5 * 'A fair jaw-cracker dwarf-language must be.' --Samwise Gamgee
7 * [p.285 of _The Lord of the Rings_, II/iii: "The Ring Goes South"]
10 /* This file contains functions for compiling a regular expression. See
11 * also regexec.c which funnily enough, contains functions for executing
12 * a regular expression.
14 * This file is also copied at build time to ext/re/re_comp.c, where
15 * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
16 * This causes the main functions to be compiled under new names and with
17 * debugging support added, which makes "use re 'debug'" work.
20 /* NOTE: this is derived from Henry Spencer's regexp code, and should not
21 * confused with the original package (see point 3 below). Thanks, Henry!
24 /* Additional note: this code is very heavily munged from Henry's version
25 * in places. In some spots I've traded clarity for efficiency, so don't
26 * blame Henry for some of the lack of readability.
29 /* The names of the functions have been changed from regcomp and
30 * regexec to pregcomp and pregexec in order to avoid conflicts
31 * with the POSIX routines of the same names.
34 #ifdef PERL_EXT_RE_BUILD
39 * pregcomp and pregexec -- regsub and regerror are not used in perl
41 * Copyright (c) 1986 by University of Toronto.
42 * Written by Henry Spencer. Not derived from licensed software.
44 * Permission is granted to anyone to use this software for any
45 * purpose on any computer system, and to redistribute it freely,
46 * subject to the following restrictions:
48 * 1. The author is not responsible for the consequences of use of
49 * this software, no matter how awful, even if they arise
52 * 2. The origin of this software must not be misrepresented, either
53 * by explicit claim or by omission.
55 * 3. Altered versions must be plainly marked as such, and must not
56 * be misrepresented as being the original software.
59 **** Alterations to Henry's code are...
61 **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
62 **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
63 **** by Larry Wall and others
65 **** You may distribute under the terms of either the GNU General Public
66 **** License or the Artistic License, as specified in the README file.
69 * Beware that some of this code is subtly aware of the way operator
70 * precedence is structured in regular expressions. Serious changes in
71 * regular-expression syntax might require a total rethink.
74 #define PERL_IN_REGCOMP_C
77 #ifndef PERL_IN_XSUB_RE
82 #ifdef PERL_IN_XSUB_RE
93 # if defined(BUGGY_MSC6)
94 /* MSC 6.00A breaks on op/regexp.t test 85 unless we turn this off */
95 # pragma optimize("a",off)
96 /* But MSC 6.00A is happy with 'w', for aliases only across function calls*/
97 # pragma optimize("w",on )
98 # endif /* BUGGY_MSC6 */
102 #define STATIC static
105 typedef struct RExC_state_t {
106 U32 flags; /* are we folding, multilining? */
107 char *precomp; /* uncompiled string. */
108 REGEXP *rx_sv; /* The SV that is the regexp. */
109 regexp *rx; /* perl core regexp structure */
110 regexp_internal *rxi; /* internal data for regexp object pprivate field */
111 char *start; /* Start of input for compile */
112 char *end; /* End of input for compile */
113 char *parse; /* Input-scan pointer. */
114 I32 whilem_seen; /* number of WHILEM in this expr */
115 regnode *emit_start; /* Start of emitted-code area */
116 regnode *emit_bound; /* First regnode outside of the allocated space */
117 regnode *emit; /* Code-emit pointer; ®dummy = don't = compiling */
118 I32 naughty; /* How bad is this pattern? */
119 I32 sawback; /* Did we see \1, ...? */
121 I32 size; /* Code size. */
122 I32 npar; /* Capture buffer count, (OPEN). */
123 I32 cpar; /* Capture buffer count, (CLOSE). */
124 I32 nestroot; /* root parens we are in - used by accept */
128 regnode **open_parens; /* pointers to open parens */
129 regnode **close_parens; /* pointers to close parens */
130 regnode *opend; /* END node in program */
131 I32 utf8; /* whether the pattern is utf8 or not */
132 I32 orig_utf8; /* whether the pattern was originally in utf8 */
133 /* XXX use this for future optimisation of case
134 * where pattern must be upgraded to utf8. */
135 HV *paren_names; /* Paren names */
137 regnode **recurse; /* Recurse regops */
138 I32 recurse_count; /* Number of recurse regops */
140 char *starttry; /* -Dr: where regtry was called. */
141 #define RExC_starttry (pRExC_state->starttry)
144 const char *lastparse;
146 AV *paren_name_list; /* idx -> name */
147 #define RExC_lastparse (pRExC_state->lastparse)
148 #define RExC_lastnum (pRExC_state->lastnum)
149 #define RExC_paren_name_list (pRExC_state->paren_name_list)
153 #define RExC_flags (pRExC_state->flags)
154 #define RExC_precomp (pRExC_state->precomp)
155 #define RExC_rx_sv (pRExC_state->rx_sv)
156 #define RExC_rx (pRExC_state->rx)
157 #define RExC_rxi (pRExC_state->rxi)
158 #define RExC_start (pRExC_state->start)
159 #define RExC_end (pRExC_state->end)
160 #define RExC_parse (pRExC_state->parse)
161 #define RExC_whilem_seen (pRExC_state->whilem_seen)
162 #ifdef RE_TRACK_PATTERN_OFFSETS
163 #define RExC_offsets (pRExC_state->rxi->u.offsets) /* I am not like the others */
165 #define RExC_emit (pRExC_state->emit)
166 #define RExC_emit_start (pRExC_state->emit_start)
167 #define RExC_emit_bound (pRExC_state->emit_bound)
168 #define RExC_naughty (pRExC_state->naughty)
169 #define RExC_sawback (pRExC_state->sawback)
170 #define RExC_seen (pRExC_state->seen)
171 #define RExC_size (pRExC_state->size)
172 #define RExC_npar (pRExC_state->npar)
173 #define RExC_nestroot (pRExC_state->nestroot)
174 #define RExC_extralen (pRExC_state->extralen)
175 #define RExC_seen_zerolen (pRExC_state->seen_zerolen)
176 #define RExC_seen_evals (pRExC_state->seen_evals)
177 #define RExC_utf8 (pRExC_state->utf8)
178 #define RExC_orig_utf8 (pRExC_state->orig_utf8)
179 #define RExC_open_parens (pRExC_state->open_parens)
180 #define RExC_close_parens (pRExC_state->close_parens)
181 #define RExC_opend (pRExC_state->opend)
182 #define RExC_paren_names (pRExC_state->paren_names)
183 #define RExC_recurse (pRExC_state->recurse)
184 #define RExC_recurse_count (pRExC_state->recurse_count)
187 #define ISMULT1(c) ((c) == '*' || (c) == '+' || (c) == '?')
188 #define ISMULT2(s) ((*s) == '*' || (*s) == '+' || (*s) == '?' || \
189 ((*s) == '{' && regcurly(s)))
192 #undef SPSTART /* dratted cpp namespace... */
195 * Flags to be passed up and down.
197 #define WORST 0 /* Worst case. */
198 #define HASWIDTH 0x01 /* Known to match non-null strings. */
199 #define SIMPLE 0x02 /* Simple enough to be STAR/PLUS operand. */
200 #define SPSTART 0x04 /* Starts with * or +. */
201 #define TRYAGAIN 0x08 /* Weeded out a declaration. */
202 #define POSTPONED 0x10 /* (?1),(?&name), (??{...}) or similar */
204 #define REG_NODE_NUM(x) ((x) ? (int)((x)-RExC_emit_start) : -1)
206 /* whether trie related optimizations are enabled */
207 #if PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION
208 #define TRIE_STUDY_OPT
209 #define FULL_TRIE_STUDY
215 #define PBYTE(u8str,paren) ((U8*)(u8str))[(paren) >> 3]
216 #define PBITVAL(paren) (1 << ((paren) & 7))
217 #define PAREN_TEST(u8str,paren) ( PBYTE(u8str,paren) & PBITVAL(paren))
218 #define PAREN_SET(u8str,paren) PBYTE(u8str,paren) |= PBITVAL(paren)
219 #define PAREN_UNSET(u8str,paren) PBYTE(u8str,paren) &= (~PBITVAL(paren))
222 /* About scan_data_t.
224 During optimisation we recurse through the regexp program performing
225 various inplace (keyhole style) optimisations. In addition study_chunk
226 and scan_commit populate this data structure with information about
227 what strings MUST appear in the pattern. We look for the longest
228 string that must appear for at a fixed location, and we look for the
229 longest string that may appear at a floating location. So for instance
234 Both 'FOO' and 'A' are fixed strings. Both 'B' and 'BAR' are floating
235 strings (because they follow a .* construct). study_chunk will identify
236 both FOO and BAR as being the longest fixed and floating strings respectively.
238 The strings can be composites, for instance
242 will result in a composite fixed substring 'foo'.
244 For each string some basic information is maintained:
246 - offset or min_offset
247 This is the position the string must appear at, or not before.
248 It also implicitly (when combined with minlenp) tells us how many
249 character must match before the string we are searching.
250 Likewise when combined with minlenp and the length of the string
251 tells us how many characters must appear after the string we have
255 Only used for floating strings. This is the rightmost point that
256 the string can appear at. Ifset to I32 max it indicates that the
257 string can occur infinitely far to the right.
260 A pointer to the minimum length of the pattern that the string
261 was found inside. This is important as in the case of positive
262 lookahead or positive lookbehind we can have multiple patterns
267 The minimum length of the pattern overall is 3, the minimum length
268 of the lookahead part is 3, but the minimum length of the part that
269 will actually match is 1. So 'FOO's minimum length is 3, but the
270 minimum length for the F is 1. This is important as the minimum length
271 is used to determine offsets in front of and behind the string being
272 looked for. Since strings can be composites this is the length of the
273 pattern at the time it was commited with a scan_commit. Note that
274 the length is calculated by study_chunk, so that the minimum lengths
275 are not known until the full pattern has been compiled, thus the
276 pointer to the value.
280 In the case of lookbehind the string being searched for can be
281 offset past the start point of the final matching string.
282 If this value was just blithely removed from the min_offset it would
283 invalidate some of the calculations for how many chars must match
284 before or after (as they are derived from min_offset and minlen and
285 the length of the string being searched for).
286 When the final pattern is compiled and the data is moved from the
287 scan_data_t structure into the regexp structure the information
288 about lookbehind is factored in, with the information that would
289 have been lost precalculated in the end_shift field for the
292 The fields pos_min and pos_delta are used to store the minimum offset
293 and the delta to the maximum offset at the current point in the pattern.
297 typedef struct scan_data_t {
298 /*I32 len_min; unused */
299 /*I32 len_delta; unused */
303 I32 last_end; /* min value, <0 unless valid. */
306 SV **longest; /* Either &l_fixed, or &l_float. */
307 SV *longest_fixed; /* longest fixed string found in pattern */
308 I32 offset_fixed; /* offset where it starts */
309 I32 *minlen_fixed; /* pointer to the minlen relevent to the string */
310 I32 lookbehind_fixed; /* is the position of the string modfied by LB */
311 SV *longest_float; /* longest floating string found in pattern */
312 I32 offset_float_min; /* earliest point in string it can appear */
313 I32 offset_float_max; /* latest point in string it can appear */
314 I32 *minlen_float; /* pointer to the minlen relevent to the string */
315 I32 lookbehind_float; /* is the position of the string modified by LB */
319 struct regnode_charclass_class *start_class;
323 * Forward declarations for pregcomp()'s friends.
326 static const scan_data_t zero_scan_data =
327 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ,0};
329 #define SF_BEFORE_EOL (SF_BEFORE_SEOL|SF_BEFORE_MEOL)
330 #define SF_BEFORE_SEOL 0x0001
331 #define SF_BEFORE_MEOL 0x0002
332 #define SF_FIX_BEFORE_EOL (SF_FIX_BEFORE_SEOL|SF_FIX_BEFORE_MEOL)
333 #define SF_FL_BEFORE_EOL (SF_FL_BEFORE_SEOL|SF_FL_BEFORE_MEOL)
336 # define SF_FIX_SHIFT_EOL (0+2)
337 # define SF_FL_SHIFT_EOL (0+4)
339 # define SF_FIX_SHIFT_EOL (+2)
340 # define SF_FL_SHIFT_EOL (+4)
343 #define SF_FIX_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FIX_SHIFT_EOL)
344 #define SF_FIX_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FIX_SHIFT_EOL)
346 #define SF_FL_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FL_SHIFT_EOL)
347 #define SF_FL_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FL_SHIFT_EOL) /* 0x20 */
348 #define SF_IS_INF 0x0040
349 #define SF_HAS_PAR 0x0080
350 #define SF_IN_PAR 0x0100
351 #define SF_HAS_EVAL 0x0200
352 #define SCF_DO_SUBSTR 0x0400
353 #define SCF_DO_STCLASS_AND 0x0800
354 #define SCF_DO_STCLASS_OR 0x1000
355 #define SCF_DO_STCLASS (SCF_DO_STCLASS_AND|SCF_DO_STCLASS_OR)
356 #define SCF_WHILEM_VISITED_POS 0x2000
358 #define SCF_TRIE_RESTUDY 0x4000 /* Do restudy? */
359 #define SCF_SEEN_ACCEPT 0x8000
361 #define UTF (RExC_utf8 != 0)
362 #define LOC ((RExC_flags & RXf_PMf_LOCALE) != 0)
363 #define FOLD ((RExC_flags & RXf_PMf_FOLD) != 0)
365 #define OOB_UNICODE 12345678
366 #define OOB_NAMEDCLASS -1
368 #define CHR_SVLEN(sv) (UTF ? sv_len_utf8(sv) : SvCUR(sv))
369 #define CHR_DIST(a,b) (UTF ? utf8_distance(a,b) : a - b)
372 /* length of regex to show in messages that don't mark a position within */
373 #define RegexLengthToShowInErrorMessages 127
376 * If MARKER[12] are adjusted, be sure to adjust the constants at the top
377 * of t/op/regmesg.t, the tests in t/op/re_tests, and those in
378 * op/pragma/warn/regcomp.
380 #define MARKER1 "<-- HERE" /* marker as it appears in the description */
381 #define MARKER2 " <-- HERE " /* marker as it appears within the regex */
383 #define REPORT_LOCATION " in regex; marked by " MARKER1 " in m/%.*s" MARKER2 "%s/"
386 * Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
387 * arg. Show regex, up to a maximum length. If it's too long, chop and add
390 #define _FAIL(code) STMT_START { \
391 const char *ellipses = ""; \
392 IV len = RExC_end - RExC_precomp; \
395 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
396 if (len > RegexLengthToShowInErrorMessages) { \
397 /* chop 10 shorter than the max, to ensure meaning of "..." */ \
398 len = RegexLengthToShowInErrorMessages - 10; \
404 #define FAIL(msg) _FAIL( \
405 Perl_croak(aTHX_ "%s in regex m/%.*s%s/", \
406 msg, (int)len, RExC_precomp, ellipses))
408 #define FAIL2(msg,arg) _FAIL( \
409 Perl_croak(aTHX_ msg " in regex m/%.*s%s/", \
410 arg, (int)len, RExC_precomp, ellipses))
413 * Simple_vFAIL -- like FAIL, but marks the current location in the scan
415 #define Simple_vFAIL(m) STMT_START { \
416 const IV offset = RExC_parse - RExC_precomp; \
417 Perl_croak(aTHX_ "%s" REPORT_LOCATION, \
418 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
422 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL()
424 #define vFAIL(m) STMT_START { \
426 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
431 * Like Simple_vFAIL(), but accepts two arguments.
433 #define Simple_vFAIL2(m,a1) STMT_START { \
434 const IV offset = RExC_parse - RExC_precomp; \
435 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, \
436 (int)offset, RExC_precomp, RExC_precomp + offset); \
440 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL2().
442 #define vFAIL2(m,a1) STMT_START { \
444 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
445 Simple_vFAIL2(m, a1); \
450 * Like Simple_vFAIL(), but accepts three arguments.
452 #define Simple_vFAIL3(m, a1, a2) STMT_START { \
453 const IV offset = RExC_parse - RExC_precomp; \
454 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, \
455 (int)offset, RExC_precomp, RExC_precomp + offset); \
459 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL3().
461 #define vFAIL3(m,a1,a2) STMT_START { \
463 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
464 Simple_vFAIL3(m, a1, a2); \
468 * Like Simple_vFAIL(), but accepts four arguments.
470 #define Simple_vFAIL4(m, a1, a2, a3) STMT_START { \
471 const IV offset = RExC_parse - RExC_precomp; \
472 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, \
473 (int)offset, RExC_precomp, RExC_precomp + offset); \
476 #define ckWARNreg(loc,m) STMT_START { \
477 const IV offset = loc - RExC_precomp; \
478 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
479 (int)offset, RExC_precomp, RExC_precomp + offset); \
482 #define ckWARNregdep(loc,m) STMT_START { \
483 const IV offset = loc - RExC_precomp; \
484 Perl_ck_warner_d(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
486 (int)offset, RExC_precomp, RExC_precomp + offset); \
489 #define ckWARN2reg(loc, m, a1) STMT_START { \
490 const IV offset = loc - RExC_precomp; \
491 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
492 a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
495 #define vWARN3(loc, m, a1, a2) STMT_START { \
496 const IV offset = loc - RExC_precomp; \
497 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
498 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
501 #define ckWARN3reg(loc, m, a1, a2) STMT_START { \
502 const IV offset = loc - RExC_precomp; \
503 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
504 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
507 #define vWARN4(loc, m, a1, a2, a3) STMT_START { \
508 const IV offset = loc - RExC_precomp; \
509 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
510 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
513 #define ckWARN4reg(loc, m, a1, a2, a3) STMT_START { \
514 const IV offset = loc - RExC_precomp; \
515 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
516 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
519 #define vWARN5(loc, m, a1, a2, a3, a4) STMT_START { \
520 const IV offset = loc - RExC_precomp; \
521 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
522 a1, a2, a3, a4, (int)offset, RExC_precomp, RExC_precomp + offset); \
526 /* Allow for side effects in s */
527 #define REGC(c,s) STMT_START { \
528 if (!SIZE_ONLY) *(s) = (c); else (void)(s); \
531 /* Macros for recording node offsets. 20001227 mjd@plover.com
532 * Nodes are numbered 1, 2, 3, 4. Node #n's position is recorded in
533 * element 2*n-1 of the array. Element #2n holds the byte length node #n.
534 * Element 0 holds the number n.
535 * Position is 1 indexed.
537 #ifndef RE_TRACK_PATTERN_OFFSETS
538 #define Set_Node_Offset_To_R(node,byte)
539 #define Set_Node_Offset(node,byte)
540 #define Set_Cur_Node_Offset
541 #define Set_Node_Length_To_R(node,len)
542 #define Set_Node_Length(node,len)
543 #define Set_Node_Cur_Length(node)
544 #define Node_Offset(n)
545 #define Node_Length(n)
546 #define Set_Node_Offset_Length(node,offset,len)
547 #define ProgLen(ri) ri->u.proglen
548 #define SetProgLen(ri,x) ri->u.proglen = x
550 #define ProgLen(ri) ri->u.offsets[0]
551 #define SetProgLen(ri,x) ri->u.offsets[0] = x
552 #define Set_Node_Offset_To_R(node,byte) STMT_START { \
554 MJD_OFFSET_DEBUG(("** (%d) offset of node %d is %d.\n", \
555 __LINE__, (int)(node), (int)(byte))); \
557 Perl_croak(aTHX_ "value of node is %d in Offset macro", (int)(node)); \
559 RExC_offsets[2*(node)-1] = (byte); \
564 #define Set_Node_Offset(node,byte) \
565 Set_Node_Offset_To_R((node)-RExC_emit_start, (byte)-RExC_start)
566 #define Set_Cur_Node_Offset Set_Node_Offset(RExC_emit, RExC_parse)
568 #define Set_Node_Length_To_R(node,len) STMT_START { \
570 MJD_OFFSET_DEBUG(("** (%d) size of node %d is %d.\n", \
571 __LINE__, (int)(node), (int)(len))); \
573 Perl_croak(aTHX_ "value of node is %d in Length macro", (int)(node)); \
575 RExC_offsets[2*(node)] = (len); \
580 #define Set_Node_Length(node,len) \
581 Set_Node_Length_To_R((node)-RExC_emit_start, len)
582 #define Set_Cur_Node_Length(len) Set_Node_Length(RExC_emit, len)
583 #define Set_Node_Cur_Length(node) \
584 Set_Node_Length(node, RExC_parse - parse_start)
586 /* Get offsets and lengths */
587 #define Node_Offset(n) (RExC_offsets[2*((n)-RExC_emit_start)-1])
588 #define Node_Length(n) (RExC_offsets[2*((n)-RExC_emit_start)])
590 #define Set_Node_Offset_Length(node,offset,len) STMT_START { \
591 Set_Node_Offset_To_R((node)-RExC_emit_start, (offset)); \
592 Set_Node_Length_To_R((node)-RExC_emit_start, (len)); \
596 #if PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS
597 #define EXPERIMENTAL_INPLACESCAN
598 #endif /*RE_TRACK_PATTERN_OFFSETS*/
600 #define DEBUG_STUDYDATA(str,data,depth) \
601 DEBUG_OPTIMISE_MORE_r(if(data){ \
602 PerlIO_printf(Perl_debug_log, \
603 "%*s" str "Pos:%"IVdf"/%"IVdf \
604 " Flags: 0x%"UVXf" Whilem_c: %"IVdf" Lcp: %"IVdf" %s", \
605 (int)(depth)*2, "", \
606 (IV)((data)->pos_min), \
607 (IV)((data)->pos_delta), \
608 (UV)((data)->flags), \
609 (IV)((data)->whilem_c), \
610 (IV)((data)->last_closep ? *((data)->last_closep) : -1), \
611 is_inf ? "INF " : "" \
613 if ((data)->last_found) \
614 PerlIO_printf(Perl_debug_log, \
615 "Last:'%s' %"IVdf":%"IVdf"/%"IVdf" %sFixed:'%s' @ %"IVdf \
616 " %sFloat: '%s' @ %"IVdf"/%"IVdf"", \
617 SvPVX_const((data)->last_found), \
618 (IV)((data)->last_end), \
619 (IV)((data)->last_start_min), \
620 (IV)((data)->last_start_max), \
621 ((data)->longest && \
622 (data)->longest==&((data)->longest_fixed)) ? "*" : "", \
623 SvPVX_const((data)->longest_fixed), \
624 (IV)((data)->offset_fixed), \
625 ((data)->longest && \
626 (data)->longest==&((data)->longest_float)) ? "*" : "", \
627 SvPVX_const((data)->longest_float), \
628 (IV)((data)->offset_float_min), \
629 (IV)((data)->offset_float_max) \
631 PerlIO_printf(Perl_debug_log,"\n"); \
634 static void clear_re(pTHX_ void *r);
636 /* Mark that we cannot extend a found fixed substring at this point.
637 Update the longest found anchored substring and the longest found
638 floating substrings if needed. */
641 S_scan_commit(pTHX_ const RExC_state_t *pRExC_state, scan_data_t *data, I32 *minlenp, int is_inf)
643 const STRLEN l = CHR_SVLEN(data->last_found);
644 const STRLEN old_l = CHR_SVLEN(*data->longest);
645 GET_RE_DEBUG_FLAGS_DECL;
647 PERL_ARGS_ASSERT_SCAN_COMMIT;
649 if ((l >= old_l) && ((l > old_l) || (data->flags & SF_BEFORE_EOL))) {
650 SvSetMagicSV(*data->longest, data->last_found);
651 if (*data->longest == data->longest_fixed) {
652 data->offset_fixed = l ? data->last_start_min : data->pos_min;
653 if (data->flags & SF_BEFORE_EOL)
655 |= ((data->flags & SF_BEFORE_EOL) << SF_FIX_SHIFT_EOL);
657 data->flags &= ~SF_FIX_BEFORE_EOL;
658 data->minlen_fixed=minlenp;
659 data->lookbehind_fixed=0;
661 else { /* *data->longest == data->longest_float */
662 data->offset_float_min = l ? data->last_start_min : data->pos_min;
663 data->offset_float_max = (l
664 ? data->last_start_max
665 : data->pos_min + data->pos_delta);
666 if (is_inf || (U32)data->offset_float_max > (U32)I32_MAX)
667 data->offset_float_max = I32_MAX;
668 if (data->flags & SF_BEFORE_EOL)
670 |= ((data->flags & SF_BEFORE_EOL) << SF_FL_SHIFT_EOL);
672 data->flags &= ~SF_FL_BEFORE_EOL;
673 data->minlen_float=minlenp;
674 data->lookbehind_float=0;
677 SvCUR_set(data->last_found, 0);
679 SV * const sv = data->last_found;
680 if (SvUTF8(sv) && SvMAGICAL(sv)) {
681 MAGIC * const mg = mg_find(sv, PERL_MAGIC_utf8);
687 data->flags &= ~SF_BEFORE_EOL;
688 DEBUG_STUDYDATA("commit: ",data,0);
691 /* Can match anything (initialization) */
693 S_cl_anything(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
695 PERL_ARGS_ASSERT_CL_ANYTHING;
697 ANYOF_CLASS_ZERO(cl);
698 ANYOF_BITMAP_SETALL(cl);
699 cl->flags = ANYOF_EOS|ANYOF_UNICODE_ALL;
701 cl->flags |= ANYOF_LOCALE;
704 /* Can match anything (initialization) */
706 S_cl_is_anything(const struct regnode_charclass_class *cl)
710 PERL_ARGS_ASSERT_CL_IS_ANYTHING;
712 for (value = 0; value <= ANYOF_MAX; value += 2)
713 if (ANYOF_CLASS_TEST(cl, value) && ANYOF_CLASS_TEST(cl, value + 1))
715 if (!(cl->flags & ANYOF_UNICODE_ALL))
717 if (!ANYOF_BITMAP_TESTALLSET((const void*)cl))
722 /* Can match anything (initialization) */
724 S_cl_init(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
726 PERL_ARGS_ASSERT_CL_INIT;
728 Zero(cl, 1, struct regnode_charclass_class);
730 cl_anything(pRExC_state, cl);
734 S_cl_init_zero(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
736 PERL_ARGS_ASSERT_CL_INIT_ZERO;
738 Zero(cl, 1, struct regnode_charclass_class);
740 cl_anything(pRExC_state, cl);
742 cl->flags |= ANYOF_LOCALE;
745 /* 'And' a given class with another one. Can create false positives */
746 /* We assume that cl is not inverted */
748 S_cl_and(struct regnode_charclass_class *cl,
749 const struct regnode_charclass_class *and_with)
751 PERL_ARGS_ASSERT_CL_AND;
753 assert(and_with->type == ANYOF);
754 if (!(and_with->flags & ANYOF_CLASS)
755 && !(cl->flags & ANYOF_CLASS)
756 && (and_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
757 && !(and_with->flags & ANYOF_FOLD)
758 && !(cl->flags & ANYOF_FOLD)) {
761 if (and_with->flags & ANYOF_INVERT)
762 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
763 cl->bitmap[i] &= ~and_with->bitmap[i];
765 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
766 cl->bitmap[i] &= and_with->bitmap[i];
767 } /* XXXX: logic is complicated otherwise, leave it along for a moment. */
768 if (!(and_with->flags & ANYOF_EOS))
769 cl->flags &= ~ANYOF_EOS;
771 if (cl->flags & ANYOF_UNICODE_ALL && and_with->flags & ANYOF_UNICODE &&
772 !(and_with->flags & ANYOF_INVERT)) {
773 cl->flags &= ~ANYOF_UNICODE_ALL;
774 cl->flags |= ANYOF_UNICODE;
775 ARG_SET(cl, ARG(and_with));
777 if (!(and_with->flags & ANYOF_UNICODE_ALL) &&
778 !(and_with->flags & ANYOF_INVERT))
779 cl->flags &= ~ANYOF_UNICODE_ALL;
780 if (!(and_with->flags & (ANYOF_UNICODE|ANYOF_UNICODE_ALL)) &&
781 !(and_with->flags & ANYOF_INVERT))
782 cl->flags &= ~ANYOF_UNICODE;
785 /* 'OR' a given class with another one. Can create false positives */
786 /* We assume that cl is not inverted */
788 S_cl_or(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, const struct regnode_charclass_class *or_with)
790 PERL_ARGS_ASSERT_CL_OR;
792 if (or_with->flags & ANYOF_INVERT) {
794 * (B1 | CL1) | (!B2 & !CL2) = (B1 | !B2 & !CL2) | (CL1 | (!B2 & !CL2))
795 * <= (B1 | !B2) | (CL1 | !CL2)
796 * which is wasteful if CL2 is small, but we ignore CL2:
797 * (B1 | CL1) | (!B2 & !CL2) <= (B1 | CL1) | !B2 = (B1 | !B2) | CL1
798 * XXXX Can we handle case-fold? Unclear:
799 * (OK1(i) | OK1(i')) | !(OK1(i) | OK1(i')) =
800 * (OK1(i) | OK1(i')) | (!OK1(i) & !OK1(i'))
802 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
803 && !(or_with->flags & ANYOF_FOLD)
804 && !(cl->flags & ANYOF_FOLD) ) {
807 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
808 cl->bitmap[i] |= ~or_with->bitmap[i];
809 } /* XXXX: logic is complicated otherwise */
811 cl_anything(pRExC_state, cl);
814 /* (B1 | CL1) | (B2 | CL2) = (B1 | B2) | (CL1 | CL2)) */
815 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
816 && (!(or_with->flags & ANYOF_FOLD)
817 || (cl->flags & ANYOF_FOLD)) ) {
820 /* OR char bitmap and class bitmap separately */
821 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
822 cl->bitmap[i] |= or_with->bitmap[i];
823 if (or_with->flags & ANYOF_CLASS) {
824 for (i = 0; i < ANYOF_CLASSBITMAP_SIZE; i++)
825 cl->classflags[i] |= or_with->classflags[i];
826 cl->flags |= ANYOF_CLASS;
829 else { /* XXXX: logic is complicated, leave it along for a moment. */
830 cl_anything(pRExC_state, cl);
833 if (or_with->flags & ANYOF_EOS)
834 cl->flags |= ANYOF_EOS;
836 if (cl->flags & ANYOF_UNICODE && or_with->flags & ANYOF_UNICODE &&
837 ARG(cl) != ARG(or_with)) {
838 cl->flags |= ANYOF_UNICODE_ALL;
839 cl->flags &= ~ANYOF_UNICODE;
841 if (or_with->flags & ANYOF_UNICODE_ALL) {
842 cl->flags |= ANYOF_UNICODE_ALL;
843 cl->flags &= ~ANYOF_UNICODE;
847 #define TRIE_LIST_ITEM(state,idx) (trie->states[state].trans.list)[ idx ]
848 #define TRIE_LIST_CUR(state) ( TRIE_LIST_ITEM( state, 0 ).forid )
849 #define TRIE_LIST_LEN(state) ( TRIE_LIST_ITEM( state, 0 ).newstate )
850 #define TRIE_LIST_USED(idx) ( trie->states[state].trans.list ? (TRIE_LIST_CUR( idx ) - 1) : 0 )
855 dump_trie(trie,widecharmap,revcharmap)
856 dump_trie_interim_list(trie,widecharmap,revcharmap,next_alloc)
857 dump_trie_interim_table(trie,widecharmap,revcharmap,next_alloc)
859 These routines dump out a trie in a somewhat readable format.
860 The _interim_ variants are used for debugging the interim
861 tables that are used to generate the final compressed
862 representation which is what dump_trie expects.
864 Part of the reason for their existance is to provide a form
865 of documentation as to how the different representations function.
870 Dumps the final compressed table form of the trie to Perl_debug_log.
871 Used for debugging make_trie().
875 S_dump_trie(pTHX_ const struct _reg_trie_data *trie, HV *widecharmap,
876 AV *revcharmap, U32 depth)
879 SV *sv=sv_newmortal();
880 int colwidth= widecharmap ? 6 : 4;
882 GET_RE_DEBUG_FLAGS_DECL;
884 PERL_ARGS_ASSERT_DUMP_TRIE;
886 PerlIO_printf( Perl_debug_log, "%*sChar : %-6s%-6s%-4s ",
887 (int)depth * 2 + 2,"",
888 "Match","Base","Ofs" );
890 for( state = 0 ; state < trie->uniquecharcount ; state++ ) {
891 SV ** const tmp = av_fetch( revcharmap, state, 0);
893 PerlIO_printf( Perl_debug_log, "%*s",
895 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
896 PL_colors[0], PL_colors[1],
897 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
898 PERL_PV_ESCAPE_FIRSTCHAR
903 PerlIO_printf( Perl_debug_log, "\n%*sState|-----------------------",
904 (int)depth * 2 + 2,"");
906 for( state = 0 ; state < trie->uniquecharcount ; state++ )
907 PerlIO_printf( Perl_debug_log, "%.*s", colwidth, "--------");
908 PerlIO_printf( Perl_debug_log, "\n");
910 for( state = 1 ; state < trie->statecount ; state++ ) {
911 const U32 base = trie->states[ state ].trans.base;
913 PerlIO_printf( Perl_debug_log, "%*s#%4"UVXf"|", (int)depth * 2 + 2,"", (UV)state);
915 if ( trie->states[ state ].wordnum ) {
916 PerlIO_printf( Perl_debug_log, " W%4X", trie->states[ state ].wordnum );
918 PerlIO_printf( Perl_debug_log, "%6s", "" );
921 PerlIO_printf( Perl_debug_log, " @%4"UVXf" ", (UV)base );
926 while( ( base + ofs < trie->uniquecharcount ) ||
927 ( base + ofs - trie->uniquecharcount < trie->lasttrans
928 && trie->trans[ base + ofs - trie->uniquecharcount ].check != state))
931 PerlIO_printf( Perl_debug_log, "+%2"UVXf"[ ", (UV)ofs);
933 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
934 if ( ( base + ofs >= trie->uniquecharcount ) &&
935 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
936 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
938 PerlIO_printf( Perl_debug_log, "%*"UVXf,
940 (UV)trie->trans[ base + ofs - trie->uniquecharcount ].next );
942 PerlIO_printf( Perl_debug_log, "%*s",colwidth," ." );
946 PerlIO_printf( Perl_debug_log, "]");
949 PerlIO_printf( Perl_debug_log, "\n" );
951 PerlIO_printf(Perl_debug_log, "%*sword_info N:(prev,len)=", (int)depth*2, "");
952 for (word=1; word <= trie->wordcount; word++) {
953 PerlIO_printf(Perl_debug_log, " %d:(%d,%d)",
954 (int)word, (int)(trie->wordinfo[word].prev),
955 (int)(trie->wordinfo[word].len));
957 PerlIO_printf(Perl_debug_log, "\n" );
960 Dumps a fully constructed but uncompressed trie in list form.
961 List tries normally only are used for construction when the number of
962 possible chars (trie->uniquecharcount) is very high.
963 Used for debugging make_trie().
966 S_dump_trie_interim_list(pTHX_ const struct _reg_trie_data *trie,
967 HV *widecharmap, AV *revcharmap, U32 next_alloc,
971 SV *sv=sv_newmortal();
972 int colwidth= widecharmap ? 6 : 4;
973 GET_RE_DEBUG_FLAGS_DECL;
975 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_LIST;
977 /* print out the table precompression. */
978 PerlIO_printf( Perl_debug_log, "%*sState :Word | Transition Data\n%*s%s",
979 (int)depth * 2 + 2,"", (int)depth * 2 + 2,"",
980 "------:-----+-----------------\n" );
982 for( state=1 ; state < next_alloc ; state ++ ) {
985 PerlIO_printf( Perl_debug_log, "%*s %4"UVXf" :",
986 (int)depth * 2 + 2,"", (UV)state );
987 if ( ! trie->states[ state ].wordnum ) {
988 PerlIO_printf( Perl_debug_log, "%5s| ","");
990 PerlIO_printf( Perl_debug_log, "W%4x| ",
991 trie->states[ state ].wordnum
994 for( charid = 1 ; charid <= TRIE_LIST_USED( state ) ; charid++ ) {
995 SV ** const tmp = av_fetch( revcharmap, TRIE_LIST_ITEM(state,charid).forid, 0);
997 PerlIO_printf( Perl_debug_log, "%*s:%3X=%4"UVXf" | ",
999 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1000 PL_colors[0], PL_colors[1],
1001 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1002 PERL_PV_ESCAPE_FIRSTCHAR
1004 TRIE_LIST_ITEM(state,charid).forid,
1005 (UV)TRIE_LIST_ITEM(state,charid).newstate
1008 PerlIO_printf(Perl_debug_log, "\n%*s| ",
1009 (int)((depth * 2) + 14), "");
1012 PerlIO_printf( Perl_debug_log, "\n");
1017 Dumps a fully constructed but uncompressed trie in table form.
1018 This is the normal DFA style state transition table, with a few
1019 twists to facilitate compression later.
1020 Used for debugging make_trie().
1023 S_dump_trie_interim_table(pTHX_ const struct _reg_trie_data *trie,
1024 HV *widecharmap, AV *revcharmap, U32 next_alloc,
1029 SV *sv=sv_newmortal();
1030 int colwidth= widecharmap ? 6 : 4;
1031 GET_RE_DEBUG_FLAGS_DECL;
1033 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_TABLE;
1036 print out the table precompression so that we can do a visual check
1037 that they are identical.
1040 PerlIO_printf( Perl_debug_log, "%*sChar : ",(int)depth * 2 + 2,"" );
1042 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1043 SV ** const tmp = av_fetch( revcharmap, charid, 0);
1045 PerlIO_printf( Perl_debug_log, "%*s",
1047 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1048 PL_colors[0], PL_colors[1],
1049 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1050 PERL_PV_ESCAPE_FIRSTCHAR
1056 PerlIO_printf( Perl_debug_log, "\n%*sState+-",(int)depth * 2 + 2,"" );
1058 for( charid=0 ; charid < trie->uniquecharcount ; charid++ ) {
1059 PerlIO_printf( Perl_debug_log, "%.*s", colwidth,"--------");
1062 PerlIO_printf( Perl_debug_log, "\n" );
1064 for( state=1 ; state < next_alloc ; state += trie->uniquecharcount ) {
1066 PerlIO_printf( Perl_debug_log, "%*s%4"UVXf" : ",
1067 (int)depth * 2 + 2,"",
1068 (UV)TRIE_NODENUM( state ) );
1070 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1071 UV v=(UV)SAFE_TRIE_NODENUM( trie->trans[ state + charid ].next );
1073 PerlIO_printf( Perl_debug_log, "%*"UVXf, colwidth, v );
1075 PerlIO_printf( Perl_debug_log, "%*s", colwidth, "." );
1077 if ( ! trie->states[ TRIE_NODENUM( state ) ].wordnum ) {
1078 PerlIO_printf( Perl_debug_log, " (%4"UVXf")\n", (UV)trie->trans[ state ].check );
1080 PerlIO_printf( Perl_debug_log, " (%4"UVXf") W%4X\n", (UV)trie->trans[ state ].check,
1081 trie->states[ TRIE_NODENUM( state ) ].wordnum );
1089 /* make_trie(startbranch,first,last,tail,word_count,flags,depth)
1090 startbranch: the first branch in the whole branch sequence
1091 first : start branch of sequence of branch-exact nodes.
1092 May be the same as startbranch
1093 last : Thing following the last branch.
1094 May be the same as tail.
1095 tail : item following the branch sequence
1096 count : words in the sequence
1097 flags : currently the OP() type we will be building one of /EXACT(|F|Fl)/
1098 depth : indent depth
1100 Inplace optimizes a sequence of 2 or more Branch-Exact nodes into a TRIE node.
1102 A trie is an N'ary tree where the branches are determined by digital
1103 decomposition of the key. IE, at the root node you look up the 1st character and
1104 follow that branch repeat until you find the end of the branches. Nodes can be
1105 marked as "accepting" meaning they represent a complete word. Eg:
1109 would convert into the following structure. Numbers represent states, letters
1110 following numbers represent valid transitions on the letter from that state, if
1111 the number is in square brackets it represents an accepting state, otherwise it
1112 will be in parenthesis.
1114 +-h->+-e->[3]-+-r->(8)-+-s->[9]
1118 (1) +-i->(6)-+-s->[7]
1120 +-s->(3)-+-h->(4)-+-e->[5]
1122 Accept Word Mapping: 3=>1 (he),5=>2 (she), 7=>3 (his), 9=>4 (hers)
1124 This shows that when matching against the string 'hers' we will begin at state 1
1125 read 'h' and move to state 2, read 'e' and move to state 3 which is accepting,
1126 then read 'r' and go to state 8 followed by 's' which takes us to state 9 which
1127 is also accepting. Thus we know that we can match both 'he' and 'hers' with a
1128 single traverse. We store a mapping from accepting to state to which word was
1129 matched, and then when we have multiple possibilities we try to complete the
1130 rest of the regex in the order in which they occured in the alternation.
1132 The only prior NFA like behaviour that would be changed by the TRIE support is
1133 the silent ignoring of duplicate alternations which are of the form:
1135 / (DUPE|DUPE) X? (?{ ... }) Y /x
1137 Thus EVAL blocks follwing a trie may be called a different number of times with
1138 and without the optimisation. With the optimisations dupes will be silently
1139 ignored. This inconsistant behaviour of EVAL type nodes is well established as
1140 the following demonstrates:
1142 'words'=~/(word|word|word)(?{ print $1 })[xyz]/
1144 which prints out 'word' three times, but
1146 'words'=~/(word|word|word)(?{ print $1 })S/
1148 which doesnt print it out at all. This is due to other optimisations kicking in.
1150 Example of what happens on a structural level:
1152 The regexp /(ac|ad|ab)+/ will produce the folowing debug output:
1154 1: CURLYM[1] {1,32767}(18)
1165 This would be optimizable with startbranch=5, first=5, last=16, tail=16
1166 and should turn into:
1168 1: CURLYM[1] {1,32767}(18)
1170 [Words:3 Chars Stored:6 Unique Chars:4 States:5 NCP:1]
1178 Cases where tail != last would be like /(?foo|bar)baz/:
1188 which would be optimizable with startbranch=1, first=1, last=7, tail=8
1189 and would end up looking like:
1192 [Words:2 Chars Stored:6 Unique Chars:5 States:7 NCP:1]
1199 d = uvuni_to_utf8_flags(d, uv, 0);
1201 is the recommended Unicode-aware way of saying
1206 #define TRIE_STORE_REVCHAR \
1209 SV *zlopp = newSV(2); \
1210 unsigned char *flrbbbbb = (unsigned char *) SvPVX(zlopp); \
1211 unsigned const char *const kapow = uvuni_to_utf8(flrbbbbb, uvc & 0xFF); \
1212 SvCUR_set(zlopp, kapow - flrbbbbb); \
1215 av_push(revcharmap, zlopp); \
1217 char ooooff = (char)uvc; \
1218 av_push(revcharmap, newSVpvn(&ooooff, 1)); \
1222 #define TRIE_READ_CHAR STMT_START { \
1226 if ( foldlen > 0 ) { \
1227 uvc = utf8n_to_uvuni( scan, UTF8_MAXLEN, &len, uniflags ); \
1232 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1233 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1234 foldlen -= UNISKIP( uvc ); \
1235 scan = foldbuf + UNISKIP( uvc ); \
1238 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1248 #define TRIE_LIST_PUSH(state,fid,ns) STMT_START { \
1249 if ( TRIE_LIST_CUR( state ) >=TRIE_LIST_LEN( state ) ) { \
1250 U32 ging = TRIE_LIST_LEN( state ) *= 2; \
1251 Renew( trie->states[ state ].trans.list, ging, reg_trie_trans_le ); \
1253 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).forid = fid; \
1254 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).newstate = ns; \
1255 TRIE_LIST_CUR( state )++; \
1258 #define TRIE_LIST_NEW(state) STMT_START { \
1259 Newxz( trie->states[ state ].trans.list, \
1260 4, reg_trie_trans_le ); \
1261 TRIE_LIST_CUR( state ) = 1; \
1262 TRIE_LIST_LEN( state ) = 4; \
1265 #define TRIE_HANDLE_WORD(state) STMT_START { \
1266 U16 dupe= trie->states[ state ].wordnum; \
1267 regnode * const noper_next = regnext( noper ); \
1270 /* store the word for dumping */ \
1272 if (OP(noper) != NOTHING) \
1273 tmp = newSVpvn_utf8(STRING(noper), STR_LEN(noper), UTF); \
1275 tmp = newSVpvn_utf8( "", 0, UTF ); \
1276 av_push( trie_words, tmp ); \
1280 trie->wordinfo[curword].prev = 0; \
1281 trie->wordinfo[curword].len = wordlen; \
1282 trie->wordinfo[curword].accept = state; \
1284 if ( noper_next < tail ) { \
1286 trie->jump = (U16 *) PerlMemShared_calloc( word_count + 1, sizeof(U16) ); \
1287 trie->jump[curword] = (U16)(noper_next - convert); \
1289 jumper = noper_next; \
1291 nextbranch= regnext(cur); \
1295 /* It's a dupe. Pre-insert into the wordinfo[].prev */\
1296 /* chain, so that when the bits of chain are later */\
1297 /* linked together, the dups appear in the chain */\
1298 trie->wordinfo[curword].prev = trie->wordinfo[dupe].prev; \
1299 trie->wordinfo[dupe].prev = curword; \
1301 /* we haven't inserted this word yet. */ \
1302 trie->states[ state ].wordnum = curword; \
1307 #define TRIE_TRANS_STATE(state,base,ucharcount,charid,special) \
1308 ( ( base + charid >= ucharcount \
1309 && base + charid < ubound \
1310 && state == trie->trans[ base - ucharcount + charid ].check \
1311 && trie->trans[ base - ucharcount + charid ].next ) \
1312 ? trie->trans[ base - ucharcount + charid ].next \
1313 : ( state==1 ? special : 0 ) \
1317 #define MADE_JUMP_TRIE 2
1318 #define MADE_EXACT_TRIE 4
1321 S_make_trie(pTHX_ RExC_state_t *pRExC_state, regnode *startbranch, regnode *first, regnode *last, regnode *tail, U32 word_count, U32 flags, U32 depth)
1324 /* first pass, loop through and scan words */
1325 reg_trie_data *trie;
1326 HV *widecharmap = NULL;
1327 AV *revcharmap = newAV();
1329 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1334 regnode *jumper = NULL;
1335 regnode *nextbranch = NULL;
1336 regnode *convert = NULL;
1337 U32 *prev_states; /* temp array mapping each state to previous one */
1338 /* we just use folder as a flag in utf8 */
1339 const U8 * const folder = ( flags == EXACTF
1341 : ( flags == EXACTFL
1348 const U32 data_slot = add_data( pRExC_state, 4, "tuuu" );
1349 AV *trie_words = NULL;
1350 /* along with revcharmap, this only used during construction but both are
1351 * useful during debugging so we store them in the struct when debugging.
1354 const U32 data_slot = add_data( pRExC_state, 2, "tu" );
1355 STRLEN trie_charcount=0;
1357 SV *re_trie_maxbuff;
1358 GET_RE_DEBUG_FLAGS_DECL;
1360 PERL_ARGS_ASSERT_MAKE_TRIE;
1362 PERL_UNUSED_ARG(depth);
1365 trie = (reg_trie_data *) PerlMemShared_calloc( 1, sizeof(reg_trie_data) );
1367 trie->startstate = 1;
1368 trie->wordcount = word_count;
1369 RExC_rxi->data->data[ data_slot ] = (void*)trie;
1370 trie->charmap = (U16 *) PerlMemShared_calloc( 256, sizeof(U16) );
1371 if (!(UTF && folder))
1372 trie->bitmap = (char *) PerlMemShared_calloc( ANYOF_BITMAP_SIZE, 1 );
1373 trie->wordinfo = (reg_trie_wordinfo *) PerlMemShared_calloc(
1374 trie->wordcount+1, sizeof(reg_trie_wordinfo));
1377 trie_words = newAV();
1380 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
1381 if (!SvIOK(re_trie_maxbuff)) {
1382 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
1385 PerlIO_printf( Perl_debug_log,
1386 "%*smake_trie start==%d, first==%d, last==%d, tail==%d depth=%d\n",
1387 (int)depth * 2 + 2, "",
1388 REG_NODE_NUM(startbranch),REG_NODE_NUM(first),
1389 REG_NODE_NUM(last), REG_NODE_NUM(tail),
1393 /* Find the node we are going to overwrite */
1394 if ( first == startbranch && OP( last ) != BRANCH ) {
1395 /* whole branch chain */
1398 /* branch sub-chain */
1399 convert = NEXTOPER( first );
1402 /* -- First loop and Setup --
1404 We first traverse the branches and scan each word to determine if it
1405 contains widechars, and how many unique chars there are, this is
1406 important as we have to build a table with at least as many columns as we
1409 We use an array of integers to represent the character codes 0..255
1410 (trie->charmap) and we use a an HV* to store Unicode characters. We use the
1411 native representation of the character value as the key and IV's for the
1414 *TODO* If we keep track of how many times each character is used we can
1415 remap the columns so that the table compression later on is more
1416 efficient in terms of memory by ensuring most common value is in the
1417 middle and the least common are on the outside. IMO this would be better
1418 than a most to least common mapping as theres a decent chance the most
1419 common letter will share a node with the least common, meaning the node
1420 will not be compressable. With a middle is most common approach the worst
1421 case is when we have the least common nodes twice.
1425 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1426 regnode * const noper = NEXTOPER( cur );
1427 const U8 *uc = (U8*)STRING( noper );
1428 const U8 * const e = uc + STR_LEN( noper );
1430 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1431 const U8 *scan = (U8*)NULL;
1432 U32 wordlen = 0; /* required init */
1434 bool set_bit = trie->bitmap ? 1 : 0; /*store the first char in the bitmap?*/
1436 if (OP(noper) == NOTHING) {
1440 if ( set_bit ) /* bitmap only alloced when !(UTF&&Folding) */
1441 TRIE_BITMAP_SET(trie,*uc); /* store the raw first byte
1442 regardless of encoding */
1444 for ( ; uc < e ; uc += len ) {
1445 TRIE_CHARCOUNT(trie)++;
1449 if ( !trie->charmap[ uvc ] ) {
1450 trie->charmap[ uvc ]=( ++trie->uniquecharcount );
1452 trie->charmap[ folder[ uvc ] ] = trie->charmap[ uvc ];
1456 /* store the codepoint in the bitmap, and if its ascii
1457 also store its folded equivelent. */
1458 TRIE_BITMAP_SET(trie,uvc);
1460 /* store the folded codepoint */
1461 if ( folder ) TRIE_BITMAP_SET(trie,folder[ uvc ]);
1464 /* store first byte of utf8 representation of
1465 codepoints in the 127 < uvc < 256 range */
1466 if (127 < uvc && uvc < 192) {
1467 TRIE_BITMAP_SET(trie,194);
1468 } else if (191 < uvc ) {
1469 TRIE_BITMAP_SET(trie,195);
1470 /* && uvc < 256 -- we know uvc is < 256 already */
1473 set_bit = 0; /* We've done our bit :-) */
1478 widecharmap = newHV();
1480 svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 1 );
1483 Perl_croak( aTHX_ "error creating/fetching widecharmap entry for 0x%"UVXf, uvc );
1485 if ( !SvTRUE( *svpp ) ) {
1486 sv_setiv( *svpp, ++trie->uniquecharcount );
1491 if( cur == first ) {
1494 } else if (chars < trie->minlen) {
1496 } else if (chars > trie->maxlen) {
1500 } /* end first pass */
1501 DEBUG_TRIE_COMPILE_r(
1502 PerlIO_printf( Perl_debug_log, "%*sTRIE(%s): W:%d C:%d Uq:%d Min:%d Max:%d\n",
1503 (int)depth * 2 + 2,"",
1504 ( widecharmap ? "UTF8" : "NATIVE" ), (int)word_count,
1505 (int)TRIE_CHARCOUNT(trie), trie->uniquecharcount,
1506 (int)trie->minlen, (int)trie->maxlen )
1510 We now know what we are dealing with in terms of unique chars and
1511 string sizes so we can calculate how much memory a naive
1512 representation using a flat table will take. If it's over a reasonable
1513 limit (as specified by ${^RE_TRIE_MAXBUF}) we use a more memory
1514 conservative but potentially much slower representation using an array
1517 At the end we convert both representations into the same compressed
1518 form that will be used in regexec.c for matching with. The latter
1519 is a form that cannot be used to construct with but has memory
1520 properties similar to the list form and access properties similar
1521 to the table form making it both suitable for fast searches and
1522 small enough that its feasable to store for the duration of a program.
1524 See the comment in the code where the compressed table is produced
1525 inplace from the flat tabe representation for an explanation of how
1526 the compression works.
1531 Newx(prev_states, TRIE_CHARCOUNT(trie) + 2, U32);
1534 if ( (IV)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1) > SvIV(re_trie_maxbuff) ) {
1536 Second Pass -- Array Of Lists Representation
1538 Each state will be represented by a list of charid:state records
1539 (reg_trie_trans_le) the first such element holds the CUR and LEN
1540 points of the allocated array. (See defines above).
1542 We build the initial structure using the lists, and then convert
1543 it into the compressed table form which allows faster lookups
1544 (but cant be modified once converted).
1547 STRLEN transcount = 1;
1549 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1550 "%*sCompiling trie using list compiler\n",
1551 (int)depth * 2 + 2, ""));
1553 trie->states = (reg_trie_state *)
1554 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1555 sizeof(reg_trie_state) );
1559 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1561 regnode * const noper = NEXTOPER( cur );
1562 U8 *uc = (U8*)STRING( noper );
1563 const U8 * const e = uc + STR_LEN( noper );
1564 U32 state = 1; /* required init */
1565 U16 charid = 0; /* sanity init */
1566 U8 *scan = (U8*)NULL; /* sanity init */
1567 STRLEN foldlen = 0; /* required init */
1568 U32 wordlen = 0; /* required init */
1569 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1571 if (OP(noper) != NOTHING) {
1572 for ( ; uc < e ; uc += len ) {
1577 charid = trie->charmap[ uvc ];
1579 SV** const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1583 charid=(U16)SvIV( *svpp );
1586 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1593 if ( !trie->states[ state ].trans.list ) {
1594 TRIE_LIST_NEW( state );
1596 for ( check = 1; check <= TRIE_LIST_USED( state ); check++ ) {
1597 if ( TRIE_LIST_ITEM( state, check ).forid == charid ) {
1598 newstate = TRIE_LIST_ITEM( state, check ).newstate;
1603 newstate = next_alloc++;
1604 prev_states[newstate] = state;
1605 TRIE_LIST_PUSH( state, charid, newstate );
1610 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1614 TRIE_HANDLE_WORD(state);
1616 } /* end second pass */
1618 /* next alloc is the NEXT state to be allocated */
1619 trie->statecount = next_alloc;
1620 trie->states = (reg_trie_state *)
1621 PerlMemShared_realloc( trie->states,
1623 * sizeof(reg_trie_state) );
1625 /* and now dump it out before we compress it */
1626 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_list(trie, widecharmap,
1627 revcharmap, next_alloc,
1631 trie->trans = (reg_trie_trans *)
1632 PerlMemShared_calloc( transcount, sizeof(reg_trie_trans) );
1639 for( state=1 ; state < next_alloc ; state ++ ) {
1643 DEBUG_TRIE_COMPILE_MORE_r(
1644 PerlIO_printf( Perl_debug_log, "tp: %d zp: %d ",tp,zp)
1648 if (trie->states[state].trans.list) {
1649 U16 minid=TRIE_LIST_ITEM( state, 1).forid;
1653 for( idx = 2 ; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1654 const U16 forid = TRIE_LIST_ITEM( state, idx).forid;
1655 if ( forid < minid ) {
1657 } else if ( forid > maxid ) {
1661 if ( transcount < tp + maxid - minid + 1) {
1663 trie->trans = (reg_trie_trans *)
1664 PerlMemShared_realloc( trie->trans,
1666 * sizeof(reg_trie_trans) );
1667 Zero( trie->trans + (transcount / 2), transcount / 2 , reg_trie_trans );
1669 base = trie->uniquecharcount + tp - minid;
1670 if ( maxid == minid ) {
1672 for ( ; zp < tp ; zp++ ) {
1673 if ( ! trie->trans[ zp ].next ) {
1674 base = trie->uniquecharcount + zp - minid;
1675 trie->trans[ zp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1676 trie->trans[ zp ].check = state;
1682 trie->trans[ tp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1683 trie->trans[ tp ].check = state;
1688 for ( idx=1; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1689 const U32 tid = base - trie->uniquecharcount + TRIE_LIST_ITEM( state, idx ).forid;
1690 trie->trans[ tid ].next = TRIE_LIST_ITEM( state, idx ).newstate;
1691 trie->trans[ tid ].check = state;
1693 tp += ( maxid - minid + 1 );
1695 Safefree(trie->states[ state ].trans.list);
1698 DEBUG_TRIE_COMPILE_MORE_r(
1699 PerlIO_printf( Perl_debug_log, " base: %d\n",base);
1702 trie->states[ state ].trans.base=base;
1704 trie->lasttrans = tp + 1;
1708 Second Pass -- Flat Table Representation.
1710 we dont use the 0 slot of either trans[] or states[] so we add 1 to each.
1711 We know that we will need Charcount+1 trans at most to store the data
1712 (one row per char at worst case) So we preallocate both structures
1713 assuming worst case.
1715 We then construct the trie using only the .next slots of the entry
1718 We use the .check field of the first entry of the node temporarily to
1719 make compression both faster and easier by keeping track of how many non
1720 zero fields are in the node.
1722 Since trans are numbered from 1 any 0 pointer in the table is a FAIL
1725 There are two terms at use here: state as a TRIE_NODEIDX() which is a
1726 number representing the first entry of the node, and state as a
1727 TRIE_NODENUM() which is the trans number. state 1 is TRIE_NODEIDX(1) and
1728 TRIE_NODENUM(1), state 2 is TRIE_NODEIDX(2) and TRIE_NODENUM(3) if there
1729 are 2 entrys per node. eg:
1737 The table is internally in the right hand, idx form. However as we also
1738 have to deal with the states array which is indexed by nodenum we have to
1739 use TRIE_NODENUM() to convert.
1742 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1743 "%*sCompiling trie using table compiler\n",
1744 (int)depth * 2 + 2, ""));
1746 trie->trans = (reg_trie_trans *)
1747 PerlMemShared_calloc( ( TRIE_CHARCOUNT(trie) + 1 )
1748 * trie->uniquecharcount + 1,
1749 sizeof(reg_trie_trans) );
1750 trie->states = (reg_trie_state *)
1751 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1752 sizeof(reg_trie_state) );
1753 next_alloc = trie->uniquecharcount + 1;
1756 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1758 regnode * const noper = NEXTOPER( cur );
1759 const U8 *uc = (U8*)STRING( noper );
1760 const U8 * const e = uc + STR_LEN( noper );
1762 U32 state = 1; /* required init */
1764 U16 charid = 0; /* sanity init */
1765 U32 accept_state = 0; /* sanity init */
1766 U8 *scan = (U8*)NULL; /* sanity init */
1768 STRLEN foldlen = 0; /* required init */
1769 U32 wordlen = 0; /* required init */
1770 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1772 if ( OP(noper) != NOTHING ) {
1773 for ( ; uc < e ; uc += len ) {
1778 charid = trie->charmap[ uvc ];
1780 SV* const * const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1781 charid = svpp ? (U16)SvIV(*svpp) : 0;
1785 if ( !trie->trans[ state + charid ].next ) {
1786 trie->trans[ state + charid ].next = next_alloc;
1787 trie->trans[ state ].check++;
1788 prev_states[TRIE_NODENUM(next_alloc)]
1789 = TRIE_NODENUM(state);
1790 next_alloc += trie->uniquecharcount;
1792 state = trie->trans[ state + charid ].next;
1794 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1796 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1799 accept_state = TRIE_NODENUM( state );
1800 TRIE_HANDLE_WORD(accept_state);
1802 } /* end second pass */
1804 /* and now dump it out before we compress it */
1805 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_table(trie, widecharmap,
1807 next_alloc, depth+1));
1811 * Inplace compress the table.*
1813 For sparse data sets the table constructed by the trie algorithm will
1814 be mostly 0/FAIL transitions or to put it another way mostly empty.
1815 (Note that leaf nodes will not contain any transitions.)
1817 This algorithm compresses the tables by eliminating most such
1818 transitions, at the cost of a modest bit of extra work during lookup:
1820 - Each states[] entry contains a .base field which indicates the
1821 index in the state[] array wheres its transition data is stored.
1823 - If .base is 0 there are no valid transitions from that node.
1825 - If .base is nonzero then charid is added to it to find an entry in
1828 -If trans[states[state].base+charid].check!=state then the
1829 transition is taken to be a 0/Fail transition. Thus if there are fail
1830 transitions at the front of the node then the .base offset will point
1831 somewhere inside the previous nodes data (or maybe even into a node
1832 even earlier), but the .check field determines if the transition is
1836 The following process inplace converts the table to the compressed
1837 table: We first do not compress the root node 1,and mark its all its
1838 .check pointers as 1 and set its .base pointer as 1 as well. This
1839 allows to do a DFA construction from the compressed table later, and
1840 ensures that any .base pointers we calculate later are greater than
1843 - We set 'pos' to indicate the first entry of the second node.
1845 - We then iterate over the columns of the node, finding the first and
1846 last used entry at l and m. We then copy l..m into pos..(pos+m-l),
1847 and set the .check pointers accordingly, and advance pos
1848 appropriately and repreat for the next node. Note that when we copy
1849 the next pointers we have to convert them from the original
1850 NODEIDX form to NODENUM form as the former is not valid post
1853 - If a node has no transitions used we mark its base as 0 and do not
1854 advance the pos pointer.
1856 - If a node only has one transition we use a second pointer into the
1857 structure to fill in allocated fail transitions from other states.
1858 This pointer is independent of the main pointer and scans forward
1859 looking for null transitions that are allocated to a state. When it
1860 finds one it writes the single transition into the "hole". If the
1861 pointer doesnt find one the single transition is appended as normal.
1863 - Once compressed we can Renew/realloc the structures to release the
1866 See "Table-Compression Methods" in sec 3.9 of the Red Dragon,
1867 specifically Fig 3.47 and the associated pseudocode.
1871 const U32 laststate = TRIE_NODENUM( next_alloc );
1874 trie->statecount = laststate;
1876 for ( state = 1 ; state < laststate ; state++ ) {
1878 const U32 stateidx = TRIE_NODEIDX( state );
1879 const U32 o_used = trie->trans[ stateidx ].check;
1880 U32 used = trie->trans[ stateidx ].check;
1881 trie->trans[ stateidx ].check = 0;
1883 for ( charid = 0 ; used && charid < trie->uniquecharcount ; charid++ ) {
1884 if ( flag || trie->trans[ stateidx + charid ].next ) {
1885 if ( trie->trans[ stateidx + charid ].next ) {
1887 for ( ; zp < pos ; zp++ ) {
1888 if ( ! trie->trans[ zp ].next ) {
1892 trie->states[ state ].trans.base = zp + trie->uniquecharcount - charid ;
1893 trie->trans[ zp ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1894 trie->trans[ zp ].check = state;
1895 if ( ++zp > pos ) pos = zp;
1902 trie->states[ state ].trans.base = pos + trie->uniquecharcount - charid ;
1904 trie->trans[ pos ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1905 trie->trans[ pos ].check = state;
1910 trie->lasttrans = pos + 1;
1911 trie->states = (reg_trie_state *)
1912 PerlMemShared_realloc( trie->states, laststate
1913 * sizeof(reg_trie_state) );
1914 DEBUG_TRIE_COMPILE_MORE_r(
1915 PerlIO_printf( Perl_debug_log,
1916 "%*sAlloc: %d Orig: %"IVdf" elements, Final:%"IVdf". Savings of %%%5.2f\n",
1917 (int)depth * 2 + 2,"",
1918 (int)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1 ),
1921 ( ( next_alloc - pos ) * 100 ) / (double)next_alloc );
1924 } /* end table compress */
1926 DEBUG_TRIE_COMPILE_MORE_r(
1927 PerlIO_printf(Perl_debug_log, "%*sStatecount:%"UVxf" Lasttrans:%"UVxf"\n",
1928 (int)depth * 2 + 2, "",
1929 (UV)trie->statecount,
1930 (UV)trie->lasttrans)
1932 /* resize the trans array to remove unused space */
1933 trie->trans = (reg_trie_trans *)
1934 PerlMemShared_realloc( trie->trans, trie->lasttrans
1935 * sizeof(reg_trie_trans) );
1937 { /* Modify the program and insert the new TRIE node*/
1938 U8 nodetype =(U8)(flags & 0xFF);
1942 regnode *optimize = NULL;
1943 #ifdef RE_TRACK_PATTERN_OFFSETS
1946 U32 mjd_nodelen = 0;
1947 #endif /* RE_TRACK_PATTERN_OFFSETS */
1948 #endif /* DEBUGGING */
1950 This means we convert either the first branch or the first Exact,
1951 depending on whether the thing following (in 'last') is a branch
1952 or not and whther first is the startbranch (ie is it a sub part of
1953 the alternation or is it the whole thing.)
1954 Assuming its a sub part we conver the EXACT otherwise we convert
1955 the whole branch sequence, including the first.
1957 /* Find the node we are going to overwrite */
1958 if ( first != startbranch || OP( last ) == BRANCH ) {
1959 /* branch sub-chain */
1960 NEXT_OFF( first ) = (U16)(last - first);
1961 #ifdef RE_TRACK_PATTERN_OFFSETS
1963 mjd_offset= Node_Offset((convert));
1964 mjd_nodelen= Node_Length((convert));
1967 /* whole branch chain */
1969 #ifdef RE_TRACK_PATTERN_OFFSETS
1972 const regnode *nop = NEXTOPER( convert );
1973 mjd_offset= Node_Offset((nop));
1974 mjd_nodelen= Node_Length((nop));
1978 PerlIO_printf(Perl_debug_log, "%*sMJD offset:%"UVuf" MJD length:%"UVuf"\n",
1979 (int)depth * 2 + 2, "",
1980 (UV)mjd_offset, (UV)mjd_nodelen)
1983 /* But first we check to see if there is a common prefix we can
1984 split out as an EXACT and put in front of the TRIE node. */
1985 trie->startstate= 1;
1986 if ( trie->bitmap && !widecharmap && !trie->jump ) {
1988 for ( state = 1 ; state < trie->statecount-1 ; state++ ) {
1992 const U32 base = trie->states[ state ].trans.base;
1994 if ( trie->states[state].wordnum )
1997 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
1998 if ( ( base + ofs >= trie->uniquecharcount ) &&
1999 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
2000 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
2002 if ( ++count > 1 ) {
2003 SV **tmp = av_fetch( revcharmap, ofs, 0);
2004 const U8 *ch = (U8*)SvPV_nolen_const( *tmp );
2005 if ( state == 1 ) break;
2007 Zero(trie->bitmap, ANYOF_BITMAP_SIZE, char);
2009 PerlIO_printf(Perl_debug_log,
2010 "%*sNew Start State=%"UVuf" Class: [",
2011 (int)depth * 2 + 2, "",
2014 SV ** const tmp = av_fetch( revcharmap, idx, 0);
2015 const U8 * const ch = (U8*)SvPV_nolen_const( *tmp );
2017 TRIE_BITMAP_SET(trie,*ch);
2019 TRIE_BITMAP_SET(trie, folder[ *ch ]);
2021 PerlIO_printf(Perl_debug_log, "%s", (char*)ch)
2025 TRIE_BITMAP_SET(trie,*ch);
2027 TRIE_BITMAP_SET(trie,folder[ *ch ]);
2028 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"%s", ch));
2034 SV **tmp = av_fetch( revcharmap, idx, 0);
2036 char *ch = SvPV( *tmp, len );
2038 SV *sv=sv_newmortal();
2039 PerlIO_printf( Perl_debug_log,
2040 "%*sPrefix State: %"UVuf" Idx:%"UVuf" Char='%s'\n",
2041 (int)depth * 2 + 2, "",
2043 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 6,
2044 PL_colors[0], PL_colors[1],
2045 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
2046 PERL_PV_ESCAPE_FIRSTCHAR
2051 OP( convert ) = nodetype;
2052 str=STRING(convert);
2055 STR_LEN(convert) += len;
2061 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"]\n"));
2066 trie->prefixlen = (state-1);
2068 regnode *n = convert+NODE_SZ_STR(convert);
2069 NEXT_OFF(convert) = NODE_SZ_STR(convert);
2070 trie->startstate = state;
2071 trie->minlen -= (state - 1);
2072 trie->maxlen -= (state - 1);
2074 /* At least the UNICOS C compiler choked on this
2075 * being argument to DEBUG_r(), so let's just have
2078 #ifdef PERL_EXT_RE_BUILD
2084 regnode *fix = convert;
2085 U32 word = trie->wordcount;
2087 Set_Node_Offset_Length(convert, mjd_offset, state - 1);
2088 while( ++fix < n ) {
2089 Set_Node_Offset_Length(fix, 0, 0);
2092 SV ** const tmp = av_fetch( trie_words, word, 0 );
2094 if ( STR_LEN(convert) <= SvCUR(*tmp) )
2095 sv_chop(*tmp, SvPV_nolen(*tmp) + STR_LEN(convert));
2097 sv_chop(*tmp, SvPV_nolen(*tmp) + SvCUR(*tmp));
2105 NEXT_OFF(convert) = (U16)(tail - convert);
2106 DEBUG_r(optimize= n);
2112 if ( trie->maxlen ) {
2113 NEXT_OFF( convert ) = (U16)(tail - convert);
2114 ARG_SET( convert, data_slot );
2115 /* Store the offset to the first unabsorbed branch in
2116 jump[0], which is otherwise unused by the jump logic.
2117 We use this when dumping a trie and during optimisation. */
2119 trie->jump[0] = (U16)(nextbranch - convert);
2122 if ( !trie->states[trie->startstate].wordnum && trie->bitmap &&
2123 ( (char *)jumper - (char *)convert) >= (int)sizeof(struct regnode_charclass) )
2125 OP( convert ) = TRIEC;
2126 Copy(trie->bitmap, ((struct regnode_charclass *)convert)->bitmap, ANYOF_BITMAP_SIZE, char);
2127 PerlMemShared_free(trie->bitmap);
2130 OP( convert ) = TRIE;
2132 /* store the type in the flags */
2133 convert->flags = nodetype;
2137 + regarglen[ OP( convert ) ];
2139 /* XXX We really should free up the resource in trie now,
2140 as we won't use them - (which resources?) dmq */
2142 /* needed for dumping*/
2143 DEBUG_r(if (optimize) {
2144 regnode *opt = convert;
2146 while ( ++opt < optimize) {
2147 Set_Node_Offset_Length(opt,0,0);
2150 Try to clean up some of the debris left after the
2153 while( optimize < jumper ) {
2154 mjd_nodelen += Node_Length((optimize));
2155 OP( optimize ) = OPTIMIZED;
2156 Set_Node_Offset_Length(optimize,0,0);
2159 Set_Node_Offset_Length(convert,mjd_offset,mjd_nodelen);
2161 } /* end node insert */
2163 /* Finish populating the prev field of the wordinfo array. Walk back
2164 * from each accept state until we find another accept state, and if
2165 * so, point the first word's .prev field at the second word. If the
2166 * second already has a .prev field set, stop now. This will be the
2167 * case either if we've already processed that word's accept state,
2168 * or that that state had multiple words, and the overspill words
2169 * were already linked up earlier.
2176 for (word=1; word <= trie->wordcount; word++) {
2178 if (trie->wordinfo[word].prev)
2180 state = trie->wordinfo[word].accept;
2182 state = prev_states[state];
2185 prev = trie->states[state].wordnum;
2189 trie->wordinfo[word].prev = prev;
2191 Safefree(prev_states);
2195 /* and now dump out the compressed format */
2196 DEBUG_TRIE_COMPILE_r(dump_trie(trie, widecharmap, revcharmap, depth+1));
2198 RExC_rxi->data->data[ data_slot + 1 ] = (void*)widecharmap;
2200 RExC_rxi->data->data[ data_slot + TRIE_WORDS_OFFSET ] = (void*)trie_words;
2201 RExC_rxi->data->data[ data_slot + 3 ] = (void*)revcharmap;
2203 SvREFCNT_dec(revcharmap);
2207 : trie->startstate>1
2213 S_make_trie_failtable(pTHX_ RExC_state_t *pRExC_state, regnode *source, regnode *stclass, U32 depth)
2215 /* The Trie is constructed and compressed now so we can build a fail array now if its needed
2217 This is basically the Aho-Corasick algorithm. Its from exercise 3.31 and 3.32 in the
2218 "Red Dragon" -- Compilers, principles, techniques, and tools. Aho, Sethi, Ullman 1985/88
2221 We find the fail state for each state in the trie, this state is the longest proper
2222 suffix of the current states 'word' that is also a proper prefix of another word in our
2223 trie. State 1 represents the word '' and is the thus the default fail state. This allows
2224 the DFA not to have to restart after its tried and failed a word at a given point, it
2225 simply continues as though it had been matching the other word in the first place.
2227 'abcdgu'=~/abcdefg|cdgu/
2228 When we get to 'd' we are still matching the first word, we would encounter 'g' which would
2229 fail, which would bring use to the state representing 'd' in the second word where we would
2230 try 'g' and succeed, prodceding to match 'cdgu'.
2232 /* add a fail transition */
2233 const U32 trie_offset = ARG(source);
2234 reg_trie_data *trie=(reg_trie_data *)RExC_rxi->data->data[trie_offset];
2236 const U32 ucharcount = trie->uniquecharcount;
2237 const U32 numstates = trie->statecount;
2238 const U32 ubound = trie->lasttrans + ucharcount;
2242 U32 base = trie->states[ 1 ].trans.base;
2245 const U32 data_slot = add_data( pRExC_state, 1, "T" );
2246 GET_RE_DEBUG_FLAGS_DECL;
2248 PERL_ARGS_ASSERT_MAKE_TRIE_FAILTABLE;
2250 PERL_UNUSED_ARG(depth);
2254 ARG_SET( stclass, data_slot );
2255 aho = (reg_ac_data *) PerlMemShared_calloc( 1, sizeof(reg_ac_data) );
2256 RExC_rxi->data->data[ data_slot ] = (void*)aho;
2257 aho->trie=trie_offset;
2258 aho->states=(reg_trie_state *)PerlMemShared_malloc( numstates * sizeof(reg_trie_state) );
2259 Copy( trie->states, aho->states, numstates, reg_trie_state );
2260 Newxz( q, numstates, U32);
2261 aho->fail = (U32 *) PerlMemShared_calloc( numstates, sizeof(U32) );
2264 /* initialize fail[0..1] to be 1 so that we always have
2265 a valid final fail state */
2266 fail[ 0 ] = fail[ 1 ] = 1;
2268 for ( charid = 0; charid < ucharcount ; charid++ ) {
2269 const U32 newstate = TRIE_TRANS_STATE( 1, base, ucharcount, charid, 0 );
2271 q[ q_write ] = newstate;
2272 /* set to point at the root */
2273 fail[ q[ q_write++ ] ]=1;
2276 while ( q_read < q_write) {
2277 const U32 cur = q[ q_read++ % numstates ];
2278 base = trie->states[ cur ].trans.base;
2280 for ( charid = 0 ; charid < ucharcount ; charid++ ) {
2281 const U32 ch_state = TRIE_TRANS_STATE( cur, base, ucharcount, charid, 1 );
2283 U32 fail_state = cur;
2286 fail_state = fail[ fail_state ];
2287 fail_base = aho->states[ fail_state ].trans.base;
2288 } while ( !TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 ) );
2290 fail_state = TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 );
2291 fail[ ch_state ] = fail_state;
2292 if ( !aho->states[ ch_state ].wordnum && aho->states[ fail_state ].wordnum )
2294 aho->states[ ch_state ].wordnum = aho->states[ fail_state ].wordnum;
2296 q[ q_write++ % numstates] = ch_state;
2300 /* restore fail[0..1] to 0 so that we "fall out" of the AC loop
2301 when we fail in state 1, this allows us to use the
2302 charclass scan to find a valid start char. This is based on the principle
2303 that theres a good chance the string being searched contains lots of stuff
2304 that cant be a start char.
2306 fail[ 0 ] = fail[ 1 ] = 0;
2307 DEBUG_TRIE_COMPILE_r({
2308 PerlIO_printf(Perl_debug_log,
2309 "%*sStclass Failtable (%"UVuf" states): 0",
2310 (int)(depth * 2), "", (UV)numstates
2312 for( q_read=1; q_read<numstates; q_read++ ) {
2313 PerlIO_printf(Perl_debug_log, ", %"UVuf, (UV)fail[q_read]);
2315 PerlIO_printf(Perl_debug_log, "\n");
2318 /*RExC_seen |= REG_SEEN_TRIEDFA;*/
2323 * There are strange code-generation bugs caused on sparc64 by gcc-2.95.2.
2324 * These need to be revisited when a newer toolchain becomes available.
2326 #if defined(__sparc64__) && defined(__GNUC__)
2327 # if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
2328 # undef SPARC64_GCC_WORKAROUND
2329 # define SPARC64_GCC_WORKAROUND 1
2333 #define DEBUG_PEEP(str,scan,depth) \
2334 DEBUG_OPTIMISE_r({if (scan){ \
2335 SV * const mysv=sv_newmortal(); \
2336 regnode *Next = regnext(scan); \
2337 regprop(RExC_rx, mysv, scan); \
2338 PerlIO_printf(Perl_debug_log, "%*s" str ">%3d: %s (%d)\n", \
2339 (int)depth*2, "", REG_NODE_NUM(scan), SvPV_nolen_const(mysv),\
2340 Next ? (REG_NODE_NUM(Next)) : 0 ); \
2347 #define JOIN_EXACT(scan,min,flags) \
2348 if (PL_regkind[OP(scan)] == EXACT) \
2349 join_exact(pRExC_state,(scan),(min),(flags),NULL,depth+1)
2352 S_join_exact(pTHX_ RExC_state_t *pRExC_state, regnode *scan, I32 *min, U32 flags,regnode *val, U32 depth) {
2353 /* Merge several consecutive EXACTish nodes into one. */
2354 regnode *n = regnext(scan);
2356 regnode *next = scan + NODE_SZ_STR(scan);
2360 regnode *stop = scan;
2361 GET_RE_DEBUG_FLAGS_DECL;
2363 PERL_UNUSED_ARG(depth);
2366 PERL_ARGS_ASSERT_JOIN_EXACT;
2367 #ifndef EXPERIMENTAL_INPLACESCAN
2368 PERL_UNUSED_ARG(flags);
2369 PERL_UNUSED_ARG(val);
2371 DEBUG_PEEP("join",scan,depth);
2373 /* Skip NOTHING, merge EXACT*. */
2375 ( PL_regkind[OP(n)] == NOTHING ||
2376 (stringok && (OP(n) == OP(scan))))
2378 && NEXT_OFF(scan) + NEXT_OFF(n) < I16_MAX) {
2380 if (OP(n) == TAIL || n > next)
2382 if (PL_regkind[OP(n)] == NOTHING) {
2383 DEBUG_PEEP("skip:",n,depth);
2384 NEXT_OFF(scan) += NEXT_OFF(n);
2385 next = n + NODE_STEP_REGNODE;
2392 else if (stringok) {
2393 const unsigned int oldl = STR_LEN(scan);
2394 regnode * const nnext = regnext(n);
2396 DEBUG_PEEP("merg",n,depth);
2399 if (oldl + STR_LEN(n) > U8_MAX)
2401 NEXT_OFF(scan) += NEXT_OFF(n);
2402 STR_LEN(scan) += STR_LEN(n);
2403 next = n + NODE_SZ_STR(n);
2404 /* Now we can overwrite *n : */
2405 Move(STRING(n), STRING(scan) + oldl, STR_LEN(n), char);
2413 #ifdef EXPERIMENTAL_INPLACESCAN
2414 if (flags && !NEXT_OFF(n)) {
2415 DEBUG_PEEP("atch", val, depth);
2416 if (reg_off_by_arg[OP(n)]) {
2417 ARG_SET(n, val - n);
2420 NEXT_OFF(n) = val - n;
2427 if (UTF && ( OP(scan) == EXACTF ) && ( STR_LEN(scan) >= 6 ) ) {
2429 Two problematic code points in Unicode casefolding of EXACT nodes:
2431 U+0390 - GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
2432 U+03B0 - GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
2438 U+03B9 U+0308 U+0301 0xCE 0xB9 0xCC 0x88 0xCC 0x81
2439 U+03C5 U+0308 U+0301 0xCF 0x85 0xCC 0x88 0xCC 0x81
2441 This means that in case-insensitive matching (or "loose matching",
2442 as Unicode calls it), an EXACTF of length six (the UTF-8 encoded byte
2443 length of the above casefolded versions) can match a target string
2444 of length two (the byte length of UTF-8 encoded U+0390 or U+03B0).
2445 This would rather mess up the minimum length computation.
2447 What we'll do is to look for the tail four bytes, and then peek
2448 at the preceding two bytes to see whether we need to decrease
2449 the minimum length by four (six minus two).
2451 Thanks to the design of UTF-8, there cannot be false matches:
2452 A sequence of valid UTF-8 bytes cannot be a subsequence of
2453 another valid sequence of UTF-8 bytes.
2456 char * const s0 = STRING(scan), *s, *t;
2457 char * const s1 = s0 + STR_LEN(scan) - 1;
2458 char * const s2 = s1 - 4;
2459 #ifdef EBCDIC /* RD tunifold greek 0390 and 03B0 */
2460 const char t0[] = "\xaf\x49\xaf\x42";
2462 const char t0[] = "\xcc\x88\xcc\x81";
2464 const char * const t1 = t0 + 3;
2467 s < s2 && (t = ninstr(s, s1, t0, t1));
2470 if (((U8)t[-1] == 0x68 && (U8)t[-2] == 0xB4) ||
2471 ((U8)t[-1] == 0x46 && (U8)t[-2] == 0xB5))
2473 if (((U8)t[-1] == 0xB9 && (U8)t[-2] == 0xCE) ||
2474 ((U8)t[-1] == 0x85 && (U8)t[-2] == 0xCF))
2482 n = scan + NODE_SZ_STR(scan);
2484 if (PL_regkind[OP(n)] != NOTHING || OP(n) == NOTHING) {
2491 DEBUG_OPTIMISE_r(if (merged){DEBUG_PEEP("finl",scan,depth)});
2495 /* REx optimizer. Converts nodes into quickier variants "in place".
2496 Finds fixed substrings. */
2498 /* Stops at toplevel WHILEM as well as at "last". At end *scanp is set
2499 to the position after last scanned or to NULL. */
2501 #define INIT_AND_WITHP \
2502 assert(!and_withp); \
2503 Newx(and_withp,1,struct regnode_charclass_class); \
2504 SAVEFREEPV(and_withp)
2506 /* this is a chain of data about sub patterns we are processing that
2507 need to be handled seperately/specially in study_chunk. Its so
2508 we can simulate recursion without losing state. */
2510 typedef struct scan_frame {
2511 regnode *last; /* last node to process in this frame */
2512 regnode *next; /* next node to process when last is reached */
2513 struct scan_frame *prev; /*previous frame*/
2514 I32 stop; /* what stopparen do we use */
2518 #define SCAN_COMMIT(s, data, m) scan_commit(s, data, m, is_inf)
2520 #define CASE_SYNST_FNC(nAmE) \
2522 if (flags & SCF_DO_STCLASS_AND) { \
2523 for (value = 0; value < 256; value++) \
2524 if (!is_ ## nAmE ## _cp(value)) \
2525 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2528 for (value = 0; value < 256; value++) \
2529 if (is_ ## nAmE ## _cp(value)) \
2530 ANYOF_BITMAP_SET(data->start_class, value); \
2534 if (flags & SCF_DO_STCLASS_AND) { \
2535 for (value = 0; value < 256; value++) \
2536 if (is_ ## nAmE ## _cp(value)) \
2537 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2540 for (value = 0; value < 256; value++) \
2541 if (!is_ ## nAmE ## _cp(value)) \
2542 ANYOF_BITMAP_SET(data->start_class, value); \
2549 S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
2550 I32 *minlenp, I32 *deltap,
2555 struct regnode_charclass_class *and_withp,
2556 U32 flags, U32 depth)
2557 /* scanp: Start here (read-write). */
2558 /* deltap: Write maxlen-minlen here. */
2559 /* last: Stop before this one. */
2560 /* data: string data about the pattern */
2561 /* stopparen: treat close N as END */
2562 /* recursed: which subroutines have we recursed into */
2563 /* and_withp: Valid if flags & SCF_DO_STCLASS_OR */
2566 I32 min = 0, pars = 0, code;
2567 regnode *scan = *scanp, *next;
2569 int is_inf = (flags & SCF_DO_SUBSTR) && (data->flags & SF_IS_INF);
2570 int is_inf_internal = 0; /* The studied chunk is infinite */
2571 I32 is_par = OP(scan) == OPEN ? ARG(scan) : 0;
2572 scan_data_t data_fake;
2573 SV *re_trie_maxbuff = NULL;
2574 regnode *first_non_open = scan;
2575 I32 stopmin = I32_MAX;
2576 scan_frame *frame = NULL;
2577 GET_RE_DEBUG_FLAGS_DECL;
2579 PERL_ARGS_ASSERT_STUDY_CHUNK;
2582 StructCopy(&zero_scan_data, &data_fake, scan_data_t);
2586 while (first_non_open && OP(first_non_open) == OPEN)
2587 first_non_open=regnext(first_non_open);
2592 while ( scan && OP(scan) != END && scan < last ){
2593 /* Peephole optimizer: */
2594 DEBUG_STUDYDATA("Peep:", data,depth);
2595 DEBUG_PEEP("Peep",scan,depth);
2596 JOIN_EXACT(scan,&min,0);
2598 /* Follow the next-chain of the current node and optimize
2599 away all the NOTHINGs from it. */
2600 if (OP(scan) != CURLYX) {
2601 const int max = (reg_off_by_arg[OP(scan)]
2603 /* I32 may be smaller than U16 on CRAYs! */
2604 : (I32_MAX < U16_MAX ? I32_MAX : U16_MAX));
2605 int off = (reg_off_by_arg[OP(scan)] ? ARG(scan) : NEXT_OFF(scan));
2609 /* Skip NOTHING and LONGJMP. */
2610 while ((n = regnext(n))
2611 && ((PL_regkind[OP(n)] == NOTHING && (noff = NEXT_OFF(n)))
2612 || ((OP(n) == LONGJMP) && (noff = ARG(n))))
2613 && off + noff < max)
2615 if (reg_off_by_arg[OP(scan)])
2618 NEXT_OFF(scan) = off;
2623 /* The principal pseudo-switch. Cannot be a switch, since we
2624 look into several different things. */
2625 if (OP(scan) == BRANCH || OP(scan) == BRANCHJ
2626 || OP(scan) == IFTHEN) {
2627 next = regnext(scan);
2629 /* demq: the op(next)==code check is to see if we have "branch-branch" AFAICT */
2631 if (OP(next) == code || code == IFTHEN) {
2632 /* NOTE - There is similar code to this block below for handling
2633 TRIE nodes on a re-study. If you change stuff here check there
2635 I32 max1 = 0, min1 = I32_MAX, num = 0;
2636 struct regnode_charclass_class accum;
2637 regnode * const startbranch=scan;
2639 if (flags & SCF_DO_SUBSTR)
2640 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot merge strings after this. */
2641 if (flags & SCF_DO_STCLASS)
2642 cl_init_zero(pRExC_state, &accum);
2644 while (OP(scan) == code) {
2645 I32 deltanext, minnext, f = 0, fake;
2646 struct regnode_charclass_class this_class;
2649 data_fake.flags = 0;
2651 data_fake.whilem_c = data->whilem_c;
2652 data_fake.last_closep = data->last_closep;
2655 data_fake.last_closep = &fake;
2657 data_fake.pos_delta = delta;
2658 next = regnext(scan);
2659 scan = NEXTOPER(scan);
2661 scan = NEXTOPER(scan);
2662 if (flags & SCF_DO_STCLASS) {
2663 cl_init(pRExC_state, &this_class);
2664 data_fake.start_class = &this_class;
2665 f = SCF_DO_STCLASS_AND;
2667 if (flags & SCF_WHILEM_VISITED_POS)
2668 f |= SCF_WHILEM_VISITED_POS;
2670 /* we suppose the run is continuous, last=next...*/
2671 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
2673 stopparen, recursed, NULL, f,depth+1);
2676 if (max1 < minnext + deltanext)
2677 max1 = minnext + deltanext;
2678 if (deltanext == I32_MAX)
2679 is_inf = is_inf_internal = 1;
2681 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
2683 if (data_fake.flags & SCF_SEEN_ACCEPT) {
2684 if ( stopmin > minnext)
2685 stopmin = min + min1;
2686 flags &= ~SCF_DO_SUBSTR;
2688 data->flags |= SCF_SEEN_ACCEPT;
2691 if (data_fake.flags & SF_HAS_EVAL)
2692 data->flags |= SF_HAS_EVAL;
2693 data->whilem_c = data_fake.whilem_c;
2695 if (flags & SCF_DO_STCLASS)
2696 cl_or(pRExC_state, &accum, &this_class);
2698 if (code == IFTHEN && num < 2) /* Empty ELSE branch */
2700 if (flags & SCF_DO_SUBSTR) {
2701 data->pos_min += min1;
2702 data->pos_delta += max1 - min1;
2703 if (max1 != min1 || is_inf)
2704 data->longest = &(data->longest_float);
2707 delta += max1 - min1;
2708 if (flags & SCF_DO_STCLASS_OR) {
2709 cl_or(pRExC_state, data->start_class, &accum);
2711 cl_and(data->start_class, and_withp);
2712 flags &= ~SCF_DO_STCLASS;
2715 else if (flags & SCF_DO_STCLASS_AND) {
2717 cl_and(data->start_class, &accum);
2718 flags &= ~SCF_DO_STCLASS;
2721 /* Switch to OR mode: cache the old value of
2722 * data->start_class */
2724 StructCopy(data->start_class, and_withp,
2725 struct regnode_charclass_class);
2726 flags &= ~SCF_DO_STCLASS_AND;
2727 StructCopy(&accum, data->start_class,
2728 struct regnode_charclass_class);
2729 flags |= SCF_DO_STCLASS_OR;
2730 data->start_class->flags |= ANYOF_EOS;
2734 if (PERL_ENABLE_TRIE_OPTIMISATION && OP( startbranch ) == BRANCH ) {
2737 Assuming this was/is a branch we are dealing with: 'scan' now
2738 points at the item that follows the branch sequence, whatever
2739 it is. We now start at the beginning of the sequence and look
2746 which would be constructed from a pattern like /A|LIST|OF|WORDS/
2748 If we can find such a subseqence we need to turn the first
2749 element into a trie and then add the subsequent branch exact
2750 strings to the trie.
2754 1. patterns where the whole set of branch can be converted.
2756 2. patterns where only a subset can be converted.
2758 In case 1 we can replace the whole set with a single regop
2759 for the trie. In case 2 we need to keep the start and end
2762 'BRANCH EXACT; BRANCH EXACT; BRANCH X'
2763 becomes BRANCH TRIE; BRANCH X;
2765 There is an additional case, that being where there is a
2766 common prefix, which gets split out into an EXACT like node
2767 preceding the TRIE node.
2769 If x(1..n)==tail then we can do a simple trie, if not we make
2770 a "jump" trie, such that when we match the appropriate word
2771 we "jump" to the appopriate tail node. Essentailly we turn
2772 a nested if into a case structure of sorts.
2777 if (!re_trie_maxbuff) {
2778 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
2779 if (!SvIOK(re_trie_maxbuff))
2780 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
2782 if ( SvIV(re_trie_maxbuff)>=0 ) {
2784 regnode *first = (regnode *)NULL;
2785 regnode *last = (regnode *)NULL;
2786 regnode *tail = scan;
2791 SV * const mysv = sv_newmortal(); /* for dumping */
2793 /* var tail is used because there may be a TAIL
2794 regop in the way. Ie, the exacts will point to the
2795 thing following the TAIL, but the last branch will
2796 point at the TAIL. So we advance tail. If we
2797 have nested (?:) we may have to move through several
2801 while ( OP( tail ) == TAIL ) {
2802 /* this is the TAIL generated by (?:) */
2803 tail = regnext( tail );
2808 regprop(RExC_rx, mysv, tail );
2809 PerlIO_printf( Perl_debug_log, "%*s%s%s\n",
2810 (int)depth * 2 + 2, "",
2811 "Looking for TRIE'able sequences. Tail node is: ",
2812 SvPV_nolen_const( mysv )
2818 step through the branches, cur represents each
2819 branch, noper is the first thing to be matched
2820 as part of that branch and noper_next is the
2821 regnext() of that node. if noper is an EXACT
2822 and noper_next is the same as scan (our current
2823 position in the regex) then the EXACT branch is
2824 a possible optimization target. Once we have
2825 two or more consequetive such branches we can
2826 create a trie of the EXACT's contents and stich
2827 it in place. If the sequence represents all of
2828 the branches we eliminate the whole thing and
2829 replace it with a single TRIE. If it is a
2830 subsequence then we need to stitch it in. This
2831 means the first branch has to remain, and needs
2832 to be repointed at the item on the branch chain
2833 following the last branch optimized. This could
2834 be either a BRANCH, in which case the
2835 subsequence is internal, or it could be the
2836 item following the branch sequence in which
2837 case the subsequence is at the end.
2841 /* dont use tail as the end marker for this traverse */
2842 for ( cur = startbranch ; cur != scan ; cur = regnext( cur ) ) {
2843 regnode * const noper = NEXTOPER( cur );
2844 #if defined(DEBUGGING) || defined(NOJUMPTRIE)
2845 regnode * const noper_next = regnext( noper );
2849 regprop(RExC_rx, mysv, cur);
2850 PerlIO_printf( Perl_debug_log, "%*s- %s (%d)",
2851 (int)depth * 2 + 2,"", SvPV_nolen_const( mysv ), REG_NODE_NUM(cur) );
2853 regprop(RExC_rx, mysv, noper);
2854 PerlIO_printf( Perl_debug_log, " -> %s",
2855 SvPV_nolen_const(mysv));
2858 regprop(RExC_rx, mysv, noper_next );
2859 PerlIO_printf( Perl_debug_log,"\t=> %s\t",
2860 SvPV_nolen_const(mysv));
2862 PerlIO_printf( Perl_debug_log, "(First==%d,Last==%d,Cur==%d)\n",
2863 REG_NODE_NUM(first), REG_NODE_NUM(last), REG_NODE_NUM(cur) );
2865 if ( (((first && optype!=NOTHING) ? OP( noper ) == optype
2866 : PL_regkind[ OP( noper ) ] == EXACT )
2867 || OP(noper) == NOTHING )
2869 && noper_next == tail
2874 if ( !first || optype == NOTHING ) {
2875 if (!first) first = cur;
2876 optype = OP( noper );
2882 Currently we do not believe that the trie logic can
2883 handle case insensitive matching properly when the
2884 pattern is not unicode (thus forcing unicode semantics).
2886 If/when this is fixed the following define can be swapped
2887 in below to fully enable trie logic.
2889 #define TRIE_TYPE_IS_SAFE 1
2892 #define TRIE_TYPE_IS_SAFE (UTF || optype==EXACT)
2894 if ( last && TRIE_TYPE_IS_SAFE ) {
2895 make_trie( pRExC_state,
2896 startbranch, first, cur, tail, count,
2899 if ( PL_regkind[ OP( noper ) ] == EXACT
2901 && noper_next == tail
2906 optype = OP( noper );
2916 regprop(RExC_rx, mysv, cur);
2917 PerlIO_printf( Perl_debug_log,
2918 "%*s- %s (%d) <SCAN FINISHED>\n", (int)depth * 2 + 2,
2919 "", SvPV_nolen_const( mysv ),REG_NODE_NUM(cur));
2923 if ( last && TRIE_TYPE_IS_SAFE ) {
2924 made= make_trie( pRExC_state, startbranch, first, scan, tail, count, optype, depth+1 );
2925 #ifdef TRIE_STUDY_OPT
2926 if ( ((made == MADE_EXACT_TRIE &&
2927 startbranch == first)
2928 || ( first_non_open == first )) &&
2930 flags |= SCF_TRIE_RESTUDY;
2931 if ( startbranch == first
2934 RExC_seen &=~REG_TOP_LEVEL_BRANCHES;
2944 else if ( code == BRANCHJ ) { /* single branch is optimized. */
2945 scan = NEXTOPER(NEXTOPER(scan));
2946 } else /* single branch is optimized. */
2947 scan = NEXTOPER(scan);
2949 } else if (OP(scan) == SUSPEND || OP(scan) == GOSUB || OP(scan) == GOSTART) {
2950 scan_frame *newframe = NULL;
2955 if (OP(scan) != SUSPEND) {
2956 /* set the pointer */
2957 if (OP(scan) == GOSUB) {
2959 RExC_recurse[ARG2L(scan)] = scan;
2960 start = RExC_open_parens[paren-1];
2961 end = RExC_close_parens[paren-1];
2964 start = RExC_rxi->program + 1;
2968 Newxz(recursed, (((RExC_npar)>>3) +1), U8);
2969 SAVEFREEPV(recursed);
2971 if (!PAREN_TEST(recursed,paren+1)) {
2972 PAREN_SET(recursed,paren+1);
2973 Newx(newframe,1,scan_frame);
2975 if (flags & SCF_DO_SUBSTR) {
2976 SCAN_COMMIT(pRExC_state,data,minlenp);
2977 data->longest = &(data->longest_float);
2979 is_inf = is_inf_internal = 1;
2980 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
2981 cl_anything(pRExC_state, data->start_class);
2982 flags &= ~SCF_DO_STCLASS;
2985 Newx(newframe,1,scan_frame);
2988 end = regnext(scan);
2993 SAVEFREEPV(newframe);
2994 newframe->next = regnext(scan);
2995 newframe->last = last;
2996 newframe->stop = stopparen;
2997 newframe->prev = frame;
3007 else if (OP(scan) == EXACT) {
3008 I32 l = STR_LEN(scan);
3011 const U8 * const s = (U8*)STRING(scan);
3012 l = utf8_length(s, s + l);
3013 uc = utf8_to_uvchr(s, NULL);
3015 uc = *((U8*)STRING(scan));
3018 if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
3019 /* The code below prefers earlier match for fixed
3020 offset, later match for variable offset. */
3021 if (data->last_end == -1) { /* Update the start info. */
3022 data->last_start_min = data->pos_min;
3023 data->last_start_max = is_inf
3024 ? I32_MAX : data->pos_min + data->pos_delta;
3026 sv_catpvn(data->last_found, STRING(scan), STR_LEN(scan));
3028 SvUTF8_on(data->last_found);
3030 SV * const sv = data->last_found;
3031 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3032 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3033 if (mg && mg->mg_len >= 0)
3034 mg->mg_len += utf8_length((U8*)STRING(scan),
3035 (U8*)STRING(scan)+STR_LEN(scan));
3037 data->last_end = data->pos_min + l;
3038 data->pos_min += l; /* As in the first entry. */
3039 data->flags &= ~SF_BEFORE_EOL;
3041 if (flags & SCF_DO_STCLASS_AND) {
3042 /* Check whether it is compatible with what we know already! */
3046 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3047 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3048 && (!(data->start_class->flags & ANYOF_FOLD)
3049 || !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
3052 ANYOF_CLASS_ZERO(data->start_class);
3053 ANYOF_BITMAP_ZERO(data->start_class);
3055 ANYOF_BITMAP_SET(data->start_class, uc);
3056 data->start_class->flags &= ~ANYOF_EOS;
3058 data->start_class->flags &= ~ANYOF_UNICODE_ALL;
3060 else if (flags & SCF_DO_STCLASS_OR) {
3061 /* false positive possible if the class is case-folded */
3063 ANYOF_BITMAP_SET(data->start_class, uc);
3065 data->start_class->flags |= ANYOF_UNICODE_ALL;
3066 data->start_class->flags &= ~ANYOF_EOS;
3067 cl_and(data->start_class, and_withp);
3069 flags &= ~SCF_DO_STCLASS;
3071 else if (PL_regkind[OP(scan)] == EXACT) { /* But OP != EXACT! */
3072 I32 l = STR_LEN(scan);
3073 UV uc = *((U8*)STRING(scan));
3075 /* Search for fixed substrings supports EXACT only. */
3076 if (flags & SCF_DO_SUBSTR) {
3078 SCAN_COMMIT(pRExC_state, data, minlenp);
3081 const U8 * const s = (U8 *)STRING(scan);
3082 l = utf8_length(s, s + l);
3083 uc = utf8_to_uvchr(s, NULL);
3086 if (flags & SCF_DO_SUBSTR)
3088 if (flags & SCF_DO_STCLASS_AND) {
3089 /* Check whether it is compatible with what we know already! */
3093 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3094 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3095 && !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
3097 ANYOF_CLASS_ZERO(data->start_class);
3098 ANYOF_BITMAP_ZERO(data->start_class);
3100 ANYOF_BITMAP_SET(data->start_class, uc);
3101 data->start_class->flags &= ~ANYOF_EOS;
3102 data->start_class->flags |= ANYOF_FOLD;
3103 if (OP(scan) == EXACTFL)
3104 data->start_class->flags |= ANYOF_LOCALE;
3107 else if (flags & SCF_DO_STCLASS_OR) {
3108 if (data->start_class->flags & ANYOF_FOLD) {
3109 /* false positive possible if the class is case-folded.
3110 Assume that the locale settings are the same... */
3112 ANYOF_BITMAP_SET(data->start_class, uc);
3113 data->start_class->flags &= ~ANYOF_EOS;
3115 cl_and(data->start_class, and_withp);
3117 flags &= ~SCF_DO_STCLASS;
3119 else if (REGNODE_VARIES(OP(scan))) {
3120 I32 mincount, maxcount, minnext, deltanext, fl = 0;
3121 I32 f = flags, pos_before = 0;
3122 regnode * const oscan = scan;
3123 struct regnode_charclass_class this_class;
3124 struct regnode_charclass_class *oclass = NULL;
3125 I32 next_is_eval = 0;
3127 switch (PL_regkind[OP(scan)]) {
3128 case WHILEM: /* End of (?:...)* . */
3129 scan = NEXTOPER(scan);
3132 if (flags & (SCF_DO_SUBSTR | SCF_DO_STCLASS)) {
3133 next = NEXTOPER(scan);
3134 if (OP(next) == EXACT || (flags & SCF_DO_STCLASS)) {
3136 maxcount = REG_INFTY;
3137 next = regnext(scan);
3138 scan = NEXTOPER(scan);
3142 if (flags & SCF_DO_SUBSTR)
3147 if (flags & SCF_DO_STCLASS) {
3149 maxcount = REG_INFTY;
3150 next = regnext(scan);
3151 scan = NEXTOPER(scan);
3154 is_inf = is_inf_internal = 1;
3155 scan = regnext(scan);
3156 if (flags & SCF_DO_SUBSTR) {
3157 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot extend fixed substrings */
3158 data->longest = &(data->longest_float);
3160 goto optimize_curly_tail;
3162 if (stopparen>0 && (OP(scan)==CURLYN || OP(scan)==CURLYM)
3163 && (scan->flags == stopparen))
3168 mincount = ARG1(scan);
3169 maxcount = ARG2(scan);
3171 next = regnext(scan);
3172 if (OP(scan) == CURLYX) {
3173 I32 lp = (data ? *(data->last_closep) : 0);
3174 scan->flags = ((lp <= (I32)U8_MAX) ? (U8)lp : U8_MAX);
3176 scan = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
3177 next_is_eval = (OP(scan) == EVAL);
3179 if (flags & SCF_DO_SUBSTR) {
3180 if (mincount == 0) SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot extend fixed substrings */
3181 pos_before = data->pos_min;
3185 data->flags &= ~(SF_HAS_PAR|SF_IN_PAR|SF_HAS_EVAL);
3187 data->flags |= SF_IS_INF;
3189 if (flags & SCF_DO_STCLASS) {
3190 cl_init(pRExC_state, &this_class);
3191 oclass = data->start_class;
3192 data->start_class = &this_class;
3193 f |= SCF_DO_STCLASS_AND;
3194 f &= ~SCF_DO_STCLASS_OR;
3196 /* These are the cases when once a subexpression
3197 fails at a particular position, it cannot succeed
3198 even after backtracking at the enclosing scope.
3200 XXXX what if minimal match and we are at the
3201 initial run of {n,m}? */
3202 if ((mincount != maxcount - 1) && (maxcount != REG_INFTY))
3203 f &= ~SCF_WHILEM_VISITED_POS;
3205 /* This will finish on WHILEM, setting scan, or on NULL: */
3206 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
3207 last, data, stopparen, recursed, NULL,
3209 ? (f & ~SCF_DO_SUBSTR) : f),depth+1);
3211 if (flags & SCF_DO_STCLASS)
3212 data->start_class = oclass;
3213 if (mincount == 0 || minnext == 0) {
3214 if (flags & SCF_DO_STCLASS_OR) {
3215 cl_or(pRExC_state, data->start_class, &this_class);
3217 else if (flags & SCF_DO_STCLASS_AND) {
3218 /* Switch to OR mode: cache the old value of
3219 * data->start_class */
3221 StructCopy(data->start_class, and_withp,
3222 struct regnode_charclass_class);
3223 flags &= ~SCF_DO_STCLASS_AND;
3224 StructCopy(&this_class, data->start_class,
3225 struct regnode_charclass_class);
3226 flags |= SCF_DO_STCLASS_OR;
3227 data->start_class->flags |= ANYOF_EOS;
3229 } else { /* Non-zero len */
3230 if (flags & SCF_DO_STCLASS_OR) {
3231 cl_or(pRExC_state, data->start_class, &this_class);
3232 cl_and(data->start_class, and_withp);
3234 else if (flags & SCF_DO_STCLASS_AND)
3235 cl_and(data->start_class, &this_class);
3236 flags &= ~SCF_DO_STCLASS;
3238 if (!scan) /* It was not CURLYX, but CURLY. */
3240 if ( /* ? quantifier ok, except for (?{ ... }) */
3241 (next_is_eval || !(mincount == 0 && maxcount == 1))
3242 && (minnext == 0) && (deltanext == 0)
3243 && data && !(data->flags & (SF_HAS_PAR|SF_IN_PAR))
3244 && maxcount <= REG_INFTY/3) /* Complement check for big count */
3246 ckWARNreg(RExC_parse,
3247 "Quantifier unexpected on zero-length expression");
3250 min += minnext * mincount;
3251 is_inf_internal |= ((maxcount == REG_INFTY
3252 && (minnext + deltanext) > 0)
3253 || deltanext == I32_MAX);
3254 is_inf |= is_inf_internal;
3255 delta += (minnext + deltanext) * maxcount - minnext * mincount;
3257 /* Try powerful optimization CURLYX => CURLYN. */
3258 if ( OP(oscan) == CURLYX && data
3259 && data->flags & SF_IN_PAR
3260 && !(data->flags & SF_HAS_EVAL)
3261 && !deltanext && minnext == 1 ) {
3262 /* Try to optimize to CURLYN. */
3263 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS;
3264 regnode * const nxt1 = nxt;
3271 if (!REGNODE_SIMPLE(OP(nxt))
3272 && !(PL_regkind[OP(nxt)] == EXACT
3273 && STR_LEN(nxt) == 1))
3279 if (OP(nxt) != CLOSE)
3281 if (RExC_open_parens) {
3282 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3283 RExC_close_parens[ARG(nxt1)-1]=nxt+2; /*close->while*/
3285 /* Now we know that nxt2 is the only contents: */
3286 oscan->flags = (U8)ARG(nxt);
3288 OP(nxt1) = NOTHING; /* was OPEN. */
3291 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3292 NEXT_OFF(nxt1+ 1) = 0; /* just for consistancy. */
3293 NEXT_OFF(nxt2) = 0; /* just for consistancy with CURLY. */
3294 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3295 OP(nxt + 1) = OPTIMIZED; /* was count. */
3296 NEXT_OFF(nxt+ 1) = 0; /* just for consistancy. */
3301 /* Try optimization CURLYX => CURLYM. */
3302 if ( OP(oscan) == CURLYX && data
3303 && !(data->flags & SF_HAS_PAR)
3304 && !(data->flags & SF_HAS_EVAL)
3305 && !deltanext /* atom is fixed width */
3306 && minnext != 0 /* CURLYM can't handle zero width */
3308 /* XXXX How to optimize if data == 0? */
3309 /* Optimize to a simpler form. */
3310 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN */
3314 while ( (nxt2 = regnext(nxt)) /* skip over embedded stuff*/
3315 && (OP(nxt2) != WHILEM))
3317 OP(nxt2) = SUCCEED; /* Whas WHILEM */
3318 /* Need to optimize away parenths. */
3319 if ((data->flags & SF_IN_PAR) && OP(nxt) == CLOSE) {
3320 /* Set the parenth number. */
3321 regnode *nxt1 = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN*/
3323 oscan->flags = (U8)ARG(nxt);
3324 if (RExC_open_parens) {
3325 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3326 RExC_close_parens[ARG(nxt1)-1]=nxt2+1; /*close->NOTHING*/
3328 OP(nxt1) = OPTIMIZED; /* was OPEN. */
3329 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3332 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3333 OP(nxt + 1) = OPTIMIZED; /* was count. */
3334 NEXT_OFF(nxt1 + 1) = 0; /* just for consistancy. */
3335 NEXT_OFF(nxt + 1) = 0; /* just for consistancy. */
3338 while ( nxt1 && (OP(nxt1) != WHILEM)) {
3339 regnode *nnxt = regnext(nxt1);
3342 if (reg_off_by_arg[OP(nxt1)])
3343 ARG_SET(nxt1, nxt2 - nxt1);
3344 else if (nxt2 - nxt1 < U16_MAX)
3345 NEXT_OFF(nxt1) = nxt2 - nxt1;
3347 OP(nxt) = NOTHING; /* Cannot beautify */
3352 /* Optimize again: */
3353 study_chunk(pRExC_state, &nxt1, minlenp, &deltanext, nxt,
3354 NULL, stopparen, recursed, NULL, 0,depth+1);
3359 else if ((OP(oscan) == CURLYX)
3360 && (flags & SCF_WHILEM_VISITED_POS)
3361 /* See the comment on a similar expression above.
3362 However, this time it not a subexpression
3363 we care about, but the expression itself. */
3364 && (maxcount == REG_INFTY)
3365 && data && ++data->whilem_c < 16) {
3366 /* This stays as CURLYX, we can put the count/of pair. */
3367 /* Find WHILEM (as in regexec.c) */
3368 regnode *nxt = oscan + NEXT_OFF(oscan);
3370 if (OP(PREVOPER(nxt)) == NOTHING) /* LONGJMP */
3372 PREVOPER(nxt)->flags = (U8)(data->whilem_c
3373 | (RExC_whilem_seen << 4)); /* On WHILEM */
3375 if (data && fl & (SF_HAS_PAR|SF_IN_PAR))
3377 if (flags & SCF_DO_SUBSTR) {
3378 SV *last_str = NULL;
3379 int counted = mincount != 0;
3381 if (data->last_end > 0 && mincount != 0) { /* Ends with a string. */
3382 #if defined(SPARC64_GCC_WORKAROUND)
3385 const char *s = NULL;
3388 if (pos_before >= data->last_start_min)
3391 b = data->last_start_min;
3394 s = SvPV_const(data->last_found, l);
3395 old = b - data->last_start_min;
3398 I32 b = pos_before >= data->last_start_min
3399 ? pos_before : data->last_start_min;
3401 const char * const s = SvPV_const(data->last_found, l);
3402 I32 old = b - data->last_start_min;
3406 old = utf8_hop((U8*)s, old) - (U8*)s;
3409 /* Get the added string: */
3410 last_str = newSVpvn_utf8(s + old, l, UTF);
3411 if (deltanext == 0 && pos_before == b) {
3412 /* What was added is a constant string */
3414 SvGROW(last_str, (mincount * l) + 1);
3415 repeatcpy(SvPVX(last_str) + l,
3416 SvPVX_const(last_str), l, mincount - 1);
3417 SvCUR_set(last_str, SvCUR(last_str) * mincount);
3418 /* Add additional parts. */
3419 SvCUR_set(data->last_found,
3420 SvCUR(data->last_found) - l);
3421 sv_catsv(data->last_found, last_str);
3423 SV * sv = data->last_found;
3425 SvUTF8(sv) && SvMAGICAL(sv) ?
3426 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3427 if (mg && mg->mg_len >= 0)
3428 mg->mg_len += CHR_SVLEN(last_str) - l;
3430 data->last_end += l * (mincount - 1);
3433 /* start offset must point into the last copy */
3434 data->last_start_min += minnext * (mincount - 1);
3435 data->last_start_max += is_inf ? I32_MAX
3436 : (maxcount - 1) * (minnext + data->pos_delta);
3439 /* It is counted once already... */
3440 data->pos_min += minnext * (mincount - counted);
3441 data->pos_delta += - counted * deltanext +
3442 (minnext + deltanext) * maxcount - minnext * mincount;
3443 if (mincount != maxcount) {
3444 /* Cannot extend fixed substrings found inside
3446 SCAN_COMMIT(pRExC_state,data,minlenp);
3447 if (mincount && last_str) {
3448 SV * const sv = data->last_found;
3449 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3450 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3454 sv_setsv(sv, last_str);
3455 data->last_end = data->pos_min;
3456 data->last_start_min =
3457 data->pos_min - CHR_SVLEN(last_str);
3458 data->last_start_max = is_inf
3460 : data->pos_min + data->pos_delta
3461 - CHR_SVLEN(last_str);
3463 data->longest = &(data->longest_float);
3465 SvREFCNT_dec(last_str);
3467 if (data && (fl & SF_HAS_EVAL))
3468 data->flags |= SF_HAS_EVAL;
3469 optimize_curly_tail:
3470 if (OP(oscan) != CURLYX) {
3471 while (PL_regkind[OP(next = regnext(oscan))] == NOTHING
3473 NEXT_OFF(oscan) += NEXT_OFF(next);
3476 default: /* REF and CLUMP only? */
3477 if (flags & SCF_DO_SUBSTR) {
3478 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3479 data->longest = &(data->longest_float);
3481 is_inf = is_inf_internal = 1;
3482 if (flags & SCF_DO_STCLASS_OR)
3483 cl_anything(pRExC_state, data->start_class);
3484 flags &= ~SCF_DO_STCLASS;
3488 else if (OP(scan) == LNBREAK) {
3489 if (flags & SCF_DO_STCLASS) {
3491 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3492 if (flags & SCF_DO_STCLASS_AND) {
3493 for (value = 0; value < 256; value++)
3494 if (!is_VERTWS_cp(value))
3495 ANYOF_BITMAP_CLEAR(data->start_class, value);
3498 for (value = 0; value < 256; value++)
3499 if (is_VERTWS_cp(value))
3500 ANYOF_BITMAP_SET(data->start_class, value);
3502 if (flags & SCF_DO_STCLASS_OR)
3503 cl_and(data->start_class, and_withp);
3504 flags &= ~SCF_DO_STCLASS;
3508 if (flags & SCF_DO_SUBSTR) {
3509 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3511 data->pos_delta += 1;
3512 data->longest = &(data->longest_float);
3516 else if (OP(scan) == FOLDCHAR) {
3517 int d = ARG(scan)==0xDF ? 1 : 2;
3518 flags &= ~SCF_DO_STCLASS;
3521 if (flags & SCF_DO_SUBSTR) {
3522 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3524 data->pos_delta += d;
3525 data->longest = &(data->longest_float);
3528 else if (REGNODE_SIMPLE(OP(scan))) {
3531 if (flags & SCF_DO_SUBSTR) {
3532 SCAN_COMMIT(pRExC_state,data,minlenp);
3536 if (flags & SCF_DO_STCLASS) {
3537 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3539 /* Some of the logic below assumes that switching
3540 locale on will only add false positives. */
3541 switch (PL_regkind[OP(scan)]) {
3545 /* Perl_croak(aTHX_ "panic: unexpected simple REx opcode %d", OP(scan)); */
3546 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3547 cl_anything(pRExC_state, data->start_class);
3550 if (OP(scan) == SANY)
3552 if (flags & SCF_DO_STCLASS_OR) { /* Everything but \n */
3553 value = (ANYOF_BITMAP_TEST(data->start_class,'\n')
3554 || (data->start_class->flags & ANYOF_CLASS));
3555 cl_anything(pRExC_state, data->start_class);
3557 if (flags & SCF_DO_STCLASS_AND || !value)
3558 ANYOF_BITMAP_CLEAR(data->start_class,'\n');
3561 if (flags & SCF_DO_STCLASS_AND)
3562 cl_and(data->start_class,
3563 (struct regnode_charclass_class*)scan);
3565 cl_or(pRExC_state, data->start_class,
3566 (struct regnode_charclass_class*)scan);
3569 if (flags & SCF_DO_STCLASS_AND) {
3570 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3571 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3572 for (value = 0; value < 256; value++)
3573 if (!isALNUM(value))
3574 ANYOF_BITMAP_CLEAR(data->start_class, value);
3578 if (data->start_class->flags & ANYOF_LOCALE)
3579 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3581 for (value = 0; value < 256; value++)
3583 ANYOF_BITMAP_SET(data->start_class, value);
3588 if (flags & SCF_DO_STCLASS_AND) {
3589 if (data->start_class->flags & ANYOF_LOCALE)
3590 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3593 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3594 data->start_class->flags |= ANYOF_LOCALE;
3598 if (flags & SCF_DO_STCLASS_AND) {
3599 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3600 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3601 for (value = 0; value < 256; value++)
3603 ANYOF_BITMAP_CLEAR(data->start_class, value);
3607 if (data->start_class->flags & ANYOF_LOCALE)
3608 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3610 for (value = 0; value < 256; value++)
3611 if (!isALNUM(value))
3612 ANYOF_BITMAP_SET(data->start_class, value);
3617 if (flags & SCF_DO_STCLASS_AND) {
3618 if (data->start_class->flags & ANYOF_LOCALE)
3619 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3622 data->start_class->flags |= ANYOF_LOCALE;
3623 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3627 if (flags & SCF_DO_STCLASS_AND) {
3628 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3629 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3630 for (value = 0; value < 256; value++)
3631 if (!isSPACE(value))
3632 ANYOF_BITMAP_CLEAR(data->start_class, value);
3636 if (data->start_class->flags & ANYOF_LOCALE)
3637 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3639 for (value = 0; value < 256; value++)
3641 ANYOF_BITMAP_SET(data->start_class, value);
3646 if (flags & SCF_DO_STCLASS_AND) {
3647 if (data->start_class->flags & ANYOF_LOCALE)
3648 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3651 data->start_class->flags |= ANYOF_LOCALE;
3652 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3656 if (flags & SCF_DO_STCLASS_AND) {
3657 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3658 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3659 for (value = 0; value < 256; value++)
3661 ANYOF_BITMAP_CLEAR(data->start_class, value);
3665 if (data->start_class->flags & ANYOF_LOCALE)
3666 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3668 for (value = 0; value < 256; value++)
3669 if (!isSPACE(value))
3670 ANYOF_BITMAP_SET(data->start_class, value);
3675 if (flags & SCF_DO_STCLASS_AND) {
3676 if (data->start_class->flags & ANYOF_LOCALE) {
3677 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3678 for (value = 0; value < 256; value++)
3679 if (!isSPACE(value))
3680 ANYOF_BITMAP_CLEAR(data->start_class, value);
3684 data->start_class->flags |= ANYOF_LOCALE;
3685 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3689 if (flags & SCF_DO_STCLASS_AND) {
3690 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NDIGIT);
3691 for (value = 0; value < 256; value++)
3692 if (!isDIGIT(value))
3693 ANYOF_BITMAP_CLEAR(data->start_class, value);
3696 if (data->start_class->flags & ANYOF_LOCALE)
3697 ANYOF_CLASS_SET(data->start_class,ANYOF_DIGIT);
3699 for (value = 0; value < 256; value++)
3701 ANYOF_BITMAP_SET(data->start_class, value);
3706 if (flags & SCF_DO_STCLASS_AND) {
3707 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_DIGIT);
3708 for (value = 0; value < 256; value++)
3710 ANYOF_BITMAP_CLEAR(data->start_class, value);
3713 if (data->start_class->flags & ANYOF_LOCALE)
3714 ANYOF_CLASS_SET(data->start_class,ANYOF_NDIGIT);
3716 for (value = 0; value < 256; value++)
3717 if (!isDIGIT(value))
3718 ANYOF_BITMAP_SET(data->start_class, value);
3722 CASE_SYNST_FNC(VERTWS);
3723 CASE_SYNST_FNC(HORIZWS);
3726 if (flags & SCF_DO_STCLASS_OR)
3727 cl_and(data->start_class, and_withp);
3728 flags &= ~SCF_DO_STCLASS;
3731 else if (PL_regkind[OP(scan)] == EOL && flags & SCF_DO_SUBSTR) {
3732 data->flags |= (OP(scan) == MEOL
3736 else if ( PL_regkind[OP(scan)] == BRANCHJ
3737 /* Lookbehind, or need to calculate parens/evals/stclass: */
3738 && (scan->flags || data || (flags & SCF_DO_STCLASS))
3739 && (OP(scan) == IFMATCH || OP(scan) == UNLESSM)) {
3740 if ( !PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3741 || OP(scan) == UNLESSM )
3743 /* Negative Lookahead/lookbehind
3744 In this case we can't do fixed string optimisation.
3747 I32 deltanext, minnext, fake = 0;
3749 struct regnode_charclass_class intrnl;
3752 data_fake.flags = 0;
3754 data_fake.whilem_c = data->whilem_c;
3755 data_fake.last_closep = data->last_closep;
3758 data_fake.last_closep = &fake;
3759 data_fake.pos_delta = delta;
3760 if ( flags & SCF_DO_STCLASS && !scan->flags
3761 && OP(scan) == IFMATCH ) { /* Lookahead */
3762 cl_init(pRExC_state, &intrnl);
3763 data_fake.start_class = &intrnl;
3764 f |= SCF_DO_STCLASS_AND;
3766 if (flags & SCF_WHILEM_VISITED_POS)
3767 f |= SCF_WHILEM_VISITED_POS;
3768 next = regnext(scan);
3769 nscan = NEXTOPER(NEXTOPER(scan));
3770 minnext = study_chunk(pRExC_state, &nscan, minlenp, &deltanext,
3771 last, &data_fake, stopparen, recursed, NULL, f, depth+1);
3774 FAIL("Variable length lookbehind not implemented");
3776 else if (minnext > (I32)U8_MAX) {
3777 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3779 scan->flags = (U8)minnext;
3782 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3784 if (data_fake.flags & SF_HAS_EVAL)
3785 data->flags |= SF_HAS_EVAL;
3786 data->whilem_c = data_fake.whilem_c;
3788 if (f & SCF_DO_STCLASS_AND) {
3789 if (flags & SCF_DO_STCLASS_OR) {
3790 /* OR before, AND after: ideally we would recurse with
3791 * data_fake to get the AND applied by study of the
3792 * remainder of the pattern, and then derecurse;
3793 * *** HACK *** for now just treat as "no information".
3794 * See [perl #56690].
3796 cl_init(pRExC_state, data->start_class);
3798 /* AND before and after: combine and continue */
3799 const int was = (data->start_class->flags & ANYOF_EOS);
3801 cl_and(data->start_class, &intrnl);
3803 data->start_class->flags |= ANYOF_EOS;
3807 #if PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3809 /* Positive Lookahead/lookbehind
3810 In this case we can do fixed string optimisation,
3811 but we must be careful about it. Note in the case of
3812 lookbehind the positions will be offset by the minimum
3813 length of the pattern, something we won't know about
3814 until after the recurse.
3816 I32 deltanext, fake = 0;
3818 struct regnode_charclass_class intrnl;
3820 /* We use SAVEFREEPV so that when the full compile
3821 is finished perl will clean up the allocated
3822 minlens when its all done. This was we don't
3823 have to worry about freeing them when we know
3824 they wont be used, which would be a pain.
3827 Newx( minnextp, 1, I32 );
3828 SAVEFREEPV(minnextp);
3831 StructCopy(data, &data_fake, scan_data_t);
3832 if ((flags & SCF_DO_SUBSTR) && data->last_found) {
3835 SCAN_COMMIT(pRExC_state, &data_fake,minlenp);
3836 data_fake.last_found=newSVsv(data->last_found);
3840 data_fake.last_closep = &fake;
3841 data_fake.flags = 0;
3842 data_fake.pos_delta = delta;
3844 data_fake.flags |= SF_IS_INF;
3845 if ( flags & SCF_DO_STCLASS && !scan->flags
3846 && OP(scan) == IFMATCH ) { /* Lookahead */
3847 cl_init(pRExC_state, &intrnl);
3848 data_fake.start_class = &intrnl;
3849 f |= SCF_DO_STCLASS_AND;
3851 if (flags & SCF_WHILEM_VISITED_POS)
3852 f |= SCF_WHILEM_VISITED_POS;
3853 next = regnext(scan);
3854 nscan = NEXTOPER(NEXTOPER(scan));
3856 *minnextp = study_chunk(pRExC_state, &nscan, minnextp, &deltanext,
3857 last, &data_fake, stopparen, recursed, NULL, f,depth+1);
3860 FAIL("Variable length lookbehind not implemented");
3862 else if (*minnextp > (I32)U8_MAX) {
3863 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3865 scan->flags = (U8)*minnextp;
3870 if (f & SCF_DO_STCLASS_AND) {
3871 const int was = (data->start_class->flags & ANYOF_EOS);
3873 cl_and(data->start_class, &intrnl);
3875 data->start_class->flags |= ANYOF_EOS;
3878 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3880 if (data_fake.flags & SF_HAS_EVAL)
3881 data->flags |= SF_HAS_EVAL;
3882 data->whilem_c = data_fake.whilem_c;
3883 if ((flags & SCF_DO_SUBSTR) && data_fake.last_found) {
3884 if (RExC_rx->minlen<*minnextp)
3885 RExC_rx->minlen=*minnextp;
3886 SCAN_COMMIT(pRExC_state, &data_fake, minnextp);
3887 SvREFCNT_dec(data_fake.last_found);
3889 if ( data_fake.minlen_fixed != minlenp )
3891 data->offset_fixed= data_fake.offset_fixed;
3892 data->minlen_fixed= data_fake.minlen_fixed;
3893 data->lookbehind_fixed+= scan->flags;
3895 if ( data_fake.minlen_float != minlenp )
3897 data->minlen_float= data_fake.minlen_float;
3898 data->offset_float_min=data_fake.offset_float_min;
3899 data->offset_float_max=data_fake.offset_float_max;
3900 data->lookbehind_float+= scan->flags;
3909 else if (OP(scan) == OPEN) {
3910 if (stopparen != (I32)ARG(scan))
3913 else if (OP(scan) == CLOSE) {
3914 if (stopparen == (I32)ARG(scan)) {
3917 if ((I32)ARG(scan) == is_par) {
3918 next = regnext(scan);
3920 if ( next && (OP(next) != WHILEM) && next < last)
3921 is_par = 0; /* Disable optimization */
3924 *(data->last_closep) = ARG(scan);
3926 else if (OP(scan) == EVAL) {
3928 data->flags |= SF_HAS_EVAL;
3930 else if ( PL_regkind[OP(scan)] == ENDLIKE ) {
3931 if (flags & SCF_DO_SUBSTR) {
3932 SCAN_COMMIT(pRExC_state,data,minlenp);
3933 flags &= ~SCF_DO_SUBSTR;
3935 if (data && OP(scan)==ACCEPT) {
3936 data->flags |= SCF_SEEN_ACCEPT;
3941 else if (OP(scan) == LOGICAL && scan->flags == 2) /* Embedded follows */
3943 if (flags & SCF_DO_SUBSTR) {
3944 SCAN_COMMIT(pRExC_state,data,minlenp);
3945 data->longest = &(data->longest_float);
3947 is_inf = is_inf_internal = 1;
3948 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3949 cl_anything(pRExC_state, data->start_class);
3950 flags &= ~SCF_DO_STCLASS;
3952 else if (OP(scan) == GPOS) {
3953 if (!(RExC_rx->extflags & RXf_GPOS_FLOAT) &&
3954 !(delta || is_inf || (data && data->pos_delta)))
3956 if (!(RExC_rx->extflags & RXf_ANCH) && (flags & SCF_DO_SUBSTR))
3957 RExC_rx->extflags |= RXf_ANCH_GPOS;
3958 if (RExC_rx->gofs < (U32)min)
3959 RExC_rx->gofs = min;
3961 RExC_rx->extflags |= RXf_GPOS_FLOAT;
3965 #ifdef TRIE_STUDY_OPT
3966 #ifdef FULL_TRIE_STUDY
3967 else if (PL_regkind[OP(scan)] == TRIE) {
3968 /* NOTE - There is similar code to this block above for handling
3969 BRANCH nodes on the initial study. If you change stuff here
3971 regnode *trie_node= scan;
3972 regnode *tail= regnext(scan);
3973 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
3974 I32 max1 = 0, min1 = I32_MAX;
3975 struct regnode_charclass_class accum;
3977 if (flags & SCF_DO_SUBSTR) /* XXXX Add !SUSPEND? */
3978 SCAN_COMMIT(pRExC_state, data,minlenp); /* Cannot merge strings after this. */
3979 if (flags & SCF_DO_STCLASS)
3980 cl_init_zero(pRExC_state, &accum);
3986 const regnode *nextbranch= NULL;
3989 for ( word=1 ; word <= trie->wordcount ; word++)
3991 I32 deltanext=0, minnext=0, f = 0, fake;
3992 struct regnode_charclass_class this_class;
3994 data_fake.flags = 0;
3996 data_fake.whilem_c = data->whilem_c;
3997 data_fake.last_closep = data->last_closep;
4000 data_fake.last_closep = &fake;
4001 data_fake.pos_delta = delta;
4002 if (flags & SCF_DO_STCLASS) {
4003 cl_init(pRExC_state, &this_class);
4004 data_fake.start_class = &this_class;
4005 f = SCF_DO_STCLASS_AND;
4007 if (flags & SCF_WHILEM_VISITED_POS)
4008 f |= SCF_WHILEM_VISITED_POS;
4010 if (trie->jump[word]) {
4012 nextbranch = trie_node + trie->jump[0];
4013 scan= trie_node + trie->jump[word];
4014 /* We go from the jump point to the branch that follows
4015 it. Note this means we need the vestigal unused branches
4016 even though they arent otherwise used.
4018 minnext = study_chunk(pRExC_state, &scan, minlenp,
4019 &deltanext, (regnode *)nextbranch, &data_fake,
4020 stopparen, recursed, NULL, f,depth+1);
4022 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
4023 nextbranch= regnext((regnode*)nextbranch);
4025 if (min1 > (I32)(minnext + trie->minlen))
4026 min1 = minnext + trie->minlen;
4027 if (max1 < (I32)(minnext + deltanext + trie->maxlen))
4028 max1 = minnext + deltanext + trie->maxlen;
4029 if (deltanext == I32_MAX)
4030 is_inf = is_inf_internal = 1;
4032 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
4034 if (data_fake.flags & SCF_SEEN_ACCEPT) {
4035 if ( stopmin > min + min1)
4036 stopmin = min + min1;
4037 flags &= ~SCF_DO_SUBSTR;
4039 data->flags |= SCF_SEEN_ACCEPT;
4042 if (data_fake.flags & SF_HAS_EVAL)
4043 data->flags |= SF_HAS_EVAL;
4044 data->whilem_c = data_fake.whilem_c;
4046 if (flags & SCF_DO_STCLASS)
4047 cl_or(pRExC_state, &accum, &this_class);
4050 if (flags & SCF_DO_SUBSTR) {
4051 data->pos_min += min1;
4052 data->pos_delta += max1 - min1;
4053 if (max1 != min1 || is_inf)
4054 data->longest = &(data->longest_float);
4057 delta += max1 - min1;
4058 if (flags & SCF_DO_STCLASS_OR) {
4059 cl_or(pRExC_state, data->start_class, &accum);
4061 cl_and(data->start_class, and_withp);
4062 flags &= ~SCF_DO_STCLASS;
4065 else if (flags & SCF_DO_STCLASS_AND) {
4067 cl_and(data->start_class, &accum);
4068 flags &= ~SCF_DO_STCLASS;
4071 /* Switch to OR mode: cache the old value of
4072 * data->start_class */
4074 StructCopy(data->start_class, and_withp,
4075 struct regnode_charclass_class);
4076 flags &= ~SCF_DO_STCLASS_AND;
4077 StructCopy(&accum, data->start_class,
4078 struct regnode_charclass_class);
4079 flags |= SCF_DO_STCLASS_OR;
4080 data->start_class->flags |= ANYOF_EOS;
4087 else if (PL_regkind[OP(scan)] == TRIE) {
4088 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
4091 min += trie->minlen;
4092 delta += (trie->maxlen - trie->minlen);
4093 flags &= ~SCF_DO_STCLASS; /* xxx */
4094 if (flags & SCF_DO_SUBSTR) {
4095 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
4096 data->pos_min += trie->minlen;
4097 data->pos_delta += (trie->maxlen - trie->minlen);
4098 if (trie->maxlen != trie->minlen)
4099 data->longest = &(data->longest_float);
4101 if (trie->jump) /* no more substrings -- for now /grr*/
4102 flags &= ~SCF_DO_SUBSTR;
4104 #endif /* old or new */
4105 #endif /* TRIE_STUDY_OPT */
4107 /* Else: zero-length, ignore. */
4108 scan = regnext(scan);
4113 stopparen = frame->stop;
4114 frame = frame->prev;
4115 goto fake_study_recurse;
4120 DEBUG_STUDYDATA("pre-fin:",data,depth);
4123 *deltap = is_inf_internal ? I32_MAX : delta;
4124 if (flags & SCF_DO_SUBSTR && is_inf)
4125 data->pos_delta = I32_MAX - data->pos_min;
4126 if (is_par > (I32)U8_MAX)
4128 if (is_par && pars==1 && data) {
4129 data->flags |= SF_IN_PAR;
4130 data->flags &= ~SF_HAS_PAR;
4132 else if (pars && data) {
4133 data->flags |= SF_HAS_PAR;
4134 data->flags &= ~SF_IN_PAR;
4136 if (flags & SCF_DO_STCLASS_OR)
4137 cl_and(data->start_class, and_withp);
4138 if (flags & SCF_TRIE_RESTUDY)
4139 data->flags |= SCF_TRIE_RESTUDY;
4141 DEBUG_STUDYDATA("post-fin:",data,depth);
4143 return min < stopmin ? min : stopmin;
4147 S_add_data(RExC_state_t *pRExC_state, U32 n, const char *s)
4149 U32 count = RExC_rxi->data ? RExC_rxi->data->count : 0;
4151 PERL_ARGS_ASSERT_ADD_DATA;
4153 Renewc(RExC_rxi->data,
4154 sizeof(*RExC_rxi->data) + sizeof(void*) * (count + n - 1),
4155 char, struct reg_data);
4157 Renew(RExC_rxi->data->what, count + n, U8);
4159 Newx(RExC_rxi->data->what, n, U8);
4160 RExC_rxi->data->count = count + n;
4161 Copy(s, RExC_rxi->data->what + count, n, U8);
4165 /*XXX: todo make this not included in a non debugging perl */
4166 #ifndef PERL_IN_XSUB_RE
4168 Perl_reginitcolors(pTHX)
4171 const char * const s = PerlEnv_getenv("PERL_RE_COLORS");
4173 char *t = savepv(s);
4177 t = strchr(t, '\t');
4183 PL_colors[i] = t = (char *)"";
4188 PL_colors[i++] = (char *)"";
4195 #ifdef TRIE_STUDY_OPT
4196 #define CHECK_RESTUDY_GOTO \
4198 (data.flags & SCF_TRIE_RESTUDY) \
4202 #define CHECK_RESTUDY_GOTO
4206 - pregcomp - compile a regular expression into internal code
4208 * We can't allocate space until we know how big the compiled form will be,
4209 * but we can't compile it (and thus know how big it is) until we've got a
4210 * place to put the code. So we cheat: we compile it twice, once with code
4211 * generation turned off and size counting turned on, and once "for real".
4212 * This also means that we don't allocate space until we are sure that the
4213 * thing really will compile successfully, and we never have to move the
4214 * code and thus invalidate pointers into it. (Note that it has to be in
4215 * one piece because free() must be able to free it all.) [NB: not true in perl]
4217 * Beware that the optimization-preparation code in here knows about some
4218 * of the structure of the compiled regexp. [I'll say.]
4223 #ifndef PERL_IN_XSUB_RE
4224 #define RE_ENGINE_PTR &PL_core_reg_engine
4226 extern const struct regexp_engine my_reg_engine;
4227 #define RE_ENGINE_PTR &my_reg_engine
4230 #ifndef PERL_IN_XSUB_RE
4232 Perl_pregcomp(pTHX_ SV * const pattern, const U32 flags)
4235 HV * const table = GvHV(PL_hintgv);
4237 PERL_ARGS_ASSERT_PREGCOMP;
4239 /* Dispatch a request to compile a regexp to correct
4242 SV **ptr= hv_fetchs(table, "regcomp", FALSE);
4243 GET_RE_DEBUG_FLAGS_DECL;
4244 if (ptr && SvIOK(*ptr) && SvIV(*ptr)) {
4245 const regexp_engine *eng=INT2PTR(regexp_engine*,SvIV(*ptr));
4247 PerlIO_printf(Perl_debug_log, "Using engine %"UVxf"\n",
4250 return CALLREGCOMP_ENG(eng, pattern, flags);
4253 return Perl_re_compile(aTHX_ pattern, flags);
4258 Perl_re_compile(pTHX_ SV * const pattern, U32 pm_flags)
4263 register regexp_internal *ri;
4265 char *exp = SvPV(pattern, plen);
4266 char* xend = exp + plen;
4273 RExC_state_t RExC_state;
4274 RExC_state_t * const pRExC_state = &RExC_state;
4275 #ifdef TRIE_STUDY_OPT
4277 RExC_state_t copyRExC_state;
4279 GET_RE_DEBUG_FLAGS_DECL;
4281 PERL_ARGS_ASSERT_RE_COMPILE;
4283 DEBUG_r(if (!PL_colorset) reginitcolors());
4285 RExC_utf8 = RExC_orig_utf8 = SvUTF8(pattern);
4288 SV *dsv= sv_newmortal();
4289 RE_PV_QUOTED_DECL(s, RExC_utf8,
4290 dsv, exp, plen, 60);
4291 PerlIO_printf(Perl_debug_log, "%sCompiling REx%s %s\n",
4292 PL_colors[4],PL_colors[5],s);
4297 RExC_flags = pm_flags;
4301 RExC_seen_zerolen = *exp == '^' ? -1 : 0;
4302 RExC_seen_evals = 0;
4305 /* First pass: determine size, legality. */
4313 RExC_emit = &PL_regdummy;
4314 RExC_whilem_seen = 0;
4315 RExC_open_parens = NULL;
4316 RExC_close_parens = NULL;
4318 RExC_paren_names = NULL;
4320 RExC_paren_name_list = NULL;
4322 RExC_recurse = NULL;
4323 RExC_recurse_count = 0;
4325 #if 0 /* REGC() is (currently) a NOP at the first pass.
4326 * Clever compilers notice this and complain. --jhi */
4327 REGC((U8)REG_MAGIC, (char*)RExC_emit);
4329 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "Starting first pass (sizing)\n"));
4330 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4331 RExC_precomp = NULL;
4334 if (RExC_utf8 && !RExC_orig_utf8) {
4335 /* It's possible to write a regexp in ascii that represents Unicode
4336 codepoints outside of the byte range, such as via \x{100}. If we
4337 detect such a sequence we have to convert the entire pattern to utf8
4338 and then recompile, as our sizing calculation will have been based
4339 on 1 byte == 1 character, but we will need to use utf8 to encode
4340 at least some part of the pattern, and therefore must convert the whole
4342 XXX: somehow figure out how to make this less expensive...
4345 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log,
4346 "UTF8 mismatch! Converting to utf8 for resizing and compile\n"));
4347 exp = (char*)Perl_bytes_to_utf8(aTHX_ (U8*)exp, &len);
4349 RExC_orig_utf8 = RExC_utf8;
4351 goto redo_first_pass;
4354 PerlIO_printf(Perl_debug_log,
4355 "Required size %"IVdf" nodes\n"
4356 "Starting second pass (creation)\n",
4359 RExC_lastparse=NULL;
4361 /* Small enough for pointer-storage convention?
4362 If extralen==0, this means that we will not need long jumps. */
4363 if (RExC_size >= 0x10000L && RExC_extralen)
4364 RExC_size += RExC_extralen;
4367 if (RExC_whilem_seen > 15)
4368 RExC_whilem_seen = 15;
4370 /* Allocate space and zero-initialize. Note, the two step process
4371 of zeroing when in debug mode, thus anything assigned has to
4372 happen after that */
4373 rx = (REGEXP*) newSV_type(SVt_REGEXP);
4374 r = (struct regexp*)SvANY(rx);
4375 Newxc(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode),
4376 char, regexp_internal);
4377 if ( r == NULL || ri == NULL )
4378 FAIL("Regexp out of space");
4380 /* avoid reading uninitialized memory in DEBUGGING code in study_chunk() */
4381 Zero(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode), char);
4383 /* bulk initialize base fields with 0. */
4384 Zero(ri, sizeof(regexp_internal), char);
4387 /* non-zero initialization begins here */
4389 r->engine= RE_ENGINE_PTR;
4390 r->extflags = pm_flags;
4392 bool has_p = ((r->extflags & RXf_PMf_KEEPCOPY) == RXf_PMf_KEEPCOPY);
4393 bool has_minus = ((r->extflags & RXf_PMf_STD_PMMOD) != RXf_PMf_STD_PMMOD);
4394 bool has_runon = ((RExC_seen & REG_SEEN_RUN_ON_COMMENT)==REG_SEEN_RUN_ON_COMMENT);
4395 U16 reganch = (U16)((r->extflags & RXf_PMf_STD_PMMOD)
4396 >> RXf_PMf_STD_PMMOD_SHIFT);
4397 const char *fptr = STD_PAT_MODS; /*"msix"*/
4399 const STRLEN wraplen = plen + has_minus + has_p + has_runon
4400 + (sizeof(STD_PAT_MODS) - 1)
4401 + (sizeof("(?:)") - 1);
4403 p = sv_grow(MUTABLE_SV(rx), wraplen + 1);
4404 SvCUR_set(rx, wraplen);
4406 SvFLAGS(rx) |= SvUTF8(pattern);
4409 *p++ = KEEPCOPY_PAT_MOD; /*'p'*/
4411 char *r = p + (sizeof(STD_PAT_MODS) - 1) + has_minus - 1;
4412 char *colon = r + 1;
4415 while((ch = *fptr++)) {
4429 Copy(RExC_precomp, p, plen, char);
4430 assert ((RX_WRAPPED(rx) - p) < 16);
4431 r->pre_prefix = p - RX_WRAPPED(rx);
4440 r->nparens = RExC_npar - 1; /* set early to validate backrefs */
4442 if (RExC_seen & REG_SEEN_RECURSE) {
4443 Newxz(RExC_open_parens, RExC_npar,regnode *);
4444 SAVEFREEPV(RExC_open_parens);
4445 Newxz(RExC_close_parens,RExC_npar,regnode *);
4446 SAVEFREEPV(RExC_close_parens);
4449 /* Useful during FAIL. */
4450 #ifdef RE_TRACK_PATTERN_OFFSETS
4451 Newxz(ri->u.offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
4452 DEBUG_OFFSETS_r(PerlIO_printf(Perl_debug_log,
4453 "%s %"UVuf" bytes for offset annotations.\n",
4454 ri->u.offsets ? "Got" : "Couldn't get",
4455 (UV)((2*RExC_size+1) * sizeof(U32))));
4457 SetProgLen(ri,RExC_size);
4461 REH_CALL_COMP_BEGIN_HOOK(pRExC_state->rx);
4463 /* Second pass: emit code. */
4464 RExC_flags = pm_flags; /* don't let top level (?i) bleed */
4469 RExC_emit_start = ri->program;
4470 RExC_emit = ri->program;
4471 RExC_emit_bound = ri->program + RExC_size + 1;
4473 /* Store the count of eval-groups for security checks: */
4474 RExC_rx->seen_evals = RExC_seen_evals;
4475 REGC((U8)REG_MAGIC, (char*) RExC_emit++);
4476 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4480 /* XXXX To minimize changes to RE engine we always allocate
4481 3-units-long substrs field. */
4482 Newx(r->substrs, 1, struct reg_substr_data);
4483 if (RExC_recurse_count) {
4484 Newxz(RExC_recurse,RExC_recurse_count,regnode *);
4485 SAVEFREEPV(RExC_recurse);
4489 r->minlen = minlen = sawplus = sawopen = 0;
4490 Zero(r->substrs, 1, struct reg_substr_data);
4492 #ifdef TRIE_STUDY_OPT
4494 StructCopy(&zero_scan_data, &data, scan_data_t);
4495 copyRExC_state = RExC_state;
4498 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log,"Restudying\n"));
4500 RExC_state = copyRExC_state;
4501 if (seen & REG_TOP_LEVEL_BRANCHES)
4502 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
4504 RExC_seen &= ~REG_TOP_LEVEL_BRANCHES;
4505 if (data.last_found) {
4506 SvREFCNT_dec(data.longest_fixed);
4507 SvREFCNT_dec(data.longest_float);
4508 SvREFCNT_dec(data.last_found);
4510 StructCopy(&zero_scan_data, &data, scan_data_t);
4513 StructCopy(&zero_scan_data, &data, scan_data_t);
4516 /* Dig out information for optimizations. */
4517 r->extflags = RExC_flags; /* was pm_op */
4518 /*dmq: removed as part of de-PMOP: pm->op_pmflags = RExC_flags; */
4521 SvUTF8_on(rx); /* Unicode in it? */
4522 ri->regstclass = NULL;
4523 if (RExC_naughty >= 10) /* Probably an expensive pattern. */
4524 r->intflags |= PREGf_NAUGHTY;
4525 scan = ri->program + 1; /* First BRANCH. */
4527 /* testing for BRANCH here tells us whether there is "must appear"
4528 data in the pattern. If there is then we can use it for optimisations */
4529 if (!(RExC_seen & REG_TOP_LEVEL_BRANCHES)) { /* Only one top-level choice. */
4531 STRLEN longest_float_length, longest_fixed_length;
4532 struct regnode_charclass_class ch_class; /* pointed to by data */
4534 I32 last_close = 0; /* pointed to by data */
4535 regnode *first= scan;
4536 regnode *first_next= regnext(first);
4539 * Skip introductions and multiplicators >= 1
4540 * so that we can extract the 'meat' of the pattern that must
4541 * match in the large if() sequence following.
4542 * NOTE that EXACT is NOT covered here, as it is normally
4543 * picked up by the optimiser separately.
4545 * This is unfortunate as the optimiser isnt handling lookahead
4546 * properly currently.
4549 while ((OP(first) == OPEN && (sawopen = 1)) ||
4550 /* An OR of *one* alternative - should not happen now. */
4551 (OP(first) == BRANCH && OP(first_next) != BRANCH) ||
4552 /* for now we can't handle lookbehind IFMATCH*/
4553 (OP(first) == IFMATCH && !first->flags) ||
4554 (OP(first) == PLUS) ||
4555 (OP(first) == MINMOD) ||
4556 /* An {n,m} with n>0 */
4557 (PL_regkind[OP(first)] == CURLY && ARG1(first) > 0) ||
4558 (OP(first) == NOTHING && PL_regkind[OP(first_next)] != END ))
4561 * the only op that could be a regnode is PLUS, all the rest
4562 * will be regnode_1 or regnode_2.
4565 if (OP(first) == PLUS)
4568 first += regarglen[OP(first)];
4570 first = NEXTOPER(first);
4571 first_next= regnext(first);
4574 /* Starting-point info. */
4576 DEBUG_PEEP("first:",first,0);
4577 /* Ignore EXACT as we deal with it later. */
4578 if (PL_regkind[OP(first)] == EXACT) {
4579 if (OP(first) == EXACT)
4580 NOOP; /* Empty, get anchored substr later. */
4581 else if ((OP(first) == EXACTF || OP(first) == EXACTFL))
4582 ri->regstclass = first;
4585 else if (PL_regkind[OP(first)] == TRIE &&
4586 ((reg_trie_data *)ri->data->data[ ARG(first) ])->minlen>0)
4589 /* this can happen only on restudy */
4590 if ( OP(first) == TRIE ) {
4591 struct regnode_1 *trieop = (struct regnode_1 *)
4592 PerlMemShared_calloc(1, sizeof(struct regnode_1));
4593 StructCopy(first,trieop,struct regnode_1);
4594 trie_op=(regnode *)trieop;
4596 struct regnode_charclass *trieop = (struct regnode_charclass *)
4597 PerlMemShared_calloc(1, sizeof(struct regnode_charclass));
4598 StructCopy(first,trieop,struct regnode_charclass);
4599 trie_op=(regnode *)trieop;
4602 make_trie_failtable(pRExC_state, (regnode *)first, trie_op, 0);
4603 ri->regstclass = trie_op;
4606 else if (REGNODE_SIMPLE(OP(first)))
4607 ri->regstclass = first;
4608 else if (PL_regkind[OP(first)] == BOUND ||
4609 PL_regkind[OP(first)] == NBOUND)
4610 ri->regstclass = first;
4611 else if (PL_regkind[OP(first)] == BOL) {
4612 r->extflags |= (OP(first) == MBOL
4614 : (OP(first) == SBOL
4617 first = NEXTOPER(first);
4620 else if (OP(first) == GPOS) {
4621 r->extflags |= RXf_ANCH_GPOS;
4622 first = NEXTOPER(first);
4625 else if ((!sawopen || !RExC_sawback) &&
4626 (OP(first) == STAR &&
4627 PL_regkind[OP(NEXTOPER(first))] == REG_ANY) &&
4628 !(r->extflags & RXf_ANCH) && !(RExC_seen & REG_SEEN_EVAL))
4630 /* turn .* into ^.* with an implied $*=1 */
4632 (OP(NEXTOPER(first)) == REG_ANY)
4635 r->extflags |= type;
4636 r->intflags |= PREGf_IMPLICIT;
4637 first = NEXTOPER(first);
4640 if (sawplus && (!sawopen || !RExC_sawback)
4641 && !(RExC_seen & REG_SEEN_EVAL)) /* May examine pos and $& */
4642 /* x+ must match at the 1st pos of run of x's */
4643 r->intflags |= PREGf_SKIP;
4645 /* Scan is after the zeroth branch, first is atomic matcher. */
4646 #ifdef TRIE_STUDY_OPT
4649 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4650 (IV)(first - scan + 1))
4654 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4655 (IV)(first - scan + 1))
4661 * If there's something expensive in the r.e., find the
4662 * longest literal string that must appear and make it the
4663 * regmust. Resolve ties in favor of later strings, since
4664 * the regstart check works with the beginning of the r.e.
4665 * and avoiding duplication strengthens checking. Not a
4666 * strong reason, but sufficient in the absence of others.
4667 * [Now we resolve ties in favor of the earlier string if
4668 * it happens that c_offset_min has been invalidated, since the
4669 * earlier string may buy us something the later one won't.]
4672 data.longest_fixed = newSVpvs("");
4673 data.longest_float = newSVpvs("");
4674 data.last_found = newSVpvs("");
4675 data.longest = &(data.longest_fixed);
4677 if (!ri->regstclass) {
4678 cl_init(pRExC_state, &ch_class);
4679 data.start_class = &ch_class;
4680 stclass_flag = SCF_DO_STCLASS_AND;
4681 } else /* XXXX Check for BOUND? */
4683 data.last_closep = &last_close;
4685 minlen = study_chunk(pRExC_state, &first, &minlen, &fake, scan + RExC_size, /* Up to end */
4686 &data, -1, NULL, NULL,
4687 SCF_DO_SUBSTR | SCF_WHILEM_VISITED_POS | stclass_flag,0);
4693 if ( RExC_npar == 1 && data.longest == &(data.longest_fixed)
4694 && data.last_start_min == 0 && data.last_end > 0
4695 && !RExC_seen_zerolen
4696 && !(RExC_seen & REG_SEEN_VERBARG)
4697 && (!(RExC_seen & REG_SEEN_GPOS) || (r->extflags & RXf_ANCH_GPOS)))
4698 r->extflags |= RXf_CHECK_ALL;
4699 scan_commit(pRExC_state, &data,&minlen,0);
4700 SvREFCNT_dec(data.last_found);
4702 /* Note that code very similar to this but for anchored string
4703 follows immediately below, changes may need to be made to both.
4706 longest_float_length = CHR_SVLEN(data.longest_float);
4707 if (longest_float_length
4708 || (data.flags & SF_FL_BEFORE_EOL
4709 && (!(data.flags & SF_FL_BEFORE_MEOL)
4710 || (RExC_flags & RXf_PMf_MULTILINE))))
4714 if (SvCUR(data.longest_fixed) /* ok to leave SvCUR */
4715 && data.offset_fixed == data.offset_float_min
4716 && SvCUR(data.longest_fixed) == SvCUR(data.longest_float))
4717 goto remove_float; /* As in (a)+. */
4719 /* copy the information about the longest float from the reg_scan_data
4720 over to the program. */
4721 if (SvUTF8(data.longest_float)) {
4722 r->float_utf8 = data.longest_float;
4723 r->float_substr = NULL;
4725 r->float_substr = data.longest_float;
4726 r->float_utf8 = NULL;
4728 /* float_end_shift is how many chars that must be matched that
4729 follow this item. We calculate it ahead of time as once the
4730 lookbehind offset is added in we lose the ability to correctly
4732 ml = data.minlen_float ? *(data.minlen_float)
4733 : (I32)longest_float_length;
4734 r->float_end_shift = ml - data.offset_float_min
4735 - longest_float_length + (SvTAIL(data.longest_float) != 0)
4736 + data.lookbehind_float;
4737 r->float_min_offset = data.offset_float_min - data.lookbehind_float;
4738 r->float_max_offset = data.offset_float_max;
4739 if (data.offset_float_max < I32_MAX) /* Don't offset infinity */
4740 r->float_max_offset -= data.lookbehind_float;
4742 t = (data.flags & SF_FL_BEFORE_EOL /* Can't have SEOL and MULTI */
4743 && (!(data.flags & SF_FL_BEFORE_MEOL)
4744 || (RExC_flags & RXf_PMf_MULTILINE)));
4745 fbm_compile(data.longest_float, t ? FBMcf_TAIL : 0);
4749 r->float_substr = r->float_utf8 = NULL;
4750 SvREFCNT_dec(data.longest_float);
4751 longest_float_length = 0;
4754 /* Note that code very similar to this but for floating string
4755 is immediately above, changes may need to be made to both.
4758 longest_fixed_length = CHR_SVLEN(data.longest_fixed);
4759 if (longest_fixed_length
4760 || (data.flags & SF_FIX_BEFORE_EOL /* Cannot have SEOL and MULTI */
4761 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4762 || (RExC_flags & RXf_PMf_MULTILINE))))
4766 /* copy the information about the longest fixed
4767 from the reg_scan_data over to the program. */
4768 if (SvUTF8(data.longest_fixed)) {
4769 r->anchored_utf8 = data.longest_fixed;
4770 r->anchored_substr = NULL;
4772 r->anchored_substr = data.longest_fixed;
4773 r->anchored_utf8 = NULL;
4775 /* fixed_end_shift is how many chars that must be matched that
4776 follow this item. We calculate it ahead of time as once the
4777 lookbehind offset is added in we lose the ability to correctly
4779 ml = data.minlen_fixed ? *(data.minlen_fixed)
4780 : (I32)longest_fixed_length;
4781 r->anchored_end_shift = ml - data.offset_fixed
4782 - longest_fixed_length + (SvTAIL(data.longest_fixed) != 0)
4783 + data.lookbehind_fixed;
4784 r->anchored_offset = data.offset_fixed - data.lookbehind_fixed;
4786 t = (data.flags & SF_FIX_BEFORE_EOL /* Can't have SEOL and MULTI */
4787 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4788 || (RExC_flags & RXf_PMf_MULTILINE)));
4789 fbm_compile(data.longest_fixed, t ? FBMcf_TAIL : 0);
4792 r->anchored_substr = r->anchored_utf8 = NULL;
4793 SvREFCNT_dec(data.longest_fixed);
4794 longest_fixed_length = 0;
4797 && (OP(ri->regstclass) == REG_ANY || OP(ri->regstclass) == SANY))
4798 ri->regstclass = NULL;
4799 if ((!(r->anchored_substr || r->anchored_utf8) || r->anchored_offset)
4801 && !(data.start_class->flags & ANYOF_EOS)
4802 && !cl_is_anything(data.start_class))
4804 const U32 n = add_data(pRExC_state, 1, "f");
4806 Newx(RExC_rxi->data->data[n], 1,
4807 struct regnode_charclass_class);
4808 StructCopy(data.start_class,
4809 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4810 struct regnode_charclass_class);
4811 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4812 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4813 DEBUG_COMPILE_r({ SV *sv = sv_newmortal();
4814 regprop(r, sv, (regnode*)data.start_class);
4815 PerlIO_printf(Perl_debug_log,
4816 "synthetic stclass \"%s\".\n",
4817 SvPVX_const(sv));});
4820 /* A temporary algorithm prefers floated substr to fixed one to dig more info. */
4821 if (longest_fixed_length > longest_float_length) {
4822 r->check_end_shift = r->anchored_end_shift;
4823 r->check_substr = r->anchored_substr;
4824 r->check_utf8 = r->anchored_utf8;
4825 r->check_offset_min = r->check_offset_max = r->anchored_offset;
4826 if (r->extflags & RXf_ANCH_SINGLE)
4827 r->extflags |= RXf_NOSCAN;
4830 r->check_end_shift = r->float_end_shift;
4831 r->check_substr = r->float_substr;
4832 r->check_utf8 = r->float_utf8;
4833 r->check_offset_min = r->float_min_offset;
4834 r->check_offset_max = r->float_max_offset;
4836 /* XXXX Currently intuiting is not compatible with ANCH_GPOS.
4837 This should be changed ASAP! */
4838 if ((r->check_substr || r->check_utf8) && !(r->extflags & RXf_ANCH_GPOS)) {
4839 r->extflags |= RXf_USE_INTUIT;
4840 if (SvTAIL(r->check_substr ? r->check_substr : r->check_utf8))
4841 r->extflags |= RXf_INTUIT_TAIL;
4843 /* XXX Unneeded? dmq (shouldn't as this is handled elsewhere)
4844 if ( (STRLEN)minlen < longest_float_length )
4845 minlen= longest_float_length;
4846 if ( (STRLEN)minlen < longest_fixed_length )
4847 minlen= longest_fixed_length;
4851 /* Several toplevels. Best we can is to set minlen. */
4853 struct regnode_charclass_class ch_class;
4856 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "\nMulti Top Level\n"));
4858 scan = ri->program + 1;
4859 cl_init(pRExC_state, &ch_class);
4860 data.start_class = &ch_class;
4861 data.last_closep = &last_close;
4864 minlen = study_chunk(pRExC_state, &scan, &minlen, &fake, scan + RExC_size,
4865 &data, -1, NULL, NULL, SCF_DO_STCLASS_AND|SCF_WHILEM_VISITED_POS,0);
4869 r->check_substr = r->check_utf8 = r->anchored_substr = r->anchored_utf8
4870 = r->float_substr = r->float_utf8 = NULL;
4871 if (!(data.start_class->flags & ANYOF_EOS)
4872 && !cl_is_anything(data.start_class))
4874 const U32 n = add_data(pRExC_state, 1, "f");
4876 Newx(RExC_rxi->data->data[n], 1,
4877 struct regnode_charclass_class);
4878 StructCopy(data.start_class,
4879 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4880 struct regnode_charclass_class);
4881 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4882 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4883 DEBUG_COMPILE_r({ SV* sv = sv_newmortal();
4884 regprop(r, sv, (regnode*)data.start_class);
4885 PerlIO_printf(Perl_debug_log,
4886 "synthetic stclass \"%s\".\n",
4887 SvPVX_const(sv));});
4891 /* Guard against an embedded (?=) or (?<=) with a longer minlen than
4892 the "real" pattern. */
4894 PerlIO_printf(Perl_debug_log,"minlen: %"IVdf" r->minlen:%"IVdf"\n",
4895 (IV)minlen, (IV)r->minlen);
4897 r->minlenret = minlen;
4898 if (r->minlen < minlen)
4901 if (RExC_seen & REG_SEEN_GPOS)
4902 r->extflags |= RXf_GPOS_SEEN;
4903 if (RExC_seen & REG_SEEN_LOOKBEHIND)
4904 r->extflags |= RXf_LOOKBEHIND_SEEN;
4905 if (RExC_seen & REG_SEEN_EVAL)
4906 r->extflags |= RXf_EVAL_SEEN;
4907 if (RExC_seen & REG_SEEN_CANY)
4908 r->extflags |= RXf_CANY_SEEN;
4909 if (RExC_seen & REG_SEEN_VERBARG)
4910 r->intflags |= PREGf_VERBARG_SEEN;
4911 if (RExC_seen & REG_SEEN_CUTGROUP)
4912 r->intflags |= PREGf_CUTGROUP_SEEN;
4913 if (RExC_paren_names)
4914 RXp_PAREN_NAMES(r) = MUTABLE_HV(SvREFCNT_inc(RExC_paren_names));
4916 RXp_PAREN_NAMES(r) = NULL;
4918 #ifdef STUPID_PATTERN_CHECKS
4919 if (RX_PRELEN(rx) == 0)
4920 r->extflags |= RXf_NULL;
4921 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
4922 /* XXX: this should happen BEFORE we compile */
4923 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
4924 else if (RX_PRELEN(rx) == 3 && memEQ("\\s+", RX_PRECOMP(rx), 3))
4925 r->extflags |= RXf_WHITE;
4926 else if (RX_PRELEN(rx) == 1 && RXp_PRECOMP(rx)[0] == '^')
4927 r->extflags |= RXf_START_ONLY;
4929 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
4930 /* XXX: this should happen BEFORE we compile */
4931 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
4933 regnode *first = ri->program + 1;
4935 U8 nop = OP(NEXTOPER(first));
4937 if (PL_regkind[fop] == NOTHING && nop == END)
4938 r->extflags |= RXf_NULL;
4939 else if (PL_regkind[fop] == BOL && nop == END)
4940 r->extflags |= RXf_START_ONLY;
4941 else if (fop == PLUS && nop ==SPACE && OP(regnext(first))==END)
4942 r->extflags |= RXf_WHITE;
4946 if (RExC_paren_names) {
4947 ri->name_list_idx = add_data( pRExC_state, 1, "a" );
4948 ri->data->data[ri->name_list_idx] = (void*)SvREFCNT_inc(RExC_paren_name_list);
4951 ri->name_list_idx = 0;
4953 if (RExC_recurse_count) {
4954 for ( ; RExC_recurse_count ; RExC_recurse_count-- ) {
4955 const regnode *scan = RExC_recurse[RExC_recurse_count-1];
4956 ARG2L_SET( scan, RExC_open_parens[ARG(scan)-1] - scan );
4959 Newxz(r->offs, RExC_npar, regexp_paren_pair);
4960 /* assume we don't need to swap parens around before we match */
4963 PerlIO_printf(Perl_debug_log,"Final program:\n");
4966 #ifdef RE_TRACK_PATTERN_OFFSETS
4967 DEBUG_OFFSETS_r(if (ri->u.offsets) {
4968 const U32 len = ri->u.offsets[0];
4970 GET_RE_DEBUG_FLAGS_DECL;
4971 PerlIO_printf(Perl_debug_log, "Offsets: [%"UVuf"]\n\t", (UV)ri->u.offsets[0]);
4972 for (i = 1; i <= len; i++) {
4973 if (ri->u.offsets[i*2-1] || ri->u.offsets[i*2])
4974 PerlIO_printf(Perl_debug_log, "%"UVuf":%"UVuf"[%"UVuf"] ",
4975 (UV)i, (UV)ri->u.offsets[i*2-1], (UV)ri->u.offsets[i*2]);
4977 PerlIO_printf(Perl_debug_log, "\n");
4983 #undef RE_ENGINE_PTR
4987 Perl_reg_named_buff(pTHX_ REGEXP * const rx, SV * const key, SV * const value,
4990 PERL_ARGS_ASSERT_REG_NAMED_BUFF;
4992 PERL_UNUSED_ARG(value);
4994 if (flags & RXapif_FETCH) {
4995 return reg_named_buff_fetch(rx, key, flags);
4996 } else if (flags & (RXapif_STORE | RXapif_DELETE | RXapif_CLEAR)) {
4997 Perl_croak_no_modify(aTHX);
4999 } else if (flags & RXapif_EXISTS) {
5000 return reg_named_buff_exists(rx, key, flags)
5003 } else if (flags & RXapif_REGNAMES) {
5004 return reg_named_buff_all(rx, flags);
5005 } else if (flags & (RXapif_SCALAR | RXapif_REGNAMES_COUNT)) {
5006 return reg_named_buff_scalar(rx, flags);
5008 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff", (int)flags);
5014 Perl_reg_named_buff_iter(pTHX_ REGEXP * const rx, const SV * const lastkey,
5017 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ITER;
5018 PERL_UNUSED_ARG(lastkey);
5020 if (flags & RXapif_FIRSTKEY)
5021 return reg_named_buff_firstkey(rx, flags);
5022 else if (flags & RXapif_NEXTKEY)
5023 return reg_named_buff_nextkey(rx, flags);
5025 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_iter", (int)flags);
5031 Perl_reg_named_buff_fetch(pTHX_ REGEXP * const r, SV * const namesv,
5034 AV *retarray = NULL;
5036 struct regexp *const rx = (struct regexp *)SvANY(r);
5038 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FETCH;
5040 if (flags & RXapif_ALL)
5043 if (rx && RXp_PAREN_NAMES(rx)) {
5044 HE *he_str = hv_fetch_ent( RXp_PAREN_NAMES(rx), namesv, 0, 0 );
5047 SV* sv_dat=HeVAL(he_str);
5048 I32 *nums=(I32*)SvPVX(sv_dat);
5049 for ( i=0; i<SvIVX(sv_dat); i++ ) {
5050 if ((I32)(rx->nparens) >= nums[i]
5051 && rx->offs[nums[i]].start != -1
5052 && rx->offs[nums[i]].end != -1)
5055 CALLREG_NUMBUF_FETCH(r,nums[i],ret);
5059 ret = newSVsv(&PL_sv_undef);
5062 av_push(retarray, ret);
5065 return newRV_noinc(MUTABLE_SV(retarray));
5072 Perl_reg_named_buff_exists(pTHX_ REGEXP * const r, SV * const key,
5075 struct regexp *const rx = (struct regexp *)SvANY(r);
5077 PERL_ARGS_ASSERT_REG_NAMED_BUFF_EXISTS;
5079 if (rx && RXp_PAREN_NAMES(rx)) {
5080 if (flags & RXapif_ALL) {
5081 return hv_exists_ent(RXp_PAREN_NAMES(rx), key, 0);
5083 SV *sv = CALLREG_NAMED_BUFF_FETCH(r, key, flags);
5097 Perl_reg_named_buff_firstkey(pTHX_ REGEXP * const r, const U32 flags)
5099 struct regexp *const rx = (struct regexp *)SvANY(r);
5101 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FIRSTKEY;
5103 if ( rx && RXp_PAREN_NAMES(rx) ) {
5104 (void)hv_iterinit(RXp_PAREN_NAMES(rx));
5106 return CALLREG_NAMED_BUFF_NEXTKEY(r, NULL, flags & ~RXapif_FIRSTKEY);
5113 Perl_reg_named_buff_nextkey(pTHX_ REGEXP * const r, const U32 flags)
5115 struct regexp *const rx = (struct regexp *)SvANY(r);
5116 GET_RE_DEBUG_FLAGS_DECL;
5118 PERL_ARGS_ASSERT_REG_NAMED_BUFF_NEXTKEY;
5120 if (rx && RXp_PAREN_NAMES(rx)) {
5121 HV *hv = RXp_PAREN_NAMES(rx);
5123 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5126 SV* sv_dat = HeVAL(temphe);
5127 I32 *nums = (I32*)SvPVX(sv_dat);
5128 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5129 if ((I32)(rx->lastparen) >= nums[i] &&
5130 rx->offs[nums[i]].start != -1 &&
5131 rx->offs[nums[i]].end != -1)
5137 if (parno || flags & RXapif_ALL) {
5138 return newSVhek(HeKEY_hek(temphe));
5146 Perl_reg_named_buff_scalar(pTHX_ REGEXP * const r, const U32 flags)
5151 struct regexp *const rx = (struct regexp *)SvANY(r);
5153 PERL_ARGS_ASSERT_REG_NAMED_BUFF_SCALAR;
5155 if (rx && RXp_PAREN_NAMES(rx)) {
5156 if (flags & (RXapif_ALL | RXapif_REGNAMES_COUNT)) {
5157 return newSViv(HvTOTALKEYS(RXp_PAREN_NAMES(rx)));
5158 } else if (flags & RXapif_ONE) {
5159 ret = CALLREG_NAMED_BUFF_ALL(r, (flags | RXapif_REGNAMES));
5160 av = MUTABLE_AV(SvRV(ret));
5161 length = av_len(av);
5163 return newSViv(length + 1);
5165 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_scalar", (int)flags);
5169 return &PL_sv_undef;
5173 Perl_reg_named_buff_all(pTHX_ REGEXP * const r, const U32 flags)
5175 struct regexp *const rx = (struct regexp *)SvANY(r);
5178 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ALL;
5180 if (rx && RXp_PAREN_NAMES(rx)) {
5181 HV *hv= RXp_PAREN_NAMES(rx);
5183 (void)hv_iterinit(hv);
5184 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5187 SV* sv_dat = HeVAL(temphe);
5188 I32 *nums = (I32*)SvPVX(sv_dat);
5189 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5190 if ((I32)(rx->lastparen) >= nums[i] &&
5191 rx->offs[nums[i]].start != -1 &&
5192 rx->offs[nums[i]].end != -1)
5198 if (parno || flags & RXapif_ALL) {
5199 av_push(av, newSVhek(HeKEY_hek(temphe)));
5204 return newRV_noinc(MUTABLE_SV(av));
5208 Perl_reg_numbered_buff_fetch(pTHX_ REGEXP * const r, const I32 paren,
5211 struct regexp *const rx = (struct regexp *)SvANY(r);
5216 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_FETCH;
5219 sv_setsv(sv,&PL_sv_undef);
5223 if (paren == RX_BUFF_IDX_PREMATCH && rx->offs[0].start != -1) {
5225 i = rx->offs[0].start;
5229 if (paren == RX_BUFF_IDX_POSTMATCH && rx->offs[0].end != -1) {
5231 s = rx->subbeg + rx->offs[0].end;
5232 i = rx->sublen - rx->offs[0].end;
5235 if ( 0 <= paren && paren <= (I32)rx->nparens &&
5236 (s1 = rx->offs[paren].start) != -1 &&
5237 (t1 = rx->offs[paren].end) != -1)
5241 s = rx->subbeg + s1;
5243 sv_setsv(sv,&PL_sv_undef);
5246 assert(rx->sublen >= (s - rx->subbeg) + i );
5248 const int oldtainted = PL_tainted;
5250 sv_setpvn(sv, s, i);
5251 PL_tainted = oldtainted;
5252 if ( (rx->extflags & RXf_CANY_SEEN)
5253 ? (RXp_MATCH_UTF8(rx)
5254 && (!i || is_utf8_string((U8*)s, i)))
5255 : (RXp_MATCH_UTF8(rx)) )
5262 if (RXp_MATCH_TAINTED(rx)) {
5263 if (SvTYPE(sv) >= SVt_PVMG) {
5264 MAGIC* const mg = SvMAGIC(sv);
5267 SvMAGIC_set(sv, mg->mg_moremagic);
5269 if ((mgt = SvMAGIC(sv))) {
5270 mg->mg_moremagic = mgt;
5271 SvMAGIC_set(sv, mg);
5281 sv_setsv(sv,&PL_sv_undef);
5287 Perl_reg_numbered_buff_store(pTHX_ REGEXP * const rx, const I32 paren,
5288 SV const * const value)
5290 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_STORE;
5292 PERL_UNUSED_ARG(rx);
5293 PERL_UNUSED_ARG(paren);
5294 PERL_UNUSED_ARG(value);
5297 Perl_croak_no_modify(aTHX);
5301 Perl_reg_numbered_buff_length(pTHX_ REGEXP * const r, const SV * const sv,
5304 struct regexp *const rx = (struct regexp *)SvANY(r);
5308 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_LENGTH;
5310 /* Some of this code was originally in C<Perl_magic_len> in F<mg.c> */
5312 /* $` / ${^PREMATCH} */
5313 case RX_BUFF_IDX_PREMATCH:
5314 if (rx->offs[0].start != -1) {
5315 i = rx->offs[0].start;
5323 /* $' / ${^POSTMATCH} */
5324 case RX_BUFF_IDX_POSTMATCH:
5325 if (rx->offs[0].end != -1) {
5326 i = rx->sublen - rx->offs[0].end;
5328 s1 = rx->offs[0].end;
5334 /* $& / ${^MATCH}, $1, $2, ... */
5336 if (paren <= (I32)rx->nparens &&
5337 (s1 = rx->offs[paren].start) != -1 &&
5338 (t1 = rx->offs[paren].end) != -1)
5343 if (ckWARN(WARN_UNINITIALIZED))
5344 report_uninit((const SV *)sv);
5349 if (i > 0 && RXp_MATCH_UTF8(rx)) {
5350 const char * const s = rx->subbeg + s1;
5355 if (is_utf8_string_loclen((U8*)s, i, &ep, &el))
5362 Perl_reg_qr_package(pTHX_ REGEXP * const rx)
5364 PERL_ARGS_ASSERT_REG_QR_PACKAGE;
5365 PERL_UNUSED_ARG(rx);
5369 return newSVpvs("Regexp");
5372 /* Scans the name of a named buffer from the pattern.
5373 * If flags is REG_RSN_RETURN_NULL returns null.
5374 * If flags is REG_RSN_RETURN_NAME returns an SV* containing the name
5375 * If flags is REG_RSN_RETURN_DATA returns the data SV* corresponding
5376 * to the parsed name as looked up in the RExC_paren_names hash.
5377 * If there is an error throws a vFAIL().. type exception.
5380 #define REG_RSN_RETURN_NULL 0
5381 #define REG_RSN_RETURN_NAME 1
5382 #define REG_RSN_RETURN_DATA 2
5385 S_reg_scan_name(pTHX_ RExC_state_t *pRExC_state, U32 flags)
5387 char *name_start = RExC_parse;
5389 PERL_ARGS_ASSERT_REG_SCAN_NAME;
5391 if (isIDFIRST_lazy_if(RExC_parse, UTF)) {
5392 /* skip IDFIRST by using do...while */
5395 RExC_parse += UTF8SKIP(RExC_parse);
5396 } while (isALNUM_utf8((U8*)RExC_parse));
5400 } while (isALNUM(*RExC_parse));
5405 = newSVpvn_flags(name_start, (int)(RExC_parse - name_start),
5406 SVs_TEMP | (UTF ? SVf_UTF8 : 0));
5407 if ( flags == REG_RSN_RETURN_NAME)
5409 else if (flags==REG_RSN_RETURN_DATA) {
5412 if ( ! sv_name ) /* should not happen*/
5413 Perl_croak(aTHX_ "panic: no svname in reg_scan_name");
5414 if (RExC_paren_names)
5415 he_str = hv_fetch_ent( RExC_paren_names, sv_name, 0, 0 );
5417 sv_dat = HeVAL(he_str);
5419 vFAIL("Reference to nonexistent named group");
5423 Perl_croak(aTHX_ "panic: bad flag in reg_scan_name");
5430 #define DEBUG_PARSE_MSG(funcname) DEBUG_PARSE_r({ \
5431 int rem=(int)(RExC_end - RExC_parse); \
5440 if (RExC_lastparse!=RExC_parse) \
5441 PerlIO_printf(Perl_debug_log," >%.*s%-*s", \
5444 iscut ? "..." : "<" \
5447 PerlIO_printf(Perl_debug_log,"%16s",""); \
5450 num = RExC_size + 1; \
5452 num=REG_NODE_NUM(RExC_emit); \
5453 if (RExC_lastnum!=num) \
5454 PerlIO_printf(Perl_debug_log,"|%4d",num); \
5456 PerlIO_printf(Perl_debug_log,"|%4s",""); \
5457 PerlIO_printf(Perl_debug_log,"|%*s%-4s", \
5458 (int)((depth*2)), "", \
5462 RExC_lastparse=RExC_parse; \
5467 #define DEBUG_PARSE(funcname) DEBUG_PARSE_r({ \
5468 DEBUG_PARSE_MSG((funcname)); \
5469 PerlIO_printf(Perl_debug_log,"%4s","\n"); \
5471 #define DEBUG_PARSE_FMT(funcname,fmt,args) DEBUG_PARSE_r({ \
5472 DEBUG_PARSE_MSG((funcname)); \
5473 PerlIO_printf(Perl_debug_log,fmt "\n",args); \
5476 - reg - regular expression, i.e. main body or parenthesized thing
5478 * Caller must absorb opening parenthesis.
5480 * Combining parenthesis handling with the base level of regular expression
5481 * is a trifle forced, but the need to tie the tails of the branches to what
5482 * follows makes it hard to avoid.
5484 #define REGTAIL(x,y,z) regtail((x),(y),(z),depth+1)
5486 #define REGTAIL_STUDY(x,y,z) regtail_study((x),(y),(z),depth+1)
5488 #define REGTAIL_STUDY(x,y,z) regtail((x),(y),(z),depth+1)
5492 S_reg(pTHX_ RExC_state_t *pRExC_state, I32 paren, I32 *flagp,U32 depth)
5493 /* paren: Parenthesized? 0=top, 1=(, inside: changed to letter. */
5496 register regnode *ret; /* Will be the head of the group. */
5497 register regnode *br;
5498 register regnode *lastbr;
5499 register regnode *ender = NULL;
5500 register I32 parno = 0;
5502 U32 oregflags = RExC_flags;
5503 bool have_branch = 0;
5505 I32 freeze_paren = 0;
5506 I32 after_freeze = 0;
5508 /* for (?g), (?gc), and (?o) warnings; warning
5509 about (?c) will warn about (?g) -- japhy */
5511 #define WASTED_O 0x01
5512 #define WASTED_G 0x02
5513 #define WASTED_C 0x04
5514 #define WASTED_GC (0x02|0x04)
5515 I32 wastedflags = 0x00;
5517 char * parse_start = RExC_parse; /* MJD */
5518 char * const oregcomp_parse = RExC_parse;
5520 GET_RE_DEBUG_FLAGS_DECL;
5522 PERL_ARGS_ASSERT_REG;
5523 DEBUG_PARSE("reg ");
5525 *flagp = 0; /* Tentatively. */
5528 /* Make an OPEN node, if parenthesized. */
5530 if ( *RExC_parse == '*') { /* (*VERB:ARG) */
5531 char *start_verb = RExC_parse;
5532 STRLEN verb_len = 0;
5533 char *start_arg = NULL;
5534 unsigned char op = 0;
5536 int internal_argval = 0; /* internal_argval is only useful if !argok */
5537 while ( *RExC_parse && *RExC_parse != ')' ) {
5538 if ( *RExC_parse == ':' ) {
5539 start_arg = RExC_parse + 1;
5545 verb_len = RExC_parse - start_verb;
5548 while ( *RExC_parse && *RExC_parse != ')' )
5550 if ( *RExC_parse != ')' )
5551 vFAIL("Unterminated verb pattern argument");
5552 if ( RExC_parse == start_arg )
5555 if ( *RExC_parse != ')' )
5556 vFAIL("Unterminated verb pattern");
5559 switch ( *start_verb ) {
5560 case 'A': /* (*ACCEPT) */
5561 if ( memEQs(start_verb,verb_len,"ACCEPT") ) {
5563 internal_argval = RExC_nestroot;
5566 case 'C': /* (*COMMIT) */
5567 if ( memEQs(start_verb,verb_len,"COMMIT") )
5570 case 'F': /* (*FAIL) */
5571 if ( verb_len==1 || memEQs(start_verb,verb_len,"FAIL") ) {
5576 case ':': /* (*:NAME) */
5577 case 'M': /* (*MARK:NAME) */
5578 if ( verb_len==0 || memEQs(start_verb,verb_len,"MARK") ) {
5583 case 'P': /* (*PRUNE) */
5584 if ( memEQs(start_verb,verb_len,"PRUNE") )
5587 case 'S': /* (*SKIP) */
5588 if ( memEQs(start_verb,verb_len,"SKIP") )
5591 case 'T': /* (*THEN) */
5592 /* [19:06] <TimToady> :: is then */
5593 if ( memEQs(start_verb,verb_len,"THEN") ) {
5595 RExC_seen |= REG_SEEN_CUTGROUP;
5601 vFAIL3("Unknown verb pattern '%.*s'",
5602 verb_len, start_verb);
5605 if ( start_arg && internal_argval ) {
5606 vFAIL3("Verb pattern '%.*s' may not have an argument",
5607 verb_len, start_verb);
5608 } else if ( argok < 0 && !start_arg ) {
5609 vFAIL3("Verb pattern '%.*s' has a mandatory argument",
5610 verb_len, start_verb);
5612 ret = reganode(pRExC_state, op, internal_argval);
5613 if ( ! internal_argval && ! SIZE_ONLY ) {
5615 SV *sv = newSVpvn( start_arg, RExC_parse - start_arg);
5616 ARG(ret) = add_data( pRExC_state, 1, "S" );
5617 RExC_rxi->data->data[ARG(ret)]=(void*)sv;
5624 if (!internal_argval)
5625 RExC_seen |= REG_SEEN_VERBARG;
5626 } else if ( start_arg ) {
5627 vFAIL3("Verb pattern '%.*s' may not have an argument",
5628 verb_len, start_verb);
5630 ret = reg_node(pRExC_state, op);
5632 nextchar(pRExC_state);
5635 if (*RExC_parse == '?') { /* (?...) */
5636 bool is_logical = 0;
5637 const char * const seqstart = RExC_parse;
5640 paren = *RExC_parse++;
5641 ret = NULL; /* For look-ahead/behind. */
5644 case 'P': /* (?P...) variants for those used to PCRE/Python */
5645 paren = *RExC_parse++;
5646 if ( paren == '<') /* (?P<...>) named capture */
5648 else if (paren == '>') { /* (?P>name) named recursion */
5649 goto named_recursion;
5651 else if (paren == '=') { /* (?P=...) named backref */
5652 /* this pretty much dupes the code for \k<NAME> in regatom(), if
5653 you change this make sure you change that */
5654 char* name_start = RExC_parse;
5656 SV *sv_dat = reg_scan_name(pRExC_state,
5657 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5658 if (RExC_parse == name_start || *RExC_parse != ')')
5659 vFAIL2("Sequence %.3s... not terminated",parse_start);
5662 num = add_data( pRExC_state, 1, "S" );
5663 RExC_rxi->data->data[num]=(void*)sv_dat;
5664 SvREFCNT_inc_simple_void(sv_dat);
5667 ret = reganode(pRExC_state,
5668 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
5672 Set_Node_Offset(ret, parse_start+1);
5673 Set_Node_Cur_Length(ret); /* MJD */
5675 nextchar(pRExC_state);
5679 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5681 case '<': /* (?<...) */
5682 if (*RExC_parse == '!')
5684 else if (*RExC_parse != '=')
5690 case '\'': /* (?'...') */
5691 name_start= RExC_parse;
5692 svname = reg_scan_name(pRExC_state,
5693 SIZE_ONLY ? /* reverse test from the others */
5694 REG_RSN_RETURN_NAME :
5695 REG_RSN_RETURN_NULL);
5696 if (RExC_parse == name_start) {
5698 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5701 if (*RExC_parse != paren)
5702 vFAIL2("Sequence (?%c... not terminated",
5703 paren=='>' ? '<' : paren);
5707 if (!svname) /* shouldnt happen */
5709 "panic: reg_scan_name returned NULL");
5710 if (!RExC_paren_names) {
5711 RExC_paren_names= newHV();
5712 sv_2mortal(MUTABLE_SV(RExC_paren_names));
5714 RExC_paren_name_list= newAV();
5715 sv_2mortal(MUTABLE_SV(RExC_paren_name_list));
5718 he_str = hv_fetch_ent( RExC_paren_names, svname, 1, 0 );
5720 sv_dat = HeVAL(he_str);
5722 /* croak baby croak */
5724 "panic: paren_name hash element allocation failed");
5725 } else if ( SvPOK(sv_dat) ) {
5726 /* (?|...) can mean we have dupes so scan to check
5727 its already been stored. Maybe a flag indicating
5728 we are inside such a construct would be useful,
5729 but the arrays are likely to be quite small, so
5730 for now we punt -- dmq */
5731 IV count = SvIV(sv_dat);
5732 I32 *pv = (I32*)SvPVX(sv_dat);
5734 for ( i = 0 ; i < count ; i++ ) {
5735 if ( pv[i] == RExC_npar ) {
5741 pv = (I32*)SvGROW(sv_dat, SvCUR(sv_dat) + sizeof(I32)+1);
5742 SvCUR_set(sv_dat, SvCUR(sv_dat) + sizeof(I32));
5743 pv[count] = RExC_npar;
5744 SvIV_set(sv_dat, SvIVX(sv_dat) + 1);
5747 (void)SvUPGRADE(sv_dat,SVt_PVNV);
5748 sv_setpvn(sv_dat, (char *)&(RExC_npar), sizeof(I32));
5750 SvIV_set(sv_dat, 1);
5753 if (!av_store(RExC_paren_name_list, RExC_npar, SvREFCNT_inc(svname)))
5754 SvREFCNT_dec(svname);
5757 /*sv_dump(sv_dat);*/
5759 nextchar(pRExC_state);
5761 goto capturing_parens;
5763 RExC_seen |= REG_SEEN_LOOKBEHIND;
5765 case '=': /* (?=...) */
5766 RExC_seen_zerolen++;
5768 case '!': /* (?!...) */
5769 RExC_seen_zerolen++;
5770 if (*RExC_parse == ')') {
5771 ret=reg_node(pRExC_state, OPFAIL);
5772 nextchar(pRExC_state);
5776 case '|': /* (?|...) */
5777 /* branch reset, behave like a (?:...) except that
5778 buffers in alternations share the same numbers */
5780 after_freeze = freeze_paren = RExC_npar;
5782 case ':': /* (?:...) */
5783 case '>': /* (?>...) */
5785 case '$': /* (?$...) */
5786 case '@': /* (?@...) */
5787 vFAIL2("Sequence (?%c...) not implemented", (int)paren);
5789 case '#': /* (?#...) */
5790 while (*RExC_parse && *RExC_parse != ')')
5792 if (*RExC_parse != ')')
5793 FAIL("Sequence (?#... not terminated");
5794 nextchar(pRExC_state);
5797 case '0' : /* (?0) */
5798 case 'R' : /* (?R) */
5799 if (*RExC_parse != ')')
5800 FAIL("Sequence (?R) not terminated");
5801 ret = reg_node(pRExC_state, GOSTART);
5802 *flagp |= POSTPONED;
5803 nextchar(pRExC_state);
5806 { /* named and numeric backreferences */
5808 case '&': /* (?&NAME) */
5809 parse_start = RExC_parse - 1;
5812 SV *sv_dat = reg_scan_name(pRExC_state,
5813 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5814 num = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5816 goto gen_recurse_regop;
5819 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5821 vFAIL("Illegal pattern");
5823 goto parse_recursion;
5825 case '-': /* (?-1) */
5826 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5827 RExC_parse--; /* rewind to let it be handled later */
5831 case '1': case '2': case '3': case '4': /* (?1) */
5832 case '5': case '6': case '7': case '8': case '9':
5835 num = atoi(RExC_parse);
5836 parse_start = RExC_parse - 1; /* MJD */
5837 if (*RExC_parse == '-')
5839 while (isDIGIT(*RExC_parse))
5841 if (*RExC_parse!=')')
5842 vFAIL("Expecting close bracket");
5845 if ( paren == '-' ) {
5847 Diagram of capture buffer numbering.
5848 Top line is the normal capture buffer numbers
5849 Botton line is the negative indexing as from
5853 /(a(x)y)(a(b(c(?-2)d)e)f)(g(h))/
5857 num = RExC_npar + num;
5860 vFAIL("Reference to nonexistent group");
5862 } else if ( paren == '+' ) {
5863 num = RExC_npar + num - 1;
5866 ret = reganode(pRExC_state, GOSUB, num);
5868 if (num > (I32)RExC_rx->nparens) {
5870 vFAIL("Reference to nonexistent group");
5872 ARG2L_SET( ret, RExC_recurse_count++);
5874 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
5875 "Recurse #%"UVuf" to %"IVdf"\n", (UV)ARG(ret), (IV)ARG2L(ret)));
5879 RExC_seen |= REG_SEEN_RECURSE;
5880 Set_Node_Length(ret, 1 + regarglen[OP(ret)]); /* MJD */
5881 Set_Node_Offset(ret, parse_start); /* MJD */
5883 *flagp |= POSTPONED;
5884 nextchar(pRExC_state);
5886 } /* named and numeric backreferences */
5889 case '?': /* (??...) */
5891 if (*RExC_parse != '{') {
5893 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5896 *flagp |= POSTPONED;
5897 paren = *RExC_parse++;
5899 case '{': /* (?{...}) */
5904 char *s = RExC_parse;
5906 RExC_seen_zerolen++;
5907 RExC_seen |= REG_SEEN_EVAL;
5908 while (count && (c = *RExC_parse)) {
5919 if (*RExC_parse != ')') {
5921 vFAIL("Sequence (?{...}) not terminated or not {}-balanced");
5925 OP_4tree *sop, *rop;
5926 SV * const sv = newSVpvn(s, RExC_parse - 1 - s);
5929 Perl_save_re_context(aTHX);
5930 rop = sv_compile_2op(sv, &sop, "re", &pad);
5931 sop->op_private |= OPpREFCOUNTED;
5932 /* re_dup will OpREFCNT_inc */
5933 OpREFCNT_set(sop, 1);
5936 n = add_data(pRExC_state, 3, "nop");
5937 RExC_rxi->data->data[n] = (void*)rop;
5938 RExC_rxi->data->data[n+1] = (void*)sop;
5939 RExC_rxi->data->data[n+2] = (void*)pad;
5942 else { /* First pass */
5943 if (PL_reginterp_cnt < ++RExC_seen_evals
5945 /* No compiled RE interpolated, has runtime
5946 components ===> unsafe. */
5947 FAIL("Eval-group not allowed at runtime, use re 'eval'");
5948 if (PL_tainting && PL_tainted)
5949 FAIL("Eval-group in insecure regular expression");
5950 #if PERL_VERSION > 8
5951 if (IN_PERL_COMPILETIME)
5956 nextchar(pRExC_state);
5958 ret = reg_node(pRExC_state, LOGICAL);
5961 REGTAIL(pRExC_state, ret, reganode(pRExC_state, EVAL, n));
5962 /* deal with the length of this later - MJD */
5965 ret = reganode(pRExC_state, EVAL, n);
5966 Set_Node_Length(ret, RExC_parse - parse_start + 1);
5967 Set_Node_Offset(ret, parse_start);
5970 case '(': /* (?(?{...})...) and (?(?=...)...) */
5973 if (RExC_parse[0] == '?') { /* (?(?...)) */
5974 if (RExC_parse[1] == '=' || RExC_parse[1] == '!'
5975 || RExC_parse[1] == '<'
5976 || RExC_parse[1] == '{') { /* Lookahead or eval. */
5979 ret = reg_node(pRExC_state, LOGICAL);
5982 REGTAIL(pRExC_state, ret, reg(pRExC_state, 1, &flag,depth+1));
5986 else if ( RExC_parse[0] == '<' /* (?(<NAME>)...) */
5987 || RExC_parse[0] == '\'' ) /* (?('NAME')...) */
5989 char ch = RExC_parse[0] == '<' ? '>' : '\'';
5990 char *name_start= RExC_parse++;
5992 SV *sv_dat=reg_scan_name(pRExC_state,
5993 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5994 if (RExC_parse == name_start || *RExC_parse != ch)
5995 vFAIL2("Sequence (?(%c... not terminated",
5996 (ch == '>' ? '<' : ch));
5999 num = add_data( pRExC_state, 1, "S" );
6000 RExC_rxi->data->data[num]=(void*)sv_dat;
6001 SvREFCNT_inc_simple_void(sv_dat);
6003 ret = reganode(pRExC_state,NGROUPP,num);
6004 goto insert_if_check_paren;
6006 else if (RExC_parse[0] == 'D' &&
6007 RExC_parse[1] == 'E' &&
6008 RExC_parse[2] == 'F' &&
6009 RExC_parse[3] == 'I' &&
6010 RExC_parse[4] == 'N' &&
6011 RExC_parse[5] == 'E')
6013 ret = reganode(pRExC_state,DEFINEP,0);
6016 goto insert_if_check_paren;
6018 else if (RExC_parse[0] == 'R') {
6021 if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
6022 parno = atoi(RExC_parse++);
6023 while (isDIGIT(*RExC_parse))
6025 } else if (RExC_parse[0] == '&') {
6028 sv_dat = reg_scan_name(pRExC_state,
6029 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
6030 parno = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
6032 ret = reganode(pRExC_state,INSUBP,parno);
6033 goto insert_if_check_paren;
6035 else if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
6038 parno = atoi(RExC_parse++);
6040 while (isDIGIT(*RExC_parse))
6042 ret = reganode(pRExC_state, GROUPP, parno);
6044 insert_if_check_paren:
6045 if ((c = *nextchar(pRExC_state)) != ')')
6046 vFAIL("Switch condition not recognized");
6048 REGTAIL(pRExC_state, ret, reganode(pRExC_state, IFTHEN, 0));
6049 br = regbranch(pRExC_state, &flags, 1,depth+1);
6051 br = reganode(pRExC_state, LONGJMP, 0);
6053 REGTAIL(pRExC_state, br, reganode(pRExC_state, LONGJMP, 0));
6054 c = *nextchar(pRExC_state);
6059 vFAIL("(?(DEFINE)....) does not allow branches");
6060 lastbr = reganode(pRExC_state, IFTHEN, 0); /* Fake one for optimizer. */
6061 regbranch(pRExC_state, &flags, 1,depth+1);
6062 REGTAIL(pRExC_state, ret, lastbr);
6065 c = *nextchar(pRExC_state);
6070 vFAIL("Switch (?(condition)... contains too many branches");
6071 ender = reg_node(pRExC_state, TAIL);
6072 REGTAIL(pRExC_state, br, ender);
6074 REGTAIL(pRExC_state, lastbr, ender);
6075 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender);
6078 REGTAIL(pRExC_state, ret, ender);
6079 RExC_size++; /* XXX WHY do we need this?!!
6080 For large programs it seems to be required
6081 but I can't figure out why. -- dmq*/
6085 vFAIL2("Unknown switch condition (?(%.2s", RExC_parse);
6089 RExC_parse--; /* for vFAIL to print correctly */
6090 vFAIL("Sequence (? incomplete");
6094 parse_flags: /* (?i) */
6096 U32 posflags = 0, negflags = 0;
6097 U32 *flagsp = &posflags;
6099 while (*RExC_parse) {
6100 /* && strchr("iogcmsx", *RExC_parse) */
6101 /* (?g), (?gc) and (?o) are useless here
6102 and must be globally applied -- japhy */
6103 switch (*RExC_parse) {
6104 CASE_STD_PMMOD_FLAGS_PARSE_SET(flagsp);
6105 case ONCE_PAT_MOD: /* 'o' */
6106 case GLOBAL_PAT_MOD: /* 'g' */
6107 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6108 const I32 wflagbit = *RExC_parse == 'o' ? WASTED_O : WASTED_G;
6109 if (! (wastedflags & wflagbit) ) {
6110 wastedflags |= wflagbit;
6113 "Useless (%s%c) - %suse /%c modifier",
6114 flagsp == &negflags ? "?-" : "?",
6116 flagsp == &negflags ? "don't " : "",
6123 case CONTINUE_PAT_MOD: /* 'c' */
6124 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6125 if (! (wastedflags & WASTED_C) ) {
6126 wastedflags |= WASTED_GC;
6129 "Useless (%sc) - %suse /gc modifier",
6130 flagsp == &negflags ? "?-" : "?",
6131 flagsp == &negflags ? "don't " : ""
6136 case KEEPCOPY_PAT_MOD: /* 'p' */
6137 if (flagsp == &negflags) {
6139 ckWARNreg(RExC_parse + 1,"Useless use of (?-p)");
6141 *flagsp |= RXf_PMf_KEEPCOPY;
6145 if (flagsp == &negflags) {
6147 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6151 wastedflags = 0; /* reset so (?g-c) warns twice */
6157 RExC_flags |= posflags;
6158 RExC_flags &= ~negflags;
6160 oregflags |= posflags;
6161 oregflags &= ~negflags;
6163 nextchar(pRExC_state);
6174 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6179 }} /* one for the default block, one for the switch */
6186 ret = reganode(pRExC_state, OPEN, parno);
6189 RExC_nestroot = parno;
6190 if (RExC_seen & REG_SEEN_RECURSE
6191 && !RExC_open_parens[parno-1])
6193 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6194 "Setting open paren #%"IVdf" to %d\n",
6195 (IV)parno, REG_NODE_NUM(ret)));
6196 RExC_open_parens[parno-1]= ret;
6199 Set_Node_Length(ret, 1); /* MJD */
6200 Set_Node_Offset(ret, RExC_parse); /* MJD */
6208 /* Pick up the branches, linking them together. */
6209 parse_start = RExC_parse; /* MJD */
6210 br = regbranch(pRExC_state, &flags, 1,depth+1);
6213 if (RExC_npar > after_freeze)
6214 after_freeze = RExC_npar;
6215 RExC_npar = freeze_paren;
6218 /* branch_len = (paren != 0); */
6222 if (*RExC_parse == '|') {
6223 if (!SIZE_ONLY && RExC_extralen) {
6224 reginsert(pRExC_state, BRANCHJ, br, depth+1);
6227 reginsert(pRExC_state, BRANCH, br, depth+1);
6228 Set_Node_Length(br, paren != 0);
6229 Set_Node_Offset_To_R(br-RExC_emit_start, parse_start-RExC_start);
6233 RExC_extralen += 1; /* For BRANCHJ-BRANCH. */
6235 else if (paren == ':') {
6236 *flagp |= flags&SIMPLE;
6238 if (is_open) { /* Starts with OPEN. */
6239 REGTAIL(pRExC_state, ret, br); /* OPEN -> first. */
6241 else if (paren != '?') /* Not Conditional */
6243 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6245 while (*RExC_parse == '|') {
6246 if (!SIZE_ONLY && RExC_extralen) {
6247 ender = reganode(pRExC_state, LONGJMP,0);
6248 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender); /* Append to the previous. */
6251 RExC_extralen += 2; /* Account for LONGJMP. */
6252 nextchar(pRExC_state);
6254 if (RExC_npar > after_freeze)
6255 after_freeze = RExC_npar;
6256 RExC_npar = freeze_paren;
6258 br = regbranch(pRExC_state, &flags, 0, depth+1);
6262 REGTAIL(pRExC_state, lastbr, br); /* BRANCH -> BRANCH. */
6264 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6267 if (have_branch || paren != ':') {
6268 /* Make a closing node, and hook it on the end. */
6271 ender = reg_node(pRExC_state, TAIL);
6274 ender = reganode(pRExC_state, CLOSE, parno);
6275 if (!SIZE_ONLY && RExC_seen & REG_SEEN_RECURSE) {
6276 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6277 "Setting close paren #%"IVdf" to %d\n",
6278 (IV)parno, REG_NODE_NUM(ender)));
6279 RExC_close_parens[parno-1]= ender;
6280 if (RExC_nestroot == parno)
6283 Set_Node_Offset(ender,RExC_parse+1); /* MJD */
6284 Set_Node_Length(ender,1); /* MJD */
6290 *flagp &= ~HASWIDTH;
6293 ender = reg_node(pRExC_state, SUCCEED);
6296 ender = reg_node(pRExC_state, END);
6298 assert(!RExC_opend); /* there can only be one! */
6303 REGTAIL(pRExC_state, lastbr, ender);
6305 if (have_branch && !SIZE_ONLY) {
6307 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
6309 /* Hook the tails of the branches to the closing node. */
6310 for (br = ret; br; br = regnext(br)) {
6311 const U8 op = PL_regkind[OP(br)];
6313 REGTAIL_STUDY(pRExC_state, NEXTOPER(br), ender);
6315 else if (op == BRANCHJ) {
6316 REGTAIL_STUDY(pRExC_state, NEXTOPER(NEXTOPER(br)), ender);
6324 static const char parens[] = "=!<,>";
6326 if (paren && (p = strchr(parens, paren))) {
6327 U8 node = ((p - parens) % 2) ? UNLESSM : IFMATCH;
6328 int flag = (p - parens) > 1;
6331 node = SUSPEND, flag = 0;
6332 reginsert(pRExC_state, node,ret, depth+1);
6333 Set_Node_Cur_Length(ret);
6334 Set_Node_Offset(ret, parse_start + 1);
6336 REGTAIL_STUDY(pRExC_state, ret, reg_node(pRExC_state, TAIL));
6340 /* Check for proper termination. */
6342 RExC_flags = oregflags;
6343 if (RExC_parse >= RExC_end || *nextchar(pRExC_state) != ')') {
6344 RExC_parse = oregcomp_parse;
6345 vFAIL("Unmatched (");
6348 else if (!paren && RExC_parse < RExC_end) {
6349 if (*RExC_parse == ')') {
6351 vFAIL("Unmatched )");
6354 FAIL("Junk on end of regexp"); /* "Can't happen". */
6358 RExC_npar = after_freeze;
6363 - regbranch - one alternative of an | operator
6365 * Implements the concatenation operator.
6368 S_regbranch(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, I32 first, U32 depth)
6371 register regnode *ret;
6372 register regnode *chain = NULL;
6373 register regnode *latest;
6374 I32 flags = 0, c = 0;
6375 GET_RE_DEBUG_FLAGS_DECL;
6377 PERL_ARGS_ASSERT_REGBRANCH;
6379 DEBUG_PARSE("brnc");
6384 if (!SIZE_ONLY && RExC_extralen)
6385 ret = reganode(pRExC_state, BRANCHJ,0);
6387 ret = reg_node(pRExC_state, BRANCH);
6388 Set_Node_Length(ret, 1);
6392 if (!first && SIZE_ONLY)
6393 RExC_extralen += 1; /* BRANCHJ */
6395 *flagp = WORST; /* Tentatively. */
6398 nextchar(pRExC_state);
6399 while (RExC_parse < RExC_end && *RExC_parse != '|' && *RExC_parse != ')') {
6401 latest = regpiece(pRExC_state, &flags,depth+1);
6402 if (latest == NULL) {
6403 if (flags & TRYAGAIN)
6407 else if (ret == NULL)
6409 *flagp |= flags&(HASWIDTH|POSTPONED);
6410 if (chain == NULL) /* First piece. */
6411 *flagp |= flags&SPSTART;
6414 REGTAIL(pRExC_state, chain, latest);
6419 if (chain == NULL) { /* Loop ran zero times. */
6420 chain = reg_node(pRExC_state, NOTHING);
6425 *flagp |= flags&SIMPLE;
6432 - regpiece - something followed by possible [*+?]
6434 * Note that the branching code sequences used for ? and the general cases
6435 * of * and + are somewhat optimized: they use the same NOTHING node as
6436 * both the endmarker for their branch list and the body of the last branch.
6437 * It might seem that this node could be dispensed with entirely, but the
6438 * endmarker role is not redundant.
6441 S_regpiece(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6444 register regnode *ret;
6446 register char *next;
6448 const char * const origparse = RExC_parse;
6450 I32 max = REG_INFTY;
6452 const char *maxpos = NULL;
6453 GET_RE_DEBUG_FLAGS_DECL;
6455 PERL_ARGS_ASSERT_REGPIECE;
6457 DEBUG_PARSE("piec");
6459 ret = regatom(pRExC_state, &flags,depth+1);
6461 if (flags & TRYAGAIN)
6468 if (op == '{' && regcurly(RExC_parse)) {
6470 parse_start = RExC_parse; /* MJD */
6471 next = RExC_parse + 1;
6472 while (isDIGIT(*next) || *next == ',') {
6481 if (*next == '}') { /* got one */
6485 min = atoi(RExC_parse);
6489 maxpos = RExC_parse;
6491 if (!max && *maxpos != '0')
6492 max = REG_INFTY; /* meaning "infinity" */
6493 else if (max >= REG_INFTY)
6494 vFAIL2("Quantifier in {,} bigger than %d", REG_INFTY - 1);
6496 nextchar(pRExC_state);
6499 if ((flags&SIMPLE)) {
6500 RExC_naughty += 2 + RExC_naughty / 2;
6501 reginsert(pRExC_state, CURLY, ret, depth+1);
6502 Set_Node_Offset(ret, parse_start+1); /* MJD */
6503 Set_Node_Cur_Length(ret);
6506 regnode * const w = reg_node(pRExC_state, WHILEM);
6509 REGTAIL(pRExC_state, ret, w);
6510 if (!SIZE_ONLY && RExC_extralen) {
6511 reginsert(pRExC_state, LONGJMP,ret, depth+1);
6512 reginsert(pRExC_state, NOTHING,ret, depth+1);
6513 NEXT_OFF(ret) = 3; /* Go over LONGJMP. */
6515 reginsert(pRExC_state, CURLYX,ret, depth+1);
6517 Set_Node_Offset(ret, parse_start+1);
6518 Set_Node_Length(ret,
6519 op == '{' ? (RExC_parse - parse_start) : 1);
6521 if (!SIZE_ONLY && RExC_extralen)
6522 NEXT_OFF(ret) = 3; /* Go over NOTHING to LONGJMP. */
6523 REGTAIL(pRExC_state, ret, reg_node(pRExC_state, NOTHING));
6525 RExC_whilem_seen++, RExC_extralen += 3;
6526 RExC_naughty += 4 + RExC_naughty; /* compound interest */
6535 vFAIL("Can't do {n,m} with n > m");
6537 ARG1_SET(ret, (U16)min);
6538 ARG2_SET(ret, (U16)max);
6550 #if 0 /* Now runtime fix should be reliable. */
6552 /* if this is reinstated, don't forget to put this back into perldiag:
6554 =item Regexp *+ operand could be empty at {#} in regex m/%s/
6556 (F) The part of the regexp subject to either the * or + quantifier
6557 could match an empty string. The {#} shows in the regular
6558 expression about where the problem was discovered.
6562 if (!(flags&HASWIDTH) && op != '?')
6563 vFAIL("Regexp *+ operand could be empty");
6566 parse_start = RExC_parse;
6567 nextchar(pRExC_state);
6569 *flagp = (op != '+') ? (WORST|SPSTART|HASWIDTH) : (WORST|HASWIDTH);
6571 if (op == '*' && (flags&SIMPLE)) {
6572 reginsert(pRExC_state, STAR, ret, depth+1);
6576 else if (op == '*') {
6580 else if (op == '+' && (flags&SIMPLE)) {
6581 reginsert(pRExC_state, PLUS, ret, depth+1);
6585 else if (op == '+') {
6589 else if (op == '?') {
6594 if (!SIZE_ONLY && !(flags&(HASWIDTH|POSTPONED)) && max > REG_INFTY/3) {
6595 ckWARN3reg(RExC_parse,
6596 "%.*s matches null string many times",
6597 (int)(RExC_parse >= origparse ? RExC_parse - origparse : 0),
6601 if (RExC_parse < RExC_end && *RExC_parse == '?') {
6602 nextchar(pRExC_state);
6603 reginsert(pRExC_state, MINMOD, ret, depth+1);
6604 REGTAIL(pRExC_state, ret, ret + NODE_STEP_REGNODE);
6606 #ifndef REG_ALLOW_MINMOD_SUSPEND
6609 if (RExC_parse < RExC_end && *RExC_parse == '+') {
6611 nextchar(pRExC_state);
6612 ender = reg_node(pRExC_state, SUCCEED);
6613 REGTAIL(pRExC_state, ret, ender);
6614 reginsert(pRExC_state, SUSPEND, ret, depth+1);
6616 ender = reg_node(pRExC_state, TAIL);
6617 REGTAIL(pRExC_state, ret, ender);
6621 if (RExC_parse < RExC_end && ISMULT2(RExC_parse)) {
6623 vFAIL("Nested quantifiers");
6630 /* reg_namedseq(pRExC_state,UVp)
6632 This is expected to be called by a parser routine that has
6633 recognized '\N' and needs to handle the rest. RExC_parse is
6634 expected to point at the first char following the N at the time
6637 The \N may be inside (indicated by valuep not being NULL) or outside a
6640 \N may begin either a named sequence, or if outside a character class, mean
6641 to match a non-newline. For non single-quoted regexes, the tokenizer has
6642 attempted to decide which, and in the case of a named sequence converted it
6643 into one of the forms: \N{} (if the sequence is null), or \N{U+c1.c2...},
6644 where c1... are the characters in the sequence. For single-quoted regexes,
6645 the tokenizer passes the \N sequence through unchanged; this code will not
6646 attempt to determine this nor expand those. The net effect is that if the
6647 beginning of the passed-in pattern isn't '{U+' or there is no '}', it
6648 signals that this \N occurrence means to match a non-newline.
6650 Only the \N{U+...} form should occur in a character class, for the same
6651 reason that '.' inside a character class means to just match a period: it
6652 just doesn't make sense.
6654 If valuep is non-null then it is assumed that we are parsing inside
6655 of a charclass definition and the first codepoint in the resolved
6656 string is returned via *valuep and the routine will return NULL.
6657 In this mode if a multichar string is returned from the charnames
6658 handler, a warning will be issued, and only the first char in the
6659 sequence will be examined. If the string returned is zero length
6660 then the value of *valuep is undefined and NON-NULL will
6661 be returned to indicate failure. (This will NOT be a valid pointer
6664 If valuep is null then it is assumed that we are parsing normal text and a
6665 new EXACT node is inserted into the program containing the resolved string,
6666 and a pointer to the new node is returned. But if the string is zero length
6667 a NOTHING node is emitted instead.
6669 On success RExC_parse is set to the char following the endbrace.
6670 Parsing failures will generate a fatal error via vFAIL(...)
6673 S_reg_namedseq(pTHX_ RExC_state_t *pRExC_state, UV *valuep, I32 *flagp)
6675 char * endbrace; /* '}' following the name */
6676 regnode *ret = NULL;
6678 char* parse_start = RExC_parse - 2; /* points to the '\N' */
6682 GET_RE_DEBUG_FLAGS_DECL;
6684 PERL_ARGS_ASSERT_REG_NAMEDSEQ;
6688 /* The [^\n] meaning of \N ignores spaces and comments under the /x
6689 * modifier. The other meaning does not */
6690 p = (RExC_flags & RXf_PMf_EXTENDED)
6691 ? regwhite( pRExC_state, RExC_parse )
6694 /* Disambiguate between \N meaning a named character versus \N meaning
6695 * [^\n]. The former is assumed when it can't be the latter. */
6696 if (*p != '{' || regcurly(p)) {
6699 /* no bare \N in a charclass */
6700 vFAIL("\\N in a character class must be a named character: \\N{...}");
6702 nextchar(pRExC_state);
6703 ret = reg_node(pRExC_state, REG_ANY);
6704 *flagp |= HASWIDTH|SIMPLE;
6707 Set_Node_Length(ret, 1); /* MJD */
6711 /* Here, we have decided it should be a named sequence */
6713 /* The test above made sure that the next real character is a '{', but
6714 * under the /x modifier, it could be separated by space (or a comment and
6715 * \n) and this is not allowed (for consistency with \x{...} and the
6716 * tokenizer handling of \N{NAME}). */
6717 if (*RExC_parse != '{') {
6718 vFAIL("Missing braces on \\N{}");
6721 RExC_parse++; /* Skip past the '{' */
6723 if (! (endbrace = strchr(RExC_parse, '}')) /* no trailing brace */
6724 || ! (endbrace == RExC_parse /* nothing between the {} */
6725 || (endbrace - RExC_parse >= 2 /* U+ (bad hex is checked below */
6726 && strnEQ(RExC_parse, "U+", 2)))) /* for a better error msg) */
6728 if (endbrace) RExC_parse = endbrace; /* position msg's '<--HERE' */
6729 vFAIL("\\N{NAME} must be resolved by the lexer");
6732 if (endbrace == RExC_parse) { /* empty: \N{} */
6734 RExC_parse = endbrace + 1;
6735 return reg_node(pRExC_state,NOTHING);
6739 ckWARNreg(RExC_parse,
6740 "Ignoring zero length \\N{} in character class"
6742 RExC_parse = endbrace + 1;
6745 return (regnode *) &RExC_parse; /* Invalid regnode pointer */
6748 RExC_utf8 = 1; /* named sequences imply Unicode semantics */
6749 RExC_parse += 2; /* Skip past the 'U+' */
6751 if (valuep) { /* In a bracketed char class */
6752 /* We only pay attention to the first char of
6753 multichar strings being returned. I kinda wonder
6754 if this makes sense as it does change the behaviour
6755 from earlier versions, OTOH that behaviour was broken
6756 as well. XXX Solution is to recharacterize as
6757 [rest-of-class]|multi1|multi2... */
6759 STRLEN length_of_hex;
6760 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
6761 | PERL_SCAN_DISALLOW_PREFIX
6762 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
6764 char * endchar = RExC_parse + strcspn(RExC_parse, ".}");
6765 if (endchar < endbrace) {
6766 ckWARNreg(endchar, "Using just the first character returned by \\N{} in character class");
6769 length_of_hex = (STRLEN)(endchar - RExC_parse);
6770 *valuep = grok_hex(RExC_parse, &length_of_hex, &flags, NULL);
6772 /* The tokenizer should have guaranteed validity, but it's possible to
6773 * bypass it by using single quoting, so check */
6774 if (length_of_hex == 0
6775 || length_of_hex != (STRLEN)(endchar - RExC_parse) )
6777 RExC_parse += length_of_hex; /* Includes all the valid */
6778 RExC_parse += (RExC_orig_utf8) /* point to after 1st invalid */
6779 ? UTF8SKIP(RExC_parse)
6781 /* Guard against malformed utf8 */
6782 if (RExC_parse >= endchar) RExC_parse = endchar;
6783 vFAIL("Invalid hexadecimal number in \\N{U+...}");
6786 RExC_parse = endbrace + 1;
6787 if (endchar == endbrace) return NULL;
6789 ret = (regnode *) &RExC_parse; /* Invalid regnode pointer */
6791 else { /* Not a char class */
6792 char *s; /* String to put in generated EXACT node */
6793 STRLEN len = 0; /* Its current length */
6794 char *endchar; /* Points to '.' or '}' ending cur char in the input
6797 ret = reg_node(pRExC_state,
6798 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
6801 /* Exact nodes can hold only a U8 length's of text = 255. Loop through
6802 * the input which is of the form now 'c1.c2.c3...}' until find the
6803 * ending brace or exeed length 255. The characters that exceed this
6804 * limit are dropped. The limit could be relaxed should it become
6805 * desirable by reparsing this as (?:\N{NAME}), so could generate
6806 * multiple EXACT nodes, as is done for just regular input. But this
6807 * is primarily a named character, and not intended to be a huge long
6808 * string, so 255 bytes should be good enough */
6810 STRLEN length_of_hex;
6811 I32 grok_flags = PERL_SCAN_ALLOW_UNDERSCORES
6812 | PERL_SCAN_DISALLOW_PREFIX
6813 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
6814 UV cp; /* Ord of current character */
6816 /* Code points are separated by dots. If none, there is only one
6817 * code point, and is terminated by the brace */
6818 endchar = RExC_parse + strcspn(RExC_parse, ".}");
6820 /* The values are Unicode even on EBCDIC machines */
6821 length_of_hex = (STRLEN)(endchar - RExC_parse);
6822 cp = grok_hex(RExC_parse, &length_of_hex, &grok_flags, NULL);
6823 if ( length_of_hex == 0
6824 || length_of_hex != (STRLEN)(endchar - RExC_parse) )
6826 RExC_parse += length_of_hex; /* Includes all the valid */
6827 RExC_parse += (RExC_orig_utf8) /* point to after 1st invalid */
6828 ? UTF8SKIP(RExC_parse)
6830 /* Guard against malformed utf8 */
6831 if (RExC_parse >= endchar) RExC_parse = endchar;
6832 vFAIL("Invalid hexadecimal number in \\N{U+...}");
6835 if (! FOLD) { /* Not folding, just append to the string */
6838 /* Quit before adding this character if would exceed limit */
6839 if (len + UNISKIP(cp) > U8_MAX) break;
6841 unilen = reguni(pRExC_state, cp, s);
6846 } else { /* Folding, output the folded equivalent */
6847 STRLEN foldlen,numlen;
6848 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
6849 cp = toFOLD_uni(cp, tmpbuf, &foldlen);
6851 /* Quit before exceeding size limit */
6852 if (len + foldlen > U8_MAX) break;
6854 for (foldbuf = tmpbuf;
6858 cp = utf8_to_uvchr(foldbuf, &numlen);
6860 const STRLEN unilen = reguni(pRExC_state, cp, s);
6863 /* In EBCDIC the numlen and unilen can differ. */
6865 if (numlen >= foldlen)
6869 break; /* "Can't happen." */
6873 /* Point to the beginning of the next character in the sequence. */
6874 RExC_parse = endchar + 1;
6876 /* Quit if no more characters */
6877 if (RExC_parse >= endbrace) break;
6882 if (RExC_parse < endbrace) {
6883 ckWARNreg(RExC_parse - 1,
6884 "Using just the first characters returned by \\N{}");
6887 RExC_size += STR_SZ(len);
6890 RExC_emit += STR_SZ(len);
6893 RExC_parse = endbrace + 1;
6895 *flagp |= HASWIDTH; /* Not SIMPLE, as that causes the engine to fail
6896 with malformed in t/re/pat_advanced.t */
6898 Set_Node_Cur_Length(ret); /* MJD */
6899 nextchar(pRExC_state);
6909 * It returns the code point in utf8 for the value in *encp.
6910 * value: a code value in the source encoding
6911 * encp: a pointer to an Encode object
6913 * If the result from Encode is not a single character,
6914 * it returns U+FFFD (Replacement character) and sets *encp to NULL.
6917 S_reg_recode(pTHX_ const char value, SV **encp)
6920 SV * const sv = newSVpvn_flags(&value, numlen, SVs_TEMP);
6921 const char * const s = *encp ? sv_recode_to_utf8(sv, *encp) : SvPVX(sv);
6922 const STRLEN newlen = SvCUR(sv);
6923 UV uv = UNICODE_REPLACEMENT;
6925 PERL_ARGS_ASSERT_REG_RECODE;
6929 ? utf8n_to_uvchr((U8*)s, newlen, &numlen, UTF8_ALLOW_DEFAULT)
6932 if (!newlen || numlen != newlen) {
6933 uv = UNICODE_REPLACEMENT;
6941 - regatom - the lowest level
6943 Try to identify anything special at the start of the pattern. If there
6944 is, then handle it as required. This may involve generating a single regop,
6945 such as for an assertion; or it may involve recursing, such as to
6946 handle a () structure.
6948 If the string doesn't start with something special then we gobble up
6949 as much literal text as we can.
6951 Once we have been able to handle whatever type of thing started the
6952 sequence, we return.
6954 Note: we have to be careful with escapes, as they can be both literal
6955 and special, and in the case of \10 and friends can either, depending
6956 on context. Specifically there are two seperate switches for handling
6957 escape sequences, with the one for handling literal escapes requiring
6958 a dummy entry for all of the special escapes that are actually handled
6963 S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6966 register regnode *ret = NULL;
6968 char *parse_start = RExC_parse;
6969 GET_RE_DEBUG_FLAGS_DECL;
6970 DEBUG_PARSE("atom");
6971 *flagp = WORST; /* Tentatively. */
6973 PERL_ARGS_ASSERT_REGATOM;
6976 switch ((U8)*RExC_parse) {
6978 RExC_seen_zerolen++;
6979 nextchar(pRExC_state);
6980 if (RExC_flags & RXf_PMf_MULTILINE)
6981 ret = reg_node(pRExC_state, MBOL);
6982 else if (RExC_flags & RXf_PMf_SINGLELINE)
6983 ret = reg_node(pRExC_state, SBOL);
6985 ret = reg_node(pRExC_state, BOL);
6986 Set_Node_Length(ret, 1); /* MJD */
6989 nextchar(pRExC_state);
6991 RExC_seen_zerolen++;
6992 if (RExC_flags & RXf_PMf_MULTILINE)
6993 ret = reg_node(pRExC_state, MEOL);
6994 else if (RExC_flags & RXf_PMf_SINGLELINE)
6995 ret = reg_node(pRExC_state, SEOL);
6997 ret = reg_node(pRExC_state, EOL);
6998 Set_Node_Length(ret, 1); /* MJD */
7001 nextchar(pRExC_state);
7002 if (RExC_flags & RXf_PMf_SINGLELINE)
7003 ret = reg_node(pRExC_state, SANY);
7005 ret = reg_node(pRExC_state, REG_ANY);
7006 *flagp |= HASWIDTH|SIMPLE;
7008 Set_Node_Length(ret, 1); /* MJD */
7012 char * const oregcomp_parse = ++RExC_parse;
7013 ret = regclass(pRExC_state,depth+1);
7014 if (*RExC_parse != ']') {
7015 RExC_parse = oregcomp_parse;
7016 vFAIL("Unmatched [");
7018 nextchar(pRExC_state);
7019 *flagp |= HASWIDTH|SIMPLE;
7020 Set_Node_Length(ret, RExC_parse - oregcomp_parse + 1); /* MJD */
7024 nextchar(pRExC_state);
7025 ret = reg(pRExC_state, 1, &flags,depth+1);
7027 if (flags & TRYAGAIN) {
7028 if (RExC_parse == RExC_end) {
7029 /* Make parent create an empty node if needed. */
7037 *flagp |= flags&(HASWIDTH|SPSTART|SIMPLE|POSTPONED);
7041 if (flags & TRYAGAIN) {
7045 vFAIL("Internal urp");
7046 /* Supposed to be caught earlier. */
7049 if (!regcurly(RExC_parse)) {
7058 vFAIL("Quantifier follows nothing");
7066 len=0; /* silence a spurious compiler warning */
7067 if ((cp = what_len_TRICKYFOLD_safe(RExC_parse,RExC_end,UTF,len))) {
7068 *flagp |= HASWIDTH; /* could be SIMPLE too, but needs a handler in regexec.regrepeat */
7069 RExC_parse+=len-1; /* we get one from nextchar() as well. :-( */
7070 ret = reganode(pRExC_state, FOLDCHAR, cp);
7071 Set_Node_Length(ret, 1); /* MJD */
7072 nextchar(pRExC_state); /* kill whitespace under /x */
7080 This switch handles escape sequences that resolve to some kind
7081 of special regop and not to literal text. Escape sequnces that
7082 resolve to literal text are handled below in the switch marked
7085 Every entry in this switch *must* have a corresponding entry
7086 in the literal escape switch. However, the opposite is not
7087 required, as the default for this switch is to jump to the
7088 literal text handling code.
7090 switch ((U8)*++RExC_parse) {
7095 /* Special Escapes */
7097 RExC_seen_zerolen++;
7098 ret = reg_node(pRExC_state, SBOL);
7100 goto finish_meta_pat;
7102 ret = reg_node(pRExC_state, GPOS);
7103 RExC_seen |= REG_SEEN_GPOS;
7105 goto finish_meta_pat;
7107 RExC_seen_zerolen++;
7108 ret = reg_node(pRExC_state, KEEPS);
7110 /* XXX:dmq : disabling in-place substitution seems to
7111 * be necessary here to avoid cases of memory corruption, as
7112 * with: C<$_="x" x 80; s/x\K/y/> -- rgs
7114 RExC_seen |= REG_SEEN_LOOKBEHIND;
7115 goto finish_meta_pat;
7117 ret = reg_node(pRExC_state, SEOL);
7119 RExC_seen_zerolen++; /* Do not optimize RE away */
7120 goto finish_meta_pat;
7122 ret = reg_node(pRExC_state, EOS);
7124 RExC_seen_zerolen++; /* Do not optimize RE away */
7125 goto finish_meta_pat;
7127 ret = reg_node(pRExC_state, CANY);
7128 RExC_seen |= REG_SEEN_CANY;
7129 *flagp |= HASWIDTH|SIMPLE;
7130 goto finish_meta_pat;
7132 ret = reg_node(pRExC_state, CLUMP);
7134 goto finish_meta_pat;
7136 ret = reg_node(pRExC_state, (U8)(LOC ? ALNUML : ALNUM));
7137 *flagp |= HASWIDTH|SIMPLE;
7138 goto finish_meta_pat;
7140 ret = reg_node(pRExC_state, (U8)(LOC ? NALNUML : NALNUM));
7141 *flagp |= HASWIDTH|SIMPLE;
7142 goto finish_meta_pat;
7144 RExC_seen_zerolen++;
7145 RExC_seen |= REG_SEEN_LOOKBEHIND;
7146 ret = reg_node(pRExC_state, (U8)(LOC ? BOUNDL : BOUND));
7148 goto finish_meta_pat;
7150 RExC_seen_zerolen++;
7151 RExC_seen |= REG_SEEN_LOOKBEHIND;
7152 ret = reg_node(pRExC_state, (U8)(LOC ? NBOUNDL : NBOUND));
7154 goto finish_meta_pat;
7156 ret = reg_node(pRExC_state, (U8)(LOC ? SPACEL : SPACE));
7157 *flagp |= HASWIDTH|SIMPLE;
7158 goto finish_meta_pat;
7160 ret = reg_node(pRExC_state, (U8)(LOC ? NSPACEL : NSPACE));
7161 *flagp |= HASWIDTH|SIMPLE;
7162 goto finish_meta_pat;
7164 ret = reg_node(pRExC_state, DIGIT);
7165 *flagp |= HASWIDTH|SIMPLE;
7166 goto finish_meta_pat;
7168 ret = reg_node(pRExC_state, NDIGIT);
7169 *flagp |= HASWIDTH|SIMPLE;
7170 goto finish_meta_pat;
7172 ret = reg_node(pRExC_state, LNBREAK);
7173 *flagp |= HASWIDTH|SIMPLE;
7174 goto finish_meta_pat;
7176 ret = reg_node(pRExC_state, HORIZWS);
7177 *flagp |= HASWIDTH|SIMPLE;
7178 goto finish_meta_pat;
7180 ret = reg_node(pRExC_state, NHORIZWS);
7181 *flagp |= HASWIDTH|SIMPLE;
7182 goto finish_meta_pat;
7184 ret = reg_node(pRExC_state, VERTWS);
7185 *flagp |= HASWIDTH|SIMPLE;
7186 goto finish_meta_pat;
7188 ret = reg_node(pRExC_state, NVERTWS);
7189 *flagp |= HASWIDTH|SIMPLE;
7191 nextchar(pRExC_state);
7192 Set_Node_Length(ret, 2); /* MJD */
7197 char* const oldregxend = RExC_end;
7199 char* parse_start = RExC_parse - 2;
7202 if (RExC_parse[1] == '{') {
7203 /* a lovely hack--pretend we saw [\pX] instead */
7204 RExC_end = strchr(RExC_parse, '}');
7206 const U8 c = (U8)*RExC_parse;
7208 RExC_end = oldregxend;
7209 vFAIL2("Missing right brace on \\%c{}", c);
7214 RExC_end = RExC_parse + 2;
7215 if (RExC_end > oldregxend)
7216 RExC_end = oldregxend;
7220 ret = regclass(pRExC_state,depth+1);
7222 RExC_end = oldregxend;
7225 Set_Node_Offset(ret, parse_start + 2);
7226 Set_Node_Cur_Length(ret);
7227 nextchar(pRExC_state);
7228 *flagp |= HASWIDTH|SIMPLE;
7232 /* Handle \N and \N{NAME} here and not below because it can be
7233 multicharacter. join_exact() will join them up later on.
7234 Also this makes sure that things like /\N{BLAH}+/ and
7235 \N{BLAH} being multi char Just Happen. dmq*/
7237 ret= reg_namedseq(pRExC_state, NULL, flagp);
7239 case 'k': /* Handle \k<NAME> and \k'NAME' */
7242 char ch= RExC_parse[1];
7243 if (ch != '<' && ch != '\'' && ch != '{') {
7245 vFAIL2("Sequence %.2s... not terminated",parse_start);
7247 /* this pretty much dupes the code for (?P=...) in reg(), if
7248 you change this make sure you change that */
7249 char* name_start = (RExC_parse += 2);
7251 SV *sv_dat = reg_scan_name(pRExC_state,
7252 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7253 ch= (ch == '<') ? '>' : (ch == '{') ? '}' : '\'';
7254 if (RExC_parse == name_start || *RExC_parse != ch)
7255 vFAIL2("Sequence %.3s... not terminated",parse_start);
7258 num = add_data( pRExC_state, 1, "S" );
7259 RExC_rxi->data->data[num]=(void*)sv_dat;
7260 SvREFCNT_inc_simple_void(sv_dat);
7264 ret = reganode(pRExC_state,
7265 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
7269 /* override incorrect value set in reganode MJD */
7270 Set_Node_Offset(ret, parse_start+1);
7271 Set_Node_Cur_Length(ret); /* MJD */
7272 nextchar(pRExC_state);
7278 case '1': case '2': case '3': case '4':
7279 case '5': case '6': case '7': case '8': case '9':
7282 bool isg = *RExC_parse == 'g';
7287 if (*RExC_parse == '{') {
7291 if (*RExC_parse == '-') {
7295 if (hasbrace && !isDIGIT(*RExC_parse)) {
7296 if (isrel) RExC_parse--;
7298 goto parse_named_seq;
7300 num = atoi(RExC_parse);
7301 if (isg && num == 0)
7302 vFAIL("Reference to invalid group 0");
7304 num = RExC_npar - num;
7306 vFAIL("Reference to nonexistent or unclosed group");
7308 if (!isg && num > 9 && num >= RExC_npar)
7311 char * const parse_start = RExC_parse - 1; /* MJD */
7312 while (isDIGIT(*RExC_parse))
7314 if (parse_start == RExC_parse - 1)
7315 vFAIL("Unterminated \\g... pattern");
7317 if (*RExC_parse != '}')
7318 vFAIL("Unterminated \\g{...} pattern");
7322 if (num > (I32)RExC_rx->nparens)
7323 vFAIL("Reference to nonexistent group");
7326 ret = reganode(pRExC_state,
7327 (U8)(FOLD ? (LOC ? REFFL : REFF) : REF),
7331 /* override incorrect value set in reganode MJD */
7332 Set_Node_Offset(ret, parse_start+1);
7333 Set_Node_Cur_Length(ret); /* MJD */
7335 nextchar(pRExC_state);
7340 if (RExC_parse >= RExC_end)
7341 FAIL("Trailing \\");
7344 /* Do not generate "unrecognized" warnings here, we fall
7345 back into the quick-grab loop below */
7352 if (RExC_flags & RXf_PMf_EXTENDED) {
7353 if ( reg_skipcomment( pRExC_state ) )
7360 register STRLEN len;
7365 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
7367 parse_start = RExC_parse - 1;
7373 ret = reg_node(pRExC_state,
7374 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
7376 for (len = 0, p = RExC_parse - 1;
7377 len < 127 && p < RExC_end;
7380 char * const oldp = p;
7382 if (RExC_flags & RXf_PMf_EXTENDED)
7383 p = regwhite( pRExC_state, p );
7388 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7389 goto normal_default;
7399 /* Literal Escapes Switch
7401 This switch is meant to handle escape sequences that
7402 resolve to a literal character.
7404 Every escape sequence that represents something
7405 else, like an assertion or a char class, is handled
7406 in the switch marked 'Special Escapes' above in this
7407 routine, but also has an entry here as anything that
7408 isn't explicitly mentioned here will be treated as
7409 an unescaped equivalent literal.
7413 /* These are all the special escapes. */
7417 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7418 goto normal_default;
7419 case 'A': /* Start assertion */
7420 case 'b': case 'B': /* Word-boundary assertion*/
7421 case 'C': /* Single char !DANGEROUS! */
7422 case 'd': case 'D': /* digit class */
7423 case 'g': case 'G': /* generic-backref, pos assertion */
7424 case 'h': case 'H': /* HORIZWS */
7425 case 'k': case 'K': /* named backref, keep marker */
7426 case 'N': /* named char sequence */
7427 case 'p': case 'P': /* Unicode property */
7428 case 'R': /* LNBREAK */
7429 case 's': case 'S': /* space class */
7430 case 'v': case 'V': /* VERTWS */
7431 case 'w': case 'W': /* word class */
7432 case 'X': /* eXtended Unicode "combining character sequence" */
7433 case 'z': case 'Z': /* End of line/string assertion */
7437 /* Anything after here is an escape that resolves to a
7438 literal. (Except digits, which may or may not)
7457 ender = ASCII_TO_NATIVE('\033');
7461 ender = ASCII_TO_NATIVE('\007');
7466 STRLEN brace_len = len;
7468 const char* error_msg;
7470 bool valid = grok_bslash_o(p,
7477 RExC_parse = p; /* going to die anyway; point
7478 to exact spot of failure */
7485 if (PL_encoding && ender < 0x100) {
7486 goto recode_encoding;
7495 char* const e = strchr(p, '}');
7499 vFAIL("Missing right brace on \\x{}");
7502 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
7503 | PERL_SCAN_DISALLOW_PREFIX;
7504 STRLEN numlen = e - p - 1;
7505 ender = grok_hex(p + 1, &numlen, &flags, NULL);
7512 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
7514 ender = grok_hex(p, &numlen, &flags, NULL);
7517 if (PL_encoding && ender < 0x100)
7518 goto recode_encoding;
7522 ender = grok_bslash_c(*p++, SIZE_ONLY);
7524 case '0': case '1': case '2': case '3':case '4':
7525 case '5': case '6': case '7': case '8':case '9':
7527 (isDIGIT(p[1]) && atoi(p) >= RExC_npar) ) {
7530 ender = grok_oct(p, &numlen, &flags, NULL);
7540 if (PL_encoding && ender < 0x100)
7541 goto recode_encoding;
7545 SV* enc = PL_encoding;
7546 ender = reg_recode((const char)(U8)ender, &enc);
7547 if (!enc && SIZE_ONLY)
7548 ckWARNreg(p, "Invalid escape in the specified encoding");
7554 FAIL("Trailing \\");
7557 if (!SIZE_ONLY&& isALPHA(*p))
7558 ckWARN2reg(p + 1, "Unrecognized escape \\%c passed through", UCHARAT(p));
7559 goto normal_default;
7564 if (UTF8_IS_START(*p) && UTF) {
7566 ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
7567 &numlen, UTF8_ALLOW_DEFAULT);
7574 if ( RExC_flags & RXf_PMf_EXTENDED)
7575 p = regwhite( pRExC_state, p );
7577 /* Prime the casefolded buffer. */
7578 ender = toFOLD_uni(ender, tmpbuf, &foldlen);
7580 if (p < RExC_end && ISMULT2(p)) { /* Back off on ?+*. */
7585 /* Emit all the Unicode characters. */
7587 for (foldbuf = tmpbuf;
7589 foldlen -= numlen) {
7590 ender = utf8_to_uvchr(foldbuf, &numlen);
7592 const STRLEN unilen = reguni(pRExC_state, ender, s);
7595 /* In EBCDIC the numlen
7596 * and unilen can differ. */
7598 if (numlen >= foldlen)
7602 break; /* "Can't happen." */
7606 const STRLEN unilen = reguni(pRExC_state, ender, s);
7615 REGC((char)ender, s++);
7621 /* Emit all the Unicode characters. */
7623 for (foldbuf = tmpbuf;
7625 foldlen -= numlen) {
7626 ender = utf8_to_uvchr(foldbuf, &numlen);
7628 const STRLEN unilen = reguni(pRExC_state, ender, s);
7631 /* In EBCDIC the numlen
7632 * and unilen can differ. */
7634 if (numlen >= foldlen)
7642 const STRLEN unilen = reguni(pRExC_state, ender, s);
7651 REGC((char)ender, s++);
7655 Set_Node_Cur_Length(ret); /* MJD */
7656 nextchar(pRExC_state);
7658 /* len is STRLEN which is unsigned, need to copy to signed */
7661 vFAIL("Internal disaster");
7665 if (len == 1 && UNI_IS_INVARIANT(ender))
7669 RExC_size += STR_SZ(len);
7672 RExC_emit += STR_SZ(len);
7682 S_regwhite( RExC_state_t *pRExC_state, char *p )
7684 const char *e = RExC_end;
7686 PERL_ARGS_ASSERT_REGWHITE;
7691 else if (*p == '#') {
7700 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
7708 /* Parse POSIX character classes: [[:foo:]], [[=foo=]], [[.foo.]].
7709 Character classes ([:foo:]) can also be negated ([:^foo:]).
7710 Returns a named class id (ANYOF_XXX) if successful, -1 otherwise.
7711 Equivalence classes ([=foo=]) and composites ([.foo.]) are parsed,
7712 but trigger failures because they are currently unimplemented. */
7714 #define POSIXCC_DONE(c) ((c) == ':')
7715 #define POSIXCC_NOTYET(c) ((c) == '=' || (c) == '.')
7716 #define POSIXCC(c) (POSIXCC_DONE(c) || POSIXCC_NOTYET(c))
7719 S_regpposixcc(pTHX_ RExC_state_t *pRExC_state, I32 value)
7722 I32 namedclass = OOB_NAMEDCLASS;
7724 PERL_ARGS_ASSERT_REGPPOSIXCC;
7726 if (value == '[' && RExC_parse + 1 < RExC_end &&
7727 /* I smell either [: or [= or [. -- POSIX has been here, right? */
7728 POSIXCC(UCHARAT(RExC_parse))) {
7729 const char c = UCHARAT(RExC_parse);
7730 char* const s = RExC_parse++;
7732 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != c)
7734 if (RExC_parse == RExC_end)
7735 /* Grandfather lone [:, [=, [. */
7738 const char* const t = RExC_parse++; /* skip over the c */
7741 if (UCHARAT(RExC_parse) == ']') {
7742 const char *posixcc = s + 1;
7743 RExC_parse++; /* skip over the ending ] */
7746 const I32 complement = *posixcc == '^' ? *posixcc++ : 0;
7747 const I32 skip = t - posixcc;
7749 /* Initially switch on the length of the name. */
7752 if (memEQ(posixcc, "word", 4)) /* this is not POSIX, this is the Perl \w */
7753 namedclass = complement ? ANYOF_NALNUM : ANYOF_ALNUM;
7756 /* Names all of length 5. */
7757 /* alnum alpha ascii blank cntrl digit graph lower
7758 print punct space upper */
7759 /* Offset 4 gives the best switch position. */
7760 switch (posixcc[4]) {
7762 if (memEQ(posixcc, "alph", 4)) /* alpha */
7763 namedclass = complement ? ANYOF_NALPHA : ANYOF_ALPHA;
7766 if (memEQ(posixcc, "spac", 4)) /* space */
7767 namedclass = complement ? ANYOF_NPSXSPC : ANYOF_PSXSPC;
7770 if (memEQ(posixcc, "grap", 4)) /* graph */
7771 namedclass = complement ? ANYOF_NGRAPH : ANYOF_GRAPH;
7774 if (memEQ(posixcc, "asci", 4)) /* ascii */
7775 namedclass = complement ? ANYOF_NASCII : ANYOF_ASCII;
7778 if (memEQ(posixcc, "blan", 4)) /* blank */
7779 namedclass = complement ? ANYOF_NBLANK : ANYOF_BLANK;
7782 if (memEQ(posixcc, "cntr", 4)) /* cntrl */
7783 namedclass = complement ? ANYOF_NCNTRL : ANYOF_CNTRL;
7786 if (memEQ(posixcc, "alnu", 4)) /* alnum */
7787 namedclass = complement ? ANYOF_NALNUMC : ANYOF_ALNUMC;
7790 if (memEQ(posixcc, "lowe", 4)) /* lower */
7791 namedclass = complement ? ANYOF_NLOWER : ANYOF_LOWER;
7792 else if (memEQ(posixcc, "uppe", 4)) /* upper */
7793 namedclass = complement ? ANYOF_NUPPER : ANYOF_UPPER;
7796 if (memEQ(posixcc, "digi", 4)) /* digit */
7797 namedclass = complement ? ANYOF_NDIGIT : ANYOF_DIGIT;
7798 else if (memEQ(posixcc, "prin", 4)) /* print */
7799 namedclass = complement ? ANYOF_NPRINT : ANYOF_PRINT;
7800 else if (memEQ(posixcc, "punc", 4)) /* punct */
7801 namedclass = complement ? ANYOF_NPUNCT : ANYOF_PUNCT;
7806 if (memEQ(posixcc, "xdigit", 6))
7807 namedclass = complement ? ANYOF_NXDIGIT : ANYOF_XDIGIT;
7811 if (namedclass == OOB_NAMEDCLASS)
7812 Simple_vFAIL3("POSIX class [:%.*s:] unknown",
7814 assert (posixcc[skip] == ':');
7815 assert (posixcc[skip+1] == ']');
7816 } else if (!SIZE_ONLY) {
7817 /* [[=foo=]] and [[.foo.]] are still future. */
7819 /* adjust RExC_parse so the warning shows after
7821 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse) != ']')
7823 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7826 /* Maternal grandfather:
7827 * "[:" ending in ":" but not in ":]" */
7837 S_checkposixcc(pTHX_ RExC_state_t *pRExC_state)
7841 PERL_ARGS_ASSERT_CHECKPOSIXCC;
7843 if (POSIXCC(UCHARAT(RExC_parse))) {
7844 const char *s = RExC_parse;
7845 const char c = *s++;
7849 if (*s && c == *s && s[1] == ']') {
7851 "POSIX syntax [%c %c] belongs inside character classes",
7854 /* [[=foo=]] and [[.foo.]] are still future. */
7855 if (POSIXCC_NOTYET(c)) {
7856 /* adjust RExC_parse so the error shows after
7858 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse++) != ']')
7860 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7867 #define _C_C_T_(NAME,TEST,WORD) \
7870 ANYOF_CLASS_SET(ret, ANYOF_##NAME); \
7872 for (value = 0; value < 256; value++) \
7874 ANYOF_BITMAP_SET(ret, value); \
7879 case ANYOF_N##NAME: \
7881 ANYOF_CLASS_SET(ret, ANYOF_N##NAME); \
7883 for (value = 0; value < 256; value++) \
7885 ANYOF_BITMAP_SET(ret, value); \
7891 #define _C_C_T_NOLOC_(NAME,TEST,WORD) \
7893 for (value = 0; value < 256; value++) \
7895 ANYOF_BITMAP_SET(ret, value); \
7899 case ANYOF_N##NAME: \
7900 for (value = 0; value < 256; value++) \
7902 ANYOF_BITMAP_SET(ret, value); \
7908 We dont use PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS as the direct test
7909 so that it is possible to override the option here without having to
7910 rebuild the entire core. as we are required to do if we change regcomp.h
7911 which is where PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS is defined.
7913 #if PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS
7914 #define BROKEN_UNICODE_CHARCLASS_MAPPINGS
7917 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
7918 #define POSIX_CC_UNI_NAME(CCNAME) CCNAME
7920 #define POSIX_CC_UNI_NAME(CCNAME) "Posix" CCNAME
7924 parse a class specification and produce either an ANYOF node that
7925 matches the pattern or if the pattern matches a single char only and
7926 that char is < 256 and we are case insensitive then we produce an
7931 S_regclass(pTHX_ RExC_state_t *pRExC_state, U32 depth)
7934 register UV nextvalue;
7935 register IV prevvalue = OOB_UNICODE;
7936 register IV range = 0;
7937 UV value = 0; /* XXX:dmq: needs to be referenceable (unfortunately) */
7938 register regnode *ret;
7941 char *rangebegin = NULL;
7942 bool need_class = 0;
7945 bool optimize_invert = TRUE;
7946 AV* unicode_alternate = NULL;
7948 UV literal_endpoint = 0;
7950 UV stored = 0; /* number of chars stored in the class */
7952 regnode * const orig_emit = RExC_emit; /* Save the original RExC_emit in
7953 case we need to change the emitted regop to an EXACT. */
7954 const char * orig_parse = RExC_parse;
7955 GET_RE_DEBUG_FLAGS_DECL;
7957 PERL_ARGS_ASSERT_REGCLASS;
7959 PERL_UNUSED_ARG(depth);
7962 DEBUG_PARSE("clas");
7964 /* Assume we are going to generate an ANYOF node. */
7965 ret = reganode(pRExC_state, ANYOF, 0);
7968 ANYOF_FLAGS(ret) = 0;
7970 if (UCHARAT(RExC_parse) == '^') { /* Complement of range. */
7974 ANYOF_FLAGS(ret) |= ANYOF_INVERT;
7978 RExC_size += ANYOF_SKIP;
7979 listsv = &PL_sv_undef; /* For code scanners: listsv always non-NULL. */
7982 RExC_emit += ANYOF_SKIP;
7984 ANYOF_FLAGS(ret) |= ANYOF_FOLD;
7986 ANYOF_FLAGS(ret) |= ANYOF_LOCALE;
7987 ANYOF_BITMAP_ZERO(ret);
7988 listsv = newSVpvs("# comment\n");
7991 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
7993 if (!SIZE_ONLY && POSIXCC(nextvalue))
7994 checkposixcc(pRExC_state);
7996 /* allow 1st char to be ] (allowing it to be - is dealt with later) */
7997 if (UCHARAT(RExC_parse) == ']')
8001 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != ']') {
8005 namedclass = OOB_NAMEDCLASS; /* initialize as illegal */
8008 rangebegin = RExC_parse;
8010 value = utf8n_to_uvchr((U8*)RExC_parse,
8011 RExC_end - RExC_parse,
8012 &numlen, UTF8_ALLOW_DEFAULT);
8013 RExC_parse += numlen;
8016 value = UCHARAT(RExC_parse++);
8018 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
8019 if (value == '[' && POSIXCC(nextvalue))
8020 namedclass = regpposixcc(pRExC_state, value);
8021 else if (value == '\\') {
8023 value = utf8n_to_uvchr((U8*)RExC_parse,
8024 RExC_end - RExC_parse,
8025 &numlen, UTF8_ALLOW_DEFAULT);
8026 RExC_parse += numlen;
8029 value = UCHARAT(RExC_parse++);
8030 /* Some compilers cannot handle switching on 64-bit integer
8031 * values, therefore value cannot be an UV. Yes, this will
8032 * be a problem later if we want switch on Unicode.
8033 * A similar issue a little bit later when switching on
8034 * namedclass. --jhi */
8035 switch ((I32)value) {
8036 case 'w': namedclass = ANYOF_ALNUM; break;
8037 case 'W': namedclass = ANYOF_NALNUM; break;
8038 case 's': namedclass = ANYOF_SPACE; break;
8039 case 'S': namedclass = ANYOF_NSPACE; break;
8040 case 'd': namedclass = ANYOF_DIGIT; break;
8041 case 'D': namedclass = ANYOF_NDIGIT; break;
8042 case 'v': namedclass = ANYOF_VERTWS; break;
8043 case 'V': namedclass = ANYOF_NVERTWS; break;
8044 case 'h': namedclass = ANYOF_HORIZWS; break;
8045 case 'H': namedclass = ANYOF_NHORIZWS; break;
8046 case 'N': /* Handle \N{NAME} in class */
8048 /* We only pay attention to the first char of
8049 multichar strings being returned. I kinda wonder
8050 if this makes sense as it does change the behaviour
8051 from earlier versions, OTOH that behaviour was broken
8053 UV v; /* value is register so we cant & it /grrr */
8054 if (reg_namedseq(pRExC_state, &v, NULL)) {
8064 if (RExC_parse >= RExC_end)
8065 vFAIL2("Empty \\%c{}", (U8)value);
8066 if (*RExC_parse == '{') {
8067 const U8 c = (U8)value;
8068 e = strchr(RExC_parse++, '}');
8070 vFAIL2("Missing right brace on \\%c{}", c);
8071 while (isSPACE(UCHARAT(RExC_parse)))
8073 if (e == RExC_parse)
8074 vFAIL2("Empty \\%c{}", c);
8076 while (isSPACE(UCHARAT(RExC_parse + n - 1)))
8084 if (UCHARAT(RExC_parse) == '^') {
8087 value = value == 'p' ? 'P' : 'p'; /* toggle */
8088 while (isSPACE(UCHARAT(RExC_parse))) {
8093 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::%.*s\n",
8094 (value=='p' ? '+' : '!'), (int)n, RExC_parse);
8097 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8098 namedclass = ANYOF_MAX; /* no official name, but it's named */
8101 case 'n': value = '\n'; break;
8102 case 'r': value = '\r'; break;
8103 case 't': value = '\t'; break;
8104 case 'f': value = '\f'; break;
8105 case 'b': value = '\b'; break;
8106 case 'e': value = ASCII_TO_NATIVE('\033');break;
8107 case 'a': value = ASCII_TO_NATIVE('\007');break;
8109 RExC_parse--; /* function expects to be pointed at the 'o' */
8111 const char* error_msg;
8112 bool valid = grok_bslash_o(RExC_parse,
8117 RExC_parse += numlen;
8122 if (PL_encoding && value < 0x100) {
8123 goto recode_encoding;
8127 if (*RExC_parse == '{') {
8128 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
8129 | PERL_SCAN_DISALLOW_PREFIX;
8130 char * const e = strchr(RExC_parse++, '}');
8132 vFAIL("Missing right brace on \\x{}");
8134 numlen = e - RExC_parse;
8135 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8139 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
8141 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8142 RExC_parse += numlen;
8144 if (PL_encoding && value < 0x100)
8145 goto recode_encoding;
8148 value = grok_bslash_c(*RExC_parse++, SIZE_ONLY);
8150 case '0': case '1': case '2': case '3': case '4':
8151 case '5': case '6': case '7': case '8': case '9':
8155 value = grok_oct(--RExC_parse, &numlen, &flags, NULL);
8156 RExC_parse += numlen;
8157 if (PL_encoding && value < 0x100)
8158 goto recode_encoding;
8163 SV* enc = PL_encoding;
8164 value = reg_recode((const char)(U8)value, &enc);
8165 if (!enc && SIZE_ONLY)
8166 ckWARNreg(RExC_parse,
8167 "Invalid escape in the specified encoding");
8171 if (!SIZE_ONLY && isALPHA(value))
8172 ckWARN2reg(RExC_parse,
8173 "Unrecognized escape \\%c in character class passed through",
8177 } /* end of \blah */
8183 if (namedclass > OOB_NAMEDCLASS) { /* this is a named class \blah */
8185 if (!SIZE_ONLY && !need_class)
8186 ANYOF_CLASS_ZERO(ret);
8190 /* a bad range like a-\d, a-[:digit:] ? */
8194 RExC_parse >= rangebegin ?
8195 RExC_parse - rangebegin : 0;
8196 ckWARN4reg(RExC_parse,
8197 "False [] range \"%*.*s\"",
8200 if (prevvalue < 256) {
8201 ANYOF_BITMAP_SET(ret, prevvalue);
8202 ANYOF_BITMAP_SET(ret, '-');
8205 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8206 Perl_sv_catpvf(aTHX_ listsv,
8207 "%04"UVxf"\n%04"UVxf"\n", (UV)prevvalue, (UV) '-');
8211 range = 0; /* this was not a true range */
8217 const char *what = NULL;
8220 if (namedclass > OOB_NAMEDCLASS)
8221 optimize_invert = FALSE;
8222 /* Possible truncation here but in some 64-bit environments
8223 * the compiler gets heartburn about switch on 64-bit values.
8224 * A similar issue a little earlier when switching on value.
8226 switch ((I32)namedclass) {
8228 case _C_C_T_(ALNUMC, isALNUMC(value), POSIX_CC_UNI_NAME("Alnum"));
8229 case _C_C_T_(ALPHA, isALPHA(value), POSIX_CC_UNI_NAME("Alpha"));
8230 case _C_C_T_(BLANK, isBLANK(value), POSIX_CC_UNI_NAME("Blank"));
8231 case _C_C_T_(CNTRL, isCNTRL(value), POSIX_CC_UNI_NAME("Cntrl"));
8232 case _C_C_T_(GRAPH, isGRAPH(value), POSIX_CC_UNI_NAME("Graph"));
8233 case _C_C_T_(LOWER, isLOWER(value), POSIX_CC_UNI_NAME("Lower"));
8234 case _C_C_T_(PRINT, isPRINT(value), POSIX_CC_UNI_NAME("Print"));
8235 case _C_C_T_(PSXSPC, isPSXSPC(value), POSIX_CC_UNI_NAME("Space"));
8236 case _C_C_T_(PUNCT, isPUNCT(value), POSIX_CC_UNI_NAME("Punct"));
8237 case _C_C_T_(UPPER, isUPPER(value), POSIX_CC_UNI_NAME("Upper"));
8238 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
8239 case _C_C_T_(ALNUM, isALNUM(value), "Word");
8240 case _C_C_T_(SPACE, isSPACE(value), "SpacePerl");
8242 case _C_C_T_(SPACE, isSPACE(value), "PerlSpace");
8243 case _C_C_T_(ALNUM, isALNUM(value), "PerlWord");
8245 case _C_C_T_(XDIGIT, isXDIGIT(value), "XDigit");
8246 case _C_C_T_NOLOC_(VERTWS, is_VERTWS_latin1(&value), "VertSpace");
8247 case _C_C_T_NOLOC_(HORIZWS, is_HORIZWS_latin1(&value), "HorizSpace");
8250 ANYOF_CLASS_SET(ret, ANYOF_ASCII);
8253 for (value = 0; value < 128; value++)
8254 ANYOF_BITMAP_SET(ret, value);
8256 for (value = 0; value < 256; value++) {
8258 ANYOF_BITMAP_SET(ret, value);
8267 ANYOF_CLASS_SET(ret, ANYOF_NASCII);
8270 for (value = 128; value < 256; value++)
8271 ANYOF_BITMAP_SET(ret, value);
8273 for (value = 0; value < 256; value++) {
8274 if (!isASCII(value))
8275 ANYOF_BITMAP_SET(ret, value);
8284 ANYOF_CLASS_SET(ret, ANYOF_DIGIT);
8286 /* consecutive digits assumed */
8287 for (value = '0'; value <= '9'; value++)
8288 ANYOF_BITMAP_SET(ret, value);
8291 what = POSIX_CC_UNI_NAME("Digit");
8295 ANYOF_CLASS_SET(ret, ANYOF_NDIGIT);
8297 /* consecutive digits assumed */
8298 for (value = 0; value < '0'; value++)
8299 ANYOF_BITMAP_SET(ret, value);
8300 for (value = '9' + 1; value < 256; value++)
8301 ANYOF_BITMAP_SET(ret, value);
8304 what = POSIX_CC_UNI_NAME("Digit");
8307 /* this is to handle \p and \P */
8310 vFAIL("Invalid [::] class");
8314 /* Strings such as "+utf8::isWord\n" */
8315 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::Is%s\n", yesno, what);
8318 ANYOF_FLAGS(ret) |= ANYOF_CLASS;
8321 } /* end of namedclass \blah */
8324 if (prevvalue > (IV)value) /* b-a */ {
8325 const int w = RExC_parse - rangebegin;
8326 Simple_vFAIL4("Invalid [] range \"%*.*s\"", w, w, rangebegin);
8327 range = 0; /* not a valid range */
8331 prevvalue = value; /* save the beginning of the range */
8332 if (*RExC_parse == '-' && RExC_parse+1 < RExC_end &&
8333 RExC_parse[1] != ']') {
8336 /* a bad range like \w-, [:word:]- ? */
8337 if (namedclass > OOB_NAMEDCLASS) {
8338 if (ckWARN(WARN_REGEXP)) {
8340 RExC_parse >= rangebegin ?
8341 RExC_parse - rangebegin : 0;
8343 "False [] range \"%*.*s\"",
8347 ANYOF_BITMAP_SET(ret, '-');
8349 range = 1; /* yeah, it's a range! */
8350 continue; /* but do it the next time */
8354 /* now is the next time */
8355 /*stored += (value - prevvalue + 1);*/
8357 if (prevvalue < 256) {
8358 const IV ceilvalue = value < 256 ? value : 255;
8361 /* In EBCDIC [\x89-\x91] should include
8362 * the \x8e but [i-j] should not. */
8363 if (literal_endpoint == 2 &&
8364 ((isLOWER(prevvalue) && isLOWER(ceilvalue)) ||
8365 (isUPPER(prevvalue) && isUPPER(ceilvalue))))
8367 if (isLOWER(prevvalue)) {
8368 for (i = prevvalue; i <= ceilvalue; i++)
8369 if (isLOWER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8371 ANYOF_BITMAP_SET(ret, i);
8374 for (i = prevvalue; i <= ceilvalue; i++)
8375 if (isUPPER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8377 ANYOF_BITMAP_SET(ret, i);
8383 for (i = prevvalue; i <= ceilvalue; i++) {
8384 if (!ANYOF_BITMAP_TEST(ret,i)) {
8386 ANYOF_BITMAP_SET(ret, i);
8390 if (value > 255 || UTF) {
8391 const UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
8392 const UV natvalue = NATIVE_TO_UNI(value);
8393 stored+=2; /* can't optimize this class */
8394 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8395 if (prevnatvalue < natvalue) { /* what about > ? */
8396 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\t%04"UVxf"\n",
8397 prevnatvalue, natvalue);
8399 else if (prevnatvalue == natvalue) {
8400 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n", natvalue);
8402 U8 foldbuf[UTF8_MAXBYTES_CASE+1];
8404 const UV f = to_uni_fold(natvalue, foldbuf, &foldlen);
8406 #ifdef EBCDIC /* RD t/uni/fold ff and 6b */
8407 if (RExC_precomp[0] == ':' &&
8408 RExC_precomp[1] == '[' &&
8409 (f == 0xDF || f == 0x92)) {
8410 f = NATIVE_TO_UNI(f);
8413 /* If folding and foldable and a single
8414 * character, insert also the folded version
8415 * to the charclass. */
8417 #ifdef EBCDIC /* RD tunifold ligatures s,t fb05, fb06 */
8418 if ((RExC_precomp[0] == ':' &&
8419 RExC_precomp[1] == '[' &&
8421 (value == 0xFB05 || value == 0xFB06))) ?
8422 foldlen == ((STRLEN)UNISKIP(f) - 1) :
8423 foldlen == (STRLEN)UNISKIP(f) )
8425 if (foldlen == (STRLEN)UNISKIP(f))
8427 Perl_sv_catpvf(aTHX_ listsv,
8430 /* Any multicharacter foldings
8431 * require the following transform:
8432 * [ABCDEF] -> (?:[ABCabcDEFd]|pq|rst)
8433 * where E folds into "pq" and F folds
8434 * into "rst", all other characters
8435 * fold to single characters. We save
8436 * away these multicharacter foldings,
8437 * to be later saved as part of the
8438 * additional "s" data. */
8441 if (!unicode_alternate)
8442 unicode_alternate = newAV();
8443 sv = newSVpvn_utf8((char*)foldbuf, foldlen,
8445 av_push(unicode_alternate, sv);
8449 /* If folding and the value is one of the Greek
8450 * sigmas insert a few more sigmas to make the
8451 * folding rules of the sigmas to work right.
8452 * Note that not all the possible combinations
8453 * are handled here: some of them are handled
8454 * by the standard folding rules, and some of
8455 * them (literal or EXACTF cases) are handled
8456 * during runtime in regexec.c:S_find_byclass(). */
8457 if (value == UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA) {
8458 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8459 (UV)UNICODE_GREEK_CAPITAL_LETTER_SIGMA);
8460 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8461 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8463 else if (value == UNICODE_GREEK_CAPITAL_LETTER_SIGMA)
8464 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8465 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8470 literal_endpoint = 0;
8474 range = 0; /* this range (if it was one) is done now */
8478 ANYOF_FLAGS(ret) |= ANYOF_LARGE;
8480 RExC_size += ANYOF_CLASS_ADD_SKIP;
8482 RExC_emit += ANYOF_CLASS_ADD_SKIP;
8488 /****** !SIZE_ONLY AFTER HERE *********/
8490 if( stored == 1 && (value < 128 || (value < 256 && !UTF))
8491 && !( ANYOF_FLAGS(ret) & ( ANYOF_FLAGS_ALL ^ ANYOF_FOLD ) )
8493 /* optimize single char class to an EXACT node
8494 but *only* when its not a UTF/high char */
8495 const char * cur_parse= RExC_parse;
8496 RExC_emit = (regnode *)orig_emit;
8497 RExC_parse = (char *)orig_parse;
8498 ret = reg_node(pRExC_state,
8499 (U8)((ANYOF_FLAGS(ret) & ANYOF_FOLD) ? EXACTF : EXACT));
8500 RExC_parse = (char *)cur_parse;
8501 *STRING(ret)= (char)value;
8503 RExC_emit += STR_SZ(1);
8504 SvREFCNT_dec(listsv);
8507 /* optimize case-insensitive simple patterns (e.g. /[a-z]/i) */
8508 if ( /* If the only flag is folding (plus possibly inversion). */
8509 ((ANYOF_FLAGS(ret) & (ANYOF_FLAGS_ALL ^ ANYOF_INVERT)) == ANYOF_FOLD)
8511 for (value = 0; value < 256; ++value) {
8512 if (ANYOF_BITMAP_TEST(ret, value)) {
8513 UV fold = PL_fold[value];
8516 ANYOF_BITMAP_SET(ret, fold);
8519 ANYOF_FLAGS(ret) &= ~ANYOF_FOLD;
8522 /* optimize inverted simple patterns (e.g. [^a-z]) */
8523 if (optimize_invert &&
8524 /* If the only flag is inversion. */
8525 (ANYOF_FLAGS(ret) & ANYOF_FLAGS_ALL) == ANYOF_INVERT) {
8526 for (value = 0; value < ANYOF_BITMAP_SIZE; ++value)
8527 ANYOF_BITMAP(ret)[value] ^= ANYOF_FLAGS_ALL;
8528 ANYOF_FLAGS(ret) = ANYOF_UNICODE_ALL;
8531 AV * const av = newAV();
8533 /* The 0th element stores the character class description
8534 * in its textual form: used later (regexec.c:Perl_regclass_swash())
8535 * to initialize the appropriate swash (which gets stored in
8536 * the 1st element), and also useful for dumping the regnode.
8537 * The 2nd element stores the multicharacter foldings,
8538 * used later (regexec.c:S_reginclass()). */
8539 av_store(av, 0, listsv);
8540 av_store(av, 1, NULL);
8541 av_store(av, 2, MUTABLE_SV(unicode_alternate));
8542 rv = newRV_noinc(MUTABLE_SV(av));
8543 n = add_data(pRExC_state, 1, "s");
8544 RExC_rxi->data->data[n] = (void*)rv;
8552 /* reg_skipcomment()
8554 Absorbs an /x style # comments from the input stream.
8555 Returns true if there is more text remaining in the stream.
8556 Will set the REG_SEEN_RUN_ON_COMMENT flag if the comment
8557 terminates the pattern without including a newline.
8559 Note its the callers responsibility to ensure that we are
8565 S_reg_skipcomment(pTHX_ RExC_state_t *pRExC_state)
8569 PERL_ARGS_ASSERT_REG_SKIPCOMMENT;
8571 while (RExC_parse < RExC_end)
8572 if (*RExC_parse++ == '\n') {
8577 /* we ran off the end of the pattern without ending
8578 the comment, so we have to add an \n when wrapping */
8579 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
8587 Advance that parse position, and optionall absorbs
8588 "whitespace" from the inputstream.
8590 Without /x "whitespace" means (?#...) style comments only,
8591 with /x this means (?#...) and # comments and whitespace proper.
8593 Returns the RExC_parse point from BEFORE the scan occurs.
8595 This is the /x friendly way of saying RExC_parse++.
8599 S_nextchar(pTHX_ RExC_state_t *pRExC_state)
8601 char* const retval = RExC_parse++;
8603 PERL_ARGS_ASSERT_NEXTCHAR;
8606 if (*RExC_parse == '(' && RExC_parse[1] == '?' &&
8607 RExC_parse[2] == '#') {
8608 while (*RExC_parse != ')') {
8609 if (RExC_parse == RExC_end)
8610 FAIL("Sequence (?#... not terminated");
8616 if (RExC_flags & RXf_PMf_EXTENDED) {
8617 if (isSPACE(*RExC_parse)) {
8621 else if (*RExC_parse == '#') {
8622 if ( reg_skipcomment( pRExC_state ) )
8631 - reg_node - emit a node
8633 STATIC regnode * /* Location. */
8634 S_reg_node(pTHX_ RExC_state_t *pRExC_state, U8 op)
8637 register regnode *ptr;
8638 regnode * const ret = RExC_emit;
8639 GET_RE_DEBUG_FLAGS_DECL;
8641 PERL_ARGS_ASSERT_REG_NODE;
8644 SIZE_ALIGN(RExC_size);
8648 if (RExC_emit >= RExC_emit_bound)
8649 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8651 NODE_ALIGN_FILL(ret);
8653 FILL_ADVANCE_NODE(ptr, op);
8654 REH_CALL_REGCOMP_HOOK(pRExC_state->rx, (ptr) - 1);
8655 #ifdef RE_TRACK_PATTERN_OFFSETS
8656 if (RExC_offsets) { /* MJD */
8657 MJD_OFFSET_DEBUG(("%s:%d: (op %s) %s %"UVuf" (len %"UVuf") (max %"UVuf").\n",
8658 "reg_node", __LINE__,
8660 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0]
8661 ? "Overwriting end of array!\n" : "OK",
8662 (UV)(RExC_emit - RExC_emit_start),
8663 (UV)(RExC_parse - RExC_start),
8664 (UV)RExC_offsets[0]));
8665 Set_Node_Offset(RExC_emit, RExC_parse + (op == END));
8673 - reganode - emit a node with an argument
8675 STATIC regnode * /* Location. */
8676 S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
8679 register regnode *ptr;
8680 regnode * const ret = RExC_emit;
8681 GET_RE_DEBUG_FLAGS_DECL;
8683 PERL_ARGS_ASSERT_REGANODE;
8686 SIZE_ALIGN(RExC_size);
8691 assert(2==regarglen[op]+1);
8693 Anything larger than this has to allocate the extra amount.
8694 If we changed this to be:
8696 RExC_size += (1 + regarglen[op]);
8698 then it wouldn't matter. Its not clear what side effect
8699 might come from that so its not done so far.
8704 if (RExC_emit >= RExC_emit_bound)
8705 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8707 NODE_ALIGN_FILL(ret);
8709 FILL_ADVANCE_NODE_ARG(ptr, op, arg);
8710 REH_CALL_REGCOMP_HOOK(pRExC_state->rx, (ptr) - 2);
8711 #ifdef RE_TRACK_PATTERN_OFFSETS
8712 if (RExC_offsets) { /* MJD */
8713 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8717 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0] ?
8718 "Overwriting end of array!\n" : "OK",
8719 (UV)(RExC_emit - RExC_emit_start),
8720 (UV)(RExC_parse - RExC_start),
8721 (UV)RExC_offsets[0]));
8722 Set_Cur_Node_Offset;
8730 - reguni - emit (if appropriate) a Unicode character
8733 S_reguni(pTHX_ const RExC_state_t *pRExC_state, UV uv, char* s)
8737 PERL_ARGS_ASSERT_REGUNI;
8739 return SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
8743 - reginsert - insert an operator in front of already-emitted operand
8745 * Means relocating the operand.
8748 S_reginsert(pTHX_ RExC_state_t *pRExC_state, U8 op, regnode *opnd, U32 depth)
8751 register regnode *src;
8752 register regnode *dst;
8753 register regnode *place;
8754 const int offset = regarglen[(U8)op];
8755 const int size = NODE_STEP_REGNODE + offset;
8756 GET_RE_DEBUG_FLAGS_DECL;
8758 PERL_ARGS_ASSERT_REGINSERT;
8759 PERL_UNUSED_ARG(depth);
8760 /* (PL_regkind[(U8)op] == CURLY ? EXTRA_STEP_2ARGS : 0); */
8761 DEBUG_PARSE_FMT("inst"," - %s",PL_reg_name[op]);
8770 if (RExC_open_parens) {
8772 /*DEBUG_PARSE_FMT("inst"," - %"IVdf, (IV)RExC_npar);*/
8773 for ( paren=0 ; paren < RExC_npar ; paren++ ) {
8774 if ( RExC_open_parens[paren] >= opnd ) {
8775 /*DEBUG_PARSE_FMT("open"," - %d",size);*/
8776 RExC_open_parens[paren] += size;
8778 /*DEBUG_PARSE_FMT("open"," - %s","ok");*/
8780 if ( RExC_close_parens[paren] >= opnd ) {
8781 /*DEBUG_PARSE_FMT("close"," - %d",size);*/
8782 RExC_close_parens[paren] += size;
8784 /*DEBUG_PARSE_FMT("close"," - %s","ok");*/
8789 while (src > opnd) {
8790 StructCopy(--src, --dst, regnode);
8791 #ifdef RE_TRACK_PATTERN_OFFSETS
8792 if (RExC_offsets) { /* MJD 20010112 */
8793 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s copy %"UVuf" -> %"UVuf" (max %"UVuf").\n",
8797 (UV)(dst - RExC_emit_start) > RExC_offsets[0]
8798 ? "Overwriting end of array!\n" : "OK",
8799 (UV)(src - RExC_emit_start),
8800 (UV)(dst - RExC_emit_start),
8801 (UV)RExC_offsets[0]));
8802 Set_Node_Offset_To_R(dst-RExC_emit_start, Node_Offset(src));
8803 Set_Node_Length_To_R(dst-RExC_emit_start, Node_Length(src));
8809 place = opnd; /* Op node, where operand used to be. */
8810 #ifdef RE_TRACK_PATTERN_OFFSETS
8811 if (RExC_offsets) { /* MJD */
8812 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8816 (UV)(place - RExC_emit_start) > RExC_offsets[0]
8817 ? "Overwriting end of array!\n" : "OK",
8818 (UV)(place - RExC_emit_start),
8819 (UV)(RExC_parse - RExC_start),
8820 (UV)RExC_offsets[0]));
8821 Set_Node_Offset(place, RExC_parse);
8822 Set_Node_Length(place, 1);
8825 src = NEXTOPER(place);
8826 FILL_ADVANCE_NODE(place, op);
8827 REH_CALL_REGCOMP_HOOK(pRExC_state->rx, (place) - 1);
8828 Zero(src, offset, regnode);
8832 - regtail - set the next-pointer at the end of a node chain of p to val.
8833 - SEE ALSO: regtail_study
8835 /* TODO: All three parms should be const */
8837 S_regtail(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8840 register regnode *scan;
8841 GET_RE_DEBUG_FLAGS_DECL;
8843 PERL_ARGS_ASSERT_REGTAIL;
8845 PERL_UNUSED_ARG(depth);
8851 /* Find last node. */
8854 regnode * const temp = regnext(scan);
8856 SV * const mysv=sv_newmortal();
8857 DEBUG_PARSE_MSG((scan==p ? "tail" : ""));
8858 regprop(RExC_rx, mysv, scan);
8859 PerlIO_printf(Perl_debug_log, "~ %s (%d) %s %s\n",
8860 SvPV_nolen_const(mysv), REG_NODE_NUM(scan),
8861 (temp == NULL ? "->" : ""),
8862 (temp == NULL ? PL_reg_name[OP(val)] : "")
8870 if (reg_off_by_arg[OP(scan)]) {
8871 ARG_SET(scan, val - scan);
8874 NEXT_OFF(scan) = val - scan;
8880 - regtail_study - set the next-pointer at the end of a node chain of p to val.
8881 - Look for optimizable sequences at the same time.
8882 - currently only looks for EXACT chains.
8884 This is expermental code. The idea is to use this routine to perform
8885 in place optimizations on branches and groups as they are constructed,
8886 with the long term intention of removing optimization from study_chunk so
8887 that it is purely analytical.
8889 Currently only used when in DEBUG mode. The macro REGTAIL_STUDY() is used
8890 to control which is which.
8893 /* TODO: All four parms should be const */
8896 S_regtail_study(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8899 register regnode *scan;
8901 #ifdef EXPERIMENTAL_INPLACESCAN
8904 GET_RE_DEBUG_FLAGS_DECL;
8906 PERL_ARGS_ASSERT_REGTAIL_STUDY;
8912 /* Find last node. */
8916 regnode * const temp = regnext(scan);
8917 #ifdef EXPERIMENTAL_INPLACESCAN
8918 if (PL_regkind[OP(scan)] == EXACT)
8919 if (join_exact(pRExC_state,scan,&min,1,val,depth+1))
8927 if( exact == PSEUDO )
8929 else if ( exact != OP(scan) )
8938 SV * const mysv=sv_newmortal();
8939 DEBUG_PARSE_MSG((scan==p ? "tsdy" : ""));
8940 regprop(RExC_rx, mysv, scan);
8941 PerlIO_printf(Perl_debug_log, "~ %s (%d) -> %s\n",
8942 SvPV_nolen_const(mysv),
8944 PL_reg_name[exact]);
8951 SV * const mysv_val=sv_newmortal();
8952 DEBUG_PARSE_MSG("");
8953 regprop(RExC_rx, mysv_val, val);
8954 PerlIO_printf(Perl_debug_log, "~ attach to %s (%"IVdf") offset to %"IVdf"\n",
8955 SvPV_nolen_const(mysv_val),
8956 (IV)REG_NODE_NUM(val),
8960 if (reg_off_by_arg[OP(scan)]) {
8961 ARG_SET(scan, val - scan);
8964 NEXT_OFF(scan) = val - scan;
8972 - regcurly - a little FSA that accepts {\d+,?\d*}
8974 #ifndef PERL_IN_XSUB_RE
8976 Perl_regcurly(register const char *s)
8978 PERL_ARGS_ASSERT_REGCURLY;
8997 - regdump - dump a regexp onto Perl_debug_log in vaguely comprehensible form
9001 S_regdump_extflags(pTHX_ const char *lead, const U32 flags)
9006 for (bit=0; bit<32; bit++) {
9007 if (flags & (1<<bit)) {
9009 PerlIO_printf(Perl_debug_log, "%s",lead);
9010 PerlIO_printf(Perl_debug_log, "%s ",PL_reg_extflags_name[bit]);
9015 PerlIO_printf(Perl_debug_log, "\n");
9017 PerlIO_printf(Perl_debug_log, "%s[none-set]\n",lead);
9023 Perl_regdump(pTHX_ const regexp *r)
9027 SV * const sv = sv_newmortal();
9028 SV *dsv= sv_newmortal();
9030 GET_RE_DEBUG_FLAGS_DECL;
9032 PERL_ARGS_ASSERT_REGDUMP;
9034 (void)dumpuntil(r, ri->program, ri->program + 1, NULL, NULL, sv, 0, 0);
9036 /* Header fields of interest. */
9037 if (r->anchored_substr) {
9038 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->anchored_substr),
9039 RE_SV_DUMPLEN(r->anchored_substr), 30);
9040 PerlIO_printf(Perl_debug_log,
9041 "anchored %s%s at %"IVdf" ",
9042 s, RE_SV_TAIL(r->anchored_substr),
9043 (IV)r->anchored_offset);
9044 } else if (r->anchored_utf8) {
9045 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->anchored_utf8),
9046 RE_SV_DUMPLEN(r->anchored_utf8), 30);
9047 PerlIO_printf(Perl_debug_log,
9048 "anchored utf8 %s%s at %"IVdf" ",
9049 s, RE_SV_TAIL(r->anchored_utf8),
9050 (IV)r->anchored_offset);
9052 if (r->float_substr) {
9053 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->float_substr),
9054 RE_SV_DUMPLEN(r->float_substr), 30);
9055 PerlIO_printf(Perl_debug_log,
9056 "floating %s%s at %"IVdf"..%"UVuf" ",
9057 s, RE_SV_TAIL(r->float_substr),
9058 (IV)r->float_min_offset, (UV)r->float_max_offset);
9059 } else if (r->float_utf8) {
9060 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->float_utf8),
9061 RE_SV_DUMPLEN(r->float_utf8), 30);
9062 PerlIO_printf(Perl_debug_log,
9063 "floating utf8 %s%s at %"IVdf"..%"UVuf" ",
9064 s, RE_SV_TAIL(r->float_utf8),
9065 (IV)r->float_min_offset, (UV)r->float_max_offset);
9067 if (r->check_substr || r->check_utf8)
9068 PerlIO_printf(Perl_debug_log,
9070 (r->check_substr == r->float_substr
9071 && r->check_utf8 == r->float_utf8
9072 ? "(checking floating" : "(checking anchored"));
9073 if (r->extflags & RXf_NOSCAN)
9074 PerlIO_printf(Perl_debug_log, " noscan");
9075 if (r->extflags & RXf_CHECK_ALL)
9076 PerlIO_printf(Perl_debug_log, " isall");
9077 if (r->check_substr || r->check_utf8)
9078 PerlIO_printf(Perl_debug_log, ") ");
9080 if (ri->regstclass) {
9081 regprop(r, sv, ri->regstclass);
9082 PerlIO_printf(Perl_debug_log, "stclass %s ", SvPVX_const(sv));
9084 if (r->extflags & RXf_ANCH) {
9085 PerlIO_printf(Perl_debug_log, "anchored");
9086 if (r->extflags & RXf_ANCH_BOL)
9087 PerlIO_printf(Perl_debug_log, "(BOL)");
9088 if (r->extflags & RXf_ANCH_MBOL)
9089 PerlIO_printf(Perl_debug_log, "(MBOL)");
9090 if (r->extflags & RXf_ANCH_SBOL)
9091 PerlIO_printf(Perl_debug_log, "(SBOL)");
9092 if (r->extflags & RXf_ANCH_GPOS)
9093 PerlIO_printf(Perl_debug_log, "(GPOS)");
9094 PerlIO_putc(Perl_debug_log, ' ');
9096 if (r->extflags & RXf_GPOS_SEEN)
9097 PerlIO_printf(Perl_debug_log, "GPOS:%"UVuf" ", (UV)r->gofs);
9098 if (r->intflags & PREGf_SKIP)
9099 PerlIO_printf(Perl_debug_log, "plus ");
9100 if (r->intflags & PREGf_IMPLICIT)
9101 PerlIO_printf(Perl_debug_log, "implicit ");
9102 PerlIO_printf(Perl_debug_log, "minlen %"IVdf" ", (IV)r->minlen);
9103 if (r->extflags & RXf_EVAL_SEEN)
9104 PerlIO_printf(Perl_debug_log, "with eval ");
9105 PerlIO_printf(Perl_debug_log, "\n");
9106 DEBUG_FLAGS_r(regdump_extflags("r->extflags: ",r->extflags));
9108 PERL_ARGS_ASSERT_REGDUMP;
9109 PERL_UNUSED_CONTEXT;
9111 #endif /* DEBUGGING */
9115 - regprop - printable representation of opcode
9117 #define EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags) \
9120 Perl_sv_catpvf(aTHX_ sv,"%s][%s",PL_colors[1],PL_colors[0]); \
9121 if (flags & ANYOF_INVERT) \
9122 /*make sure the invert info is in each */ \
9123 sv_catpvs(sv, "^"); \
9129 Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
9134 RXi_GET_DECL(prog,progi);
9135 GET_RE_DEBUG_FLAGS_DECL;
9137 PERL_ARGS_ASSERT_REGPROP;
9141 if (OP(o) > REGNODE_MAX) /* regnode.type is unsigned */
9142 /* It would be nice to FAIL() here, but this may be called from
9143 regexec.c, and it would be hard to supply pRExC_state. */
9144 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(o), (int)REGNODE_MAX);
9145 sv_catpv(sv, PL_reg_name[OP(o)]); /* Take off const! */
9147 k = PL_regkind[OP(o)];
9151 /* Using is_utf8_string() (via PERL_PV_UNI_DETECT)
9152 * is a crude hack but it may be the best for now since
9153 * we have no flag "this EXACTish node was UTF-8"
9155 pv_pretty(sv, STRING(o), STR_LEN(o), 60, PL_colors[0], PL_colors[1],
9156 PERL_PV_ESCAPE_UNI_DETECT |
9157 PERL_PV_PRETTY_ELLIPSES |
9158 PERL_PV_PRETTY_LTGT |
9159 PERL_PV_PRETTY_NOCLEAR
9161 } else if (k == TRIE) {
9162 /* print the details of the trie in dumpuntil instead, as
9163 * progi->data isn't available here */
9164 const char op = OP(o);
9165 const U32 n = ARG(o);
9166 const reg_ac_data * const ac = IS_TRIE_AC(op) ?
9167 (reg_ac_data *)progi->data->data[n] :
9169 const reg_trie_data * const trie
9170 = (reg_trie_data*)progi->data->data[!IS_TRIE_AC(op) ? n : ac->trie];
9172 Perl_sv_catpvf(aTHX_ sv, "-%s",PL_reg_name[o->flags]);
9173 DEBUG_TRIE_COMPILE_r(
9174 Perl_sv_catpvf(aTHX_ sv,
9175 "<S:%"UVuf"/%"IVdf" W:%"UVuf" L:%"UVuf"/%"UVuf" C:%"UVuf"/%"UVuf">",
9176 (UV)trie->startstate,
9177 (IV)trie->statecount-1, /* -1 because of the unused 0 element */
9178 (UV)trie->wordcount,
9181 (UV)TRIE_CHARCOUNT(trie),
9182 (UV)trie->uniquecharcount
9185 if ( IS_ANYOF_TRIE(op) || trie->bitmap ) {
9187 int rangestart = -1;
9188 U8* bitmap = IS_ANYOF_TRIE(op) ? (U8*)ANYOF_BITMAP(o) : (U8*)TRIE_BITMAP(trie);
9190 for (i = 0; i <= 256; i++) {
9191 if (i < 256 && BITMAP_TEST(bitmap,i)) {
9192 if (rangestart == -1)
9194 } else if (rangestart != -1) {
9195 if (i <= rangestart + 3)
9196 for (; rangestart < i; rangestart++)
9197 put_byte(sv, rangestart);
9199 put_byte(sv, rangestart);
9201 put_byte(sv, i - 1);
9209 } else if (k == CURLY) {
9210 if (OP(o) == CURLYM || OP(o) == CURLYN || OP(o) == CURLYX)
9211 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* Parenth number */
9212 Perl_sv_catpvf(aTHX_ sv, " {%d,%d}", ARG1(o), ARG2(o));
9214 else if (k == WHILEM && o->flags) /* Ordinal/of */
9215 Perl_sv_catpvf(aTHX_ sv, "[%d/%d]", o->flags & 0xf, o->flags>>4);
9216 else if (k == REF || k == OPEN || k == CLOSE || k == GROUPP || OP(o)==ACCEPT) {
9217 Perl_sv_catpvf(aTHX_ sv, "%d", (int)ARG(o)); /* Parenth number */
9218 if ( RXp_PAREN_NAMES(prog) ) {
9219 if ( k != REF || OP(o) < NREF) {
9220 AV *list= MUTABLE_AV(progi->data->data[progi->name_list_idx]);
9221 SV **name= av_fetch(list, ARG(o), 0 );
9223 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9226 AV *list= MUTABLE_AV(progi->data->data[ progi->name_list_idx ]);
9227 SV *sv_dat= MUTABLE_SV(progi->data->data[ ARG( o ) ]);
9228 I32 *nums=(I32*)SvPVX(sv_dat);
9229 SV **name= av_fetch(list, nums[0], 0 );
9232 for ( n=0; n<SvIVX(sv_dat); n++ ) {
9233 Perl_sv_catpvf(aTHX_ sv, "%s%"IVdf,
9234 (n ? "," : ""), (IV)nums[n]);
9236 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9240 } else if (k == GOSUB)
9241 Perl_sv_catpvf(aTHX_ sv, "%d[%+d]", (int)ARG(o),(int)ARG2L(o)); /* Paren and offset */
9242 else if (k == VERB) {
9244 Perl_sv_catpvf(aTHX_ sv, ":%"SVf,
9245 SVfARG((MUTABLE_SV(progi->data->data[ ARG( o ) ]))));
9246 } else if (k == LOGICAL)
9247 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* 2: embedded, otherwise 1 */
9248 else if (k == FOLDCHAR)
9249 Perl_sv_catpvf(aTHX_ sv, "[0x%"UVXf"]", PTR2UV(ARG(o)) );
9250 else if (k == ANYOF) {
9251 int i, rangestart = -1;
9252 const U8 flags = ANYOF_FLAGS(o);
9255 /* Should be synchronized with * ANYOF_ #xdefines in regcomp.h */
9256 static const char * const anyofs[] = {
9289 if (flags & ANYOF_LOCALE)
9290 sv_catpvs(sv, "{loc}");
9291 if (flags & ANYOF_FOLD)
9292 sv_catpvs(sv, "{i}");
9293 Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
9294 if (flags & ANYOF_INVERT)
9297 /* output what the standard cp 0-255 bitmap matches */
9298 for (i = 0; i <= 256; i++) {
9299 if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
9300 if (rangestart == -1)
9302 } else if (rangestart != -1) {
9303 if (i <= rangestart + 3)
9304 for (; rangestart < i; rangestart++)
9305 put_byte(sv, rangestart);
9307 put_byte(sv, rangestart);
9309 put_byte(sv, i - 1);
9316 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9317 /* output any special charclass tests (used mostly under use locale) */
9318 if (o->flags & ANYOF_CLASS)
9319 for (i = 0; i < (int)(sizeof(anyofs)/sizeof(char*)); i++)
9320 if (ANYOF_CLASS_TEST(o,i)) {
9321 sv_catpv(sv, anyofs[i]);
9325 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9327 /* output information about the unicode matching */
9328 if (flags & ANYOF_UNICODE)
9329 sv_catpvs(sv, "{unicode}");
9330 else if (flags & ANYOF_UNICODE_ALL)
9331 sv_catpvs(sv, "{unicode_all}");
9335 SV * const sw = regclass_swash(prog, o, FALSE, &lv, 0);
9339 U8 s[UTF8_MAXBYTES_CASE+1];
9341 for (i = 0; i <= 256; i++) { /* just the first 256 */
9342 uvchr_to_utf8(s, i);
9344 if (i < 256 && swash_fetch(sw, s, TRUE)) {
9345 if (rangestart == -1)
9347 } else if (rangestart != -1) {
9348 if (i <= rangestart + 3)
9349 for (; rangestart < i; rangestart++) {
9350 const U8 * const e = uvchr_to_utf8(s,rangestart);
9352 for(p = s; p < e; p++)
9356 const U8 *e = uvchr_to_utf8(s,rangestart);
9358 for (p = s; p < e; p++)
9361 e = uvchr_to_utf8(s, i-1);
9362 for (p = s; p < e; p++)
9369 sv_catpvs(sv, "..."); /* et cetera */
9373 char *s = savesvpv(lv);
9374 char * const origs = s;
9376 while (*s && *s != '\n')
9380 const char * const t = ++s;
9398 Perl_sv_catpvf(aTHX_ sv, "%s]", PL_colors[1]);
9400 else if (k == BRANCHJ && (OP(o) == UNLESSM || OP(o) == IFMATCH))
9401 Perl_sv_catpvf(aTHX_ sv, "[%d]", -(o->flags));
9403 PERL_UNUSED_CONTEXT;
9404 PERL_UNUSED_ARG(sv);
9406 PERL_UNUSED_ARG(prog);
9407 #endif /* DEBUGGING */
9411 Perl_re_intuit_string(pTHX_ REGEXP * const r)
9412 { /* Assume that RE_INTUIT is set */
9414 struct regexp *const prog = (struct regexp *)SvANY(r);
9415 GET_RE_DEBUG_FLAGS_DECL;
9417 PERL_ARGS_ASSERT_RE_INTUIT_STRING;
9418 PERL_UNUSED_CONTEXT;
9422 const char * const s = SvPV_nolen_const(prog->check_substr
9423 ? prog->check_substr : prog->check_utf8);
9425 if (!PL_colorset) reginitcolors();
9426 PerlIO_printf(Perl_debug_log,
9427 "%sUsing REx %ssubstr:%s \"%s%.60s%s%s\"\n",
9429 prog->check_substr ? "" : "utf8 ",
9430 PL_colors[5],PL_colors[0],
9433 (strlen(s) > 60 ? "..." : ""));
9436 return prog->check_substr ? prog->check_substr : prog->check_utf8;
9442 handles refcounting and freeing the perl core regexp structure. When
9443 it is necessary to actually free the structure the first thing it
9444 does is call the 'free' method of the regexp_engine associated to to
9445 the regexp, allowing the handling of the void *pprivate; member
9446 first. (This routine is not overridable by extensions, which is why
9447 the extensions free is called first.)
9449 See regdupe and regdupe_internal if you change anything here.
9451 #ifndef PERL_IN_XSUB_RE
9453 Perl_pregfree(pTHX_ REGEXP *r)
9459 Perl_pregfree2(pTHX_ REGEXP *rx)
9462 struct regexp *const r = (struct regexp *)SvANY(rx);
9463 GET_RE_DEBUG_FLAGS_DECL;
9465 PERL_ARGS_ASSERT_PREGFREE2;
9468 ReREFCNT_dec(r->mother_re);
9470 CALLREGFREE_PVT(rx); /* free the private data */
9471 SvREFCNT_dec(RXp_PAREN_NAMES(r));
9474 SvREFCNT_dec(r->anchored_substr);
9475 SvREFCNT_dec(r->anchored_utf8);
9476 SvREFCNT_dec(r->float_substr);
9477 SvREFCNT_dec(r->float_utf8);
9478 Safefree(r->substrs);
9480 RX_MATCH_COPY_FREE(rx);
9481 #ifdef PERL_OLD_COPY_ON_WRITE
9482 SvREFCNT_dec(r->saved_copy);
9489 This is a hacky workaround to the structural issue of match results
9490 being stored in the regexp structure which is in turn stored in
9491 PL_curpm/PL_reg_curpm. The problem is that due to qr// the pattern
9492 could be PL_curpm in multiple contexts, and could require multiple
9493 result sets being associated with the pattern simultaneously, such
9494 as when doing a recursive match with (??{$qr})
9496 The solution is to make a lightweight copy of the regexp structure
9497 when a qr// is returned from the code executed by (??{$qr}) this
9498 lightweight copy doesnt actually own any of its data except for
9499 the starp/end and the actual regexp structure itself.
9505 Perl_reg_temp_copy (pTHX_ REGEXP *ret_x, REGEXP *rx)
9508 struct regexp *const r = (struct regexp *)SvANY(rx);
9509 register const I32 npar = r->nparens+1;
9511 PERL_ARGS_ASSERT_REG_TEMP_COPY;
9514 ret_x = (REGEXP*) newSV_type(SVt_REGEXP);
9515 ret = (struct regexp *)SvANY(ret_x);
9517 (void)ReREFCNT_inc(rx);
9518 /* We can take advantage of the existing "copied buffer" mechanism in SVs
9519 by pointing directly at the buffer, but flagging that the allocated
9520 space in the copy is zero. As we've just done a struct copy, it's now
9521 a case of zero-ing that, rather than copying the current length. */
9522 SvPV_set(ret_x, RX_WRAPPED(rx));
9523 SvFLAGS(ret_x) |= SvFLAGS(rx) & (SVf_POK|SVp_POK|SVf_UTF8);
9524 memcpy(&(ret->xpv_cur), &(r->xpv_cur),
9525 sizeof(regexp) - STRUCT_OFFSET(regexp, xpv_cur));
9526 SvLEN_set(ret_x, 0);
9527 SvSTASH_set(ret_x, NULL);
9528 SvMAGIC_set(ret_x, NULL);
9529 Newx(ret->offs, npar, regexp_paren_pair);
9530 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9532 Newx(ret->substrs, 1, struct reg_substr_data);
9533 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9535 SvREFCNT_inc_void(ret->anchored_substr);
9536 SvREFCNT_inc_void(ret->anchored_utf8);
9537 SvREFCNT_inc_void(ret->float_substr);
9538 SvREFCNT_inc_void(ret->float_utf8);
9540 /* check_substr and check_utf8, if non-NULL, point to either their
9541 anchored or float namesakes, and don't hold a second reference. */
9543 RX_MATCH_COPIED_off(ret_x);
9544 #ifdef PERL_OLD_COPY_ON_WRITE
9545 ret->saved_copy = NULL;
9547 ret->mother_re = rx;
9553 /* regfree_internal()
9555 Free the private data in a regexp. This is overloadable by
9556 extensions. Perl takes care of the regexp structure in pregfree(),
9557 this covers the *pprivate pointer which technically perldoesnt
9558 know about, however of course we have to handle the
9559 regexp_internal structure when no extension is in use.
9561 Note this is called before freeing anything in the regexp
9566 Perl_regfree_internal(pTHX_ REGEXP * const rx)
9569 struct regexp *const r = (struct regexp *)SvANY(rx);
9571 GET_RE_DEBUG_FLAGS_DECL;
9573 PERL_ARGS_ASSERT_REGFREE_INTERNAL;
9579 SV *dsv= sv_newmortal();
9580 RE_PV_QUOTED_DECL(s, RX_UTF8(rx),
9581 dsv, RX_PRECOMP(rx), RX_PRELEN(rx), 60);
9582 PerlIO_printf(Perl_debug_log,"%sFreeing REx:%s %s\n",
9583 PL_colors[4],PL_colors[5],s);
9586 #ifdef RE_TRACK_PATTERN_OFFSETS
9588 Safefree(ri->u.offsets); /* 20010421 MJD */
9591 int n = ri->data->count;
9592 PAD* new_comppad = NULL;
9597 /* If you add a ->what type here, update the comment in regcomp.h */
9598 switch (ri->data->what[n]) {
9603 SvREFCNT_dec(MUTABLE_SV(ri->data->data[n]));
9606 Safefree(ri->data->data[n]);
9609 new_comppad = MUTABLE_AV(ri->data->data[n]);
9612 if (new_comppad == NULL)
9613 Perl_croak(aTHX_ "panic: pregfree comppad");
9614 PAD_SAVE_LOCAL(old_comppad,
9615 /* Watch out for global destruction's random ordering. */
9616 (SvTYPE(new_comppad) == SVt_PVAV) ? new_comppad : NULL
9619 refcnt = OpREFCNT_dec((OP_4tree*)ri->data->data[n]);
9622 op_free((OP_4tree*)ri->data->data[n]);
9624 PAD_RESTORE_LOCAL(old_comppad);
9625 SvREFCNT_dec(MUTABLE_SV(new_comppad));
9631 { /* Aho Corasick add-on structure for a trie node.
9632 Used in stclass optimization only */
9634 reg_ac_data *aho=(reg_ac_data*)ri->data->data[n];
9636 refcount = --aho->refcount;
9639 PerlMemShared_free(aho->states);
9640 PerlMemShared_free(aho->fail);
9641 /* do this last!!!! */
9642 PerlMemShared_free(ri->data->data[n]);
9643 PerlMemShared_free(ri->regstclass);
9649 /* trie structure. */
9651 reg_trie_data *trie=(reg_trie_data*)ri->data->data[n];
9653 refcount = --trie->refcount;
9656 PerlMemShared_free(trie->charmap);
9657 PerlMemShared_free(trie->states);
9658 PerlMemShared_free(trie->trans);
9660 PerlMemShared_free(trie->bitmap);
9662 PerlMemShared_free(trie->jump);
9663 PerlMemShared_free(trie->wordinfo);
9664 /* do this last!!!! */
9665 PerlMemShared_free(ri->data->data[n]);
9670 Perl_croak(aTHX_ "panic: regfree data code '%c'", ri->data->what[n]);
9673 Safefree(ri->data->what);
9680 #define av_dup_inc(s,t) MUTABLE_AV(sv_dup_inc((const SV *)s,t))
9681 #define hv_dup_inc(s,t) MUTABLE_HV(sv_dup_inc((const SV *)s,t))
9682 #define SAVEPVN(p,n) ((p) ? savepvn(p,n) : NULL)
9685 re_dup - duplicate a regexp.
9687 This routine is expected to clone a given regexp structure. It is only
9688 compiled under USE_ITHREADS.
9690 After all of the core data stored in struct regexp is duplicated
9691 the regexp_engine.dupe method is used to copy any private data
9692 stored in the *pprivate pointer. This allows extensions to handle
9693 any duplication it needs to do.
9695 See pregfree() and regfree_internal() if you change anything here.
9697 #if defined(USE_ITHREADS)
9698 #ifndef PERL_IN_XSUB_RE
9700 Perl_re_dup_guts(pTHX_ const REGEXP *sstr, REGEXP *dstr, CLONE_PARAMS *param)
9704 const struct regexp *r = (const struct regexp *)SvANY(sstr);
9705 struct regexp *ret = (struct regexp *)SvANY(dstr);
9707 PERL_ARGS_ASSERT_RE_DUP_GUTS;
9709 npar = r->nparens+1;
9710 Newx(ret->offs, npar, regexp_paren_pair);
9711 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9713 /* no need to copy these */
9714 Newx(ret->swap, npar, regexp_paren_pair);
9718 /* Do it this way to avoid reading from *r after the StructCopy().
9719 That way, if any of the sv_dup_inc()s dislodge *r from the L1
9720 cache, it doesn't matter. */
9721 const bool anchored = r->check_substr
9722 ? r->check_substr == r->anchored_substr
9723 : r->check_utf8 == r->anchored_utf8;
9724 Newx(ret->substrs, 1, struct reg_substr_data);
9725 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9727 ret->anchored_substr = sv_dup_inc(ret->anchored_substr, param);
9728 ret->anchored_utf8 = sv_dup_inc(ret->anchored_utf8, param);
9729 ret->float_substr = sv_dup_inc(ret->float_substr, param);
9730 ret->float_utf8 = sv_dup_inc(ret->float_utf8, param);
9732 /* check_substr and check_utf8, if non-NULL, point to either their
9733 anchored or float namesakes, and don't hold a second reference. */
9735 if (ret->check_substr) {
9737 assert(r->check_utf8 == r->anchored_utf8);
9738 ret->check_substr = ret->anchored_substr;
9739 ret->check_utf8 = ret->anchored_utf8;
9741 assert(r->check_substr == r->float_substr);
9742 assert(r->check_utf8 == r->float_utf8);
9743 ret->check_substr = ret->float_substr;
9744 ret->check_utf8 = ret->float_utf8;
9746 } else if (ret->check_utf8) {
9748 ret->check_utf8 = ret->anchored_utf8;
9750 ret->check_utf8 = ret->float_utf8;
9755 RXp_PAREN_NAMES(ret) = hv_dup_inc(RXp_PAREN_NAMES(ret), param);
9758 RXi_SET(ret,CALLREGDUPE_PVT(dstr,param));
9760 if (RX_MATCH_COPIED(dstr))
9761 ret->subbeg = SAVEPVN(ret->subbeg, ret->sublen);
9764 #ifdef PERL_OLD_COPY_ON_WRITE
9765 ret->saved_copy = NULL;
9768 if (ret->mother_re) {
9769 if (SvPVX_const(dstr) == SvPVX_const(ret->mother_re)) {
9770 /* Our storage points directly to our mother regexp, but that's
9771 1: a buffer in a different thread
9772 2: something we no longer hold a reference on
9773 so we need to copy it locally. */
9774 /* Note we need to sue SvCUR() on our mother_re, because it, in
9775 turn, may well be pointing to its own mother_re. */
9776 SvPV_set(dstr, SAVEPVN(SvPVX_const(ret->mother_re),
9777 SvCUR(ret->mother_re)+1));
9778 SvLEN_set(dstr, SvCUR(ret->mother_re)+1);
9780 ret->mother_re = NULL;
9784 #endif /* PERL_IN_XSUB_RE */
9789 This is the internal complement to regdupe() which is used to copy
9790 the structure pointed to by the *pprivate pointer in the regexp.
9791 This is the core version of the extension overridable cloning hook.
9792 The regexp structure being duplicated will be copied by perl prior
9793 to this and will be provided as the regexp *r argument, however
9794 with the /old/ structures pprivate pointer value. Thus this routine
9795 may override any copying normally done by perl.
9797 It returns a pointer to the new regexp_internal structure.
9801 Perl_regdupe_internal(pTHX_ REGEXP * const rx, CLONE_PARAMS *param)
9804 struct regexp *const r = (struct regexp *)SvANY(rx);
9805 regexp_internal *reti;
9809 PERL_ARGS_ASSERT_REGDUPE_INTERNAL;
9811 npar = r->nparens+1;
9814 Newxc(reti, sizeof(regexp_internal) + len*sizeof(regnode), char, regexp_internal);
9815 Copy(ri->program, reti->program, len+1, regnode);
9818 reti->regstclass = NULL;
9822 const int count = ri->data->count;
9825 Newxc(d, sizeof(struct reg_data) + count*sizeof(void *),
9826 char, struct reg_data);
9827 Newx(d->what, count, U8);
9830 for (i = 0; i < count; i++) {
9831 d->what[i] = ri->data->what[i];
9832 switch (d->what[i]) {
9833 /* legal options are one of: sSfpontTua
9834 see also regcomp.h and pregfree() */
9835 case 'a': /* actually an AV, but the dup function is identical. */
9838 case 'p': /* actually an AV, but the dup function is identical. */
9839 case 'u': /* actually an HV, but the dup function is identical. */
9840 d->data[i] = sv_dup_inc((const SV *)ri->data->data[i], param);
9843 /* This is cheating. */
9844 Newx(d->data[i], 1, struct regnode_charclass_class);
9845 StructCopy(ri->data->data[i], d->data[i],
9846 struct regnode_charclass_class);
9847 reti->regstclass = (regnode*)d->data[i];
9850 /* Compiled op trees are readonly and in shared memory,
9851 and can thus be shared without duplication. */
9853 d->data[i] = (void*)OpREFCNT_inc((OP*)ri->data->data[i]);
9857 /* Trie stclasses are readonly and can thus be shared
9858 * without duplication. We free the stclass in pregfree
9859 * when the corresponding reg_ac_data struct is freed.
9861 reti->regstclass= ri->regstclass;
9865 ((reg_trie_data*)ri->data->data[i])->refcount++;
9869 d->data[i] = ri->data->data[i];
9872 Perl_croak(aTHX_ "panic: re_dup unknown data code '%c'", ri->data->what[i]);
9881 reti->name_list_idx = ri->name_list_idx;
9883 #ifdef RE_TRACK_PATTERN_OFFSETS
9884 if (ri->u.offsets) {
9885 Newx(reti->u.offsets, 2*len+1, U32);
9886 Copy(ri->u.offsets, reti->u.offsets, 2*len+1, U32);
9889 SetProgLen(reti,len);
9895 #endif /* USE_ITHREADS */
9897 #ifndef PERL_IN_XSUB_RE
9900 - regnext - dig the "next" pointer out of a node
9903 Perl_regnext(pTHX_ register regnode *p)
9906 register I32 offset;
9911 if (OP(p) > REGNODE_MAX) { /* regnode.type is unsigned */
9912 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(p), (int)REGNODE_MAX);
9915 offset = (reg_off_by_arg[OP(p)] ? ARG(p) : NEXT_OFF(p));
9924 S_re_croak2(pTHX_ const char* pat1,const char* pat2,...)
9927 STRLEN l1 = strlen(pat1);
9928 STRLEN l2 = strlen(pat2);
9931 const char *message;
9933 PERL_ARGS_ASSERT_RE_CROAK2;
9939 Copy(pat1, buf, l1 , char);
9940 Copy(pat2, buf + l1, l2 , char);
9941 buf[l1 + l2] = '\n';
9942 buf[l1 + l2 + 1] = '\0';
9944 /* ANSI variant takes additional second argument */
9945 va_start(args, pat2);
9949 msv = vmess(buf, &args);
9951 message = SvPV_const(msv,l1);
9954 Copy(message, buf, l1 , char);
9955 buf[l1-1] = '\0'; /* Overwrite \n */
9956 Perl_croak(aTHX_ "%s", buf);
9959 /* XXX Here's a total kludge. But we need to re-enter for swash routines. */
9961 #ifndef PERL_IN_XSUB_RE
9963 Perl_save_re_context(pTHX)
9967 struct re_save_state *state;
9969 SAVEVPTR(PL_curcop);
9970 SSGROW(SAVESTACK_ALLOC_FOR_RE_SAVE_STATE + 1);
9972 state = (struct re_save_state *)(PL_savestack + PL_savestack_ix);
9973 PL_savestack_ix += SAVESTACK_ALLOC_FOR_RE_SAVE_STATE;
9974 SSPUSHUV(SAVEt_RE_STATE);
9976 Copy(&PL_reg_state, state, 1, struct re_save_state);
9978 PL_reg_start_tmp = 0;
9979 PL_reg_start_tmpl = 0;
9980 PL_reg_oldsaved = NULL;
9981 PL_reg_oldsavedlen = 0;
9983 PL_reg_leftiter = 0;
9984 PL_reg_poscache = NULL;
9985 PL_reg_poscache_size = 0;
9986 #ifdef PERL_OLD_COPY_ON_WRITE
9990 /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
9992 const REGEXP * const rx = PM_GETRE(PL_curpm);
9995 for (i = 1; i <= RX_NPARENS(rx); i++) {
9996 char digits[TYPE_CHARS(long)];
9997 const STRLEN len = my_snprintf(digits, sizeof(digits), "%lu", (long)i);
9998 GV *const *const gvp
9999 = (GV**)hv_fetch(PL_defstash, digits, len, 0);
10002 GV * const gv = *gvp;
10003 if (SvTYPE(gv) == SVt_PVGV && GvSV(gv))
10013 clear_re(pTHX_ void *r)
10016 ReREFCNT_dec((REGEXP *)r);
10022 S_put_byte(pTHX_ SV *sv, int c)
10024 PERL_ARGS_ASSERT_PUT_BYTE;
10026 /* Our definition of isPRINT() ignores locales, so only bytes that are
10027 not part of UTF-8 are considered printable. I assume that the same
10028 holds for UTF-EBCDIC.
10029 Also, code point 255 is not printable in either (it's E0 in EBCDIC,
10030 which Wikipedia says:
10032 EO, or Eight Ones, is an 8-bit EBCDIC character code represented as all
10033 ones (binary 1111 1111, hexadecimal FF). It is similar, but not
10034 identical, to the ASCII delete (DEL) or rubout control character.
10035 ) So the old condition can be simplified to !isPRINT(c) */
10037 Perl_sv_catpvf(aTHX_ sv, "\\%o", c);
10039 const char string = c;
10040 if (c == '-' || c == ']' || c == '\\' || c == '^')
10041 sv_catpvs(sv, "\\");
10042 sv_catpvn(sv, &string, 1);
10047 #define CLEAR_OPTSTART \
10048 if (optstart) STMT_START { \
10049 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log, " (%"IVdf" nodes)\n", (IV)(node - optstart))); \
10053 #define DUMPUNTIL(b,e) CLEAR_OPTSTART; node=dumpuntil(r,start,(b),(e),last,sv,indent+1,depth+1);
10055 STATIC const regnode *
10056 S_dumpuntil(pTHX_ const regexp *r, const regnode *start, const regnode *node,
10057 const regnode *last, const regnode *plast,
10058 SV* sv, I32 indent, U32 depth)
10061 register U8 op = PSEUDO; /* Arbitrary non-END op. */
10062 register const regnode *next;
10063 const regnode *optstart= NULL;
10065 RXi_GET_DECL(r,ri);
10066 GET_RE_DEBUG_FLAGS_DECL;
10068 PERL_ARGS_ASSERT_DUMPUNTIL;
10070 #ifdef DEBUG_DUMPUNTIL
10071 PerlIO_printf(Perl_debug_log, "--- %d : %d - %d - %d\n",indent,node-start,
10072 last ? last-start : 0,plast ? plast-start : 0);
10075 if (plast && plast < last)
10078 while (PL_regkind[op] != END && (!last || node < last)) {
10079 /* While that wasn't END last time... */
10082 if (op == CLOSE || op == WHILEM)
10084 next = regnext((regnode *)node);
10087 if (OP(node) == OPTIMIZED) {
10088 if (!optstart && RE_DEBUG_FLAG(RE_DEBUG_COMPILE_OPTIMISE))
10095 regprop(r, sv, node);
10096 PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
10097 (int)(2*indent + 1), "", SvPVX_const(sv));
10099 if (OP(node) != OPTIMIZED) {
10100 if (next == NULL) /* Next ptr. */
10101 PerlIO_printf(Perl_debug_log, " (0)");
10102 else if (PL_regkind[(U8)op] == BRANCH && PL_regkind[OP(next)] != BRANCH )
10103 PerlIO_printf(Perl_debug_log, " (FAIL)");
10105 PerlIO_printf(Perl_debug_log, " (%"IVdf")", (IV)(next - start));
10106 (void)PerlIO_putc(Perl_debug_log, '\n');
10110 if (PL_regkind[(U8)op] == BRANCHJ) {
10113 register const regnode *nnode = (OP(next) == LONGJMP
10114 ? regnext((regnode *)next)
10116 if (last && nnode > last)
10118 DUMPUNTIL(NEXTOPER(NEXTOPER(node)), nnode);
10121 else if (PL_regkind[(U8)op] == BRANCH) {
10123 DUMPUNTIL(NEXTOPER(node), next);
10125 else if ( PL_regkind[(U8)op] == TRIE ) {
10126 const regnode *this_trie = node;
10127 const char op = OP(node);
10128 const U32 n = ARG(node);
10129 const reg_ac_data * const ac = op>=AHOCORASICK ?
10130 (reg_ac_data *)ri->data->data[n] :
10132 const reg_trie_data * const trie =
10133 (reg_trie_data*)ri->data->data[op<AHOCORASICK ? n : ac->trie];
10135 AV *const trie_words = MUTABLE_AV(ri->data->data[n + TRIE_WORDS_OFFSET]);
10137 const regnode *nextbranch= NULL;
10140 for (word_idx= 0; word_idx < (I32)trie->wordcount; word_idx++) {
10141 SV ** const elem_ptr = av_fetch(trie_words,word_idx,0);
10143 PerlIO_printf(Perl_debug_log, "%*s%s ",
10144 (int)(2*(indent+3)), "",
10145 elem_ptr ? pv_pretty(sv, SvPV_nolen_const(*elem_ptr), SvCUR(*elem_ptr), 60,
10146 PL_colors[0], PL_colors[1],
10147 (SvUTF8(*elem_ptr) ? PERL_PV_ESCAPE_UNI : 0) |
10148 PERL_PV_PRETTY_ELLIPSES |
10149 PERL_PV_PRETTY_LTGT
10154 U16 dist= trie->jump[word_idx+1];
10155 PerlIO_printf(Perl_debug_log, "(%"UVuf")\n",
10156 (UV)((dist ? this_trie + dist : next) - start));
10159 nextbranch= this_trie + trie->jump[0];
10160 DUMPUNTIL(this_trie + dist, nextbranch);
10162 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
10163 nextbranch= regnext((regnode *)nextbranch);
10165 PerlIO_printf(Perl_debug_log, "\n");
10168 if (last && next > last)
10173 else if ( op == CURLY ) { /* "next" might be very big: optimizer */
10174 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS,
10175 NEXTOPER(node) + EXTRA_STEP_2ARGS + 1);
10177 else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
10179 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS, next);
10181 else if ( op == PLUS || op == STAR) {
10182 DUMPUNTIL(NEXTOPER(node), NEXTOPER(node) + 1);
10184 else if (op == ANYOF) {
10185 /* arglen 1 + class block */
10186 node += 1 + ((ANYOF_FLAGS(node) & ANYOF_LARGE)
10187 ? ANYOF_CLASS_SKIP : ANYOF_SKIP);
10188 node = NEXTOPER(node);
10190 else if (PL_regkind[(U8)op] == EXACT) {
10191 /* Literal string, where present. */
10192 node += NODE_SZ_STR(node) - 1;
10193 node = NEXTOPER(node);
10196 node = NEXTOPER(node);
10197 node += regarglen[(U8)op];
10199 if (op == CURLYX || op == OPEN)
10203 #ifdef DEBUG_DUMPUNTIL
10204 PerlIO_printf(Perl_debug_log, "--- %d\n", (int)indent);
10209 #endif /* DEBUGGING */
10213 * c-indentation-style: bsd
10214 * c-basic-offset: 4
10215 * indent-tabs-mode: t
10218 * ex: set ts=8 sts=4 sw=4 noet:
projects / perl / modules / re-engine-Hooks.git / blob