->float_substr);
8882 SvREFCNT_dec(r->float_utf8);
8883 Safefree(r->substrs);
8885 RX_MATCH_COPY_FREE(r);
8886 #ifdef PERL_OLD_COPY_ON_WRITE
8888 SvREFCNT_dec(r->saved_copy);
8897 This is a hacky workaround to the structural issue of match results
8898 being stored in the regexp structure which is in turn stored in
8899 PL_curpm/PL_reg_curpm. The problem is that due to qr// the pattern
8900 could be PL_curpm in multiple contexts, and could require multiple
8901 result sets being associated with the pattern simultaneously, such
8902 as when doing a recursive match with (??{$qr})
8904 The solution is to make a lightweight copy of the regexp structure
8905 when a qr// is returned from the code executed by (??{$qr}) this
8906 lightweight copy doesnt actually own any of its data except for
8907 the starp/end and the actual regexp structure itself.
8913 Perl_reg_temp_copy (pTHX_ struct regexp *r) {
8915 register const I32 npar = r->nparens+1;
8916 (void)ReREFCNT_inc(r);
8917 Newx(ret, 1, regexp);
8918 StructCopy(r, ret, regexp);
8919 Newx(ret->offs, npar, regexp_paren_pair);
8920 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
8923 Newx(ret->substrs, 1, struct reg_substr_data);
8924 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
8926 SvREFCNT_inc_void(ret->anchored_substr);
8927 SvREFCNT_inc_void(ret->anchored_utf8);
8928 SvREFCNT_inc_void(ret->float_substr);
8929 SvREFCNT_inc_void(ret->float_utf8);
8931 /* check_substr and check_utf8, if non-NULL, point to either their
8932 anchored or float namesakes, and don't hold a second reference. */
8934 RX_MATCH_COPIED_off(ret);
8935 #ifdef PERL_OLD_COPY_ON_WRITE
8936 ret->saved_copy = NULL;
8945 /* regfree_internal()
8947 Free the private data in a regexp. This is overloadable by
8948 extensions. Perl takes care of the regexp structure in pregfree(),
8949 this covers the *pprivate pointer which technically perldoesnt
8950 know about, however of course we have to handle the
8951 regexp_internal structure when no extension is in use.
8953 Note this is called before freeing anything in the regexp
8958 Perl_regfree_internal(pTHX_ REGEXP * const r)
8962 GET_RE_DEBUG_FLAGS_DECL;
8968 SV *dsv= sv_newmortal();
8969 RE_PV_QUOTED_DECL(s, (r->extflags & RXf_UTF8),
8970 dsv, r->precomp, r->prelen, 60);
8971 PerlIO_printf(Perl_debug_log,"%sFreeing REx:%s %s\n",
8972 PL_colors[4],PL_colors[5],s);
8975 #ifdef RE_TRACK_PATTERN_OFFSETS
8977 Safefree(ri->u.offsets); /* 20010421 MJD */
8980 int n = ri->data->count;
8981 PAD* new_comppad = NULL;
8986 /* If you add a ->what type here, update the comment in regcomp.h */
8987 switch (ri->data->what[n]) {
8991 SvREFCNT_dec((SV*)ri->data->data[n]);
8994 Safefree(ri->data->data[n]);
8997 new_comppad = (AV*)ri->data->data[n];
9000 if (new_comppad == NULL)
9001 Perl_croak(aTHX_ "panic: pregfree comppad");
9002 PAD_SAVE_LOCAL(old_comppad,
9003 /* Watch out for global destruction's random ordering. */
9004 (SvTYPE(new_comppad) == SVt_PVAV) ? new_comppad : NULL
9007 refcnt = OpREFCNT_dec((OP_4tree*)ri->data->data[n]);
9010 op_free((OP_4tree*)ri->data->data[n]);
9012 PAD_RESTORE_LOCAL(old_comppad);
9013 SvREFCNT_dec((SV*)new_comppad);
9019 { /* Aho Corasick add-on structure for a trie node.
9020 Used in stclass optimization only */
9022 reg_ac_data *aho=(reg_ac_data*)ri->data->data[n];
9024 refcount = --aho->refcount;
9027 PerlMemShared_free(aho->states);
9028 PerlMemShared_free(aho->fail);
9029 /* do this last!!!! */
9030 PerlMemShared_free(ri->data->data[n]);
9031 PerlMemShared_free(ri->regstclass);
9037 /* trie structure. */
9039 reg_trie_data *trie=(reg_trie_data*)ri->data->data[n];
9041 refcount = --trie->refcount;
9044 PerlMemShared_free(trie->charmap);
9045 PerlMemShared_free(trie->states);
9046 PerlMemShared_free(trie->trans);
9048 PerlMemShared_free(trie->bitmap);
9050 PerlMemShared_free(trie->wordlen);
9052 PerlMemShared_free(trie->jump);
9054 PerlMemShared_free(trie->nextword);
9055 /* do this last!!!! */
9056 PerlMemShared_free(ri->data->data[n]);
9061 Perl_croak(aTHX_ "panic: regfree data code '%c'", ri->data->what[n]);
9064 Safefree(ri->data->what);
9071 #define sv_dup_inc(s,t) SvREFCNT_inc(sv_dup(s,t))
9072 #define av_dup_inc(s,t) (AV*)SvREFCNT_inc(sv_dup((SV*)s,t))
9073 #define hv_dup_inc(s,t) (HV*)SvREFCNT_inc(sv_dup((SV*)s,t))
9074 #define SAVEPVN(p,n) ((p) ? savepvn(p,n) : NULL)
9077 re_dup - duplicate a regexp.
9079 This routine is expected to clone a given regexp structure. It is not
9080 compiler under USE_ITHREADS.
9082 After all of the core data stored in struct regexp is duplicated
9083 the regexp_engine.dupe method is used to copy any private data
9084 stored in the *pprivate pointer. This allows extensions to handle
9085 any duplication it needs to do.
9087 See pregfree() and regfree_internal() if you change anything here.
9089 #if defined(USE_ITHREADS)
9090 #ifndef PERL_IN_XSUB_RE
9092 Perl_re_dup(pTHX_ const regexp *r, CLONE_PARAMS *param)
9099 return (REGEXP *)NULL;
9101 if ((ret = (REGEXP *)ptr_table_fetch(PL_ptr_table, r)))
9105 npar = r->nparens+1;
9106 Newx(ret, 1, regexp);
9107 StructCopy(r, ret, regexp);
9108 Newx(ret->offs, npar, regexp_paren_pair);
9109 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9111 /* no need to copy these */
9112 Newx(ret->swap, npar, regexp_paren_pair);
9116 /* Do it this way to avoid reading from *r after the StructCopy().
9117 That way, if any of the sv_dup_inc()s dislodge *r from the L1
9118 cache, it doesn't matter. */
9119 const bool anchored = r->check_substr == r->anchored_substr;
9120 Newx(ret->substrs, 1, struct reg_substr_data);
9121 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9123 ret->anchored_substr = sv_dup_inc(ret->anchored_substr, param);
9124 ret->anchored_utf8 = sv_dup_inc(ret->anchored_utf8, param);
9125 ret->float_substr = sv_dup_inc(ret->float_substr, param);
9126 ret->float_utf8 = sv_dup_inc(ret->float_utf8, param);
9128 /* check_substr and check_utf8, if non-NULL, point to either their
9129 anchored or float namesakes, and don't hold a second reference. */
9131 if (ret->check_substr) {
9133 assert(r->check_utf8 == r->anchored_utf8);
9134 ret->check_substr = ret->anchored_substr;
9135 ret->check_utf8 = ret->anchored_utf8;
9137 assert(r->check_substr == r->float_substr);
9138 assert(r->check_utf8 == r->float_utf8);
9139 ret->check_substr = ret->float_substr;
9140 ret->check_utf8 = ret->float_utf8;
9145 ret->wrapped = SAVEPVN(ret->wrapped, ret->wraplen+1);
9146 ret->precomp = ret->wrapped + (ret->precomp - ret->wrapped);
9147 ret->paren_names = hv_dup_inc(ret->paren_names, param);
9150 RXi_SET(ret,CALLREGDUPE_PVT(ret,param));
9152 if (RX_MATCH_COPIED(ret))
9153 ret->subbeg = SAVEPVN(ret->subbeg, ret->sublen);
9156 #ifdef PERL_OLD_COPY_ON_WRITE
9157 ret->saved_copy = NULL;
9160 ret->mother_re = NULL;
9162 ret->seen_evals = 0;
9164 ptr_table_store(PL_ptr_table, r, ret);
9167 #endif /* PERL_IN_XSUB_RE */
9172 This is the internal complement to regdupe() which is used to copy
9173 the structure pointed to by the *pprivate pointer in the regexp.
9174 This is the core version of the extension overridable cloning hook.
9175 The regexp structure being duplicated will be copied by perl prior
9176 to this and will be provided as the regexp *r argument, however
9177 with the /old/ structures pprivate pointer value. Thus this routine
9178 may override any copying normally done by perl.
9180 It returns a pointer to the new regexp_internal structure.
9184 Perl_regdupe_internal(pTHX_ REGEXP * const r, CLONE_PARAMS *param)
9187 regexp_internal *reti;
9191 npar = r->nparens+1;
9194 Newxc(reti, sizeof(regexp_internal) + (len+1)*sizeof(regnode), char, regexp_internal);
9195 Copy(ri->program, reti->program, len+1, regnode);
9198 reti->regstclass = NULL;
9202 const int count = ri->data->count;
9205 Newxc(d, sizeof(struct reg_data) + count*sizeof(void *),
9206 char, struct reg_data);
9207 Newx(d->what, count, U8);
9210 for (i = 0; i < count; i++) {
9211 d->what[i] = ri->data->what[i];
9212 switch (d->what[i]) {
9213 /* legal options are one of: sSfpontTu
9214 see also regcomp.h and pregfree() */
9217 case 'p': /* actually an AV, but the dup function is identical. */
9218 case 'u': /* actually an HV, but the dup function is identical. */
9219 d->data[i] = sv_dup_inc((SV *)ri->data->data[i], param);
9222 /* This is cheating. */
9223 Newx(d->data[i], 1, struct regnode_charclass_class);
9224 StructCopy(ri->data->data[i], d->data[i],
9225 struct regnode_charclass_class);
9226 reti->regstclass = (regnode*)d->data[i];
9229 /* Compiled op trees are readonly and in shared memory,
9230 and can thus be shared without duplication. */
9232 d->data[i] = (void*)OpREFCNT_inc((OP*)ri->data->data[i]);
9236 /* Trie stclasses are readonly and can thus be shared
9237 * without duplication. We free the stclass in pregfree
9238 * when the corresponding reg_ac_data struct is freed.
9240 reti->regstclass= ri->regstclass;
9244 ((reg_trie_data*)ri->data->data[i])->refcount++;
9248 d->data[i] = ri->data->data[i];
9251 Perl_croak(aTHX_ "panic: re_dup unknown data code '%c'", ri->data->what[i]);
9260 reti->name_list_idx = ri->name_list_idx;
9262 #ifdef RE_TRACK_PATTERN_OFFSETS
9263 if (ri->u.offsets) {
9264 Newx(reti->u.offsets, 2*len+1, U32);
9265 Copy(ri->u.offsets, reti->u.offsets, 2*len+1, U32);
9268 SetProgLen(reti,len);
9274 #endif /* USE_ITHREADS */
9279 converts a regexp embedded in a MAGIC struct to its stringified form,
9280 caching the converted form in the struct and returns the cached
9283 If lp is nonnull then it is used to return the length of the
9286 If flags is nonnull and the returned string contains UTF8 then
9287 (*flags & 1) will be true.
9289 If haseval is nonnull then it is used to return whether the pattern
9292 Normally called via macro:
9294 CALLREG_STRINGIFY(mg,&len,&utf8);
9298 CALLREG_AS_STR(mg,&lp,&flags,&haseval)
9300 See sv_2pv_flags() in sv.c for an example of internal usage.
9303 #ifndef PERL_IN_XSUB_RE
9306 Perl_reg_stringify(pTHX_ MAGIC *mg, STRLEN *lp, U32 *flags, I32 *haseval ) {
9308 const regexp * const re = (regexp *)mg->mg_obj;
9310 *haseval = re->seen_evals;
9312 *flags = ((re->extflags & RXf_UTF8) ? 1 : 0);
9319 - regnext - dig the "next" pointer out of a node
9322 Perl_regnext(pTHX_ register regnode *p)
9325 register I32 offset;
9330 offset = (reg_off_by_arg[OP(p)] ? ARG(p) : NEXT_OFF(p));
9339 S_re_croak2(pTHX_ const char* pat1,const char* pat2,...)
9342 STRLEN l1 = strlen(pat1);
9343 STRLEN l2 = strlen(pat2);
9346 const char *message;
9352 Copy(pat1, buf, l1 , char);
9353 Copy(pat2, buf + l1, l2 , char);
9354 buf[l1 + l2] = '\n';
9355 buf[l1 + l2 + 1] = '\0';
9357 /* ANSI variant takes additional second argument */
9358 va_start(args, pat2);
9362 msv = vmess(buf, &args);
9364 message = SvPV_const(msv,l1);
9367 Copy(message, buf, l1 , char);
9368 buf[l1-1] = '\0'; /* Overwrite \n */
9369 Perl_croak(aTHX_ "%s", buf);
9372 /* XXX Here's a total kludge. But we need to re-enter for swash routines. */
9374 #ifndef PERL_IN_XSUB_RE
9376 Perl_save_re_context(pTHX)
9380 struct re_save_state *state;
9382 SAVEVPTR(PL_curcop);
9383 SSGROW(SAVESTACK_ALLOC_FOR_RE_SAVE_STATE + 1);
9385 state = (struct re_save_state *)(PL_savestack + PL_savestack_ix);
9386 PL_savestack_ix += SAVESTACK_ALLOC_FOR_RE_SAVE_STATE;
9387 SSPUSHINT(SAVEt_RE_STATE);
9389 Copy(&PL_reg_state, state, 1, struct re_save_state);
9391 PL_reg_start_tmp = 0;
9392 PL_reg_start_tmpl = 0;
9393 PL_reg_oldsaved = NULL;
9394 PL_reg_oldsavedlen = 0;
9396 PL_reg_leftiter = 0;
9397 PL_reg_poscache = NULL;
9398 PL_reg_poscache_size = 0;
9399 #ifdef PERL_OLD_COPY_ON_WRITE
9403 /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
9405 const REGEXP * const rx = PM_GETRE(PL_curpm);
9408 for (i = 1; i <= rx->nparens; i++) {
9409 char digits[TYPE_CHARS(long)];
9410 const STRLEN len = my_snprintf(digits, sizeof(digits), "%lu", (long)i);
9411 GV *const *const gvp
9412 = (GV**)hv_fetch(PL_defstash, digits, len, 0);
9415 GV * const gv = *gvp;
9416 if (SvTYPE(gv) == SVt_PVGV && GvSV(gv))
9426 clear_re(pTHX_ void *r)
9429 ReREFCNT_dec((regexp *)r);
9435 S_put_byte(pTHX_ SV *sv, int c)
9437 if (isCNTRL(c) || c == 255 || !isPRINT(c))
9438 Perl_sv_catpvf(aTHX_ sv, "\\%o", c);
9439 else if (c == '-' || c == ']' || c == '\\' || c == '^')
9440 Perl_sv_catpvf(aTHX_ sv, "\\%c", c);
9442 Perl_sv_catpvf(aTHX_ sv, "%c", c);
9446 #define CLEAR_OPTSTART \
9447 if (optstart) STMT_START { \
9448 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log, " (%"IVdf" nodes)\n", (IV)(node - optstart))); \
9452 #define DUMPUNTIL(b,e) CLEAR_OPTSTART; node=dumpuntil(r,start,(b),(e),last,sv,indent+1,depth+1);
9454 STATIC const regnode *
9455 S_dumpuntil(pTHX_ const regexp *r, const regnode *start, const regnode *node,
9456 const regnode *last, const regnode *plast,
9457 SV* sv, I32 indent, U32 depth)
9460 register U8 op = PSEUDO; /* Arbitrary non-END op. */
9461 register const regnode *next;
9462 const regnode *optstart= NULL;
9465 GET_RE_DEBUG_FLAGS_DECL;
9467 #ifdef DEBUG_DUMPUNTIL
9468 PerlIO_printf(Perl_debug_log, "--- %d : %d - %d - %d\n",indent,node-start,
9469 last ? last-start : 0,plast ? plast-start : 0);
9472 if (plast && plast < last)
9475 while (PL_regkind[op] != END && (!last || node < last)) {
9476 /* While that wasn't END last time... */
9479 if (op == CLOSE || op == WHILEM)
9481 next = regnext((regnode *)node);
9484 if (OP(node) == OPTIMIZED) {
9485 if (!optstart && RE_DEBUG_FLAG(RE_DEBUG_COMPILE_OPTIMISE))
9492 regprop(r, sv, node);
9493 PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
9494 (int)(2*indent + 1), "", SvPVX_const(sv));
9496 if (OP(node) != OPTIMIZED) {
9497 if (next == NULL) /* Next ptr. */
9498 PerlIO_printf(Perl_debug_log, " (0)");
9499 else if (PL_regkind[(U8)op] == BRANCH && PL_regkind[OP(next)] != BRANCH )
9500 PerlIO_printf(Perl_debug_log, " (FAIL)");
9502 PerlIO_printf(Perl_debug_log, " (%"IVdf")", (IV)(next - start));
9503 (void)PerlIO_putc(Perl_debug_log, '\n');
9507 if (PL_regkind[(U8)op] == BRANCHJ) {
9510 register const regnode *nnode = (OP(next) == LONGJMP
9511 ? regnext((regnode *)next)
9513 if (last && nnode > last)
9515 DUMPUNTIL(NEXTOPER(NEXTOPER(node)), nnode);
9518 else if (PL_regkind[(U8)op] == BRANCH) {
9520 DUMPUNTIL(NEXTOPER(node), next);
9522 else if ( PL_regkind[(U8)op] == TRIE ) {
9523 const regnode *this_trie = node;
9524 const char op = OP(node);
9525 const U32 n = ARG(node);
9526 const reg_ac_data * const ac = op>=AHOCORASICK ?
9527 (reg_ac_data *)ri->data->data[n] :
9529 const reg_trie_data * const trie =
9530 (reg_trie_data*)ri->data->data[op<AHOCORASICK ? n : ac->trie];
9532 AV *const trie_words = (AV *) ri->data->data[n + TRIE_WORDS_OFFSET];
9534 const regnode *nextbranch= NULL;
9536 sv_setpvn(sv, "", 0);
9537 for (word_idx= 0; word_idx < (I32)trie->wordcount; word_idx++) {
9538 SV ** const elem_ptr = av_fetch(trie_words,word_idx,0);
9540 PerlIO_printf(Perl_debug_log, "%*s%s ",
9541 (int)(2*(indent+3)), "",
9542 elem_ptr ? pv_pretty(sv, SvPV_nolen_const(*elem_ptr), SvCUR(*elem_ptr), 60,
9543 PL_colors[0], PL_colors[1],
9544 (SvUTF8(*elem_ptr) ? PERL_PV_ESCAPE_UNI : 0) |
9545 PERL_PV_PRETTY_ELIPSES |
9551 U16 dist= trie->jump[word_idx+1];
9552 PerlIO_printf(Perl_debug_log, "(%"UVuf")\n",
9553 (UV)((dist ? this_trie + dist : next) - start));
9556 nextbranch= this_trie + trie->jump[0];
9557 DUMPUNTIL(this_trie + dist, nextbranch);
9559 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
9560 nextbranch= regnext((regnode *)nextbranch);
9562 PerlIO_printf(Perl_debug_log, "\n");
9565 if (last && next > last)
9570 else if ( op == CURLY ) { /* "next" might be very big: optimizer */
9571 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS,
9572 NEXTOPER(node) + EXTRA_STEP_2ARGS + 1);
9574 else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
9576 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS, next);
9578 else if ( op == PLUS || op == STAR) {
9579 DUMPUNTIL(NEXTOPER(node), NEXTOPER(node) + 1);
9581 else if (op == ANYOF) {
9582 /* arglen 1 + class block */
9583 node += 1 + ((ANYOF_FLAGS(node) & ANYOF_LARGE)
9584 ? ANYOF_CLASS_SKIP : ANYOF_SKIP);
9585 node = NEXTOPER(node);
9587 else if (PL_regkind[(U8)op] == EXACT) {
9588 /* Literal string, where present. */
9589 node += NODE_SZ_STR(node) - 1;
9590 node = NEXTOPER(node);
9593 node = NEXTOPER(node);
9594 node += regarglen[(U8)op];
9596 if (op == CURLYX || op == OPEN)
9600 #ifdef DEBUG_DUMPUNTIL
9601 PerlIO_printf(Perl_debug_log, "--- %d\n", (int)indent);
9606 #endif /* DEBUGGING */
9610 * c-indentation-style: bsd
9612 * indent-tabs-mode: t
9615 * ex: set ts=8 sts=4 sw=4 noet:
projects / p5sagit/p5-mst-13.2.git / blob