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
88 #include "dquote_static.c"
95 # if defined(BUGGY_MSC6)
96 /* MSC 6.00A breaks on op/regexp.t test 85 unless we turn this off */
97 # pragma optimize("a",off)
98 /* But MSC 6.00A is happy with 'w', for aliases only across function calls*/
99 # pragma optimize("w",on )
100 # endif /* BUGGY_MSC6 */
104 #define STATIC static
107 typedef struct RExC_state_t {
108 U32 flags; /* are we folding, multilining? */
109 char *precomp; /* uncompiled string. */
110 REGEXP *rx_sv; /* The SV that is the regexp. */
111 regexp *rx; /* perl core regexp structure */
112 regexp_internal *rxi; /* internal data for regexp object pprivate field */
113 char *start; /* Start of input for compile */
114 char *end; /* End of input for compile */
115 char *parse; /* Input-scan pointer. */
116 I32 whilem_seen; /* number of WHILEM in this expr */
117 regnode *emit_start; /* Start of emitted-code area */
118 regnode *emit_bound; /* First regnode outside of the allocated space */
119 regnode *emit; /* Code-emit pointer; ®dummy = don't = compiling */
120 I32 naughty; /* How bad is this pattern? */
121 I32 sawback; /* Did we see \1, ...? */
123 I32 size; /* Code size. */
124 I32 npar; /* Capture buffer count, (OPEN). */
125 I32 cpar; /* Capture buffer count, (CLOSE). */
126 I32 nestroot; /* root parens we are in - used by accept */
130 regnode **open_parens; /* pointers to open parens */
131 regnode **close_parens; /* pointers to close parens */
132 regnode *opend; /* END node in program */
133 I32 utf8; /* whether the pattern is utf8 or not */
134 I32 orig_utf8; /* whether the pattern was originally in utf8 */
135 /* XXX use this for future optimisation of case
136 * where pattern must be upgraded to utf8. */
137 HV *paren_names; /* Paren names */
139 regnode **recurse; /* Recurse regops */
140 I32 recurse_count; /* Number of recurse regops */
142 char *starttry; /* -Dr: where regtry was called. */
143 #define RExC_starttry (pRExC_state->starttry)
146 const char *lastparse;
148 AV *paren_name_list; /* idx -> name */
149 #define RExC_lastparse (pRExC_state->lastparse)
150 #define RExC_lastnum (pRExC_state->lastnum)
151 #define RExC_paren_name_list (pRExC_state->paren_name_list)
155 #define RExC_flags (pRExC_state->flags)
156 #define RExC_precomp (pRExC_state->precomp)
157 #define RExC_rx_sv (pRExC_state->rx_sv)
158 #define RExC_rx (pRExC_state->rx)
159 #define RExC_rxi (pRExC_state->rxi)
160 #define RExC_start (pRExC_state->start)
161 #define RExC_end (pRExC_state->end)
162 #define RExC_parse (pRExC_state->parse)
163 #define RExC_whilem_seen (pRExC_state->whilem_seen)
164 #ifdef RE_TRACK_PATTERN_OFFSETS
165 #define RExC_offsets (pRExC_state->rxi->u.offsets) /* I am not like the others */
167 #define RExC_emit (pRExC_state->emit)
168 #define RExC_emit_start (pRExC_state->emit_start)
169 #define RExC_emit_bound (pRExC_state->emit_bound)
170 #define RExC_naughty (pRExC_state->naughty)
171 #define RExC_sawback (pRExC_state->sawback)
172 #define RExC_seen (pRExC_state->seen)
173 #define RExC_size (pRExC_state->size)
174 #define RExC_npar (pRExC_state->npar)
175 #define RExC_nestroot (pRExC_state->nestroot)
176 #define RExC_extralen (pRExC_state->extralen)
177 #define RExC_seen_zerolen (pRExC_state->seen_zerolen)
178 #define RExC_seen_evals (pRExC_state->seen_evals)
179 #define RExC_utf8 (pRExC_state->utf8)
180 #define RExC_orig_utf8 (pRExC_state->orig_utf8)
181 #define RExC_open_parens (pRExC_state->open_parens)
182 #define RExC_close_parens (pRExC_state->close_parens)
183 #define RExC_opend (pRExC_state->opend)
184 #define RExC_paren_names (pRExC_state->paren_names)
185 #define RExC_recurse (pRExC_state->recurse)
186 #define RExC_recurse_count (pRExC_state->recurse_count)
189 #define ISMULT1(c) ((c) == '*' || (c) == '+' || (c) == '?')
190 #define ISMULT2(s) ((*s) == '*' || (*s) == '+' || (*s) == '?' || \
191 ((*s) == '{' && regcurly(s)))
194 #undef SPSTART /* dratted cpp namespace... */
197 * Flags to be passed up and down.
199 #define WORST 0 /* Worst case. */
200 #define HASWIDTH 0x01 /* Known to match non-null strings. */
202 /* Simple enough to be STAR/PLUS operand, in an EXACT node must be a single
203 * character, and if utf8, must be invariant. */
205 #define SPSTART 0x04 /* Starts with * or +. */
206 #define TRYAGAIN 0x08 /* Weeded out a declaration. */
207 #define POSTPONED 0x10 /* (?1),(?&name), (??{...}) or similar */
209 #define REG_NODE_NUM(x) ((x) ? (int)((x)-RExC_emit_start) : -1)
211 /* whether trie related optimizations are enabled */
212 #if PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION
213 #define TRIE_STUDY_OPT
214 #define FULL_TRIE_STUDY
220 #define PBYTE(u8str,paren) ((U8*)(u8str))[(paren) >> 3]
221 #define PBITVAL(paren) (1 << ((paren) & 7))
222 #define PAREN_TEST(u8str,paren) ( PBYTE(u8str,paren) & PBITVAL(paren))
223 #define PAREN_SET(u8str,paren) PBYTE(u8str,paren) |= PBITVAL(paren)
224 #define PAREN_UNSET(u8str,paren) PBYTE(u8str,paren) &= (~PBITVAL(paren))
226 /* If not already in utf8, do a longjmp back to the beginning */
227 #define UTF8_LONGJMP 42 /* Choose a value not likely to ever conflict */
228 #define REQUIRE_UTF8 STMT_START { \
229 if (! UTF) JMPENV_JUMP(UTF8_LONGJMP); \
232 /* About scan_data_t.
234 During optimisation we recurse through the regexp program performing
235 various inplace (keyhole style) optimisations. In addition study_chunk
236 and scan_commit populate this data structure with information about
237 what strings MUST appear in the pattern. We look for the longest
238 string that must appear at a fixed location, and we look for the
239 longest string that may appear at a floating location. So for instance
244 Both 'FOO' and 'A' are fixed strings. Both 'B' and 'BAR' are floating
245 strings (because they follow a .* construct). study_chunk will identify
246 both FOO and BAR as being the longest fixed and floating strings respectively.
248 The strings can be composites, for instance
252 will result in a composite fixed substring 'foo'.
254 For each string some basic information is maintained:
256 - offset or min_offset
257 This is the position the string must appear at, or not before.
258 It also implicitly (when combined with minlenp) tells us how many
259 characters must match before the string we are searching for.
260 Likewise when combined with minlenp and the length of the string it
261 tells us how many characters must appear after the string we have
265 Only used for floating strings. This is the rightmost point that
266 the string can appear at. If set to I32 max it indicates that the
267 string can occur infinitely far to the right.
270 A pointer to the minimum length of the pattern that the string
271 was found inside. This is important as in the case of positive
272 lookahead or positive lookbehind we can have multiple patterns
277 The minimum length of the pattern overall is 3, the minimum length
278 of the lookahead part is 3, but the minimum length of the part that
279 will actually match is 1. So 'FOO's minimum length is 3, but the
280 minimum length for the F is 1. This is important as the minimum length
281 is used to determine offsets in front of and behind the string being
282 looked for. Since strings can be composites this is the length of the
283 pattern at the time it was commited with a scan_commit. Note that
284 the length is calculated by study_chunk, so that the minimum lengths
285 are not known until the full pattern has been compiled, thus the
286 pointer to the value.
290 In the case of lookbehind the string being searched for can be
291 offset past the start point of the final matching string.
292 If this value was just blithely removed from the min_offset it would
293 invalidate some of the calculations for how many chars must match
294 before or after (as they are derived from min_offset and minlen and
295 the length of the string being searched for).
296 When the final pattern is compiled and the data is moved from the
297 scan_data_t structure into the regexp structure the information
298 about lookbehind is factored in, with the information that would
299 have been lost precalculated in the end_shift field for the
302 The fields pos_min and pos_delta are used to store the minimum offset
303 and the delta to the maximum offset at the current point in the pattern.
307 typedef struct scan_data_t {
308 /*I32 len_min; unused */
309 /*I32 len_delta; unused */
313 I32 last_end; /* min value, <0 unless valid. */
316 SV **longest; /* Either &l_fixed, or &l_float. */
317 SV *longest_fixed; /* longest fixed string found in pattern */
318 I32 offset_fixed; /* offset where it starts */
319 I32 *minlen_fixed; /* pointer to the minlen relevent to the string */
320 I32 lookbehind_fixed; /* is the position of the string modfied by LB */
321 SV *longest_float; /* longest floating string found in pattern */
322 I32 offset_float_min; /* earliest point in string it can appear */
323 I32 offset_float_max; /* latest point in string it can appear */
324 I32 *minlen_float; /* pointer to the minlen relevent to the string */
325 I32 lookbehind_float; /* is the position of the string modified by LB */
329 struct regnode_charclass_class *start_class;
333 * Forward declarations for pregcomp()'s friends.
336 static const scan_data_t zero_scan_data =
337 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ,0};
339 #define SF_BEFORE_EOL (SF_BEFORE_SEOL|SF_BEFORE_MEOL)
340 #define SF_BEFORE_SEOL 0x0001
341 #define SF_BEFORE_MEOL 0x0002
342 #define SF_FIX_BEFORE_EOL (SF_FIX_BEFORE_SEOL|SF_FIX_BEFORE_MEOL)
343 #define SF_FL_BEFORE_EOL (SF_FL_BEFORE_SEOL|SF_FL_BEFORE_MEOL)
346 # define SF_FIX_SHIFT_EOL (0+2)
347 # define SF_FL_SHIFT_EOL (0+4)
349 # define SF_FIX_SHIFT_EOL (+2)
350 # define SF_FL_SHIFT_EOL (+4)
353 #define SF_FIX_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FIX_SHIFT_EOL)
354 #define SF_FIX_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FIX_SHIFT_EOL)
356 #define SF_FL_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FL_SHIFT_EOL)
357 #define SF_FL_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FL_SHIFT_EOL) /* 0x20 */
358 #define SF_IS_INF 0x0040
359 #define SF_HAS_PAR 0x0080
360 #define SF_IN_PAR 0x0100
361 #define SF_HAS_EVAL 0x0200
362 #define SCF_DO_SUBSTR 0x0400
363 #define SCF_DO_STCLASS_AND 0x0800
364 #define SCF_DO_STCLASS_OR 0x1000
365 #define SCF_DO_STCLASS (SCF_DO_STCLASS_AND|SCF_DO_STCLASS_OR)
366 #define SCF_WHILEM_VISITED_POS 0x2000
368 #define SCF_TRIE_RESTUDY 0x4000 /* Do restudy? */
369 #define SCF_SEEN_ACCEPT 0x8000
371 #define UTF cBOOL(RExC_utf8)
372 #define LOC cBOOL(RExC_flags & RXf_PMf_LOCALE)
373 #define UNI_SEMANTICS cBOOL(RExC_flags & RXf_PMf_UNICODE)
374 #define FOLD cBOOL(RExC_flags & RXf_PMf_FOLD)
376 #define OOB_UNICODE 12345678
377 #define OOB_NAMEDCLASS -1
379 #define CHR_SVLEN(sv) (UTF ? sv_len_utf8(sv) : SvCUR(sv))
380 #define CHR_DIST(a,b) (UTF ? utf8_distance(a,b) : a - b)
383 /* length of regex to show in messages that don't mark a position within */
384 #define RegexLengthToShowInErrorMessages 127
387 * If MARKER[12] are adjusted, be sure to adjust the constants at the top
388 * of t/op/regmesg.t, the tests in t/op/re_tests, and those in
389 * op/pragma/warn/regcomp.
391 #define MARKER1 "<-- HERE" /* marker as it appears in the description */
392 #define MARKER2 " <-- HERE " /* marker as it appears within the regex */
394 #define REPORT_LOCATION " in regex; marked by " MARKER1 " in m/%.*s" MARKER2 "%s/"
397 * Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
398 * arg. Show regex, up to a maximum length. If it's too long, chop and add
401 #define _FAIL(code) STMT_START { \
402 const char *ellipses = ""; \
403 IV len = RExC_end - RExC_precomp; \
406 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
407 if (len > RegexLengthToShowInErrorMessages) { \
408 /* chop 10 shorter than the max, to ensure meaning of "..." */ \
409 len = RegexLengthToShowInErrorMessages - 10; \
415 #define FAIL(msg) _FAIL( \
416 Perl_croak(aTHX_ "%s in regex m/%.*s%s/", \
417 msg, (int)len, RExC_precomp, ellipses))
419 #define FAIL2(msg,arg) _FAIL( \
420 Perl_croak(aTHX_ msg " in regex m/%.*s%s/", \
421 arg, (int)len, RExC_precomp, ellipses))
424 * Simple_vFAIL -- like FAIL, but marks the current location in the scan
426 #define Simple_vFAIL(m) STMT_START { \
427 const IV offset = RExC_parse - RExC_precomp; \
428 Perl_croak(aTHX_ "%s" REPORT_LOCATION, \
429 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
433 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL()
435 #define vFAIL(m) STMT_START { \
437 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
442 * Like Simple_vFAIL(), but accepts two arguments.
444 #define Simple_vFAIL2(m,a1) STMT_START { \
445 const IV offset = RExC_parse - RExC_precomp; \
446 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, \
447 (int)offset, RExC_precomp, RExC_precomp + offset); \
451 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL2().
453 #define vFAIL2(m,a1) STMT_START { \
455 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
456 Simple_vFAIL2(m, a1); \
461 * Like Simple_vFAIL(), but accepts three arguments.
463 #define Simple_vFAIL3(m, a1, a2) STMT_START { \
464 const IV offset = RExC_parse - RExC_precomp; \
465 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, \
466 (int)offset, RExC_precomp, RExC_precomp + offset); \
470 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL3().
472 #define vFAIL3(m,a1,a2) STMT_START { \
474 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
475 Simple_vFAIL3(m, a1, a2); \
479 * Like Simple_vFAIL(), but accepts four arguments.
481 #define Simple_vFAIL4(m, a1, a2, a3) STMT_START { \
482 const IV offset = RExC_parse - RExC_precomp; \
483 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, \
484 (int)offset, RExC_precomp, RExC_precomp + offset); \
487 #define ckWARNreg(loc,m) STMT_START { \
488 const IV offset = loc - RExC_precomp; \
489 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
490 (int)offset, RExC_precomp, RExC_precomp + offset); \
493 #define ckWARNregdep(loc,m) STMT_START { \
494 const IV offset = loc - RExC_precomp; \
495 Perl_ck_warner_d(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
497 (int)offset, RExC_precomp, RExC_precomp + offset); \
500 #define ckWARN2reg(loc, m, a1) STMT_START { \
501 const IV offset = loc - RExC_precomp; \
502 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
503 a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
506 #define vWARN3(loc, m, a1, a2) STMT_START { \
507 const IV offset = loc - RExC_precomp; \
508 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
509 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
512 #define ckWARN3reg(loc, m, a1, a2) STMT_START { \
513 const IV offset = loc - RExC_precomp; \
514 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
515 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
518 #define vWARN4(loc, m, a1, a2, a3) STMT_START { \
519 const IV offset = loc - RExC_precomp; \
520 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
521 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
524 #define ckWARN4reg(loc, m, a1, a2, a3) STMT_START { \
525 const IV offset = loc - RExC_precomp; \
526 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
527 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
530 #define vWARN5(loc, m, a1, a2, a3, a4) STMT_START { \
531 const IV offset = loc - RExC_precomp; \
532 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
533 a1, a2, a3, a4, (int)offset, RExC_precomp, RExC_precomp + offset); \
537 /* Allow for side effects in s */
538 #define REGC(c,s) STMT_START { \
539 if (!SIZE_ONLY) *(s) = (c); else (void)(s); \
542 /* Macros for recording node offsets. 20001227 mjd@plover.com
543 * Nodes are numbered 1, 2, 3, 4. Node #n's position is recorded in
544 * element 2*n-1 of the array. Element #2n holds the byte length node #n.
545 * Element 0 holds the number n.
546 * Position is 1 indexed.
548 #ifndef RE_TRACK_PATTERN_OFFSETS
549 #define Set_Node_Offset_To_R(node,byte)
550 #define Set_Node_Offset(node,byte)
551 #define Set_Cur_Node_Offset
552 #define Set_Node_Length_To_R(node,len)
553 #define Set_Node_Length(node,len)
554 #define Set_Node_Cur_Length(node)
555 #define Node_Offset(n)
556 #define Node_Length(n)
557 #define Set_Node_Offset_Length(node,offset,len)
558 #define ProgLen(ri) ri->u.proglen
559 #define SetProgLen(ri,x) ri->u.proglen = x
561 #define ProgLen(ri) ri->u.offsets[0]
562 #define SetProgLen(ri,x) ri->u.offsets[0] = x
563 #define Set_Node_Offset_To_R(node,byte) STMT_START { \
565 MJD_OFFSET_DEBUG(("** (%d) offset of node %d is %d.\n", \
566 __LINE__, (int)(node), (int)(byte))); \
568 Perl_croak(aTHX_ "value of node is %d in Offset macro", (int)(node)); \
570 RExC_offsets[2*(node)-1] = (byte); \
575 #define Set_Node_Offset(node,byte) \
576 Set_Node_Offset_To_R((node)-RExC_emit_start, (byte)-RExC_start)
577 #define Set_Cur_Node_Offset Set_Node_Offset(RExC_emit, RExC_parse)
579 #define Set_Node_Length_To_R(node,len) STMT_START { \
581 MJD_OFFSET_DEBUG(("** (%d) size of node %d is %d.\n", \
582 __LINE__, (int)(node), (int)(len))); \
584 Perl_croak(aTHX_ "value of node is %d in Length macro", (int)(node)); \
586 RExC_offsets[2*(node)] = (len); \
591 #define Set_Node_Length(node,len) \
592 Set_Node_Length_To_R((node)-RExC_emit_start, len)
593 #define Set_Cur_Node_Length(len) Set_Node_Length(RExC_emit, len)
594 #define Set_Node_Cur_Length(node) \
595 Set_Node_Length(node, RExC_parse - parse_start)
597 /* Get offsets and lengths */
598 #define Node_Offset(n) (RExC_offsets[2*((n)-RExC_emit_start)-1])
599 #define Node_Length(n) (RExC_offsets[2*((n)-RExC_emit_start)])
601 #define Set_Node_Offset_Length(node,offset,len) STMT_START { \
602 Set_Node_Offset_To_R((node)-RExC_emit_start, (offset)); \
603 Set_Node_Length_To_R((node)-RExC_emit_start, (len)); \
607 #if PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS
608 #define EXPERIMENTAL_INPLACESCAN
609 #endif /*PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS*/
611 #define DEBUG_STUDYDATA(str,data,depth) \
612 DEBUG_OPTIMISE_MORE_r(if(data){ \
613 PerlIO_printf(Perl_debug_log, \
614 "%*s" str "Pos:%"IVdf"/%"IVdf \
615 " Flags: 0x%"UVXf" Whilem_c: %"IVdf" Lcp: %"IVdf" %s", \
616 (int)(depth)*2, "", \
617 (IV)((data)->pos_min), \
618 (IV)((data)->pos_delta), \
619 (UV)((data)->flags), \
620 (IV)((data)->whilem_c), \
621 (IV)((data)->last_closep ? *((data)->last_closep) : -1), \
622 is_inf ? "INF " : "" \
624 if ((data)->last_found) \
625 PerlIO_printf(Perl_debug_log, \
626 "Last:'%s' %"IVdf":%"IVdf"/%"IVdf" %sFixed:'%s' @ %"IVdf \
627 " %sFloat: '%s' @ %"IVdf"/%"IVdf"", \
628 SvPVX_const((data)->last_found), \
629 (IV)((data)->last_end), \
630 (IV)((data)->last_start_min), \
631 (IV)((data)->last_start_max), \
632 ((data)->longest && \
633 (data)->longest==&((data)->longest_fixed)) ? "*" : "", \
634 SvPVX_const((data)->longest_fixed), \
635 (IV)((data)->offset_fixed), \
636 ((data)->longest && \
637 (data)->longest==&((data)->longest_float)) ? "*" : "", \
638 SvPVX_const((data)->longest_float), \
639 (IV)((data)->offset_float_min), \
640 (IV)((data)->offset_float_max) \
642 PerlIO_printf(Perl_debug_log,"\n"); \
645 static void clear_re(pTHX_ void *r);
647 /* Mark that we cannot extend a found fixed substring at this point.
648 Update the longest found anchored substring and the longest found
649 floating substrings if needed. */
652 S_scan_commit(pTHX_ const RExC_state_t *pRExC_state, scan_data_t *data, I32 *minlenp, int is_inf)
654 const STRLEN l = CHR_SVLEN(data->last_found);
655 const STRLEN old_l = CHR_SVLEN(*data->longest);
656 GET_RE_DEBUG_FLAGS_DECL;
658 PERL_ARGS_ASSERT_SCAN_COMMIT;
660 if ((l >= old_l) && ((l > old_l) || (data->flags & SF_BEFORE_EOL))) {
661 SvSetMagicSV(*data->longest, data->last_found);
662 if (*data->longest == data->longest_fixed) {
663 data->offset_fixed = l ? data->last_start_min : data->pos_min;
664 if (data->flags & SF_BEFORE_EOL)
666 |= ((data->flags & SF_BEFORE_EOL) << SF_FIX_SHIFT_EOL);
668 data->flags &= ~SF_FIX_BEFORE_EOL;
669 data->minlen_fixed=minlenp;
670 data->lookbehind_fixed=0;
672 else { /* *data->longest == data->longest_float */
673 data->offset_float_min = l ? data->last_start_min : data->pos_min;
674 data->offset_float_max = (l
675 ? data->last_start_max
676 : data->pos_min + data->pos_delta);
677 if (is_inf || (U32)data->offset_float_max > (U32)I32_MAX)
678 data->offset_float_max = I32_MAX;
679 if (data->flags & SF_BEFORE_EOL)
681 |= ((data->flags & SF_BEFORE_EOL) << SF_FL_SHIFT_EOL);
683 data->flags &= ~SF_FL_BEFORE_EOL;
684 data->minlen_float=minlenp;
685 data->lookbehind_float=0;
688 SvCUR_set(data->last_found, 0);
690 SV * const sv = data->last_found;
691 if (SvUTF8(sv) && SvMAGICAL(sv)) {
692 MAGIC * const mg = mg_find(sv, PERL_MAGIC_utf8);
698 data->flags &= ~SF_BEFORE_EOL;
699 DEBUG_STUDYDATA("commit: ",data,0);
702 /* Can match anything (initialization) */
704 S_cl_anything(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
706 PERL_ARGS_ASSERT_CL_ANYTHING;
708 ANYOF_CLASS_ZERO(cl);
709 ANYOF_BITMAP_SETALL(cl);
710 cl->flags = ANYOF_EOS|ANYOF_UNICODE_ALL;
712 cl->flags |= ANYOF_LOCALE;
713 cl->flags |= ANYOF_FOLD;
716 /* Can match anything (initialization) */
718 S_cl_is_anything(const struct regnode_charclass_class *cl)
722 PERL_ARGS_ASSERT_CL_IS_ANYTHING;
724 for (value = 0; value <= ANYOF_MAX; value += 2)
725 if (ANYOF_CLASS_TEST(cl, value) && ANYOF_CLASS_TEST(cl, value + 1))
727 if (!(cl->flags & ANYOF_UNICODE_ALL))
729 if (!ANYOF_BITMAP_TESTALLSET((const void*)cl))
734 /* Can match anything (initialization) */
736 S_cl_init(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
738 PERL_ARGS_ASSERT_CL_INIT;
740 Zero(cl, 1, struct regnode_charclass_class);
742 cl_anything(pRExC_state, cl);
746 S_cl_init_zero(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
748 PERL_ARGS_ASSERT_CL_INIT_ZERO;
750 Zero(cl, 1, struct regnode_charclass_class);
752 cl_anything(pRExC_state, cl);
754 cl->flags |= ANYOF_LOCALE;
757 /* 'And' a given class with another one. Can create false positives */
758 /* We assume that cl is not inverted */
760 S_cl_and(struct regnode_charclass_class *cl,
761 const struct regnode_charclass_class *and_with)
763 PERL_ARGS_ASSERT_CL_AND;
765 assert(and_with->type == ANYOF);
766 if (!(and_with->flags & ANYOF_CLASS)
767 && !(cl->flags & ANYOF_CLASS)
768 && (and_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
769 && !(and_with->flags & ANYOF_FOLD)
770 && !(cl->flags & ANYOF_FOLD)) {
773 if (and_with->flags & ANYOF_INVERT)
774 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
775 cl->bitmap[i] &= ~and_with->bitmap[i];
777 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
778 cl->bitmap[i] &= and_with->bitmap[i];
779 } /* XXXX: logic is complicated otherwise, leave it along for a moment. */
780 if (!(and_with->flags & ANYOF_EOS))
781 cl->flags &= ~ANYOF_EOS;
783 if (!(and_with->flags & ANYOF_FOLD))
784 cl->flags &= ~ANYOF_FOLD;
786 if (cl->flags & ANYOF_UNICODE_ALL && and_with->flags & ANYOF_UNICODE &&
787 !(and_with->flags & ANYOF_INVERT)) {
788 cl->flags &= ~ANYOF_UNICODE_ALL;
789 cl->flags |= ANYOF_UNICODE;
790 ARG_SET(cl, ARG(and_with));
792 if (!(and_with->flags & ANYOF_UNICODE_ALL) &&
793 !(and_with->flags & ANYOF_INVERT))
794 cl->flags &= ~ANYOF_UNICODE_ALL;
795 if (!(and_with->flags & (ANYOF_UNICODE|ANYOF_UNICODE_ALL)) &&
796 !(and_with->flags & ANYOF_INVERT))
797 cl->flags &= ~ANYOF_UNICODE;
800 /* 'OR' a given class with another one. Can create false positives */
801 /* We assume that cl is not inverted */
803 S_cl_or(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, const struct regnode_charclass_class *or_with)
805 PERL_ARGS_ASSERT_CL_OR;
807 if (or_with->flags & ANYOF_INVERT) {
809 * (B1 | CL1) | (!B2 & !CL2) = (B1 | !B2 & !CL2) | (CL1 | (!B2 & !CL2))
810 * <= (B1 | !B2) | (CL1 | !CL2)
811 * which is wasteful if CL2 is small, but we ignore CL2:
812 * (B1 | CL1) | (!B2 & !CL2) <= (B1 | CL1) | !B2 = (B1 | !B2) | CL1
813 * XXXX Can we handle case-fold? Unclear:
814 * (OK1(i) | OK1(i')) | !(OK1(i) | OK1(i')) =
815 * (OK1(i) | OK1(i')) | (!OK1(i) & !OK1(i'))
817 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
818 && !(or_with->flags & ANYOF_FOLD)
819 && !(cl->flags & ANYOF_FOLD) ) {
822 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
823 cl->bitmap[i] |= ~or_with->bitmap[i];
824 } /* XXXX: logic is complicated otherwise */
826 cl_anything(pRExC_state, cl);
829 /* (B1 | CL1) | (B2 | CL2) = (B1 | B2) | (CL1 | CL2)) */
830 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
831 && (!(or_with->flags & ANYOF_FOLD)
832 || (cl->flags & ANYOF_FOLD)) ) {
835 /* OR char bitmap and class bitmap separately */
836 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
837 cl->bitmap[i] |= or_with->bitmap[i];
838 if (or_with->flags & ANYOF_CLASS) {
839 for (i = 0; i < ANYOF_CLASSBITMAP_SIZE; i++)
840 cl->classflags[i] |= or_with->classflags[i];
841 cl->flags |= ANYOF_CLASS;
844 else { /* XXXX: logic is complicated, leave it along for a moment. */
845 cl_anything(pRExC_state, cl);
848 if (or_with->flags & ANYOF_EOS)
849 cl->flags |= ANYOF_EOS;
851 if (or_with->flags & ANYOF_FOLD)
852 cl->flags |= ANYOF_FOLD;
854 if (cl->flags & ANYOF_UNICODE && or_with->flags & ANYOF_UNICODE &&
855 ARG(cl) != ARG(or_with)) {
856 cl->flags |= ANYOF_UNICODE_ALL;
857 cl->flags &= ~ANYOF_UNICODE;
859 if (or_with->flags & ANYOF_UNICODE_ALL) {
860 cl->flags |= ANYOF_UNICODE_ALL;
861 cl->flags &= ~ANYOF_UNICODE;
865 #define TRIE_LIST_ITEM(state,idx) (trie->states[state].trans.list)[ idx ]
866 #define TRIE_LIST_CUR(state) ( TRIE_LIST_ITEM( state, 0 ).forid )
867 #define TRIE_LIST_LEN(state) ( TRIE_LIST_ITEM( state, 0 ).newstate )
868 #define TRIE_LIST_USED(idx) ( trie->states[state].trans.list ? (TRIE_LIST_CUR( idx ) - 1) : 0 )
873 dump_trie(trie,widecharmap,revcharmap)
874 dump_trie_interim_list(trie,widecharmap,revcharmap,next_alloc)
875 dump_trie_interim_table(trie,widecharmap,revcharmap,next_alloc)
877 These routines dump out a trie in a somewhat readable format.
878 The _interim_ variants are used for debugging the interim
879 tables that are used to generate the final compressed
880 representation which is what dump_trie expects.
882 Part of the reason for their existance is to provide a form
883 of documentation as to how the different representations function.
888 Dumps the final compressed table form of the trie to Perl_debug_log.
889 Used for debugging make_trie().
893 S_dump_trie(pTHX_ const struct _reg_trie_data *trie, HV *widecharmap,
894 AV *revcharmap, U32 depth)
897 SV *sv=sv_newmortal();
898 int colwidth= widecharmap ? 6 : 4;
900 GET_RE_DEBUG_FLAGS_DECL;
902 PERL_ARGS_ASSERT_DUMP_TRIE;
904 PerlIO_printf( Perl_debug_log, "%*sChar : %-6s%-6s%-4s ",
905 (int)depth * 2 + 2,"",
906 "Match","Base","Ofs" );
908 for( state = 0 ; state < trie->uniquecharcount ; state++ ) {
909 SV ** const tmp = av_fetch( revcharmap, state, 0);
911 PerlIO_printf( Perl_debug_log, "%*s",
913 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
914 PL_colors[0], PL_colors[1],
915 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
916 PERL_PV_ESCAPE_FIRSTCHAR
921 PerlIO_printf( Perl_debug_log, "\n%*sState|-----------------------",
922 (int)depth * 2 + 2,"");
924 for( state = 0 ; state < trie->uniquecharcount ; state++ )
925 PerlIO_printf( Perl_debug_log, "%.*s", colwidth, "--------");
926 PerlIO_printf( Perl_debug_log, "\n");
928 for( state = 1 ; state < trie->statecount ; state++ ) {
929 const U32 base = trie->states[ state ].trans.base;
931 PerlIO_printf( Perl_debug_log, "%*s#%4"UVXf"|", (int)depth * 2 + 2,"", (UV)state);
933 if ( trie->states[ state ].wordnum ) {
934 PerlIO_printf( Perl_debug_log, " W%4X", trie->states[ state ].wordnum );
936 PerlIO_printf( Perl_debug_log, "%6s", "" );
939 PerlIO_printf( Perl_debug_log, " @%4"UVXf" ", (UV)base );
944 while( ( base + ofs < trie->uniquecharcount ) ||
945 ( base + ofs - trie->uniquecharcount < trie->lasttrans
946 && trie->trans[ base + ofs - trie->uniquecharcount ].check != state))
949 PerlIO_printf( Perl_debug_log, "+%2"UVXf"[ ", (UV)ofs);
951 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
952 if ( ( base + ofs >= trie->uniquecharcount ) &&
953 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
954 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
956 PerlIO_printf( Perl_debug_log, "%*"UVXf,
958 (UV)trie->trans[ base + ofs - trie->uniquecharcount ].next );
960 PerlIO_printf( Perl_debug_log, "%*s",colwidth," ." );
964 PerlIO_printf( Perl_debug_log, "]");
967 PerlIO_printf( Perl_debug_log, "\n" );
969 PerlIO_printf(Perl_debug_log, "%*sword_info N:(prev,len)=", (int)depth*2, "");
970 for (word=1; word <= trie->wordcount; word++) {
971 PerlIO_printf(Perl_debug_log, " %d:(%d,%d)",
972 (int)word, (int)(trie->wordinfo[word].prev),
973 (int)(trie->wordinfo[word].len));
975 PerlIO_printf(Perl_debug_log, "\n" );
978 Dumps a fully constructed but uncompressed trie in list form.
979 List tries normally only are used for construction when the number of
980 possible chars (trie->uniquecharcount) is very high.
981 Used for debugging make_trie().
984 S_dump_trie_interim_list(pTHX_ const struct _reg_trie_data *trie,
985 HV *widecharmap, AV *revcharmap, U32 next_alloc,
989 SV *sv=sv_newmortal();
990 int colwidth= widecharmap ? 6 : 4;
991 GET_RE_DEBUG_FLAGS_DECL;
993 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_LIST;
995 /* print out the table precompression. */
996 PerlIO_printf( Perl_debug_log, "%*sState :Word | Transition Data\n%*s%s",
997 (int)depth * 2 + 2,"", (int)depth * 2 + 2,"",
998 "------:-----+-----------------\n" );
1000 for( state=1 ; state < next_alloc ; state ++ ) {
1003 PerlIO_printf( Perl_debug_log, "%*s %4"UVXf" :",
1004 (int)depth * 2 + 2,"", (UV)state );
1005 if ( ! trie->states[ state ].wordnum ) {
1006 PerlIO_printf( Perl_debug_log, "%5s| ","");
1008 PerlIO_printf( Perl_debug_log, "W%4x| ",
1009 trie->states[ state ].wordnum
1012 for( charid = 1 ; charid <= TRIE_LIST_USED( state ) ; charid++ ) {
1013 SV ** const tmp = av_fetch( revcharmap, TRIE_LIST_ITEM(state,charid).forid, 0);
1015 PerlIO_printf( Perl_debug_log, "%*s:%3X=%4"UVXf" | ",
1017 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1018 PL_colors[0], PL_colors[1],
1019 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1020 PERL_PV_ESCAPE_FIRSTCHAR
1022 TRIE_LIST_ITEM(state,charid).forid,
1023 (UV)TRIE_LIST_ITEM(state,charid).newstate
1026 PerlIO_printf(Perl_debug_log, "\n%*s| ",
1027 (int)((depth * 2) + 14), "");
1030 PerlIO_printf( Perl_debug_log, "\n");
1035 Dumps a fully constructed but uncompressed trie in table form.
1036 This is the normal DFA style state transition table, with a few
1037 twists to facilitate compression later.
1038 Used for debugging make_trie().
1041 S_dump_trie_interim_table(pTHX_ const struct _reg_trie_data *trie,
1042 HV *widecharmap, AV *revcharmap, U32 next_alloc,
1047 SV *sv=sv_newmortal();
1048 int colwidth= widecharmap ? 6 : 4;
1049 GET_RE_DEBUG_FLAGS_DECL;
1051 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_TABLE;
1054 print out the table precompression so that we can do a visual check
1055 that they are identical.
1058 PerlIO_printf( Perl_debug_log, "%*sChar : ",(int)depth * 2 + 2,"" );
1060 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1061 SV ** const tmp = av_fetch( revcharmap, charid, 0);
1063 PerlIO_printf( Perl_debug_log, "%*s",
1065 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1066 PL_colors[0], PL_colors[1],
1067 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1068 PERL_PV_ESCAPE_FIRSTCHAR
1074 PerlIO_printf( Perl_debug_log, "\n%*sState+-",(int)depth * 2 + 2,"" );
1076 for( charid=0 ; charid < trie->uniquecharcount ; charid++ ) {
1077 PerlIO_printf( Perl_debug_log, "%.*s", colwidth,"--------");
1080 PerlIO_printf( Perl_debug_log, "\n" );
1082 for( state=1 ; state < next_alloc ; state += trie->uniquecharcount ) {
1084 PerlIO_printf( Perl_debug_log, "%*s%4"UVXf" : ",
1085 (int)depth * 2 + 2,"",
1086 (UV)TRIE_NODENUM( state ) );
1088 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1089 UV v=(UV)SAFE_TRIE_NODENUM( trie->trans[ state + charid ].next );
1091 PerlIO_printf( Perl_debug_log, "%*"UVXf, colwidth, v );
1093 PerlIO_printf( Perl_debug_log, "%*s", colwidth, "." );
1095 if ( ! trie->states[ TRIE_NODENUM( state ) ].wordnum ) {
1096 PerlIO_printf( Perl_debug_log, " (%4"UVXf")\n", (UV)trie->trans[ state ].check );
1098 PerlIO_printf( Perl_debug_log, " (%4"UVXf") W%4X\n", (UV)trie->trans[ state ].check,
1099 trie->states[ TRIE_NODENUM( state ) ].wordnum );
1107 /* make_trie(startbranch,first,last,tail,word_count,flags,depth)
1108 startbranch: the first branch in the whole branch sequence
1109 first : start branch of sequence of branch-exact nodes.
1110 May be the same as startbranch
1111 last : Thing following the last branch.
1112 May be the same as tail.
1113 tail : item following the branch sequence
1114 count : words in the sequence
1115 flags : currently the OP() type we will be building one of /EXACT(|F|Fl)/
1116 depth : indent depth
1118 Inplace optimizes a sequence of 2 or more Branch-Exact nodes into a TRIE node.
1120 A trie is an N'ary tree where the branches are determined by digital
1121 decomposition of the key. IE, at the root node you look up the 1st character and
1122 follow that branch repeat until you find the end of the branches. Nodes can be
1123 marked as "accepting" meaning they represent a complete word. Eg:
1127 would convert into the following structure. Numbers represent states, letters
1128 following numbers represent valid transitions on the letter from that state, if
1129 the number is in square brackets it represents an accepting state, otherwise it
1130 will be in parenthesis.
1132 +-h->+-e->[3]-+-r->(8)-+-s->[9]
1136 (1) +-i->(6)-+-s->[7]
1138 +-s->(3)-+-h->(4)-+-e->[5]
1140 Accept Word Mapping: 3=>1 (he),5=>2 (she), 7=>3 (his), 9=>4 (hers)
1142 This shows that when matching against the string 'hers' we will begin at state 1
1143 read 'h' and move to state 2, read 'e' and move to state 3 which is accepting,
1144 then read 'r' and go to state 8 followed by 's' which takes us to state 9 which
1145 is also accepting. Thus we know that we can match both 'he' and 'hers' with a
1146 single traverse. We store a mapping from accepting to state to which word was
1147 matched, and then when we have multiple possibilities we try to complete the
1148 rest of the regex in the order in which they occured in the alternation.
1150 The only prior NFA like behaviour that would be changed by the TRIE support is
1151 the silent ignoring of duplicate alternations which are of the form:
1153 / (DUPE|DUPE) X? (?{ ... }) Y /x
1155 Thus EVAL blocks follwing a trie may be called a different number of times with
1156 and without the optimisation. With the optimisations dupes will be silently
1157 ignored. This inconsistant behaviour of EVAL type nodes is well established as
1158 the following demonstrates:
1160 'words'=~/(word|word|word)(?{ print $1 })[xyz]/
1162 which prints out 'word' three times, but
1164 'words'=~/(word|word|word)(?{ print $1 })S/
1166 which doesnt print it out at all. This is due to other optimisations kicking in.
1168 Example of what happens on a structural level:
1170 The regexp /(ac|ad|ab)+/ will produce the folowing debug output:
1172 1: CURLYM[1] {1,32767}(18)
1183 This would be optimizable with startbranch=5, first=5, last=16, tail=16
1184 and should turn into:
1186 1: CURLYM[1] {1,32767}(18)
1188 [Words:3 Chars Stored:6 Unique Chars:4 States:5 NCP:1]
1196 Cases where tail != last would be like /(?foo|bar)baz/:
1206 which would be optimizable with startbranch=1, first=1, last=7, tail=8
1207 and would end up looking like:
1210 [Words:2 Chars Stored:6 Unique Chars:5 States:7 NCP:1]
1217 d = uvuni_to_utf8_flags(d, uv, 0);
1219 is the recommended Unicode-aware way of saying
1224 #define TRIE_STORE_REVCHAR \
1227 SV *zlopp = newSV(2); \
1228 unsigned char *flrbbbbb = (unsigned char *) SvPVX(zlopp); \
1229 unsigned const char *const kapow = uvuni_to_utf8(flrbbbbb, uvc & 0xFF); \
1230 SvCUR_set(zlopp, kapow - flrbbbbb); \
1233 av_push(revcharmap, zlopp); \
1235 char ooooff = (char)uvc; \
1236 av_push(revcharmap, newSVpvn(&ooooff, 1)); \
1240 #define TRIE_READ_CHAR STMT_START { \
1244 if ( foldlen > 0 ) { \
1245 uvc = utf8n_to_uvuni( scan, UTF8_MAXLEN, &len, uniflags ); \
1250 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1251 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1252 foldlen -= UNISKIP( uvc ); \
1253 scan = foldbuf + UNISKIP( uvc ); \
1256 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1266 #define TRIE_LIST_PUSH(state,fid,ns) STMT_START { \
1267 if ( TRIE_LIST_CUR( state ) >=TRIE_LIST_LEN( state ) ) { \
1268 U32 ging = TRIE_LIST_LEN( state ) *= 2; \
1269 Renew( trie->states[ state ].trans.list, ging, reg_trie_trans_le ); \
1271 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).forid = fid; \
1272 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).newstate = ns; \
1273 TRIE_LIST_CUR( state )++; \
1276 #define TRIE_LIST_NEW(state) STMT_START { \
1277 Newxz( trie->states[ state ].trans.list, \
1278 4, reg_trie_trans_le ); \
1279 TRIE_LIST_CUR( state ) = 1; \
1280 TRIE_LIST_LEN( state ) = 4; \
1283 #define TRIE_HANDLE_WORD(state) STMT_START { \
1284 U16 dupe= trie->states[ state ].wordnum; \
1285 regnode * const noper_next = regnext( noper ); \
1288 /* store the word for dumping */ \
1290 if (OP(noper) != NOTHING) \
1291 tmp = newSVpvn_utf8(STRING(noper), STR_LEN(noper), UTF); \
1293 tmp = newSVpvn_utf8( "", 0, UTF ); \
1294 av_push( trie_words, tmp ); \
1298 trie->wordinfo[curword].prev = 0; \
1299 trie->wordinfo[curword].len = wordlen; \
1300 trie->wordinfo[curword].accept = state; \
1302 if ( noper_next < tail ) { \
1304 trie->jump = (U16 *) PerlMemShared_calloc( word_count + 1, sizeof(U16) ); \
1305 trie->jump[curword] = (U16)(noper_next - convert); \
1307 jumper = noper_next; \
1309 nextbranch= regnext(cur); \
1313 /* It's a dupe. Pre-insert into the wordinfo[].prev */\
1314 /* chain, so that when the bits of chain are later */\
1315 /* linked together, the dups appear in the chain */\
1316 trie->wordinfo[curword].prev = trie->wordinfo[dupe].prev; \
1317 trie->wordinfo[dupe].prev = curword; \
1319 /* we haven't inserted this word yet. */ \
1320 trie->states[ state ].wordnum = curword; \
1325 #define TRIE_TRANS_STATE(state,base,ucharcount,charid,special) \
1326 ( ( base + charid >= ucharcount \
1327 && base + charid < ubound \
1328 && state == trie->trans[ base - ucharcount + charid ].check \
1329 && trie->trans[ base - ucharcount + charid ].next ) \
1330 ? trie->trans[ base - ucharcount + charid ].next \
1331 : ( state==1 ? special : 0 ) \
1335 #define MADE_JUMP_TRIE 2
1336 #define MADE_EXACT_TRIE 4
1339 S_make_trie(pTHX_ RExC_state_t *pRExC_state, regnode *startbranch, regnode *first, regnode *last, regnode *tail, U32 word_count, U32 flags, U32 depth)
1342 /* first pass, loop through and scan words */
1343 reg_trie_data *trie;
1344 HV *widecharmap = NULL;
1345 AV *revcharmap = newAV();
1347 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1352 regnode *jumper = NULL;
1353 regnode *nextbranch = NULL;
1354 regnode *convert = NULL;
1355 U32 *prev_states; /* temp array mapping each state to previous one */
1356 /* we just use folder as a flag in utf8 */
1357 const U8 * const folder = ( flags == EXACTF
1359 : ( flags == EXACTFL
1366 const U32 data_slot = add_data( pRExC_state, 4, "tuuu" );
1367 AV *trie_words = NULL;
1368 /* along with revcharmap, this only used during construction but both are
1369 * useful during debugging so we store them in the struct when debugging.
1372 const U32 data_slot = add_data( pRExC_state, 2, "tu" );
1373 STRLEN trie_charcount=0;
1375 SV *re_trie_maxbuff;
1376 GET_RE_DEBUG_FLAGS_DECL;
1378 PERL_ARGS_ASSERT_MAKE_TRIE;
1380 PERL_UNUSED_ARG(depth);
1383 trie = (reg_trie_data *) PerlMemShared_calloc( 1, sizeof(reg_trie_data) );
1385 trie->startstate = 1;
1386 trie->wordcount = word_count;
1387 RExC_rxi->data->data[ data_slot ] = (void*)trie;
1388 trie->charmap = (U16 *) PerlMemShared_calloc( 256, sizeof(U16) );
1389 if (!(UTF && folder))
1390 trie->bitmap = (char *) PerlMemShared_calloc( ANYOF_BITMAP_SIZE, 1 );
1391 trie->wordinfo = (reg_trie_wordinfo *) PerlMemShared_calloc(
1392 trie->wordcount+1, sizeof(reg_trie_wordinfo));
1395 trie_words = newAV();
1398 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
1399 if (!SvIOK(re_trie_maxbuff)) {
1400 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
1403 PerlIO_printf( Perl_debug_log,
1404 "%*smake_trie start==%d, first==%d, last==%d, tail==%d depth=%d\n",
1405 (int)depth * 2 + 2, "",
1406 REG_NODE_NUM(startbranch),REG_NODE_NUM(first),
1407 REG_NODE_NUM(last), REG_NODE_NUM(tail),
1411 /* Find the node we are going to overwrite */
1412 if ( first == startbranch && OP( last ) != BRANCH ) {
1413 /* whole branch chain */
1416 /* branch sub-chain */
1417 convert = NEXTOPER( first );
1420 /* -- First loop and Setup --
1422 We first traverse the branches and scan each word to determine if it
1423 contains widechars, and how many unique chars there are, this is
1424 important as we have to build a table with at least as many columns as we
1427 We use an array of integers to represent the character codes 0..255
1428 (trie->charmap) and we use a an HV* to store Unicode characters. We use the
1429 native representation of the character value as the key and IV's for the
1432 *TODO* If we keep track of how many times each character is used we can
1433 remap the columns so that the table compression later on is more
1434 efficient in terms of memory by ensuring the most common value is in the
1435 middle and the least common are on the outside. IMO this would be better
1436 than a most to least common mapping as theres a decent chance the most
1437 common letter will share a node with the least common, meaning the node
1438 will not be compressable. With a middle is most common approach the worst
1439 case is when we have the least common nodes twice.
1443 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1444 regnode * const noper = NEXTOPER( cur );
1445 const U8 *uc = (U8*)STRING( noper );
1446 const U8 * const e = uc + STR_LEN( noper );
1448 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1449 const U8 *scan = (U8*)NULL;
1450 U32 wordlen = 0; /* required init */
1452 bool set_bit = trie->bitmap ? 1 : 0; /*store the first char in the bitmap?*/
1454 if (OP(noper) == NOTHING) {
1458 if ( set_bit ) /* bitmap only alloced when !(UTF&&Folding) */
1459 TRIE_BITMAP_SET(trie,*uc); /* store the raw first byte
1460 regardless of encoding */
1462 for ( ; uc < e ; uc += len ) {
1463 TRIE_CHARCOUNT(trie)++;
1467 if ( !trie->charmap[ uvc ] ) {
1468 trie->charmap[ uvc ]=( ++trie->uniquecharcount );
1470 trie->charmap[ folder[ uvc ] ] = trie->charmap[ uvc ];
1474 /* store the codepoint in the bitmap, and if its ascii
1475 also store its folded equivelent. */
1476 TRIE_BITMAP_SET(trie,uvc);
1478 /* store the folded codepoint */
1479 if ( folder ) TRIE_BITMAP_SET(trie,folder[ uvc ]);
1482 /* store first byte of utf8 representation of
1483 codepoints in the 127 < uvc < 256 range */
1484 if (127 < uvc && uvc < 192) {
1485 TRIE_BITMAP_SET(trie,194);
1486 } else if (191 < uvc ) {
1487 TRIE_BITMAP_SET(trie,195);
1488 /* && uvc < 256 -- we know uvc is < 256 already */
1491 set_bit = 0; /* We've done our bit :-) */
1496 widecharmap = newHV();
1498 svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 1 );
1501 Perl_croak( aTHX_ "error creating/fetching widecharmap entry for 0x%"UVXf, uvc );
1503 if ( !SvTRUE( *svpp ) ) {
1504 sv_setiv( *svpp, ++trie->uniquecharcount );
1509 if( cur == first ) {
1512 } else if (chars < trie->minlen) {
1514 } else if (chars > trie->maxlen) {
1518 } /* end first pass */
1519 DEBUG_TRIE_COMPILE_r(
1520 PerlIO_printf( Perl_debug_log, "%*sTRIE(%s): W:%d C:%d Uq:%d Min:%d Max:%d\n",
1521 (int)depth * 2 + 2,"",
1522 ( widecharmap ? "UTF8" : "NATIVE" ), (int)word_count,
1523 (int)TRIE_CHARCOUNT(trie), trie->uniquecharcount,
1524 (int)trie->minlen, (int)trie->maxlen )
1528 We now know what we are dealing with in terms of unique chars and
1529 string sizes so we can calculate how much memory a naive
1530 representation using a flat table will take. If it's over a reasonable
1531 limit (as specified by ${^RE_TRIE_MAXBUF}) we use a more memory
1532 conservative but potentially much slower representation using an array
1535 At the end we convert both representations into the same compressed
1536 form that will be used in regexec.c for matching with. The latter
1537 is a form that cannot be used to construct with but has memory
1538 properties similar to the list form and access properties similar
1539 to the table form making it both suitable for fast searches and
1540 small enough that its feasable to store for the duration of a program.
1542 See the comment in the code where the compressed table is produced
1543 inplace from the flat tabe representation for an explanation of how
1544 the compression works.
1549 Newx(prev_states, TRIE_CHARCOUNT(trie) + 2, U32);
1552 if ( (IV)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1) > SvIV(re_trie_maxbuff) ) {
1554 Second Pass -- Array Of Lists Representation
1556 Each state will be represented by a list of charid:state records
1557 (reg_trie_trans_le) the first such element holds the CUR and LEN
1558 points of the allocated array. (See defines above).
1560 We build the initial structure using the lists, and then convert
1561 it into the compressed table form which allows faster lookups
1562 (but cant be modified once converted).
1565 STRLEN transcount = 1;
1567 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1568 "%*sCompiling trie using list compiler\n",
1569 (int)depth * 2 + 2, ""));
1571 trie->states = (reg_trie_state *)
1572 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1573 sizeof(reg_trie_state) );
1577 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1579 regnode * const noper = NEXTOPER( cur );
1580 U8 *uc = (U8*)STRING( noper );
1581 const U8 * const e = uc + STR_LEN( noper );
1582 U32 state = 1; /* required init */
1583 U16 charid = 0; /* sanity init */
1584 U8 *scan = (U8*)NULL; /* sanity init */
1585 STRLEN foldlen = 0; /* required init */
1586 U32 wordlen = 0; /* required init */
1587 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1589 if (OP(noper) != NOTHING) {
1590 for ( ; uc < e ; uc += len ) {
1595 charid = trie->charmap[ uvc ];
1597 SV** const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1601 charid=(U16)SvIV( *svpp );
1604 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1611 if ( !trie->states[ state ].trans.list ) {
1612 TRIE_LIST_NEW( state );
1614 for ( check = 1; check <= TRIE_LIST_USED( state ); check++ ) {
1615 if ( TRIE_LIST_ITEM( state, check ).forid == charid ) {
1616 newstate = TRIE_LIST_ITEM( state, check ).newstate;
1621 newstate = next_alloc++;
1622 prev_states[newstate] = state;
1623 TRIE_LIST_PUSH( state, charid, newstate );
1628 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1632 TRIE_HANDLE_WORD(state);
1634 } /* end second pass */
1636 /* next alloc is the NEXT state to be allocated */
1637 trie->statecount = next_alloc;
1638 trie->states = (reg_trie_state *)
1639 PerlMemShared_realloc( trie->states,
1641 * sizeof(reg_trie_state) );
1643 /* and now dump it out before we compress it */
1644 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_list(trie, widecharmap,
1645 revcharmap, next_alloc,
1649 trie->trans = (reg_trie_trans *)
1650 PerlMemShared_calloc( transcount, sizeof(reg_trie_trans) );
1657 for( state=1 ; state < next_alloc ; state ++ ) {
1661 DEBUG_TRIE_COMPILE_MORE_r(
1662 PerlIO_printf( Perl_debug_log, "tp: %d zp: %d ",tp,zp)
1666 if (trie->states[state].trans.list) {
1667 U16 minid=TRIE_LIST_ITEM( state, 1).forid;
1671 for( idx = 2 ; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1672 const U16 forid = TRIE_LIST_ITEM( state, idx).forid;
1673 if ( forid < minid ) {
1675 } else if ( forid > maxid ) {
1679 if ( transcount < tp + maxid - minid + 1) {
1681 trie->trans = (reg_trie_trans *)
1682 PerlMemShared_realloc( trie->trans,
1684 * sizeof(reg_trie_trans) );
1685 Zero( trie->trans + (transcount / 2), transcount / 2 , reg_trie_trans );
1687 base = trie->uniquecharcount + tp - minid;
1688 if ( maxid == minid ) {
1690 for ( ; zp < tp ; zp++ ) {
1691 if ( ! trie->trans[ zp ].next ) {
1692 base = trie->uniquecharcount + zp - minid;
1693 trie->trans[ zp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1694 trie->trans[ zp ].check = state;
1700 trie->trans[ tp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1701 trie->trans[ tp ].check = state;
1706 for ( idx=1; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1707 const U32 tid = base - trie->uniquecharcount + TRIE_LIST_ITEM( state, idx ).forid;
1708 trie->trans[ tid ].next = TRIE_LIST_ITEM( state, idx ).newstate;
1709 trie->trans[ tid ].check = state;
1711 tp += ( maxid - minid + 1 );
1713 Safefree(trie->states[ state ].trans.list);
1716 DEBUG_TRIE_COMPILE_MORE_r(
1717 PerlIO_printf( Perl_debug_log, " base: %d\n",base);
1720 trie->states[ state ].trans.base=base;
1722 trie->lasttrans = tp + 1;
1726 Second Pass -- Flat Table Representation.
1728 we dont use the 0 slot of either trans[] or states[] so we add 1 to each.
1729 We know that we will need Charcount+1 trans at most to store the data
1730 (one row per char at worst case) So we preallocate both structures
1731 assuming worst case.
1733 We then construct the trie using only the .next slots of the entry
1736 We use the .check field of the first entry of the node temporarily to
1737 make compression both faster and easier by keeping track of how many non
1738 zero fields are in the node.
1740 Since trans are numbered from 1 any 0 pointer in the table is a FAIL
1743 There are two terms at use here: state as a TRIE_NODEIDX() which is a
1744 number representing the first entry of the node, and state as a
1745 TRIE_NODENUM() which is the trans number. state 1 is TRIE_NODEIDX(1) and
1746 TRIE_NODENUM(1), state 2 is TRIE_NODEIDX(2) and TRIE_NODENUM(3) if there
1747 are 2 entrys per node. eg:
1755 The table is internally in the right hand, idx form. However as we also
1756 have to deal with the states array which is indexed by nodenum we have to
1757 use TRIE_NODENUM() to convert.
1760 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1761 "%*sCompiling trie using table compiler\n",
1762 (int)depth * 2 + 2, ""));
1764 trie->trans = (reg_trie_trans *)
1765 PerlMemShared_calloc( ( TRIE_CHARCOUNT(trie) + 1 )
1766 * trie->uniquecharcount + 1,
1767 sizeof(reg_trie_trans) );
1768 trie->states = (reg_trie_state *)
1769 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1770 sizeof(reg_trie_state) );
1771 next_alloc = trie->uniquecharcount + 1;
1774 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1776 regnode * const noper = NEXTOPER( cur );
1777 const U8 *uc = (U8*)STRING( noper );
1778 const U8 * const e = uc + STR_LEN( noper );
1780 U32 state = 1; /* required init */
1782 U16 charid = 0; /* sanity init */
1783 U32 accept_state = 0; /* sanity init */
1784 U8 *scan = (U8*)NULL; /* sanity init */
1786 STRLEN foldlen = 0; /* required init */
1787 U32 wordlen = 0; /* required init */
1788 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1790 if ( OP(noper) != NOTHING ) {
1791 for ( ; uc < e ; uc += len ) {
1796 charid = trie->charmap[ uvc ];
1798 SV* const * const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1799 charid = svpp ? (U16)SvIV(*svpp) : 0;
1803 if ( !trie->trans[ state + charid ].next ) {
1804 trie->trans[ state + charid ].next = next_alloc;
1805 trie->trans[ state ].check++;
1806 prev_states[TRIE_NODENUM(next_alloc)]
1807 = TRIE_NODENUM(state);
1808 next_alloc += trie->uniquecharcount;
1810 state = trie->trans[ state + charid ].next;
1812 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1814 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1817 accept_state = TRIE_NODENUM( state );
1818 TRIE_HANDLE_WORD(accept_state);
1820 } /* end second pass */
1822 /* and now dump it out before we compress it */
1823 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_table(trie, widecharmap,
1825 next_alloc, depth+1));
1829 * Inplace compress the table.*
1831 For sparse data sets the table constructed by the trie algorithm will
1832 be mostly 0/FAIL transitions or to put it another way mostly empty.
1833 (Note that leaf nodes will not contain any transitions.)
1835 This algorithm compresses the tables by eliminating most such
1836 transitions, at the cost of a modest bit of extra work during lookup:
1838 - Each states[] entry contains a .base field which indicates the
1839 index in the state[] array wheres its transition data is stored.
1841 - If .base is 0 there are no valid transitions from that node.
1843 - If .base is nonzero then charid is added to it to find an entry in
1846 -If trans[states[state].base+charid].check!=state then the
1847 transition is taken to be a 0/Fail transition. Thus if there are fail
1848 transitions at the front of the node then the .base offset will point
1849 somewhere inside the previous nodes data (or maybe even into a node
1850 even earlier), but the .check field determines if the transition is
1854 The following process inplace converts the table to the compressed
1855 table: We first do not compress the root node 1,and mark all its
1856 .check pointers as 1 and set its .base pointer as 1 as well. This
1857 allows us to do a DFA construction from the compressed table later,
1858 and ensures that any .base pointers we calculate later are greater
1861 - We set 'pos' to indicate the first entry of the second node.
1863 - We then iterate over the columns of the node, finding the first and
1864 last used entry at l and m. We then copy l..m into pos..(pos+m-l),
1865 and set the .check pointers accordingly, and advance pos
1866 appropriately and repreat for the next node. Note that when we copy
1867 the next pointers we have to convert them from the original
1868 NODEIDX form to NODENUM form as the former is not valid post
1871 - If a node has no transitions used we mark its base as 0 and do not
1872 advance the pos pointer.
1874 - If a node only has one transition we use a second pointer into the
1875 structure to fill in allocated fail transitions from other states.
1876 This pointer is independent of the main pointer and scans forward
1877 looking for null transitions that are allocated to a state. When it
1878 finds one it writes the single transition into the "hole". If the
1879 pointer doesnt find one the single transition is appended as normal.
1881 - Once compressed we can Renew/realloc the structures to release the
1884 See "Table-Compression Methods" in sec 3.9 of the Red Dragon,
1885 specifically Fig 3.47 and the associated pseudocode.
1889 const U32 laststate = TRIE_NODENUM( next_alloc );
1892 trie->statecount = laststate;
1894 for ( state = 1 ; state < laststate ; state++ ) {
1896 const U32 stateidx = TRIE_NODEIDX( state );
1897 const U32 o_used = trie->trans[ stateidx ].check;
1898 U32 used = trie->trans[ stateidx ].check;
1899 trie->trans[ stateidx ].check = 0;
1901 for ( charid = 0 ; used && charid < trie->uniquecharcount ; charid++ ) {
1902 if ( flag || trie->trans[ stateidx + charid ].next ) {
1903 if ( trie->trans[ stateidx + charid ].next ) {
1905 for ( ; zp < pos ; zp++ ) {
1906 if ( ! trie->trans[ zp ].next ) {
1910 trie->states[ state ].trans.base = zp + trie->uniquecharcount - charid ;
1911 trie->trans[ zp ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1912 trie->trans[ zp ].check = state;
1913 if ( ++zp > pos ) pos = zp;
1920 trie->states[ state ].trans.base = pos + trie->uniquecharcount - charid ;
1922 trie->trans[ pos ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1923 trie->trans[ pos ].check = state;
1928 trie->lasttrans = pos + 1;
1929 trie->states = (reg_trie_state *)
1930 PerlMemShared_realloc( trie->states, laststate
1931 * sizeof(reg_trie_state) );
1932 DEBUG_TRIE_COMPILE_MORE_r(
1933 PerlIO_printf( Perl_debug_log,
1934 "%*sAlloc: %d Orig: %"IVdf" elements, Final:%"IVdf". Savings of %%%5.2f\n",
1935 (int)depth * 2 + 2,"",
1936 (int)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1 ),
1939 ( ( next_alloc - pos ) * 100 ) / (double)next_alloc );
1942 } /* end table compress */
1944 DEBUG_TRIE_COMPILE_MORE_r(
1945 PerlIO_printf(Perl_debug_log, "%*sStatecount:%"UVxf" Lasttrans:%"UVxf"\n",
1946 (int)depth * 2 + 2, "",
1947 (UV)trie->statecount,
1948 (UV)trie->lasttrans)
1950 /* resize the trans array to remove unused space */
1951 trie->trans = (reg_trie_trans *)
1952 PerlMemShared_realloc( trie->trans, trie->lasttrans
1953 * sizeof(reg_trie_trans) );
1955 { /* Modify the program and insert the new TRIE node */
1956 U8 nodetype =(U8)(flags & 0xFF);
1960 regnode *optimize = NULL;
1961 #ifdef RE_TRACK_PATTERN_OFFSETS
1964 U32 mjd_nodelen = 0;
1965 #endif /* RE_TRACK_PATTERN_OFFSETS */
1966 #endif /* DEBUGGING */
1968 This means we convert either the first branch or the first Exact,
1969 depending on whether the thing following (in 'last') is a branch
1970 or not and whther first is the startbranch (ie is it a sub part of
1971 the alternation or is it the whole thing.)
1972 Assuming its a sub part we convert the EXACT otherwise we convert
1973 the whole branch sequence, including the first.
1975 /* Find the node we are going to overwrite */
1976 if ( first != startbranch || OP( last ) == BRANCH ) {
1977 /* branch sub-chain */
1978 NEXT_OFF( first ) = (U16)(last - first);
1979 #ifdef RE_TRACK_PATTERN_OFFSETS
1981 mjd_offset= Node_Offset((convert));
1982 mjd_nodelen= Node_Length((convert));
1985 /* whole branch chain */
1987 #ifdef RE_TRACK_PATTERN_OFFSETS
1990 const regnode *nop = NEXTOPER( convert );
1991 mjd_offset= Node_Offset((nop));
1992 mjd_nodelen= Node_Length((nop));
1996 PerlIO_printf(Perl_debug_log, "%*sMJD offset:%"UVuf" MJD length:%"UVuf"\n",
1997 (int)depth * 2 + 2, "",
1998 (UV)mjd_offset, (UV)mjd_nodelen)
2001 /* But first we check to see if there is a common prefix we can
2002 split out as an EXACT and put in front of the TRIE node. */
2003 trie->startstate= 1;
2004 if ( trie->bitmap && !widecharmap && !trie->jump ) {
2006 for ( state = 1 ; state < trie->statecount-1 ; state++ ) {
2010 const U32 base = trie->states[ state ].trans.base;
2012 if ( trie->states[state].wordnum )
2015 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
2016 if ( ( base + ofs >= trie->uniquecharcount ) &&
2017 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
2018 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
2020 if ( ++count > 1 ) {
2021 SV **tmp = av_fetch( revcharmap, ofs, 0);
2022 const U8 *ch = (U8*)SvPV_nolen_const( *tmp );
2023 if ( state == 1 ) break;
2025 Zero(trie->bitmap, ANYOF_BITMAP_SIZE, char);
2027 PerlIO_printf(Perl_debug_log,
2028 "%*sNew Start State=%"UVuf" Class: [",
2029 (int)depth * 2 + 2, "",
2032 SV ** const tmp = av_fetch( revcharmap, idx, 0);
2033 const U8 * const ch = (U8*)SvPV_nolen_const( *tmp );
2035 TRIE_BITMAP_SET(trie,*ch);
2037 TRIE_BITMAP_SET(trie, folder[ *ch ]);
2039 PerlIO_printf(Perl_debug_log, "%s", (char*)ch)
2043 TRIE_BITMAP_SET(trie,*ch);
2045 TRIE_BITMAP_SET(trie,folder[ *ch ]);
2046 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"%s", ch));
2052 SV **tmp = av_fetch( revcharmap, idx, 0);
2054 char *ch = SvPV( *tmp, len );
2056 SV *sv=sv_newmortal();
2057 PerlIO_printf( Perl_debug_log,
2058 "%*sPrefix State: %"UVuf" Idx:%"UVuf" Char='%s'\n",
2059 (int)depth * 2 + 2, "",
2061 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 6,
2062 PL_colors[0], PL_colors[1],
2063 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
2064 PERL_PV_ESCAPE_FIRSTCHAR
2069 OP( convert ) = nodetype;
2070 str=STRING(convert);
2073 STR_LEN(convert) += len;
2079 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"]\n"));
2084 trie->prefixlen = (state-1);
2086 regnode *n = convert+NODE_SZ_STR(convert);
2087 NEXT_OFF(convert) = NODE_SZ_STR(convert);
2088 trie->startstate = state;
2089 trie->minlen -= (state - 1);
2090 trie->maxlen -= (state - 1);
2092 /* At least the UNICOS C compiler choked on this
2093 * being argument to DEBUG_r(), so let's just have
2096 #ifdef PERL_EXT_RE_BUILD
2102 regnode *fix = convert;
2103 U32 word = trie->wordcount;
2105 Set_Node_Offset_Length(convert, mjd_offset, state - 1);
2106 while( ++fix < n ) {
2107 Set_Node_Offset_Length(fix, 0, 0);
2110 SV ** const tmp = av_fetch( trie_words, word, 0 );
2112 if ( STR_LEN(convert) <= SvCUR(*tmp) )
2113 sv_chop(*tmp, SvPV_nolen(*tmp) + STR_LEN(convert));
2115 sv_chop(*tmp, SvPV_nolen(*tmp) + SvCUR(*tmp));
2123 NEXT_OFF(convert) = (U16)(tail - convert);
2124 DEBUG_r(optimize= n);
2130 if ( trie->maxlen ) {
2131 NEXT_OFF( convert ) = (U16)(tail - convert);
2132 ARG_SET( convert, data_slot );
2133 /* Store the offset to the first unabsorbed branch in
2134 jump[0], which is otherwise unused by the jump logic.
2135 We use this when dumping a trie and during optimisation. */
2137 trie->jump[0] = (U16)(nextbranch - convert);
2139 /* If the start state is not accepting (meaning there is no empty string/NOTHING)
2140 * and there is a bitmap
2141 * and the first "jump target" node we found leaves enough room
2142 * then convert the TRIE node into a TRIEC node, with the bitmap
2143 * embedded inline in the opcode - this is hypothetically faster.
2145 if ( !trie->states[trie->startstate].wordnum
2147 && ( (char *)jumper - (char *)convert) >= (int)sizeof(struct regnode_charclass) )
2149 OP( convert ) = TRIEC;
2150 Copy(trie->bitmap, ((struct regnode_charclass *)convert)->bitmap, ANYOF_BITMAP_SIZE, char);
2151 PerlMemShared_free(trie->bitmap);
2154 OP( convert ) = TRIE;
2156 /* store the type in the flags */
2157 convert->flags = nodetype;
2161 + regarglen[ OP( convert ) ];
2163 /* XXX We really should free up the resource in trie now,
2164 as we won't use them - (which resources?) dmq */
2166 /* needed for dumping*/
2167 DEBUG_r(if (optimize) {
2168 regnode *opt = convert;
2170 while ( ++opt < optimize) {
2171 Set_Node_Offset_Length(opt,0,0);
2174 Try to clean up some of the debris left after the
2177 while( optimize < jumper ) {
2178 mjd_nodelen += Node_Length((optimize));
2179 OP( optimize ) = OPTIMIZED;
2180 Set_Node_Offset_Length(optimize,0,0);
2183 Set_Node_Offset_Length(convert,mjd_offset,mjd_nodelen);
2185 } /* end node insert */
2186 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, convert);
2188 /* Finish populating the prev field of the wordinfo array. Walk back
2189 * from each accept state until we find another accept state, and if
2190 * so, point the first word's .prev field at the second word. If the
2191 * second already has a .prev field set, stop now. This will be the
2192 * case either if we've already processed that word's accept state,
2193 * or that state had multiple words, and the overspill words were
2194 * already linked up earlier.
2201 for (word=1; word <= trie->wordcount; word++) {
2203 if (trie->wordinfo[word].prev)
2205 state = trie->wordinfo[word].accept;
2207 state = prev_states[state];
2210 prev = trie->states[state].wordnum;
2214 trie->wordinfo[word].prev = prev;
2216 Safefree(prev_states);
2220 /* and now dump out the compressed format */
2221 DEBUG_TRIE_COMPILE_r(dump_trie(trie, widecharmap, revcharmap, depth+1));
2223 RExC_rxi->data->data[ data_slot + 1 ] = (void*)widecharmap;
2225 RExC_rxi->data->data[ data_slot + TRIE_WORDS_OFFSET ] = (void*)trie_words;
2226 RExC_rxi->data->data[ data_slot + 3 ] = (void*)revcharmap;
2228 SvREFCNT_dec(revcharmap);
2232 : trie->startstate>1
2238 S_make_trie_failtable(pTHX_ RExC_state_t *pRExC_state, regnode *source, regnode *stclass, U32 depth)
2240 /* The Trie is constructed and compressed now so we can build a fail array if it's needed
2242 This is basically the Aho-Corasick algorithm. Its from exercise 3.31 and 3.32 in the
2243 "Red Dragon" -- Compilers, principles, techniques, and tools. Aho, Sethi, Ullman 1985/88
2246 We find the fail state for each state in the trie, this state is the longest proper
2247 suffix of the current state's 'word' that is also a proper prefix of another word in our
2248 trie. State 1 represents the word '' and is thus the default fail state. This allows
2249 the DFA not to have to restart after its tried and failed a word at a given point, it
2250 simply continues as though it had been matching the other word in the first place.
2252 'abcdgu'=~/abcdefg|cdgu/
2253 When we get to 'd' we are still matching the first word, we would encounter 'g' which would
2254 fail, which would bring us to the state representing 'd' in the second word where we would
2255 try 'g' and succeed, proceeding to match 'cdgu'.
2257 /* add a fail transition */
2258 const U32 trie_offset = ARG(source);
2259 reg_trie_data *trie=(reg_trie_data *)RExC_rxi->data->data[trie_offset];
2261 const U32 ucharcount = trie->uniquecharcount;
2262 const U32 numstates = trie->statecount;
2263 const U32 ubound = trie->lasttrans + ucharcount;
2267 U32 base = trie->states[ 1 ].trans.base;
2270 const U32 data_slot = add_data( pRExC_state, 1, "T" );
2271 GET_RE_DEBUG_FLAGS_DECL;
2273 PERL_ARGS_ASSERT_MAKE_TRIE_FAILTABLE;
2275 PERL_UNUSED_ARG(depth);
2279 ARG_SET( stclass, data_slot );
2280 aho = (reg_ac_data *) PerlMemShared_calloc( 1, sizeof(reg_ac_data) );
2281 RExC_rxi->data->data[ data_slot ] = (void*)aho;
2282 aho->trie=trie_offset;
2283 aho->states=(reg_trie_state *)PerlMemShared_malloc( numstates * sizeof(reg_trie_state) );
2284 Copy( trie->states, aho->states, numstates, reg_trie_state );
2285 Newxz( q, numstates, U32);
2286 aho->fail = (U32 *) PerlMemShared_calloc( numstates, sizeof(U32) );
2289 /* initialize fail[0..1] to be 1 so that we always have
2290 a valid final fail state */
2291 fail[ 0 ] = fail[ 1 ] = 1;
2293 for ( charid = 0; charid < ucharcount ; charid++ ) {
2294 const U32 newstate = TRIE_TRANS_STATE( 1, base, ucharcount, charid, 0 );
2296 q[ q_write ] = newstate;
2297 /* set to point at the root */
2298 fail[ q[ q_write++ ] ]=1;
2301 while ( q_read < q_write) {
2302 const U32 cur = q[ q_read++ % numstates ];
2303 base = trie->states[ cur ].trans.base;
2305 for ( charid = 0 ; charid < ucharcount ; charid++ ) {
2306 const U32 ch_state = TRIE_TRANS_STATE( cur, base, ucharcount, charid, 1 );
2308 U32 fail_state = cur;
2311 fail_state = fail[ fail_state ];
2312 fail_base = aho->states[ fail_state ].trans.base;
2313 } while ( !TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 ) );
2315 fail_state = TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 );
2316 fail[ ch_state ] = fail_state;
2317 if ( !aho->states[ ch_state ].wordnum && aho->states[ fail_state ].wordnum )
2319 aho->states[ ch_state ].wordnum = aho->states[ fail_state ].wordnum;
2321 q[ q_write++ % numstates] = ch_state;
2325 /* restore fail[0..1] to 0 so that we "fall out" of the AC loop
2326 when we fail in state 1, this allows us to use the
2327 charclass scan to find a valid start char. This is based on the principle
2328 that theres a good chance the string being searched contains lots of stuff
2329 that cant be a start char.
2331 fail[ 0 ] = fail[ 1 ] = 0;
2332 DEBUG_TRIE_COMPILE_r({
2333 PerlIO_printf(Perl_debug_log,
2334 "%*sStclass Failtable (%"UVuf" states): 0",
2335 (int)(depth * 2), "", (UV)numstates
2337 for( q_read=1; q_read<numstates; q_read++ ) {
2338 PerlIO_printf(Perl_debug_log, ", %"UVuf, (UV)fail[q_read]);
2340 PerlIO_printf(Perl_debug_log, "\n");
2343 /*RExC_seen |= REG_SEEN_TRIEDFA;*/
2348 * There are strange code-generation bugs caused on sparc64 by gcc-2.95.2.
2349 * These need to be revisited when a newer toolchain becomes available.
2351 #if defined(__sparc64__) && defined(__GNUC__)
2352 # if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
2353 # undef SPARC64_GCC_WORKAROUND
2354 # define SPARC64_GCC_WORKAROUND 1
2358 #define DEBUG_PEEP(str,scan,depth) \
2359 DEBUG_OPTIMISE_r({if (scan){ \
2360 SV * const mysv=sv_newmortal(); \
2361 regnode *Next = regnext(scan); \
2362 regprop(RExC_rx, mysv, scan); \
2363 PerlIO_printf(Perl_debug_log, "%*s" str ">%3d: %s (%d)\n", \
2364 (int)depth*2, "", REG_NODE_NUM(scan), SvPV_nolen_const(mysv),\
2365 Next ? (REG_NODE_NUM(Next)) : 0 ); \
2372 #define JOIN_EXACT(scan,min,flags) \
2373 if (PL_regkind[OP(scan)] == EXACT) \
2374 join_exact(pRExC_state,(scan),(min),(flags),NULL,depth+1)
2377 S_join_exact(pTHX_ RExC_state_t *pRExC_state, regnode *scan, I32 *min, U32 flags,regnode *val, U32 depth) {
2378 /* Merge several consecutive EXACTish nodes into one. */
2379 regnode *n = regnext(scan);
2381 regnode *next = scan + NODE_SZ_STR(scan);
2385 regnode *stop = scan;
2386 GET_RE_DEBUG_FLAGS_DECL;
2388 PERL_UNUSED_ARG(depth);
2391 PERL_ARGS_ASSERT_JOIN_EXACT;
2392 #ifndef EXPERIMENTAL_INPLACESCAN
2393 PERL_UNUSED_ARG(flags);
2394 PERL_UNUSED_ARG(val);
2396 DEBUG_PEEP("join",scan,depth);
2398 /* Skip NOTHING, merge EXACT*. */
2400 ( PL_regkind[OP(n)] == NOTHING ||
2401 (stringok && (OP(n) == OP(scan))))
2403 && NEXT_OFF(scan) + NEXT_OFF(n) < I16_MAX) {
2405 if (OP(n) == TAIL || n > next)
2407 if (PL_regkind[OP(n)] == NOTHING) {
2408 DEBUG_PEEP("skip:",n,depth);
2409 NEXT_OFF(scan) += NEXT_OFF(n);
2410 next = n + NODE_STEP_REGNODE;
2417 else if (stringok) {
2418 const unsigned int oldl = STR_LEN(scan);
2419 regnode * const nnext = regnext(n);
2421 DEBUG_PEEP("merg",n,depth);
2424 if (oldl + STR_LEN(n) > U8_MAX)
2426 NEXT_OFF(scan) += NEXT_OFF(n);
2427 STR_LEN(scan) += STR_LEN(n);
2428 next = n + NODE_SZ_STR(n);
2429 /* Now we can overwrite *n : */
2430 Move(STRING(n), STRING(scan) + oldl, STR_LEN(n), char);
2438 #ifdef EXPERIMENTAL_INPLACESCAN
2439 if (flags && !NEXT_OFF(n)) {
2440 DEBUG_PEEP("atch", val, depth);
2441 if (reg_off_by_arg[OP(n)]) {
2442 ARG_SET(n, val - n);
2445 NEXT_OFF(n) = val - n;
2452 if (UTF && ( OP(scan) == EXACTF ) && ( STR_LEN(scan) >= 6 ) ) {
2454 Two problematic code points in Unicode casefolding of EXACT nodes:
2456 U+0390 - GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
2457 U+03B0 - GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
2463 U+03B9 U+0308 U+0301 0xCE 0xB9 0xCC 0x88 0xCC 0x81
2464 U+03C5 U+0308 U+0301 0xCF 0x85 0xCC 0x88 0xCC 0x81
2466 This means that in case-insensitive matching (or "loose matching",
2467 as Unicode calls it), an EXACTF of length six (the UTF-8 encoded byte
2468 length of the above casefolded versions) can match a target string
2469 of length two (the byte length of UTF-8 encoded U+0390 or U+03B0).
2470 This would rather mess up the minimum length computation.
2472 What we'll do is to look for the tail four bytes, and then peek
2473 at the preceding two bytes to see whether we need to decrease
2474 the minimum length by four (six minus two).
2476 Thanks to the design of UTF-8, there cannot be false matches:
2477 A sequence of valid UTF-8 bytes cannot be a subsequence of
2478 another valid sequence of UTF-8 bytes.
2481 char * const s0 = STRING(scan), *s, *t;
2482 char * const s1 = s0 + STR_LEN(scan) - 1;
2483 char * const s2 = s1 - 4;
2484 #ifdef EBCDIC /* RD tunifold greek 0390 and 03B0 */
2485 const char t0[] = "\xaf\x49\xaf\x42";
2487 const char t0[] = "\xcc\x88\xcc\x81";
2489 const char * const t1 = t0 + 3;
2492 s < s2 && (t = ninstr(s, s1, t0, t1));
2495 if (((U8)t[-1] == 0x68 && (U8)t[-2] == 0xB4) ||
2496 ((U8)t[-1] == 0x46 && (U8)t[-2] == 0xB5))
2498 if (((U8)t[-1] == 0xB9 && (U8)t[-2] == 0xCE) ||
2499 ((U8)t[-1] == 0x85 && (U8)t[-2] == 0xCF))
2507 n = scan + NODE_SZ_STR(scan);
2509 if (PL_regkind[OP(n)] != NOTHING || OP(n) == NOTHING) {
2516 DEBUG_OPTIMISE_r(if (merged){DEBUG_PEEP("finl",scan,depth)});
2520 /* REx optimizer. Converts nodes into quickier variants "in place".
2521 Finds fixed substrings. */
2523 /* Stops at toplevel WHILEM as well as at "last". At end *scanp is set
2524 to the position after last scanned or to NULL. */
2526 #define INIT_AND_WITHP \
2527 assert(!and_withp); \
2528 Newx(and_withp,1,struct regnode_charclass_class); \
2529 SAVEFREEPV(and_withp)
2531 /* this is a chain of data about sub patterns we are processing that
2532 need to be handled seperately/specially in study_chunk. Its so
2533 we can simulate recursion without losing state. */
2535 typedef struct scan_frame {
2536 regnode *last; /* last node to process in this frame */
2537 regnode *next; /* next node to process when last is reached */
2538 struct scan_frame *prev; /*previous frame*/
2539 I32 stop; /* what stopparen do we use */
2543 #define SCAN_COMMIT(s, data, m) scan_commit(s, data, m, is_inf)
2545 #define CASE_SYNST_FNC(nAmE) \
2547 if (flags & SCF_DO_STCLASS_AND) { \
2548 for (value = 0; value < 256; value++) \
2549 if (!is_ ## nAmE ## _cp(value)) \
2550 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2553 for (value = 0; value < 256; value++) \
2554 if (is_ ## nAmE ## _cp(value)) \
2555 ANYOF_BITMAP_SET(data->start_class, value); \
2559 if (flags & SCF_DO_STCLASS_AND) { \
2560 for (value = 0; value < 256; value++) \
2561 if (is_ ## nAmE ## _cp(value)) \
2562 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2565 for (value = 0; value < 256; value++) \
2566 if (!is_ ## nAmE ## _cp(value)) \
2567 ANYOF_BITMAP_SET(data->start_class, value); \
2574 S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
2575 I32 *minlenp, I32 *deltap,
2580 struct regnode_charclass_class *and_withp,
2581 U32 flags, U32 depth)
2582 /* scanp: Start here (read-write). */
2583 /* deltap: Write maxlen-minlen here. */
2584 /* last: Stop before this one. */
2585 /* data: string data about the pattern */
2586 /* stopparen: treat close N as END */
2587 /* recursed: which subroutines have we recursed into */
2588 /* and_withp: Valid if flags & SCF_DO_STCLASS_OR */
2591 I32 min = 0, pars = 0, code;
2592 regnode *scan = *scanp, *next;
2594 int is_inf = (flags & SCF_DO_SUBSTR) && (data->flags & SF_IS_INF);
2595 int is_inf_internal = 0; /* The studied chunk is infinite */
2596 I32 is_par = OP(scan) == OPEN ? ARG(scan) : 0;
2597 scan_data_t data_fake;
2598 SV *re_trie_maxbuff = NULL;
2599 regnode *first_non_open = scan;
2600 I32 stopmin = I32_MAX;
2601 scan_frame *frame = NULL;
2602 GET_RE_DEBUG_FLAGS_DECL;
2604 PERL_ARGS_ASSERT_STUDY_CHUNK;
2607 StructCopy(&zero_scan_data, &data_fake, scan_data_t);
2611 while (first_non_open && OP(first_non_open) == OPEN)
2612 first_non_open=regnext(first_non_open);
2617 while ( scan && OP(scan) != END && scan < last ){
2618 /* Peephole optimizer: */
2619 DEBUG_STUDYDATA("Peep:", data,depth);
2620 DEBUG_PEEP("Peep",scan,depth);
2621 JOIN_EXACT(scan,&min,0);
2623 /* Follow the next-chain of the current node and optimize
2624 away all the NOTHINGs from it. */
2625 if (OP(scan) != CURLYX) {
2626 const int max = (reg_off_by_arg[OP(scan)]
2628 /* I32 may be smaller than U16 on CRAYs! */
2629 : (I32_MAX < U16_MAX ? I32_MAX : U16_MAX));
2630 int off = (reg_off_by_arg[OP(scan)] ? ARG(scan) : NEXT_OFF(scan));
2634 /* Skip NOTHING and LONGJMP. */
2635 while ((n = regnext(n))
2636 && ((PL_regkind[OP(n)] == NOTHING && (noff = NEXT_OFF(n)))
2637 || ((OP(n) == LONGJMP) && (noff = ARG(n))))
2638 && off + noff < max)
2640 if (reg_off_by_arg[OP(scan)])
2643 NEXT_OFF(scan) = off;
2648 /* The principal pseudo-switch. Cannot be a switch, since we
2649 look into several different things. */
2650 if (OP(scan) == BRANCH || OP(scan) == BRANCHJ
2651 || OP(scan) == IFTHEN) {
2652 next = regnext(scan);
2654 /* demq: the op(next)==code check is to see if we have "branch-branch" AFAICT */
2656 if (OP(next) == code || code == IFTHEN) {
2657 /* NOTE - There is similar code to this block below for handling
2658 TRIE nodes on a re-study. If you change stuff here check there
2660 I32 max1 = 0, min1 = I32_MAX, num = 0;
2661 struct regnode_charclass_class accum;
2662 regnode * const startbranch=scan;
2664 if (flags & SCF_DO_SUBSTR)
2665 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot merge strings after this. */
2666 if (flags & SCF_DO_STCLASS)
2667 cl_init_zero(pRExC_state, &accum);
2669 while (OP(scan) == code) {
2670 I32 deltanext, minnext, f = 0, fake;
2671 struct regnode_charclass_class this_class;
2674 data_fake.flags = 0;
2676 data_fake.whilem_c = data->whilem_c;
2677 data_fake.last_closep = data->last_closep;
2680 data_fake.last_closep = &fake;
2682 data_fake.pos_delta = delta;
2683 next = regnext(scan);
2684 scan = NEXTOPER(scan);
2686 scan = NEXTOPER(scan);
2687 if (flags & SCF_DO_STCLASS) {
2688 cl_init(pRExC_state, &this_class);
2689 data_fake.start_class = &this_class;
2690 f = SCF_DO_STCLASS_AND;
2692 if (flags & SCF_WHILEM_VISITED_POS)
2693 f |= SCF_WHILEM_VISITED_POS;
2695 /* we suppose the run is continuous, last=next...*/
2696 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
2698 stopparen, recursed, NULL, f,depth+1);
2701 if (max1 < minnext + deltanext)
2702 max1 = minnext + deltanext;
2703 if (deltanext == I32_MAX)
2704 is_inf = is_inf_internal = 1;
2706 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
2708 if (data_fake.flags & SCF_SEEN_ACCEPT) {
2709 if ( stopmin > minnext)
2710 stopmin = min + min1;
2711 flags &= ~SCF_DO_SUBSTR;
2713 data->flags |= SCF_SEEN_ACCEPT;
2716 if (data_fake.flags & SF_HAS_EVAL)
2717 data->flags |= SF_HAS_EVAL;
2718 data->whilem_c = data_fake.whilem_c;
2720 if (flags & SCF_DO_STCLASS)
2721 cl_or(pRExC_state, &accum, &this_class);
2723 if (code == IFTHEN && num < 2) /* Empty ELSE branch */
2725 if (flags & SCF_DO_SUBSTR) {
2726 data->pos_min += min1;
2727 data->pos_delta += max1 - min1;
2728 if (max1 != min1 || is_inf)
2729 data->longest = &(data->longest_float);
2732 delta += max1 - min1;
2733 if (flags & SCF_DO_STCLASS_OR) {
2734 cl_or(pRExC_state, data->start_class, &accum);
2736 cl_and(data->start_class, and_withp);
2737 flags &= ~SCF_DO_STCLASS;
2740 else if (flags & SCF_DO_STCLASS_AND) {
2742 cl_and(data->start_class, &accum);
2743 flags &= ~SCF_DO_STCLASS;
2746 /* Switch to OR mode: cache the old value of
2747 * data->start_class */
2749 StructCopy(data->start_class, and_withp,
2750 struct regnode_charclass_class);
2751 flags &= ~SCF_DO_STCLASS_AND;
2752 StructCopy(&accum, data->start_class,
2753 struct regnode_charclass_class);
2754 flags |= SCF_DO_STCLASS_OR;
2755 data->start_class->flags |= ANYOF_EOS;
2759 if (PERL_ENABLE_TRIE_OPTIMISATION && OP( startbranch ) == BRANCH ) {
2762 Assuming this was/is a branch we are dealing with: 'scan' now
2763 points at the item that follows the branch sequence, whatever
2764 it is. We now start at the beginning of the sequence and look
2771 which would be constructed from a pattern like /A|LIST|OF|WORDS/
2773 If we can find such a subseqence we need to turn the first
2774 element into a trie and then add the subsequent branch exact
2775 strings to the trie.
2779 1. patterns where the whole set of branches can be converted.
2781 2. patterns where only a subset can be converted.
2783 In case 1 we can replace the whole set with a single regop
2784 for the trie. In case 2 we need to keep the start and end
2787 'BRANCH EXACT; BRANCH EXACT; BRANCH X'
2788 becomes BRANCH TRIE; BRANCH X;
2790 There is an additional case, that being where there is a
2791 common prefix, which gets split out into an EXACT like node
2792 preceding the TRIE node.
2794 If x(1..n)==tail then we can do a simple trie, if not we make
2795 a "jump" trie, such that when we match the appropriate word
2796 we "jump" to the appopriate tail node. Essentailly we turn
2797 a nested if into a case structure of sorts.
2802 if (!re_trie_maxbuff) {
2803 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
2804 if (!SvIOK(re_trie_maxbuff))
2805 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
2807 if ( SvIV(re_trie_maxbuff)>=0 ) {
2809 regnode *first = (regnode *)NULL;
2810 regnode *last = (regnode *)NULL;
2811 regnode *tail = scan;
2816 SV * const mysv = sv_newmortal(); /* for dumping */
2818 /* var tail is used because there may be a TAIL
2819 regop in the way. Ie, the exacts will point to the
2820 thing following the TAIL, but the last branch will
2821 point at the TAIL. So we advance tail. If we
2822 have nested (?:) we may have to move through several
2826 while ( OP( tail ) == TAIL ) {
2827 /* this is the TAIL generated by (?:) */
2828 tail = regnext( tail );
2833 regprop(RExC_rx, mysv, tail );
2834 PerlIO_printf( Perl_debug_log, "%*s%s%s\n",
2835 (int)depth * 2 + 2, "",
2836 "Looking for TRIE'able sequences. Tail node is: ",
2837 SvPV_nolen_const( mysv )
2843 step through the branches, cur represents each
2844 branch, noper is the first thing to be matched
2845 as part of that branch and noper_next is the
2846 regnext() of that node. if noper is an EXACT
2847 and noper_next is the same as scan (our current
2848 position in the regex) then the EXACT branch is
2849 a possible optimization target. Once we have
2850 two or more consequetive such branches we can
2851 create a trie of the EXACT's contents and stich
2852 it in place. If the sequence represents all of
2853 the branches we eliminate the whole thing and
2854 replace it with a single TRIE. If it is a
2855 subsequence then we need to stitch it in. This
2856 means the first branch has to remain, and needs
2857 to be repointed at the item on the branch chain
2858 following the last branch optimized. This could
2859 be either a BRANCH, in which case the
2860 subsequence is internal, or it could be the
2861 item following the branch sequence in which
2862 case the subsequence is at the end.
2866 /* dont use tail as the end marker for this traverse */
2867 for ( cur = startbranch ; cur != scan ; cur = regnext( cur ) ) {
2868 regnode * const noper = NEXTOPER( cur );
2869 #if defined(DEBUGGING) || defined(NOJUMPTRIE)
2870 regnode * const noper_next = regnext( noper );
2874 regprop(RExC_rx, mysv, cur);
2875 PerlIO_printf( Perl_debug_log, "%*s- %s (%d)",
2876 (int)depth * 2 + 2,"", SvPV_nolen_const( mysv ), REG_NODE_NUM(cur) );
2878 regprop(RExC_rx, mysv, noper);
2879 PerlIO_printf( Perl_debug_log, " -> %s",
2880 SvPV_nolen_const(mysv));
2883 regprop(RExC_rx, mysv, noper_next );
2884 PerlIO_printf( Perl_debug_log,"\t=> %s\t",
2885 SvPV_nolen_const(mysv));
2887 PerlIO_printf( Perl_debug_log, "(First==%d,Last==%d,Cur==%d)\n",
2888 REG_NODE_NUM(first), REG_NODE_NUM(last), REG_NODE_NUM(cur) );
2890 if ( (((first && optype!=NOTHING) ? OP( noper ) == optype
2891 : PL_regkind[ OP( noper ) ] == EXACT )
2892 || OP(noper) == NOTHING )
2894 && noper_next == tail
2899 if ( !first || optype == NOTHING ) {
2900 if (!first) first = cur;
2901 optype = OP( noper );
2907 Currently we do not believe that the trie logic can
2908 handle case insensitive matching properly when the
2909 pattern is not unicode (thus forcing unicode semantics).
2911 If/when this is fixed the following define can be swapped
2912 in below to fully enable trie logic.
2914 #define TRIE_TYPE_IS_SAFE 1
2917 #define TRIE_TYPE_IS_SAFE (UTF || optype==EXACT)
2919 if ( last && TRIE_TYPE_IS_SAFE ) {
2920 make_trie( pRExC_state,
2921 startbranch, first, cur, tail, count,
2924 if ( PL_regkind[ OP( noper ) ] == EXACT
2926 && noper_next == tail
2931 optype = OP( noper );
2941 regprop(RExC_rx, mysv, cur);
2942 PerlIO_printf( Perl_debug_log,
2943 "%*s- %s (%d) <SCAN FINISHED>\n", (int)depth * 2 + 2,
2944 "", SvPV_nolen_const( mysv ),REG_NODE_NUM(cur));
2948 if ( last && TRIE_TYPE_IS_SAFE ) {
2949 made= make_trie( pRExC_state, startbranch, first, scan, tail, count, optype, depth+1 );
2950 #ifdef TRIE_STUDY_OPT
2951 if ( ((made == MADE_EXACT_TRIE &&
2952 startbranch == first)
2953 || ( first_non_open == first )) &&
2955 flags |= SCF_TRIE_RESTUDY;
2956 if ( startbranch == first
2959 RExC_seen &=~REG_TOP_LEVEL_BRANCHES;
2969 else if ( code == BRANCHJ ) { /* single branch is optimized. */
2970 scan = NEXTOPER(NEXTOPER(scan));
2971 } else /* single branch is optimized. */
2972 scan = NEXTOPER(scan);
2974 } else if (OP(scan) == SUSPEND || OP(scan) == GOSUB || OP(scan) == GOSTART) {
2975 scan_frame *newframe = NULL;
2980 if (OP(scan) != SUSPEND) {
2981 /* set the pointer */
2982 if (OP(scan) == GOSUB) {
2984 RExC_recurse[ARG2L(scan)] = scan;
2985 start = RExC_open_parens[paren-1];
2986 end = RExC_close_parens[paren-1];
2989 start = RExC_rxi->program + 1;
2993 Newxz(recursed, (((RExC_npar)>>3) +1), U8);
2994 SAVEFREEPV(recursed);
2996 if (!PAREN_TEST(recursed,paren+1)) {
2997 PAREN_SET(recursed,paren+1);
2998 Newx(newframe,1,scan_frame);
3000 if (flags & SCF_DO_SUBSTR) {
3001 SCAN_COMMIT(pRExC_state,data,minlenp);
3002 data->longest = &(data->longest_float);
3004 is_inf = is_inf_internal = 1;
3005 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3006 cl_anything(pRExC_state, data->start_class);
3007 flags &= ~SCF_DO_STCLASS;
3010 Newx(newframe,1,scan_frame);
3013 end = regnext(scan);
3018 SAVEFREEPV(newframe);
3019 newframe->next = regnext(scan);
3020 newframe->last = last;
3021 newframe->stop = stopparen;
3022 newframe->prev = frame;
3032 else if (OP(scan) == EXACT) {
3033 I32 l = STR_LEN(scan);
3036 const U8 * const s = (U8*)STRING(scan);
3037 l = utf8_length(s, s + l);
3038 uc = utf8_to_uvchr(s, NULL);
3040 uc = *((U8*)STRING(scan));
3043 if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
3044 /* The code below prefers earlier match for fixed
3045 offset, later match for variable offset. */
3046 if (data->last_end == -1) { /* Update the start info. */
3047 data->last_start_min = data->pos_min;
3048 data->last_start_max = is_inf
3049 ? I32_MAX : data->pos_min + data->pos_delta;
3051 sv_catpvn(data->last_found, STRING(scan), STR_LEN(scan));
3053 SvUTF8_on(data->last_found);
3055 SV * const sv = data->last_found;
3056 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3057 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3058 if (mg && mg->mg_len >= 0)
3059 mg->mg_len += utf8_length((U8*)STRING(scan),
3060 (U8*)STRING(scan)+STR_LEN(scan));
3062 data->last_end = data->pos_min + l;
3063 data->pos_min += l; /* As in the first entry. */
3064 data->flags &= ~SF_BEFORE_EOL;
3066 if (flags & SCF_DO_STCLASS_AND) {
3067 /* Check whether it is compatible with what we know already! */
3071 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3072 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3073 && (!(data->start_class->flags & ANYOF_FOLD)
3074 || !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
3077 ANYOF_CLASS_ZERO(data->start_class);
3078 ANYOF_BITMAP_ZERO(data->start_class);
3080 ANYOF_BITMAP_SET(data->start_class, uc);
3081 data->start_class->flags &= ~ANYOF_EOS;
3083 data->start_class->flags &= ~ANYOF_UNICODE_ALL;
3085 else if (flags & SCF_DO_STCLASS_OR) {
3086 /* false positive possible if the class is case-folded */
3088 ANYOF_BITMAP_SET(data->start_class, uc);
3090 data->start_class->flags |= ANYOF_UNICODE_ALL;
3091 data->start_class->flags &= ~ANYOF_EOS;
3092 cl_and(data->start_class, and_withp);
3094 flags &= ~SCF_DO_STCLASS;
3096 else if (PL_regkind[OP(scan)] == EXACT) { /* But OP != EXACT! */
3097 I32 l = STR_LEN(scan);
3098 UV uc = *((U8*)STRING(scan));
3100 /* Search for fixed substrings supports EXACT only. */
3101 if (flags & SCF_DO_SUBSTR) {
3103 SCAN_COMMIT(pRExC_state, data, minlenp);
3106 const U8 * const s = (U8 *)STRING(scan);
3107 l = utf8_length(s, s + l);
3108 uc = utf8_to_uvchr(s, NULL);
3111 if (flags & SCF_DO_SUBSTR)
3113 if (flags & SCF_DO_STCLASS_AND) {
3114 /* Check whether it is compatible with what we know already! */
3118 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3119 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3120 && !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
3122 ANYOF_CLASS_ZERO(data->start_class);
3123 ANYOF_BITMAP_ZERO(data->start_class);
3125 ANYOF_BITMAP_SET(data->start_class, uc);
3126 data->start_class->flags &= ~ANYOF_EOS;
3127 data->start_class->flags |= ANYOF_FOLD;
3128 if (OP(scan) == EXACTFL)
3129 data->start_class->flags |= ANYOF_LOCALE;
3132 else if (flags & SCF_DO_STCLASS_OR) {
3133 if (data->start_class->flags & ANYOF_FOLD) {
3134 /* false positive possible if the class is case-folded.
3135 Assume that the locale settings are the same... */
3137 ANYOF_BITMAP_SET(data->start_class, uc);
3138 data->start_class->flags &= ~ANYOF_EOS;
3140 cl_and(data->start_class, and_withp);
3142 flags &= ~SCF_DO_STCLASS;
3144 else if (REGNODE_VARIES(OP(scan))) {
3145 I32 mincount, maxcount, minnext, deltanext, fl = 0;
3146 I32 f = flags, pos_before = 0;
3147 regnode * const oscan = scan;
3148 struct regnode_charclass_class this_class;
3149 struct regnode_charclass_class *oclass = NULL;
3150 I32 next_is_eval = 0;
3152 switch (PL_regkind[OP(scan)]) {
3153 case WHILEM: /* End of (?:...)* . */
3154 scan = NEXTOPER(scan);
3157 if (flags & (SCF_DO_SUBSTR | SCF_DO_STCLASS)) {
3158 next = NEXTOPER(scan);
3159 if (OP(next) == EXACT || (flags & SCF_DO_STCLASS)) {
3161 maxcount = REG_INFTY;
3162 next = regnext(scan);
3163 scan = NEXTOPER(scan);
3167 if (flags & SCF_DO_SUBSTR)
3172 if (flags & SCF_DO_STCLASS) {
3174 maxcount = REG_INFTY;
3175 next = regnext(scan);
3176 scan = NEXTOPER(scan);
3179 is_inf = is_inf_internal = 1;
3180 scan = regnext(scan);
3181 if (flags & SCF_DO_SUBSTR) {
3182 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot extend fixed substrings */
3183 data->longest = &(data->longest_float);
3185 goto optimize_curly_tail;
3187 if (stopparen>0 && (OP(scan)==CURLYN || OP(scan)==CURLYM)
3188 && (scan->flags == stopparen))
3193 mincount = ARG1(scan);
3194 maxcount = ARG2(scan);
3196 next = regnext(scan);
3197 if (OP(scan) == CURLYX) {
3198 I32 lp = (data ? *(data->last_closep) : 0);
3199 scan->flags = ((lp <= (I32)U8_MAX) ? (U8)lp : U8_MAX);
3201 scan = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
3202 next_is_eval = (OP(scan) == EVAL);
3204 if (flags & SCF_DO_SUBSTR) {
3205 if (mincount == 0) SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot extend fixed substrings */
3206 pos_before = data->pos_min;
3210 data->flags &= ~(SF_HAS_PAR|SF_IN_PAR|SF_HAS_EVAL);
3212 data->flags |= SF_IS_INF;
3214 if (flags & SCF_DO_STCLASS) {
3215 cl_init(pRExC_state, &this_class);
3216 oclass = data->start_class;
3217 data->start_class = &this_class;
3218 f |= SCF_DO_STCLASS_AND;
3219 f &= ~SCF_DO_STCLASS_OR;
3221 /* These are the cases when once a subexpression
3222 fails at a particular position, it cannot succeed
3223 even after backtracking at the enclosing scope.
3225 XXXX what if minimal match and we are at the
3226 initial run of {n,m}? */
3227 if ((mincount != maxcount - 1) && (maxcount != REG_INFTY))
3228 f &= ~SCF_WHILEM_VISITED_POS;
3230 /* This will finish on WHILEM, setting scan, or on NULL: */
3231 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
3232 last, data, stopparen, recursed, NULL,
3234 ? (f & ~SCF_DO_SUBSTR) : f),depth+1);
3236 if (flags & SCF_DO_STCLASS)
3237 data->start_class = oclass;
3238 if (mincount == 0 || minnext == 0) {
3239 if (flags & SCF_DO_STCLASS_OR) {
3240 cl_or(pRExC_state, data->start_class, &this_class);
3242 else if (flags & SCF_DO_STCLASS_AND) {
3243 /* Switch to OR mode: cache the old value of
3244 * data->start_class */
3246 StructCopy(data->start_class, and_withp,
3247 struct regnode_charclass_class);
3248 flags &= ~SCF_DO_STCLASS_AND;
3249 StructCopy(&this_class, data->start_class,
3250 struct regnode_charclass_class);
3251 flags |= SCF_DO_STCLASS_OR;
3252 data->start_class->flags |= ANYOF_EOS;
3254 } else { /* Non-zero len */
3255 if (flags & SCF_DO_STCLASS_OR) {
3256 cl_or(pRExC_state, data->start_class, &this_class);
3257 cl_and(data->start_class, and_withp);
3259 else if (flags & SCF_DO_STCLASS_AND)
3260 cl_and(data->start_class, &this_class);
3261 flags &= ~SCF_DO_STCLASS;
3263 if (!scan) /* It was not CURLYX, but CURLY. */
3265 if ( /* ? quantifier ok, except for (?{ ... }) */
3266 (next_is_eval || !(mincount == 0 && maxcount == 1))
3267 && (minnext == 0) && (deltanext == 0)
3268 && data && !(data->flags & (SF_HAS_PAR|SF_IN_PAR))
3269 && maxcount <= REG_INFTY/3) /* Complement check for big count */
3271 ckWARNreg(RExC_parse,
3272 "Quantifier unexpected on zero-length expression");
3275 min += minnext * mincount;
3276 is_inf_internal |= ((maxcount == REG_INFTY
3277 && (minnext + deltanext) > 0)
3278 || deltanext == I32_MAX);
3279 is_inf |= is_inf_internal;
3280 delta += (minnext + deltanext) * maxcount - minnext * mincount;
3282 /* Try powerful optimization CURLYX => CURLYN. */
3283 if ( OP(oscan) == CURLYX && data
3284 && data->flags & SF_IN_PAR
3285 && !(data->flags & SF_HAS_EVAL)
3286 && !deltanext && minnext == 1 ) {
3287 /* Try to optimize to CURLYN. */
3288 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS;
3289 regnode * const nxt1 = nxt;
3296 if (!REGNODE_SIMPLE(OP(nxt))
3297 && !(PL_regkind[OP(nxt)] == EXACT
3298 && STR_LEN(nxt) == 1))
3304 if (OP(nxt) != CLOSE)
3306 if (RExC_open_parens) {
3307 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3308 RExC_close_parens[ARG(nxt1)-1]=nxt+2; /*close->while*/
3310 /* Now we know that nxt2 is the only contents: */
3311 oscan->flags = (U8)ARG(nxt);
3313 OP(nxt1) = NOTHING; /* was OPEN. */
3316 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3317 NEXT_OFF(nxt1+ 1) = 0; /* just for consistency. */
3318 NEXT_OFF(nxt2) = 0; /* just for consistency with CURLY. */
3319 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3320 OP(nxt + 1) = OPTIMIZED; /* was count. */
3321 NEXT_OFF(nxt+ 1) = 0; /* just for consistency. */
3326 /* Try optimization CURLYX => CURLYM. */
3327 if ( OP(oscan) == CURLYX && data
3328 && !(data->flags & SF_HAS_PAR)
3329 && !(data->flags & SF_HAS_EVAL)
3330 && !deltanext /* atom is fixed width */
3331 && minnext != 0 /* CURLYM can't handle zero width */
3333 /* XXXX How to optimize if data == 0? */
3334 /* Optimize to a simpler form. */
3335 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN */
3339 while ( (nxt2 = regnext(nxt)) /* skip over embedded stuff*/
3340 && (OP(nxt2) != WHILEM))
3342 OP(nxt2) = SUCCEED; /* Whas WHILEM */
3343 /* Need to optimize away parenths. */
3344 if ((data->flags & SF_IN_PAR) && OP(nxt) == CLOSE) {
3345 /* Set the parenth number. */
3346 regnode *nxt1 = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN*/
3348 oscan->flags = (U8)ARG(nxt);
3349 if (RExC_open_parens) {
3350 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3351 RExC_close_parens[ARG(nxt1)-1]=nxt2+1; /*close->NOTHING*/
3353 OP(nxt1) = OPTIMIZED; /* was OPEN. */
3354 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3357 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3358 OP(nxt + 1) = OPTIMIZED; /* was count. */
3359 NEXT_OFF(nxt1 + 1) = 0; /* just for consistancy. */
3360 NEXT_OFF(nxt + 1) = 0; /* just for consistancy. */
3363 while ( nxt1 && (OP(nxt1) != WHILEM)) {
3364 regnode *nnxt = regnext(nxt1);
3366 if (reg_off_by_arg[OP(nxt1)])
3367 ARG_SET(nxt1, nxt2 - nxt1);
3368 else if (nxt2 - nxt1 < U16_MAX)
3369 NEXT_OFF(nxt1) = nxt2 - nxt1;
3371 OP(nxt) = NOTHING; /* Cannot beautify */
3376 /* Optimize again: */
3377 study_chunk(pRExC_state, &nxt1, minlenp, &deltanext, nxt,
3378 NULL, stopparen, recursed, NULL, 0,depth+1);
3383 else if ((OP(oscan) == CURLYX)
3384 && (flags & SCF_WHILEM_VISITED_POS)
3385 /* See the comment on a similar expression above.
3386 However, this time it's not a subexpression
3387 we care about, but the expression itself. */
3388 && (maxcount == REG_INFTY)
3389 && data && ++data->whilem_c < 16) {
3390 /* This stays as CURLYX, we can put the count/of pair. */
3391 /* Find WHILEM (as in regexec.c) */
3392 regnode *nxt = oscan + NEXT_OFF(oscan);
3394 if (OP(PREVOPER(nxt)) == NOTHING) /* LONGJMP */
3396 PREVOPER(nxt)->flags = (U8)(data->whilem_c
3397 | (RExC_whilem_seen << 4)); /* On WHILEM */
3399 if (data && fl & (SF_HAS_PAR|SF_IN_PAR))
3401 if (flags & SCF_DO_SUBSTR) {
3402 SV *last_str = NULL;
3403 int counted = mincount != 0;
3405 if (data->last_end > 0 && mincount != 0) { /* Ends with a string. */
3406 #if defined(SPARC64_GCC_WORKAROUND)
3409 const char *s = NULL;
3412 if (pos_before >= data->last_start_min)
3415 b = data->last_start_min;
3418 s = SvPV_const(data->last_found, l);
3419 old = b - data->last_start_min;
3422 I32 b = pos_before >= data->last_start_min
3423 ? pos_before : data->last_start_min;
3425 const char * const s = SvPV_const(data->last_found, l);
3426 I32 old = b - data->last_start_min;
3430 old = utf8_hop((U8*)s, old) - (U8*)s;
3432 /* Get the added string: */
3433 last_str = newSVpvn_utf8(s + old, l, UTF);
3434 if (deltanext == 0 && pos_before == b) {
3435 /* What was added is a constant string */
3437 SvGROW(last_str, (mincount * l) + 1);
3438 repeatcpy(SvPVX(last_str) + l,
3439 SvPVX_const(last_str), l, mincount - 1);
3440 SvCUR_set(last_str, SvCUR(last_str) * mincount);
3441 /* Add additional parts. */
3442 SvCUR_set(data->last_found,
3443 SvCUR(data->last_found) - l);
3444 sv_catsv(data->last_found, last_str);
3446 SV * sv = data->last_found;
3448 SvUTF8(sv) && SvMAGICAL(sv) ?
3449 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3450 if (mg && mg->mg_len >= 0)
3451 mg->mg_len += CHR_SVLEN(last_str) - l;
3453 data->last_end += l * (mincount - 1);
3456 /* start offset must point into the last copy */
3457 data->last_start_min += minnext * (mincount - 1);
3458 data->last_start_max += is_inf ? I32_MAX
3459 : (maxcount - 1) * (minnext + data->pos_delta);
3462 /* It is counted once already... */
3463 data->pos_min += minnext * (mincount - counted);
3464 data->pos_delta += - counted * deltanext +
3465 (minnext + deltanext) * maxcount - minnext * mincount;
3466 if (mincount != maxcount) {
3467 /* Cannot extend fixed substrings found inside
3469 SCAN_COMMIT(pRExC_state,data,minlenp);
3470 if (mincount && last_str) {
3471 SV * const sv = data->last_found;
3472 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3473 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3477 sv_setsv(sv, last_str);
3478 data->last_end = data->pos_min;
3479 data->last_start_min =
3480 data->pos_min - CHR_SVLEN(last_str);
3481 data->last_start_max = is_inf
3483 : data->pos_min + data->pos_delta
3484 - CHR_SVLEN(last_str);
3486 data->longest = &(data->longest_float);
3488 SvREFCNT_dec(last_str);
3490 if (data && (fl & SF_HAS_EVAL))
3491 data->flags |= SF_HAS_EVAL;
3492 optimize_curly_tail:
3493 if (OP(oscan) != CURLYX) {
3494 while (PL_regkind[OP(next = regnext(oscan))] == NOTHING
3496 NEXT_OFF(oscan) += NEXT_OFF(next);
3499 default: /* REF and CLUMP only? */
3500 if (flags & SCF_DO_SUBSTR) {
3501 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3502 data->longest = &(data->longest_float);
3504 is_inf = is_inf_internal = 1;
3505 if (flags & SCF_DO_STCLASS_OR)
3506 cl_anything(pRExC_state, data->start_class);
3507 flags &= ~SCF_DO_STCLASS;
3511 else if (OP(scan) == LNBREAK) {
3512 if (flags & SCF_DO_STCLASS) {
3514 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3515 if (flags & SCF_DO_STCLASS_AND) {
3516 for (value = 0; value < 256; value++)
3517 if (!is_VERTWS_cp(value))
3518 ANYOF_BITMAP_CLEAR(data->start_class, value);
3521 for (value = 0; value < 256; value++)
3522 if (is_VERTWS_cp(value))
3523 ANYOF_BITMAP_SET(data->start_class, value);
3525 if (flags & SCF_DO_STCLASS_OR)
3526 cl_and(data->start_class, and_withp);
3527 flags &= ~SCF_DO_STCLASS;
3531 if (flags & SCF_DO_SUBSTR) {
3532 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3534 data->pos_delta += 1;
3535 data->longest = &(data->longest_float);
3538 else if (OP(scan) == FOLDCHAR) {
3539 int d = ARG(scan)==0xDF ? 1 : 2;
3540 flags &= ~SCF_DO_STCLASS;
3543 if (flags & SCF_DO_SUBSTR) {
3544 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3546 data->pos_delta += d;
3547 data->longest = &(data->longest_float);
3550 else if (REGNODE_SIMPLE(OP(scan))) {
3553 if (flags & SCF_DO_SUBSTR) {
3554 SCAN_COMMIT(pRExC_state,data,minlenp);
3558 if (flags & SCF_DO_STCLASS) {
3559 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3561 /* Some of the logic below assumes that switching
3562 locale on will only add false positives. */
3563 switch (PL_regkind[OP(scan)]) {
3567 /* Perl_croak(aTHX_ "panic: unexpected simple REx opcode %d", OP(scan)); */
3568 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3569 cl_anything(pRExC_state, data->start_class);
3572 if (OP(scan) == SANY)
3574 if (flags & SCF_DO_STCLASS_OR) { /* Everything but \n */
3575 value = (ANYOF_BITMAP_TEST(data->start_class,'\n')
3576 || (data->start_class->flags & ANYOF_CLASS));
3577 cl_anything(pRExC_state, data->start_class);
3579 if (flags & SCF_DO_STCLASS_AND || !value)
3580 ANYOF_BITMAP_CLEAR(data->start_class,'\n');
3583 if (flags & SCF_DO_STCLASS_AND)
3584 cl_and(data->start_class,
3585 (struct regnode_charclass_class*)scan);
3587 cl_or(pRExC_state, data->start_class,
3588 (struct regnode_charclass_class*)scan);
3591 if (flags & SCF_DO_STCLASS_AND) {
3592 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3593 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3594 if (FLAGS(scan) & USE_UNI) {
3595 for (value = 0; value < 256; value++) {
3596 if (!isWORDCHAR_L1(value)) {
3597 ANYOF_BITMAP_CLEAR(data->start_class, value);
3601 for (value = 0; value < 256; value++) {
3602 if (!isALNUM(value)) {
3603 ANYOF_BITMAP_CLEAR(data->start_class, value);
3610 if (data->start_class->flags & ANYOF_LOCALE)
3611 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3612 else if (FLAGS(scan) & USE_UNI) {
3613 for (value = 0; value < 256; value++) {
3614 if (isWORDCHAR_L1(value)) {
3615 ANYOF_BITMAP_SET(data->start_class, value);
3619 for (value = 0; value < 256; value++) {
3620 if (isALNUM(value)) {
3621 ANYOF_BITMAP_SET(data->start_class, value);
3628 if (flags & SCF_DO_STCLASS_AND) {
3629 if (data->start_class->flags & ANYOF_LOCALE)
3630 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3633 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3634 data->start_class->flags |= ANYOF_LOCALE;
3638 if (flags & SCF_DO_STCLASS_AND) {
3639 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3640 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3641 if (FLAGS(scan) & USE_UNI) {
3642 for (value = 0; value < 256; value++) {
3643 if (isWORDCHAR_L1(value)) {
3644 ANYOF_BITMAP_CLEAR(data->start_class, value);
3648 for (value = 0; value < 256; value++) {
3649 if (isALNUM(value)) {
3650 ANYOF_BITMAP_CLEAR(data->start_class, value);
3657 if (data->start_class->flags & ANYOF_LOCALE)
3658 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3660 for (value = 0; value < 256; value++)
3661 if (!isALNUM(value))
3662 ANYOF_BITMAP_SET(data->start_class, value);
3667 if (flags & SCF_DO_STCLASS_AND) {
3668 if (data->start_class->flags & ANYOF_LOCALE)
3669 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3672 data->start_class->flags |= ANYOF_LOCALE;
3673 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3677 if (flags & SCF_DO_STCLASS_AND) {
3678 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3679 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3680 if (FLAGS(scan) & USE_UNI) {
3681 for (value = 0; value < 256; value++) {
3682 if (!isSPACE_L1(value)) {
3683 ANYOF_BITMAP_CLEAR(data->start_class, value);
3687 for (value = 0; value < 256; value++) {
3688 if (!isSPACE(value)) {
3689 ANYOF_BITMAP_CLEAR(data->start_class, value);
3696 if (data->start_class->flags & ANYOF_LOCALE) {
3697 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3699 else if (FLAGS(scan) & USE_UNI) {
3700 for (value = 0; value < 256; value++) {
3701 if (isSPACE_L1(value)) {
3702 ANYOF_BITMAP_SET(data->start_class, value);
3706 for (value = 0; value < 256; value++) {
3707 if (isSPACE(value)) {
3708 ANYOF_BITMAP_SET(data->start_class, value);
3715 if (flags & SCF_DO_STCLASS_AND) {
3716 if (data->start_class->flags & ANYOF_LOCALE)
3717 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3720 data->start_class->flags |= ANYOF_LOCALE;
3721 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3725 if (flags & SCF_DO_STCLASS_AND) {
3726 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3727 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3728 if (FLAGS(scan) & USE_UNI) {
3729 for (value = 0; value < 256; value++) {
3730 if (isSPACE_L1(value)) {
3731 ANYOF_BITMAP_CLEAR(data->start_class, value);
3735 for (value = 0; value < 256; value++) {
3736 if (isSPACE(value)) {
3737 ANYOF_BITMAP_CLEAR(data->start_class, value);
3744 if (data->start_class->flags & ANYOF_LOCALE)
3745 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3746 else if (FLAGS(scan) & USE_UNI) {
3747 for (value = 0; value < 256; value++) {
3748 if (!isSPACE_L1(value)) {
3749 ANYOF_BITMAP_SET(data->start_class, value);
3754 for (value = 0; value < 256; value++) {
3755 if (!isSPACE(value)) {
3756 ANYOF_BITMAP_SET(data->start_class, value);
3763 if (flags & SCF_DO_STCLASS_AND) {
3764 if (data->start_class->flags & ANYOF_LOCALE) {
3765 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3766 for (value = 0; value < 256; value++)
3767 if (!isSPACE(value))
3768 ANYOF_BITMAP_CLEAR(data->start_class, value);
3772 data->start_class->flags |= ANYOF_LOCALE;
3773 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3777 if (flags & SCF_DO_STCLASS_AND) {
3778 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NDIGIT);
3779 for (value = 0; value < 256; value++)
3780 if (!isDIGIT(value))
3781 ANYOF_BITMAP_CLEAR(data->start_class, value);
3784 if (data->start_class->flags & ANYOF_LOCALE)
3785 ANYOF_CLASS_SET(data->start_class,ANYOF_DIGIT);
3787 for (value = 0; value < 256; value++)
3789 ANYOF_BITMAP_SET(data->start_class, value);
3794 if (flags & SCF_DO_STCLASS_AND) {
3795 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_DIGIT);
3796 for (value = 0; value < 256; value++)
3798 ANYOF_BITMAP_CLEAR(data->start_class, value);
3801 if (data->start_class->flags & ANYOF_LOCALE)
3802 ANYOF_CLASS_SET(data->start_class,ANYOF_NDIGIT);
3804 for (value = 0; value < 256; value++)
3805 if (!isDIGIT(value))
3806 ANYOF_BITMAP_SET(data->start_class, value);
3810 CASE_SYNST_FNC(VERTWS);
3811 CASE_SYNST_FNC(HORIZWS);
3814 if (flags & SCF_DO_STCLASS_OR)
3815 cl_and(data->start_class, and_withp);
3816 flags &= ~SCF_DO_STCLASS;
3819 else if (PL_regkind[OP(scan)] == EOL && flags & SCF_DO_SUBSTR) {
3820 data->flags |= (OP(scan) == MEOL
3824 else if ( PL_regkind[OP(scan)] == BRANCHJ
3825 /* Lookbehind, or need to calculate parens/evals/stclass: */
3826 && (scan->flags || data || (flags & SCF_DO_STCLASS))
3827 && (OP(scan) == IFMATCH || OP(scan) == UNLESSM)) {
3828 if ( !PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3829 || OP(scan) == UNLESSM )
3831 /* Negative Lookahead/lookbehind
3832 In this case we can't do fixed string optimisation.
3835 I32 deltanext, minnext, fake = 0;
3837 struct regnode_charclass_class intrnl;
3840 data_fake.flags = 0;
3842 data_fake.whilem_c = data->whilem_c;
3843 data_fake.last_closep = data->last_closep;
3846 data_fake.last_closep = &fake;
3847 data_fake.pos_delta = delta;
3848 if ( flags & SCF_DO_STCLASS && !scan->flags
3849 && OP(scan) == IFMATCH ) { /* Lookahead */
3850 cl_init(pRExC_state, &intrnl);
3851 data_fake.start_class = &intrnl;
3852 f |= SCF_DO_STCLASS_AND;
3854 if (flags & SCF_WHILEM_VISITED_POS)
3855 f |= SCF_WHILEM_VISITED_POS;
3856 next = regnext(scan);
3857 nscan = NEXTOPER(NEXTOPER(scan));
3858 minnext = study_chunk(pRExC_state, &nscan, minlenp, &deltanext,
3859 last, &data_fake, stopparen, recursed, NULL, f, depth+1);
3862 FAIL("Variable length lookbehind not implemented");
3864 else if (minnext > (I32)U8_MAX) {
3865 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3867 scan->flags = (U8)minnext;
3870 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3872 if (data_fake.flags & SF_HAS_EVAL)
3873 data->flags |= SF_HAS_EVAL;
3874 data->whilem_c = data_fake.whilem_c;
3876 if (f & SCF_DO_STCLASS_AND) {
3877 if (flags & SCF_DO_STCLASS_OR) {
3878 /* OR before, AND after: ideally we would recurse with
3879 * data_fake to get the AND applied by study of the
3880 * remainder of the pattern, and then derecurse;
3881 * *** HACK *** for now just treat as "no information".
3882 * See [perl #56690].
3884 cl_init(pRExC_state, data->start_class);
3886 /* AND before and after: combine and continue */
3887 const int was = (data->start_class->flags & ANYOF_EOS);
3889 cl_and(data->start_class, &intrnl);
3891 data->start_class->flags |= ANYOF_EOS;
3895 #if PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3897 /* Positive Lookahead/lookbehind
3898 In this case we can do fixed string optimisation,
3899 but we must be careful about it. Note in the case of
3900 lookbehind the positions will be offset by the minimum
3901 length of the pattern, something we won't know about
3902 until after the recurse.
3904 I32 deltanext, fake = 0;
3906 struct regnode_charclass_class intrnl;
3908 /* We use SAVEFREEPV so that when the full compile
3909 is finished perl will clean up the allocated
3910 minlens when it's all done. This way we don't
3911 have to worry about freeing them when we know
3912 they wont be used, which would be a pain.
3915 Newx( minnextp, 1, I32 );
3916 SAVEFREEPV(minnextp);
3919 StructCopy(data, &data_fake, scan_data_t);
3920 if ((flags & SCF_DO_SUBSTR) && data->last_found) {
3923 SCAN_COMMIT(pRExC_state, &data_fake,minlenp);
3924 data_fake.last_found=newSVsv(data->last_found);
3928 data_fake.last_closep = &fake;
3929 data_fake.flags = 0;
3930 data_fake.pos_delta = delta;
3932 data_fake.flags |= SF_IS_INF;
3933 if ( flags & SCF_DO_STCLASS && !scan->flags
3934 && OP(scan) == IFMATCH ) { /* Lookahead */
3935 cl_init(pRExC_state, &intrnl);
3936 data_fake.start_class = &intrnl;
3937 f |= SCF_DO_STCLASS_AND;
3939 if (flags & SCF_WHILEM_VISITED_POS)
3940 f |= SCF_WHILEM_VISITED_POS;
3941 next = regnext(scan);
3942 nscan = NEXTOPER(NEXTOPER(scan));
3944 *minnextp = study_chunk(pRExC_state, &nscan, minnextp, &deltanext,
3945 last, &data_fake, stopparen, recursed, NULL, f,depth+1);
3948 FAIL("Variable length lookbehind not implemented");
3950 else if (*minnextp > (I32)U8_MAX) {
3951 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3953 scan->flags = (U8)*minnextp;
3958 if (f & SCF_DO_STCLASS_AND) {
3959 const int was = (data->start_class->flags & ANYOF_EOS);
3961 cl_and(data->start_class, &intrnl);
3963 data->start_class->flags |= ANYOF_EOS;
3966 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3968 if (data_fake.flags & SF_HAS_EVAL)
3969 data->flags |= SF_HAS_EVAL;
3970 data->whilem_c = data_fake.whilem_c;
3971 if ((flags & SCF_DO_SUBSTR) && data_fake.last_found) {
3972 if (RExC_rx->minlen<*minnextp)
3973 RExC_rx->minlen=*minnextp;
3974 SCAN_COMMIT(pRExC_state, &data_fake, minnextp);
3975 SvREFCNT_dec(data_fake.last_found);
3977 if ( data_fake.minlen_fixed != minlenp )
3979 data->offset_fixed= data_fake.offset_fixed;
3980 data->minlen_fixed= data_fake.minlen_fixed;
3981 data->lookbehind_fixed+= scan->flags;
3983 if ( data_fake.minlen_float != minlenp )
3985 data->minlen_float= data_fake.minlen_float;
3986 data->offset_float_min=data_fake.offset_float_min;
3987 data->offset_float_max=data_fake.offset_float_max;
3988 data->lookbehind_float+= scan->flags;
3997 else if (OP(scan) == OPEN) {
3998 if (stopparen != (I32)ARG(scan))
4001 else if (OP(scan) == CLOSE) {
4002 if (stopparen == (I32)ARG(scan)) {
4005 if ((I32)ARG(scan) == is_par) {
4006 next = regnext(scan);
4008 if ( next && (OP(next) != WHILEM) && next < last)
4009 is_par = 0; /* Disable optimization */
4012 *(data->last_closep) = ARG(scan);
4014 else if (OP(scan) == EVAL) {
4016 data->flags |= SF_HAS_EVAL;
4018 else if ( PL_regkind[OP(scan)] == ENDLIKE ) {
4019 if (flags & SCF_DO_SUBSTR) {
4020 SCAN_COMMIT(pRExC_state,data,minlenp);
4021 flags &= ~SCF_DO_SUBSTR;
4023 if (data && OP(scan)==ACCEPT) {
4024 data->flags |= SCF_SEEN_ACCEPT;
4029 else if (OP(scan) == LOGICAL && scan->flags == 2) /* Embedded follows */
4031 if (flags & SCF_DO_SUBSTR) {
4032 SCAN_COMMIT(pRExC_state,data,minlenp);
4033 data->longest = &(data->longest_float);
4035 is_inf = is_inf_internal = 1;
4036 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
4037 cl_anything(pRExC_state, data->start_class);
4038 flags &= ~SCF_DO_STCLASS;
4040 else if (OP(scan) == GPOS) {
4041 if (!(RExC_rx->extflags & RXf_GPOS_FLOAT) &&
4042 !(delta || is_inf || (data && data->pos_delta)))
4044 if (!(RExC_rx->extflags & RXf_ANCH) && (flags & SCF_DO_SUBSTR))
4045 RExC_rx->extflags |= RXf_ANCH_GPOS;
4046 if (RExC_rx->gofs < (U32)min)
4047 RExC_rx->gofs = min;
4049 RExC_rx->extflags |= RXf_GPOS_FLOAT;
4053 #ifdef TRIE_STUDY_OPT
4054 #ifdef FULL_TRIE_STUDY
4055 else if (PL_regkind[OP(scan)] == TRIE) {
4056 /* NOTE - There is similar code to this block above for handling
4057 BRANCH nodes on the initial study. If you change stuff here
4059 regnode *trie_node= scan;
4060 regnode *tail= regnext(scan);
4061 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
4062 I32 max1 = 0, min1 = I32_MAX;
4063 struct regnode_charclass_class accum;
4065 if (flags & SCF_DO_SUBSTR) /* XXXX Add !SUSPEND? */
4066 SCAN_COMMIT(pRExC_state, data,minlenp); /* Cannot merge strings after this. */
4067 if (flags & SCF_DO_STCLASS)
4068 cl_init_zero(pRExC_state, &accum);
4074 const regnode *nextbranch= NULL;
4077 for ( word=1 ; word <= trie->wordcount ; word++)
4079 I32 deltanext=0, minnext=0, f = 0, fake;
4080 struct regnode_charclass_class this_class;
4082 data_fake.flags = 0;
4084 data_fake.whilem_c = data->whilem_c;
4085 data_fake.last_closep = data->last_closep;
4088 data_fake.last_closep = &fake;
4089 data_fake.pos_delta = delta;
4090 if (flags & SCF_DO_STCLASS) {
4091 cl_init(pRExC_state, &this_class);
4092 data_fake.start_class = &this_class;
4093 f = SCF_DO_STCLASS_AND;
4095 if (flags & SCF_WHILEM_VISITED_POS)
4096 f |= SCF_WHILEM_VISITED_POS;
4098 if (trie->jump[word]) {
4100 nextbranch = trie_node + trie->jump[0];
4101 scan= trie_node + trie->jump[word];
4102 /* We go from the jump point to the branch that follows
4103 it. Note this means we need the vestigal unused branches
4104 even though they arent otherwise used.
4106 minnext = study_chunk(pRExC_state, &scan, minlenp,
4107 &deltanext, (regnode *)nextbranch, &data_fake,
4108 stopparen, recursed, NULL, f,depth+1);
4110 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
4111 nextbranch= regnext((regnode*)nextbranch);
4113 if (min1 > (I32)(minnext + trie->minlen))
4114 min1 = minnext + trie->minlen;
4115 if (max1 < (I32)(minnext + deltanext + trie->maxlen))
4116 max1 = minnext + deltanext + trie->maxlen;
4117 if (deltanext == I32_MAX)
4118 is_inf = is_inf_internal = 1;
4120 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
4122 if (data_fake.flags & SCF_SEEN_ACCEPT) {
4123 if ( stopmin > min + min1)
4124 stopmin = min + min1;
4125 flags &= ~SCF_DO_SUBSTR;
4127 data->flags |= SCF_SEEN_ACCEPT;
4130 if (data_fake.flags & SF_HAS_EVAL)
4131 data->flags |= SF_HAS_EVAL;
4132 data->whilem_c = data_fake.whilem_c;
4134 if (flags & SCF_DO_STCLASS)
4135 cl_or(pRExC_state, &accum, &this_class);
4138 if (flags & SCF_DO_SUBSTR) {
4139 data->pos_min += min1;
4140 data->pos_delta += max1 - min1;
4141 if (max1 != min1 || is_inf)
4142 data->longest = &(data->longest_float);
4145 delta += max1 - min1;
4146 if (flags & SCF_DO_STCLASS_OR) {
4147 cl_or(pRExC_state, data->start_class, &accum);
4149 cl_and(data->start_class, and_withp);
4150 flags &= ~SCF_DO_STCLASS;
4153 else if (flags & SCF_DO_STCLASS_AND) {
4155 cl_and(data->start_class, &accum);
4156 flags &= ~SCF_DO_STCLASS;
4159 /* Switch to OR mode: cache the old value of
4160 * data->start_class */
4162 StructCopy(data->start_class, and_withp,
4163 struct regnode_charclass_class);
4164 flags &= ~SCF_DO_STCLASS_AND;
4165 StructCopy(&accum, data->start_class,
4166 struct regnode_charclass_class);
4167 flags |= SCF_DO_STCLASS_OR;
4168 data->start_class->flags |= ANYOF_EOS;
4175 else if (PL_regkind[OP(scan)] == TRIE) {
4176 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
4179 min += trie->minlen;
4180 delta += (trie->maxlen - trie->minlen);
4181 flags &= ~SCF_DO_STCLASS; /* xxx */
4182 if (flags & SCF_DO_SUBSTR) {
4183 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
4184 data->pos_min += trie->minlen;
4185 data->pos_delta += (trie->maxlen - trie->minlen);
4186 if (trie->maxlen != trie->minlen)
4187 data->longest = &(data->longest_float);
4189 if (trie->jump) /* no more substrings -- for now /grr*/
4190 flags &= ~SCF_DO_SUBSTR;
4192 #endif /* old or new */
4193 #endif /* TRIE_STUDY_OPT */
4195 /* Else: zero-length, ignore. */
4196 scan = regnext(scan);
4201 stopparen = frame->stop;
4202 frame = frame->prev;
4203 goto fake_study_recurse;
4208 DEBUG_STUDYDATA("pre-fin:",data,depth);
4211 *deltap = is_inf_internal ? I32_MAX : delta;
4212 if (flags & SCF_DO_SUBSTR && is_inf)
4213 data->pos_delta = I32_MAX - data->pos_min;
4214 if (is_par > (I32)U8_MAX)
4216 if (is_par && pars==1 && data) {
4217 data->flags |= SF_IN_PAR;
4218 data->flags &= ~SF_HAS_PAR;
4220 else if (pars && data) {
4221 data->flags |= SF_HAS_PAR;
4222 data->flags &= ~SF_IN_PAR;
4224 if (flags & SCF_DO_STCLASS_OR)
4225 cl_and(data->start_class, and_withp);
4226 if (flags & SCF_TRIE_RESTUDY)
4227 data->flags |= SCF_TRIE_RESTUDY;
4229 DEBUG_STUDYDATA("post-fin:",data,depth);
4231 return min < stopmin ? min : stopmin;
4235 S_add_data(RExC_state_t *pRExC_state, U32 n, const char *s)
4237 U32 count = RExC_rxi->data ? RExC_rxi->data->count : 0;
4239 PERL_ARGS_ASSERT_ADD_DATA;
4241 Renewc(RExC_rxi->data,
4242 sizeof(*RExC_rxi->data) + sizeof(void*) * (count + n - 1),
4243 char, struct reg_data);
4245 Renew(RExC_rxi->data->what, count + n, U8);
4247 Newx(RExC_rxi->data->what, n, U8);
4248 RExC_rxi->data->count = count + n;
4249 Copy(s, RExC_rxi->data->what + count, n, U8);
4253 /*XXX: todo make this not included in a non debugging perl */
4254 #ifndef PERL_IN_XSUB_RE
4256 Perl_reginitcolors(pTHX)
4259 const char * const s = PerlEnv_getenv("PERL_RE_COLORS");
4261 char *t = savepv(s);
4265 t = strchr(t, '\t');
4271 PL_colors[i] = t = (char *)"";
4276 PL_colors[i++] = (char *)"";
4283 #ifdef TRIE_STUDY_OPT
4284 #define CHECK_RESTUDY_GOTO \
4286 (data.flags & SCF_TRIE_RESTUDY) \
4290 #define CHECK_RESTUDY_GOTO
4294 - pregcomp - compile a regular expression into internal code
4296 * We can't allocate space until we know how big the compiled form will be,
4297 * but we can't compile it (and thus know how big it is) until we've got a
4298 * place to put the code. So we cheat: we compile it twice, once with code
4299 * generation turned off and size counting turned on, and once "for real".
4300 * This also means that we don't allocate space until we are sure that the
4301 * thing really will compile successfully, and we never have to move the
4302 * code and thus invalidate pointers into it. (Note that it has to be in
4303 * one piece because free() must be able to free it all.) [NB: not true in perl]
4305 * Beware that the optimization-preparation code in here knows about some
4306 * of the structure of the compiled regexp. [I'll say.]
4311 #ifndef PERL_IN_XSUB_RE
4312 #define RE_ENGINE_PTR &reh_regexp_engine
4314 extern const struct regexp_engine my_reg_engine;
4315 #define RE_ENGINE_PTR &my_reg_engine
4318 #ifndef PERL_IN_XSUB_RE
4320 Perl_pregcomp(pTHX_ SV * const pattern, const U32 flags)
4323 HV * const table = GvHV(PL_hintgv);
4325 PERL_ARGS_ASSERT_PREGCOMP;
4327 /* Dispatch a request to compile a regexp to correct
4330 SV **ptr= hv_fetchs(table, "regcomp", FALSE);
4331 GET_RE_DEBUG_FLAGS_DECL;
4332 if (ptr && SvIOK(*ptr) && SvIV(*ptr)) {
4333 const regexp_engine *eng=INT2PTR(regexp_engine*,SvIV(*ptr));
4335 PerlIO_printf(Perl_debug_log, "Using engine %"UVxf"\n",
4338 return CALLREGCOMP_ENG(eng, pattern, flags);
4341 return Perl_re_compile(aTHX_ pattern, flags);
4346 Perl_re_compile(pTHX_ SV * const pattern, U32 pm_flags)
4351 register regexp_internal *ri;
4359 /* these are all flags - maybe they should be turned
4360 * into a single int with different bit masks */
4361 I32 sawlookahead = 0;
4368 RExC_state_t RExC_state;
4369 RExC_state_t * const pRExC_state = &RExC_state;
4370 #ifdef TRIE_STUDY_OPT
4372 RExC_state_t copyRExC_state;
4374 GET_RE_DEBUG_FLAGS_DECL;
4376 PERL_ARGS_ASSERT_RE_COMPILE;
4378 DEBUG_r(if (!PL_colorset) reginitcolors());
4380 RExC_utf8 = RExC_orig_utf8 = SvUTF8(pattern);
4383 /* Longjmp back to here if have to switch in midstream to utf8 */
4384 if (! RExC_orig_utf8) {
4385 JMPENV_PUSH(jump_ret);
4388 if (jump_ret == 0) { /* First time through */
4389 exp = SvPV(pattern, plen);
4393 SV *dsv= sv_newmortal();
4394 RE_PV_QUOTED_DECL(s, RExC_utf8,
4395 dsv, exp, plen, 60);
4396 PerlIO_printf(Perl_debug_log, "%sCompiling REx%s %s\n",
4397 PL_colors[4],PL_colors[5],s);
4400 else { /* longjumped back */
4403 /* If the cause for the longjmp was other than changing to utf8, pop
4404 * our own setjmp, and longjmp to the correct handler */
4405 if (jump_ret != UTF8_LONGJMP) {
4407 JMPENV_JUMP(jump_ret);
4412 /* It's possible to write a regexp in ascii that represents Unicode
4413 codepoints outside of the byte range, such as via \x{100}. If we
4414 detect such a sequence we have to convert the entire pattern to utf8
4415 and then recompile, as our sizing calculation will have been based
4416 on 1 byte == 1 character, but we will need to use utf8 to encode
4417 at least some part of the pattern, and therefore must convert the whole
4420 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log,
4421 "UTF8 mismatch! Converting to utf8 for resizing and compile\n"));
4422 exp = (char*)Perl_bytes_to_utf8(aTHX_ (U8*)SvPV(pattern, plen), &len);
4424 RExC_orig_utf8 = RExC_utf8 = 1;
4428 #ifdef TRIE_STUDY_OPT
4433 RExC_flags = pm_flags;
4437 RExC_seen_zerolen = *exp == '^' ? -1 : 0;
4438 RExC_seen_evals = 0;
4441 /* First pass: determine size, legality. */
4449 RExC_emit = &PL_regdummy;
4450 RExC_whilem_seen = 0;
4451 RExC_open_parens = NULL;
4452 RExC_close_parens = NULL;
4454 RExC_paren_names = NULL;
4456 RExC_paren_name_list = NULL;
4458 RExC_recurse = NULL;
4459 RExC_recurse_count = 0;
4461 #if 0 /* REGC() is (currently) a NOP at the first pass.
4462 * Clever compilers notice this and complain. --jhi */
4463 REGC((U8)REG_MAGIC, (char*)RExC_emit);
4465 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "Starting first pass (sizing)\n"));
4466 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4467 RExC_precomp = NULL;
4471 /* Here, finished first pass. Get rid of our setjmp, which we added for
4472 * efficiency only if the passed-in string wasn't in utf8, as shown by
4473 * RExC_orig_utf8. But if the first pass was redone, that variable will be
4474 * 1 here even though the original string wasn't utf8, but in this case
4475 * there will have been a long jump */
4476 if (jump_ret == UTF8_LONGJMP || ! RExC_orig_utf8) {
4480 PerlIO_printf(Perl_debug_log,
4481 "Required size %"IVdf" nodes\n"
4482 "Starting second pass (creation)\n",
4485 RExC_lastparse=NULL;
4487 /* Small enough for pointer-storage convention?
4488 If extralen==0, this means that we will not need long jumps. */
4489 if (RExC_size >= 0x10000L && RExC_extralen)
4490 RExC_size += RExC_extralen;
4493 if (RExC_whilem_seen > 15)
4494 RExC_whilem_seen = 15;
4496 /* Allocate space and zero-initialize. Note, the two step process
4497 of zeroing when in debug mode, thus anything assigned has to
4498 happen after that */
4499 rx = (REGEXP*) newSV_type(SVt_REGEXP);
4500 r = (struct regexp*)SvANY(rx);
4501 Newxc(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode),
4502 char, regexp_internal);
4503 if ( r == NULL || ri == NULL )
4504 FAIL("Regexp out of space");
4506 /* avoid reading uninitialized memory in DEBUGGING code in study_chunk() */
4507 Zero(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode), char);
4509 /* bulk initialize base fields with 0. */
4510 Zero(ri, sizeof(regexp_internal), char);
4513 /* non-zero initialization begins here */
4515 r->engine= RE_ENGINE_PTR;
4516 r->extflags = pm_flags;
4518 bool has_p = ((r->extflags & RXf_PMf_KEEPCOPY) == RXf_PMf_KEEPCOPY);
4519 bool has_charset = cBOOL(r->extflags & (RXf_PMf_LOCALE|RXf_PMf_UNICODE));
4521 /* The caret is output if there are any defaults: if not all the STD
4522 * flags are set, or if no character set specifier is needed */
4524 (((r->extflags & RXf_PMf_STD_PMMOD) != RXf_PMf_STD_PMMOD)
4526 bool has_runon = ((RExC_seen & REG_SEEN_RUN_ON_COMMENT)==REG_SEEN_RUN_ON_COMMENT);
4527 U16 reganch = (U16)((r->extflags & RXf_PMf_STD_PMMOD)
4528 >> RXf_PMf_STD_PMMOD_SHIFT);
4529 const char *fptr = STD_PAT_MODS; /*"msix"*/
4531 /* Allocate for the worst case, which is all the std flags are turned
4532 * on. If more precision is desired, we could do a population count of
4533 * the flags set. This could be done with a small lookup table, or by
4534 * shifting, masking and adding, or even, when available, assembly
4535 * language for a machine-language population count.
4536 * We never output a minus, as all those are defaults, so are
4537 * covered by the caret */
4538 const STRLEN wraplen = plen + has_p + has_runon
4539 + has_default /* If needs a caret */
4540 + has_charset /* If needs a character set specifier */
4541 + (sizeof(STD_PAT_MODS) - 1)
4542 + (sizeof("(?:)") - 1);
4544 p = sv_grow(MUTABLE_SV(rx), wraplen + 1); /* +1 for the ending NUL */
4546 SvFLAGS(rx) |= SvUTF8(pattern);
4549 /* If a default, cover it using the caret */
4551 *p++= DEFAULT_PAT_MOD;
4554 if (r->extflags & RXf_PMf_LOCALE) {
4555 *p++ = LOCALE_PAT_MOD;
4557 *p++ = UNICODE_PAT_MOD;
4561 *p++ = KEEPCOPY_PAT_MOD; /*'p'*/
4564 while((ch = *fptr++)) {
4572 Copy(RExC_precomp, p, plen, char);
4573 assert ((RX_WRAPPED(rx) - p) < 16);
4574 r->pre_prefix = p - RX_WRAPPED(rx);
4580 SvCUR_set(rx, p - SvPVX_const(rx));
4584 r->nparens = RExC_npar - 1; /* set early to validate backrefs */
4586 if (RExC_seen & REG_SEEN_RECURSE) {
4587 Newxz(RExC_open_parens, RExC_npar,regnode *);
4588 SAVEFREEPV(RExC_open_parens);
4589 Newxz(RExC_close_parens,RExC_npar,regnode *);
4590 SAVEFREEPV(RExC_close_parens);
4593 /* Useful during FAIL. */
4594 #ifdef RE_TRACK_PATTERN_OFFSETS
4595 Newxz(ri->u.offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
4596 DEBUG_OFFSETS_r(PerlIO_printf(Perl_debug_log,
4597 "%s %"UVuf" bytes for offset annotations.\n",
4598 ri->u.offsets ? "Got" : "Couldn't get",
4599 (UV)((2*RExC_size+1) * sizeof(U32))));
4601 SetProgLen(ri,RExC_size);
4605 REH_CALL_COMP_BEGIN_HOOK(pRExC_state->rx);
4607 /* Second pass: emit code. */
4608 RExC_flags = pm_flags; /* don't let top level (?i) bleed */
4613 RExC_emit_start = ri->program;
4614 RExC_emit = ri->program;
4615 RExC_emit_bound = ri->program + RExC_size + 1;
4617 /* Store the count of eval-groups for security checks: */
4618 RExC_rx->seen_evals = RExC_seen_evals;
4619 REGC((U8)REG_MAGIC, (char*) RExC_emit++);
4620 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4624 /* XXXX To minimize changes to RE engine we always allocate
4625 3-units-long substrs field. */
4626 Newx(r->substrs, 1, struct reg_substr_data);
4627 if (RExC_recurse_count) {
4628 Newxz(RExC_recurse,RExC_recurse_count,regnode *);
4629 SAVEFREEPV(RExC_recurse);
4633 r->minlen = minlen = sawlookahead = sawplus = sawopen = 0;
4634 Zero(r->substrs, 1, struct reg_substr_data);
4636 #ifdef TRIE_STUDY_OPT
4638 StructCopy(&zero_scan_data, &data, scan_data_t);
4639 copyRExC_state = RExC_state;
4642 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log,"Restudying\n"));
4644 RExC_state = copyRExC_state;
4645 if (seen & REG_TOP_LEVEL_BRANCHES)
4646 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
4648 RExC_seen &= ~REG_TOP_LEVEL_BRANCHES;
4649 if (data.last_found) {
4650 SvREFCNT_dec(data.longest_fixed);
4651 SvREFCNT_dec(data.longest_float);
4652 SvREFCNT_dec(data.last_found);
4654 StructCopy(&zero_scan_data, &data, scan_data_t);
4657 StructCopy(&zero_scan_data, &data, scan_data_t);
4660 /* Dig out information for optimizations. */
4661 r->extflags = RExC_flags; /* was pm_op */
4662 /*dmq: removed as part of de-PMOP: pm->op_pmflags = RExC_flags; */
4665 SvUTF8_on(rx); /* Unicode in it? */
4666 ri->regstclass = NULL;
4667 if (RExC_naughty >= 10) /* Probably an expensive pattern. */
4668 r->intflags |= PREGf_NAUGHTY;
4669 scan = ri->program + 1; /* First BRANCH. */
4671 /* testing for BRANCH here tells us whether there is "must appear"
4672 data in the pattern. If there is then we can use it for optimisations */
4673 if (!(RExC_seen & REG_TOP_LEVEL_BRANCHES)) { /* Only one top-level choice. */
4675 STRLEN longest_float_length, longest_fixed_length;
4676 struct regnode_charclass_class ch_class; /* pointed to by data */
4678 I32 last_close = 0; /* pointed to by data */
4679 regnode *first= scan;
4680 regnode *first_next= regnext(first);
4682 * Skip introductions and multiplicators >= 1
4683 * so that we can extract the 'meat' of the pattern that must
4684 * match in the large if() sequence following.
4685 * NOTE that EXACT is NOT covered here, as it is normally
4686 * picked up by the optimiser separately.
4688 * This is unfortunate as the optimiser isnt handling lookahead
4689 * properly currently.
4692 while ((OP(first) == OPEN && (sawopen = 1)) ||
4693 /* An OR of *one* alternative - should not happen now. */
4694 (OP(first) == BRANCH && OP(first_next) != BRANCH) ||
4695 /* for now we can't handle lookbehind IFMATCH*/
4696 (OP(first) == IFMATCH && !first->flags && (sawlookahead = 1)) ||
4697 (OP(first) == PLUS) ||
4698 (OP(first) == MINMOD) ||
4699 /* An {n,m} with n>0 */
4700 (PL_regkind[OP(first)] == CURLY && ARG1(first) > 0) ||
4701 (OP(first) == NOTHING && PL_regkind[OP(first_next)] != END ))
4704 * the only op that could be a regnode is PLUS, all the rest
4705 * will be regnode_1 or regnode_2.
4708 if (OP(first) == PLUS)
4711 first += regarglen[OP(first)];
4713 first = NEXTOPER(first);
4714 first_next= regnext(first);
4717 /* Starting-point info. */
4719 DEBUG_PEEP("first:",first,0);
4720 /* Ignore EXACT as we deal with it later. */
4721 if (PL_regkind[OP(first)] == EXACT) {
4722 if (OP(first) == EXACT)
4723 NOOP; /* Empty, get anchored substr later. */
4724 else if ((OP(first) == EXACTF || OP(first) == EXACTFL))
4725 ri->regstclass = first;
4728 else if (PL_regkind[OP(first)] == TRIE &&
4729 ((reg_trie_data *)ri->data->data[ ARG(first) ])->minlen>0)
4732 /* this can happen only on restudy */
4733 if ( OP(first) == TRIE ) {
4734 struct regnode_1 *trieop = (struct regnode_1 *)
4735 PerlMemShared_calloc(1, sizeof(struct regnode_1));
4736 StructCopy(first,trieop,struct regnode_1);
4737 trie_op=(regnode *)trieop;
4739 struct regnode_charclass *trieop = (struct regnode_charclass *)
4740 PerlMemShared_calloc(1, sizeof(struct regnode_charclass));
4741 StructCopy(first,trieop,struct regnode_charclass);
4742 trie_op=(regnode *)trieop;
4745 make_trie_failtable(pRExC_state, (regnode *)first, trie_op, 0);
4746 ri->regstclass = trie_op;
4749 else if (REGNODE_SIMPLE(OP(first)))
4750 ri->regstclass = first;
4751 else if (PL_regkind[OP(first)] == BOUND ||
4752 PL_regkind[OP(first)] == NBOUND)
4753 ri->regstclass = first;
4754 else if (PL_regkind[OP(first)] == BOL) {
4755 r->extflags |= (OP(first) == MBOL
4757 : (OP(first) == SBOL
4760 first = NEXTOPER(first);
4763 else if (OP(first) == GPOS) {
4764 r->extflags |= RXf_ANCH_GPOS;
4765 first = NEXTOPER(first);
4768 else if ((!sawopen || !RExC_sawback) &&
4769 (OP(first) == STAR &&
4770 PL_regkind[OP(NEXTOPER(first))] == REG_ANY) &&
4771 !(r->extflags & RXf_ANCH) && !(RExC_seen & REG_SEEN_EVAL))
4773 /* turn .* into ^.* with an implied $*=1 */
4775 (OP(NEXTOPER(first)) == REG_ANY)
4778 r->extflags |= type;
4779 r->intflags |= PREGf_IMPLICIT;
4780 first = NEXTOPER(first);
4783 if (sawplus && !sawlookahead && (!sawopen || !RExC_sawback)
4784 && !(RExC_seen & REG_SEEN_EVAL)) /* May examine pos and $& */
4785 /* x+ must match at the 1st pos of run of x's */
4786 r->intflags |= PREGf_SKIP;
4788 /* Scan is after the zeroth branch, first is atomic matcher. */
4789 #ifdef TRIE_STUDY_OPT
4792 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4793 (IV)(first - scan + 1))
4797 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4798 (IV)(first - scan + 1))
4804 * If there's something expensive in the r.e., find the
4805 * longest literal string that must appear and make it the
4806 * regmust. Resolve ties in favor of later strings, since
4807 * the regstart check works with the beginning of the r.e.
4808 * and avoiding duplication strengthens checking. Not a
4809 * strong reason, but sufficient in the absence of others.
4810 * [Now we resolve ties in favor of the earlier string if
4811 * it happens that c_offset_min has been invalidated, since the
4812 * earlier string may buy us something the later one won't.]
4815 data.longest_fixed = newSVpvs("");
4816 data.longest_float = newSVpvs("");
4817 data.last_found = newSVpvs("");
4818 data.longest = &(data.longest_fixed);
4820 if (!ri->regstclass) {
4821 cl_init(pRExC_state, &ch_class);
4822 data.start_class = &ch_class;
4823 stclass_flag = SCF_DO_STCLASS_AND;
4824 } else /* XXXX Check for BOUND? */
4826 data.last_closep = &last_close;
4828 minlen = study_chunk(pRExC_state, &first, &minlen, &fake, scan + RExC_size, /* Up to end */
4829 &data, -1, NULL, NULL,
4830 SCF_DO_SUBSTR | SCF_WHILEM_VISITED_POS | stclass_flag,0);
4836 if ( RExC_npar == 1 && data.longest == &(data.longest_fixed)
4837 && data.last_start_min == 0 && data.last_end > 0
4838 && !RExC_seen_zerolen
4839 && !(RExC_seen & REG_SEEN_VERBARG)
4840 && (!(RExC_seen & REG_SEEN_GPOS) || (r->extflags & RXf_ANCH_GPOS)))
4841 r->extflags |= RXf_CHECK_ALL;
4842 scan_commit(pRExC_state, &data,&minlen,0);
4843 SvREFCNT_dec(data.last_found);
4845 /* Note that code very similar to this but for anchored string
4846 follows immediately below, changes may need to be made to both.
4849 longest_float_length = CHR_SVLEN(data.longest_float);
4850 if (longest_float_length
4851 || (data.flags & SF_FL_BEFORE_EOL
4852 && (!(data.flags & SF_FL_BEFORE_MEOL)
4853 || (RExC_flags & RXf_PMf_MULTILINE))))
4857 if (SvCUR(data.longest_fixed) /* ok to leave SvCUR */
4858 && data.offset_fixed == data.offset_float_min
4859 && SvCUR(data.longest_fixed) == SvCUR(data.longest_float))
4860 goto remove_float; /* As in (a)+. */
4862 /* copy the information about the longest float from the reg_scan_data
4863 over to the program. */
4864 if (SvUTF8(data.longest_float)) {
4865 r->float_utf8 = data.longest_float;
4866 r->float_substr = NULL;
4868 r->float_substr = data.longest_float;
4869 r->float_utf8 = NULL;
4871 /* float_end_shift is how many chars that must be matched that
4872 follow this item. We calculate it ahead of time as once the
4873 lookbehind offset is added in we lose the ability to correctly
4875 ml = data.minlen_float ? *(data.minlen_float)
4876 : (I32)longest_float_length;
4877 r->float_end_shift = ml - data.offset_float_min
4878 - longest_float_length + (SvTAIL(data.longest_float) != 0)
4879 + data.lookbehind_float;
4880 r->float_min_offset = data.offset_float_min - data.lookbehind_float;
4881 r->float_max_offset = data.offset_float_max;
4882 if (data.offset_float_max < I32_MAX) /* Don't offset infinity */
4883 r->float_max_offset -= data.lookbehind_float;
4885 t = (data.flags & SF_FL_BEFORE_EOL /* Can't have SEOL and MULTI */
4886 && (!(data.flags & SF_FL_BEFORE_MEOL)
4887 || (RExC_flags & RXf_PMf_MULTILINE)));
4888 fbm_compile(data.longest_float, t ? FBMcf_TAIL : 0);
4892 r->float_substr = r->float_utf8 = NULL;
4893 SvREFCNT_dec(data.longest_float);
4894 longest_float_length = 0;
4897 /* Note that code very similar to this but for floating string
4898 is immediately above, changes may need to be made to both.
4901 longest_fixed_length = CHR_SVLEN(data.longest_fixed);
4902 if (longest_fixed_length
4903 || (data.flags & SF_FIX_BEFORE_EOL /* Cannot have SEOL and MULTI */
4904 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4905 || (RExC_flags & RXf_PMf_MULTILINE))))
4909 /* copy the information about the longest fixed
4910 from the reg_scan_data over to the program. */
4911 if (SvUTF8(data.longest_fixed)) {
4912 r->anchored_utf8 = data.longest_fixed;
4913 r->anchored_substr = NULL;
4915 r->anchored_substr = data.longest_fixed;
4916 r->anchored_utf8 = NULL;
4918 /* fixed_end_shift is how many chars that must be matched that
4919 follow this item. We calculate it ahead of time as once the
4920 lookbehind offset is added in we lose the ability to correctly
4922 ml = data.minlen_fixed ? *(data.minlen_fixed)
4923 : (I32)longest_fixed_length;
4924 r->anchored_end_shift = ml - data.offset_fixed
4925 - longest_fixed_length + (SvTAIL(data.longest_fixed) != 0)
4926 + data.lookbehind_fixed;
4927 r->anchored_offset = data.offset_fixed - data.lookbehind_fixed;
4929 t = (data.flags & SF_FIX_BEFORE_EOL /* Can't have SEOL and MULTI */
4930 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4931 || (RExC_flags & RXf_PMf_MULTILINE)));
4932 fbm_compile(data.longest_fixed, t ? FBMcf_TAIL : 0);
4935 r->anchored_substr = r->anchored_utf8 = NULL;
4936 SvREFCNT_dec(data.longest_fixed);
4937 longest_fixed_length = 0;
4940 && (OP(ri->regstclass) == REG_ANY || OP(ri->regstclass) == SANY))
4941 ri->regstclass = NULL;
4942 if ((!(r->anchored_substr || r->anchored_utf8) || r->anchored_offset)
4944 && !(data.start_class->flags & ANYOF_EOS)
4945 && !cl_is_anything(data.start_class))
4947 const U32 n = add_data(pRExC_state, 1, "f");
4949 Newx(RExC_rxi->data->data[n], 1,
4950 struct regnode_charclass_class);
4951 StructCopy(data.start_class,
4952 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4953 struct regnode_charclass_class);
4954 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4955 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4956 DEBUG_COMPILE_r({ SV *sv = sv_newmortal();
4957 regprop(r, sv, (regnode*)data.start_class);
4958 PerlIO_printf(Perl_debug_log,
4959 "synthetic stclass \"%s\".\n",
4960 SvPVX_const(sv));});
4963 /* A temporary algorithm prefers floated substr to fixed one to dig more info. */
4964 if (longest_fixed_length > longest_float_length) {
4965 r->check_end_shift = r->anchored_end_shift;
4966 r->check_substr = r->anchored_substr;
4967 r->check_utf8 = r->anchored_utf8;
4968 r->check_offset_min = r->check_offset_max = r->anchored_offset;
4969 if (r->extflags & RXf_ANCH_SINGLE)
4970 r->extflags |= RXf_NOSCAN;
4973 r->check_end_shift = r->float_end_shift;
4974 r->check_substr = r->float_substr;
4975 r->check_utf8 = r->float_utf8;
4976 r->check_offset_min = r->float_min_offset;
4977 r->check_offset_max = r->float_max_offset;
4979 /* XXXX Currently intuiting is not compatible with ANCH_GPOS.
4980 This should be changed ASAP! */
4981 if ((r->check_substr || r->check_utf8) && !(r->extflags & RXf_ANCH_GPOS)) {
4982 r->extflags |= RXf_USE_INTUIT;
4983 if (SvTAIL(r->check_substr ? r->check_substr : r->check_utf8))
4984 r->extflags |= RXf_INTUIT_TAIL;
4986 /* XXX Unneeded? dmq (shouldn't as this is handled elsewhere)
4987 if ( (STRLEN)minlen < longest_float_length )
4988 minlen= longest_float_length;
4989 if ( (STRLEN)minlen < longest_fixed_length )
4990 minlen= longest_fixed_length;
4994 /* Several toplevels. Best we can is to set minlen. */
4996 struct regnode_charclass_class ch_class;
4999 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "\nMulti Top Level\n"));
5001 scan = ri->program + 1;
5002 cl_init(pRExC_state, &ch_class);
5003 data.start_class = &ch_class;
5004 data.last_closep = &last_close;
5007 minlen = study_chunk(pRExC_state, &scan, &minlen, &fake, scan + RExC_size,
5008 &data, -1, NULL, NULL, SCF_DO_STCLASS_AND|SCF_WHILEM_VISITED_POS,0);
5012 r->check_substr = r->check_utf8 = r->anchored_substr = r->anchored_utf8
5013 = r->float_substr = r->float_utf8 = NULL;
5014 if (!(data.start_class->flags & ANYOF_EOS)
5015 && !cl_is_anything(data.start_class))
5017 const U32 n = add_data(pRExC_state, 1, "f");
5019 Newx(RExC_rxi->data->data[n], 1,
5020 struct regnode_charclass_class);
5021 StructCopy(data.start_class,
5022 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
5023 struct regnode_charclass_class);
5024 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
5025 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
5026 DEBUG_COMPILE_r({ SV* sv = sv_newmortal();
5027 regprop(r, sv, (regnode*)data.start_class);
5028 PerlIO_printf(Perl_debug_log,
5029 "synthetic stclass \"%s\".\n",
5030 SvPVX_const(sv));});
5034 /* Guard against an embedded (?=) or (?<=) with a longer minlen than
5035 the "real" pattern. */
5037 PerlIO_printf(Perl_debug_log,"minlen: %"IVdf" r->minlen:%"IVdf"\n",
5038 (IV)minlen, (IV)r->minlen);
5040 r->minlenret = minlen;
5041 if (r->minlen < minlen)
5044 if (RExC_seen & REG_SEEN_GPOS)
5045 r->extflags |= RXf_GPOS_SEEN;
5046 if (RExC_seen & REG_SEEN_LOOKBEHIND)
5047 r->extflags |= RXf_LOOKBEHIND_SEEN;
5048 if (RExC_seen & REG_SEEN_EVAL)
5049 r->extflags |= RXf_EVAL_SEEN;
5050 if (RExC_seen & REG_SEEN_CANY)
5051 r->extflags |= RXf_CANY_SEEN;
5052 if (RExC_seen & REG_SEEN_VERBARG)
5053 r->intflags |= PREGf_VERBARG_SEEN;
5054 if (RExC_seen & REG_SEEN_CUTGROUP)
5055 r->intflags |= PREGf_CUTGROUP_SEEN;
5056 if (RExC_paren_names)
5057 RXp_PAREN_NAMES(r) = MUTABLE_HV(SvREFCNT_inc(RExC_paren_names));
5059 RXp_PAREN_NAMES(r) = NULL;
5061 #ifdef STUPID_PATTERN_CHECKS
5062 if (RX_PRELEN(rx) == 0)
5063 r->extflags |= RXf_NULL;
5064 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
5065 /* XXX: this should happen BEFORE we compile */
5066 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
5067 else if (RX_PRELEN(rx) == 3 && memEQ("\\s+", RX_PRECOMP(rx), 3))
5068 r->extflags |= RXf_WHITE;
5069 else if (RX_PRELEN(rx) == 1 && RXp_PRECOMP(rx)[0] == '^')
5070 r->extflags |= RXf_START_ONLY;
5072 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
5073 /* XXX: this should happen BEFORE we compile */
5074 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
5076 regnode *first = ri->program + 1;
5078 U8 nop = OP(NEXTOPER(first));
5080 if (PL_regkind[fop] == NOTHING && nop == END)
5081 r->extflags |= RXf_NULL;
5082 else if (PL_regkind[fop] == BOL && nop == END)
5083 r->extflags |= RXf_START_ONLY;
5084 else if (fop == PLUS && nop ==SPACE && OP(regnext(first))==END)
5085 r->extflags |= RXf_WHITE;
5089 if (RExC_paren_names) {
5090 ri->name_list_idx = add_data( pRExC_state, 1, "a" );
5091 ri->data->data[ri->name_list_idx] = (void*)SvREFCNT_inc(RExC_paren_name_list);
5094 ri->name_list_idx = 0;
5096 if (RExC_recurse_count) {
5097 for ( ; RExC_recurse_count ; RExC_recurse_count-- ) {
5098 const regnode *scan = RExC_recurse[RExC_recurse_count-1];
5099 ARG2L_SET( scan, RExC_open_parens[ARG(scan)-1] - scan );
5102 Newxz(r->offs, RExC_npar, regexp_paren_pair);
5103 /* assume we don't need to swap parens around before we match */
5106 PerlIO_printf(Perl_debug_log,"Final program:\n");
5109 #ifdef RE_TRACK_PATTERN_OFFSETS
5110 DEBUG_OFFSETS_r(if (ri->u.offsets) {
5111 const U32 len = ri->u.offsets[0];
5113 GET_RE_DEBUG_FLAGS_DECL;
5114 PerlIO_printf(Perl_debug_log, "Offsets: [%"UVuf"]\n\t", (UV)ri->u.offsets[0]);
5115 for (i = 1; i <= len; i++) {
5116 if (ri->u.offsets[i*2-1] || ri->u.offsets[i*2])
5117 PerlIO_printf(Perl_debug_log, "%"UVuf":%"UVuf"[%"UVuf"] ",
5118 (UV)i, (UV)ri->u.offsets[i*2-1], (UV)ri->u.offsets[i*2]);
5120 PerlIO_printf(Perl_debug_log, "\n");
5126 #undef RE_ENGINE_PTR
5130 Perl_reg_named_buff(pTHX_ REGEXP * const rx, SV * const key, SV * const value,
5133 PERL_ARGS_ASSERT_REG_NAMED_BUFF;
5135 PERL_UNUSED_ARG(value);
5137 if (flags & RXapif_FETCH) {
5138 return reg_named_buff_fetch(rx, key, flags);
5139 } else if (flags & (RXapif_STORE | RXapif_DELETE | RXapif_CLEAR)) {
5140 Perl_croak_no_modify(aTHX);
5142 } else if (flags & RXapif_EXISTS) {
5143 return reg_named_buff_exists(rx, key, flags)
5146 } else if (flags & RXapif_REGNAMES) {
5147 return reg_named_buff_all(rx, flags);
5148 } else if (flags & (RXapif_SCALAR | RXapif_REGNAMES_COUNT)) {
5149 return reg_named_buff_scalar(rx, flags);
5151 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff", (int)flags);
5157 Perl_reg_named_buff_iter(pTHX_ REGEXP * const rx, const SV * const lastkey,
5160 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ITER;
5161 PERL_UNUSED_ARG(lastkey);
5163 if (flags & RXapif_FIRSTKEY)
5164 return reg_named_buff_firstkey(rx, flags);
5165 else if (flags & RXapif_NEXTKEY)
5166 return reg_named_buff_nextkey(rx, flags);
5168 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_iter", (int)flags);
5174 Perl_reg_named_buff_fetch(pTHX_ REGEXP * const r, SV * const namesv,
5177 AV *retarray = NULL;
5179 struct regexp *const rx = (struct regexp *)SvANY(r);
5181 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FETCH;
5183 if (flags & RXapif_ALL)
5186 if (rx && RXp_PAREN_NAMES(rx)) {
5187 HE *he_str = hv_fetch_ent( RXp_PAREN_NAMES(rx), namesv, 0, 0 );
5190 SV* sv_dat=HeVAL(he_str);
5191 I32 *nums=(I32*)SvPVX(sv_dat);
5192 for ( i=0; i<SvIVX(sv_dat); i++ ) {
5193 if ((I32)(rx->nparens) >= nums[i]
5194 && rx->offs[nums[i]].start != -1
5195 && rx->offs[nums[i]].end != -1)
5198 CALLREG_NUMBUF_FETCH(r,nums[i],ret);
5202 ret = newSVsv(&PL_sv_undef);
5205 av_push(retarray, ret);
5208 return newRV_noinc(MUTABLE_SV(retarray));
5215 Perl_reg_named_buff_exists(pTHX_ REGEXP * const r, SV * const key,
5218 struct regexp *const rx = (struct regexp *)SvANY(r);
5220 PERL_ARGS_ASSERT_REG_NAMED_BUFF_EXISTS;
5222 if (rx && RXp_PAREN_NAMES(rx)) {
5223 if (flags & RXapif_ALL) {
5224 return hv_exists_ent(RXp_PAREN_NAMES(rx), key, 0);
5226 SV *sv = CALLREG_NAMED_BUFF_FETCH(r, key, flags);
5240 Perl_reg_named_buff_firstkey(pTHX_ REGEXP * const r, const U32 flags)
5242 struct regexp *const rx = (struct regexp *)SvANY(r);
5244 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FIRSTKEY;
5246 if ( rx && RXp_PAREN_NAMES(rx) ) {
5247 (void)hv_iterinit(RXp_PAREN_NAMES(rx));
5249 return CALLREG_NAMED_BUFF_NEXTKEY(r, NULL, flags & ~RXapif_FIRSTKEY);
5256 Perl_reg_named_buff_nextkey(pTHX_ REGEXP * const r, const U32 flags)
5258 struct regexp *const rx = (struct regexp *)SvANY(r);
5259 GET_RE_DEBUG_FLAGS_DECL;
5261 PERL_ARGS_ASSERT_REG_NAMED_BUFF_NEXTKEY;
5263 if (rx && RXp_PAREN_NAMES(rx)) {
5264 HV *hv = RXp_PAREN_NAMES(rx);
5266 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5269 SV* sv_dat = HeVAL(temphe);
5270 I32 *nums = (I32*)SvPVX(sv_dat);
5271 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5272 if ((I32)(rx->lastparen) >= nums[i] &&
5273 rx->offs[nums[i]].start != -1 &&
5274 rx->offs[nums[i]].end != -1)
5280 if (parno || flags & RXapif_ALL) {
5281 return newSVhek(HeKEY_hek(temphe));
5289 Perl_reg_named_buff_scalar(pTHX_ REGEXP * const r, const U32 flags)
5294 struct regexp *const rx = (struct regexp *)SvANY(r);
5296 PERL_ARGS_ASSERT_REG_NAMED_BUFF_SCALAR;
5298 if (rx && RXp_PAREN_NAMES(rx)) {
5299 if (flags & (RXapif_ALL | RXapif_REGNAMES_COUNT)) {
5300 return newSViv(HvTOTALKEYS(RXp_PAREN_NAMES(rx)));
5301 } else if (flags & RXapif_ONE) {
5302 ret = CALLREG_NAMED_BUFF_ALL(r, (flags | RXapif_REGNAMES));
5303 av = MUTABLE_AV(SvRV(ret));
5304 length = av_len(av);
5306 return newSViv(length + 1);
5308 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_scalar", (int)flags);
5312 return &PL_sv_undef;
5316 Perl_reg_named_buff_all(pTHX_ REGEXP * const r, const U32 flags)
5318 struct regexp *const rx = (struct regexp *)SvANY(r);
5321 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ALL;
5323 if (rx && RXp_PAREN_NAMES(rx)) {
5324 HV *hv= RXp_PAREN_NAMES(rx);
5326 (void)hv_iterinit(hv);
5327 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5330 SV* sv_dat = HeVAL(temphe);
5331 I32 *nums = (I32*)SvPVX(sv_dat);
5332 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5333 if ((I32)(rx->lastparen) >= nums[i] &&
5334 rx->offs[nums[i]].start != -1 &&
5335 rx->offs[nums[i]].end != -1)
5341 if (parno || flags & RXapif_ALL) {
5342 av_push(av, newSVhek(HeKEY_hek(temphe)));
5347 return newRV_noinc(MUTABLE_SV(av));
5351 Perl_reg_numbered_buff_fetch(pTHX_ REGEXP * const r, const I32 paren,
5354 struct regexp *const rx = (struct regexp *)SvANY(r);
5359 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_FETCH;
5362 sv_setsv(sv,&PL_sv_undef);
5366 if (paren == RX_BUFF_IDX_PREMATCH && rx->offs[0].start != -1) {
5368 i = rx->offs[0].start;
5372 if (paren == RX_BUFF_IDX_POSTMATCH && rx->offs[0].end != -1) {
5374 s = rx->subbeg + rx->offs[0].end;
5375 i = rx->sublen - rx->offs[0].end;
5378 if ( 0 <= paren && paren <= (I32)rx->nparens &&
5379 (s1 = rx->offs[paren].start) != -1 &&
5380 (t1 = rx->offs[paren].end) != -1)
5384 s = rx->subbeg + s1;
5386 sv_setsv(sv,&PL_sv_undef);
5389 assert(rx->sublen >= (s - rx->subbeg) + i );
5391 const int oldtainted = PL_tainted;
5393 sv_setpvn(sv, s, i);
5394 PL_tainted = oldtainted;
5395 if ( (rx->extflags & RXf_CANY_SEEN)
5396 ? (RXp_MATCH_UTF8(rx)
5397 && (!i || is_utf8_string((U8*)s, i)))
5398 : (RXp_MATCH_UTF8(rx)) )
5405 if (RXp_MATCH_TAINTED(rx)) {
5406 if (SvTYPE(sv) >= SVt_PVMG) {
5407 MAGIC* const mg = SvMAGIC(sv);
5410 SvMAGIC_set(sv, mg->mg_moremagic);
5412 if ((mgt = SvMAGIC(sv))) {
5413 mg->mg_moremagic = mgt;
5414 SvMAGIC_set(sv, mg);
5424 sv_setsv(sv,&PL_sv_undef);
5430 Perl_reg_numbered_buff_store(pTHX_ REGEXP * const rx, const I32 paren,
5431 SV const * const value)
5433 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_STORE;
5435 PERL_UNUSED_ARG(rx);
5436 PERL_UNUSED_ARG(paren);
5437 PERL_UNUSED_ARG(value);
5440 Perl_croak_no_modify(aTHX);
5444 Perl_reg_numbered_buff_length(pTHX_ REGEXP * const r, const SV * const sv,
5447 struct regexp *const rx = (struct regexp *)SvANY(r);
5451 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_LENGTH;
5453 /* Some of this code was originally in C<Perl_magic_len> in F<mg.c> */
5455 /* $` / ${^PREMATCH} */
5456 case RX_BUFF_IDX_PREMATCH:
5457 if (rx->offs[0].start != -1) {
5458 i = rx->offs[0].start;
5466 /* $' / ${^POSTMATCH} */
5467 case RX_BUFF_IDX_POSTMATCH:
5468 if (rx->offs[0].end != -1) {
5469 i = rx->sublen - rx->offs[0].end;
5471 s1 = rx->offs[0].end;
5477 /* $& / ${^MATCH}, $1, $2, ... */
5479 if (paren <= (I32)rx->nparens &&
5480 (s1 = rx->offs[paren].start) != -1 &&
5481 (t1 = rx->offs[paren].end) != -1)
5486 if (ckWARN(WARN_UNINITIALIZED))
5487 report_uninit((const SV *)sv);
5492 if (i > 0 && RXp_MATCH_UTF8(rx)) {
5493 const char * const s = rx->subbeg + s1;
5498 if (is_utf8_string_loclen((U8*)s, i, &ep, &el))
5505 Perl_reg_qr_package(pTHX_ REGEXP * const rx)
5507 PERL_ARGS_ASSERT_REG_QR_PACKAGE;
5508 PERL_UNUSED_ARG(rx);
5512 return newSVpvs("Regexp");
5515 /* Scans the name of a named buffer from the pattern.
5516 * If flags is REG_RSN_RETURN_NULL returns null.
5517 * If flags is REG_RSN_RETURN_NAME returns an SV* containing the name
5518 * If flags is REG_RSN_RETURN_DATA returns the data SV* corresponding
5519 * to the parsed name as looked up in the RExC_paren_names hash.
5520 * If there is an error throws a vFAIL().. type exception.
5523 #define REG_RSN_RETURN_NULL 0
5524 #define REG_RSN_RETURN_NAME 1
5525 #define REG_RSN_RETURN_DATA 2
5528 S_reg_scan_name(pTHX_ RExC_state_t *pRExC_state, U32 flags)
5530 char *name_start = RExC_parse;
5532 PERL_ARGS_ASSERT_REG_SCAN_NAME;
5534 if (isIDFIRST_lazy_if(RExC_parse, UTF)) {
5535 /* skip IDFIRST by using do...while */
5538 RExC_parse += UTF8SKIP(RExC_parse);
5539 } while (isALNUM_utf8((U8*)RExC_parse));
5543 } while (isALNUM(*RExC_parse));
5548 = newSVpvn_flags(name_start, (int)(RExC_parse - name_start),
5549 SVs_TEMP | (UTF ? SVf_UTF8 : 0));
5550 if ( flags == REG_RSN_RETURN_NAME)
5552 else if (flags==REG_RSN_RETURN_DATA) {
5555 if ( ! sv_name ) /* should not happen*/
5556 Perl_croak(aTHX_ "panic: no svname in reg_scan_name");
5557 if (RExC_paren_names)
5558 he_str = hv_fetch_ent( RExC_paren_names, sv_name, 0, 0 );
5560 sv_dat = HeVAL(he_str);
5562 vFAIL("Reference to nonexistent named group");
5566 Perl_croak(aTHX_ "panic: bad flag in reg_scan_name");
5573 #define DEBUG_PARSE_MSG(funcname) DEBUG_PARSE_r({ \
5574 int rem=(int)(RExC_end - RExC_parse); \
5583 if (RExC_lastparse!=RExC_parse) \
5584 PerlIO_printf(Perl_debug_log," >%.*s%-*s", \
5587 iscut ? "..." : "<" \
5590 PerlIO_printf(Perl_debug_log,"%16s",""); \
5593 num = RExC_size + 1; \
5595 num=REG_NODE_NUM(RExC_emit); \
5596 if (RExC_lastnum!=num) \
5597 PerlIO_printf(Perl_debug_log,"|%4d",num); \
5599 PerlIO_printf(Perl_debug_log,"|%4s",""); \
5600 PerlIO_printf(Perl_debug_log,"|%*s%-4s", \
5601 (int)((depth*2)), "", \
5605 RExC_lastparse=RExC_parse; \
5610 #define DEBUG_PARSE(funcname) DEBUG_PARSE_r({ \
5611 DEBUG_PARSE_MSG((funcname)); \
5612 PerlIO_printf(Perl_debug_log,"%4s","\n"); \
5614 #define DEBUG_PARSE_FMT(funcname,fmt,args) DEBUG_PARSE_r({ \
5615 DEBUG_PARSE_MSG((funcname)); \
5616 PerlIO_printf(Perl_debug_log,fmt "\n",args); \
5619 - reg - regular expression, i.e. main body or parenthesized thing
5621 * Caller must absorb opening parenthesis.
5623 * Combining parenthesis handling with the base level of regular expression
5624 * is a trifle forced, but the need to tie the tails of the branches to what
5625 * follows makes it hard to avoid.
5627 #define REGTAIL(x,y,z) regtail((x),(y),(z),depth+1)
5629 #define REGTAIL_STUDY(x,y,z) regtail_study((x),(y),(z),depth+1)
5631 #define REGTAIL_STUDY(x,y,z) regtail((x),(y),(z),depth+1)
5635 S_reg(pTHX_ RExC_state_t *pRExC_state, I32 paren, I32 *flagp,U32 depth)
5636 /* paren: Parenthesized? 0=top, 1=(, inside: changed to letter. */
5639 register regnode *ret; /* Will be the head of the group. */
5640 register regnode *br;
5641 register regnode *lastbr;
5642 register regnode *ender = NULL;
5643 register I32 parno = 0;
5645 U32 oregflags = RExC_flags;
5646 bool have_branch = 0;
5648 I32 freeze_paren = 0;
5649 I32 after_freeze = 0;
5651 /* for (?g), (?gc), and (?o) warnings; warning
5652 about (?c) will warn about (?g) -- japhy */
5654 #define WASTED_O 0x01
5655 #define WASTED_G 0x02
5656 #define WASTED_C 0x04
5657 #define WASTED_GC (0x02|0x04)
5658 I32 wastedflags = 0x00;
5660 char * parse_start = RExC_parse; /* MJD */
5661 char * const oregcomp_parse = RExC_parse;
5663 GET_RE_DEBUG_FLAGS_DECL;
5665 PERL_ARGS_ASSERT_REG;
5666 DEBUG_PARSE("reg ");
5668 *flagp = 0; /* Tentatively. */
5671 /* Make an OPEN node, if parenthesized. */
5673 if ( *RExC_parse == '*') { /* (*VERB:ARG) */
5674 char *start_verb = RExC_parse;
5675 STRLEN verb_len = 0;
5676 char *start_arg = NULL;
5677 unsigned char op = 0;
5679 int internal_argval = 0; /* internal_argval is only useful if !argok */
5680 while ( *RExC_parse && *RExC_parse != ')' ) {
5681 if ( *RExC_parse == ':' ) {
5682 start_arg = RExC_parse + 1;
5688 verb_len = RExC_parse - start_verb;
5691 while ( *RExC_parse && *RExC_parse != ')' )
5693 if ( *RExC_parse != ')' )
5694 vFAIL("Unterminated verb pattern argument");
5695 if ( RExC_parse == start_arg )
5698 if ( *RExC_parse != ')' )
5699 vFAIL("Unterminated verb pattern");
5702 switch ( *start_verb ) {
5703 case 'A': /* (*ACCEPT) */
5704 if ( memEQs(start_verb,verb_len,"ACCEPT") ) {
5706 internal_argval = RExC_nestroot;
5709 case 'C': /* (*COMMIT) */
5710 if ( memEQs(start_verb,verb_len,"COMMIT") )
5713 case 'F': /* (*FAIL) */
5714 if ( verb_len==1 || memEQs(start_verb,verb_len,"FAIL") ) {
5719 case ':': /* (*:NAME) */
5720 case 'M': /* (*MARK:NAME) */
5721 if ( verb_len==0 || memEQs(start_verb,verb_len,"MARK") ) {
5726 case 'P': /* (*PRUNE) */
5727 if ( memEQs(start_verb,verb_len,"PRUNE") )
5730 case 'S': /* (*SKIP) */
5731 if ( memEQs(start_verb,verb_len,"SKIP") )
5734 case 'T': /* (*THEN) */
5735 /* [19:06] <TimToady> :: is then */
5736 if ( memEQs(start_verb,verb_len,"THEN") ) {
5738 RExC_seen |= REG_SEEN_CUTGROUP;
5744 vFAIL3("Unknown verb pattern '%.*s'",
5745 verb_len, start_verb);
5748 if ( start_arg && internal_argval ) {
5749 vFAIL3("Verb pattern '%.*s' may not have an argument",
5750 verb_len, start_verb);
5751 } else if ( argok < 0 && !start_arg ) {
5752 vFAIL3("Verb pattern '%.*s' has a mandatory argument",
5753 verb_len, start_verb);
5755 ret = reganode(pRExC_state, op, internal_argval);
5756 if ( ! internal_argval && ! SIZE_ONLY ) {
5758 SV *sv = newSVpvn( start_arg, RExC_parse - start_arg);
5759 ARG(ret) = add_data( pRExC_state, 1, "S" );
5760 RExC_rxi->data->data[ARG(ret)]=(void*)sv;
5767 if (!internal_argval)
5768 RExC_seen |= REG_SEEN_VERBARG;
5769 } else if ( start_arg ) {
5770 vFAIL3("Verb pattern '%.*s' may not have an argument",
5771 verb_len, start_verb);
5773 ret = reg_node(pRExC_state, op);
5775 nextchar(pRExC_state);
5778 if (*RExC_parse == '?') { /* (?...) */
5779 bool is_logical = 0;
5780 const char * const seqstart = RExC_parse;
5781 bool has_use_defaults = FALSE;
5784 paren = *RExC_parse++;
5785 ret = NULL; /* For look-ahead/behind. */
5788 case 'P': /* (?P...) variants for those used to PCRE/Python */
5789 paren = *RExC_parse++;
5790 if ( paren == '<') /* (?P<...>) named capture */
5792 else if (paren == '>') { /* (?P>name) named recursion */
5793 goto named_recursion;
5795 else if (paren == '=') { /* (?P=...) named backref */
5796 /* this pretty much dupes the code for \k<NAME> in regatom(), if
5797 you change this make sure you change that */
5798 char* name_start = RExC_parse;
5800 SV *sv_dat = reg_scan_name(pRExC_state,
5801 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5802 if (RExC_parse == name_start || *RExC_parse != ')')
5803 vFAIL2("Sequence %.3s... not terminated",parse_start);
5806 num = add_data( pRExC_state, 1, "S" );
5807 RExC_rxi->data->data[num]=(void*)sv_dat;
5808 SvREFCNT_inc_simple_void(sv_dat);
5811 ret = reganode(pRExC_state,
5812 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
5816 Set_Node_Offset(ret, parse_start+1);
5817 Set_Node_Cur_Length(ret); /* MJD */
5819 nextchar(pRExC_state);
5823 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5825 case '<': /* (?<...) */
5826 if (*RExC_parse == '!')
5828 else if (*RExC_parse != '=')
5834 case '\'': /* (?'...') */
5835 name_start= RExC_parse;
5836 svname = reg_scan_name(pRExC_state,
5837 SIZE_ONLY ? /* reverse test from the others */
5838 REG_RSN_RETURN_NAME :
5839 REG_RSN_RETURN_NULL);
5840 if (RExC_parse == name_start) {
5842 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5845 if (*RExC_parse != paren)
5846 vFAIL2("Sequence (?%c... not terminated",
5847 paren=='>' ? '<' : paren);
5851 if (!svname) /* shouldnt happen */
5853 "panic: reg_scan_name returned NULL");
5854 if (!RExC_paren_names) {
5855 RExC_paren_names= newHV();
5856 sv_2mortal(MUTABLE_SV(RExC_paren_names));
5858 RExC_paren_name_list= newAV();
5859 sv_2mortal(MUTABLE_SV(RExC_paren_name_list));
5862 he_str = hv_fetch_ent( RExC_paren_names, svname, 1, 0 );
5864 sv_dat = HeVAL(he_str);
5866 /* croak baby croak */
5868 "panic: paren_name hash element allocation failed");
5869 } else if ( SvPOK(sv_dat) ) {
5870 /* (?|...) can mean we have dupes so scan to check
5871 its already been stored. Maybe a flag indicating
5872 we are inside such a construct would be useful,
5873 but the arrays are likely to be quite small, so
5874 for now we punt -- dmq */
5875 IV count = SvIV(sv_dat);
5876 I32 *pv = (I32*)SvPVX(sv_dat);
5878 for ( i = 0 ; i < count ; i++ ) {
5879 if ( pv[i] == RExC_npar ) {
5885 pv = (I32*)SvGROW(sv_dat, SvCUR(sv_dat) + sizeof(I32)+1);
5886 SvCUR_set(sv_dat, SvCUR(sv_dat) + sizeof(I32));
5887 pv[count] = RExC_npar;
5888 SvIV_set(sv_dat, SvIVX(sv_dat) + 1);
5891 (void)SvUPGRADE(sv_dat,SVt_PVNV);
5892 sv_setpvn(sv_dat, (char *)&(RExC_npar), sizeof(I32));
5894 SvIV_set(sv_dat, 1);
5897 if (!av_store(RExC_paren_name_list, RExC_npar, SvREFCNT_inc(svname)))
5898 SvREFCNT_dec(svname);
5901 /*sv_dump(sv_dat);*/
5903 nextchar(pRExC_state);
5905 goto capturing_parens;
5907 RExC_seen |= REG_SEEN_LOOKBEHIND;
5909 case '=': /* (?=...) */
5910 RExC_seen_zerolen++;
5912 case '!': /* (?!...) */
5913 RExC_seen_zerolen++;
5914 if (*RExC_parse == ')') {
5915 ret=reg_node(pRExC_state, OPFAIL);
5916 nextchar(pRExC_state);
5920 case '|': /* (?|...) */
5921 /* branch reset, behave like a (?:...) except that
5922 buffers in alternations share the same numbers */
5924 after_freeze = freeze_paren = RExC_npar;
5926 case ':': /* (?:...) */
5927 case '>': /* (?>...) */
5929 case '$': /* (?$...) */
5930 case '@': /* (?@...) */
5931 vFAIL2("Sequence (?%c...) not implemented", (int)paren);
5933 case '#': /* (?#...) */
5934 while (*RExC_parse && *RExC_parse != ')')
5936 if (*RExC_parse != ')')
5937 FAIL("Sequence (?#... not terminated");
5938 nextchar(pRExC_state);
5941 case '0' : /* (?0) */
5942 case 'R' : /* (?R) */
5943 if (*RExC_parse != ')')
5944 FAIL("Sequence (?R) not terminated");
5945 ret = reg_node(pRExC_state, GOSTART);
5946 *flagp |= POSTPONED;
5947 nextchar(pRExC_state);
5950 { /* named and numeric backreferences */
5952 case '&': /* (?&NAME) */
5953 parse_start = RExC_parse - 1;
5956 SV *sv_dat = reg_scan_name(pRExC_state,
5957 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5958 num = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5960 goto gen_recurse_regop;
5963 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5965 vFAIL("Illegal pattern");
5967 goto parse_recursion;
5969 case '-': /* (?-1) */
5970 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5971 RExC_parse--; /* rewind to let it be handled later */
5975 case '1': case '2': case '3': case '4': /* (?1) */
5976 case '5': case '6': case '7': case '8': case '9':
5979 num = atoi(RExC_parse);
5980 parse_start = RExC_parse - 1; /* MJD */
5981 if (*RExC_parse == '-')
5983 while (isDIGIT(*RExC_parse))
5985 if (*RExC_parse!=')')
5986 vFAIL("Expecting close bracket");
5989 if ( paren == '-' ) {
5991 Diagram of capture buffer numbering.
5992 Top line is the normal capture buffer numbers
5993 Bottom line is the negative indexing as from
5997 /(a(x)y)(a(b(c(?-2)d)e)f)(g(h))/
6001 num = RExC_npar + num;
6004 vFAIL("Reference to nonexistent group");
6006 } else if ( paren == '+' ) {
6007 num = RExC_npar + num - 1;
6010 ret = reganode(pRExC_state, GOSUB, num);
6012 if (num > (I32)RExC_rx->nparens) {
6014 vFAIL("Reference to nonexistent group");
6016 ARG2L_SET( ret, RExC_recurse_count++);
6018 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6019 "Recurse #%"UVuf" to %"IVdf"\n", (UV)ARG(ret), (IV)ARG2L(ret)));
6023 RExC_seen |= REG_SEEN_RECURSE;
6024 Set_Node_Length(ret, 1 + regarglen[OP(ret)]); /* MJD */
6025 Set_Node_Offset(ret, parse_start); /* MJD */
6027 *flagp |= POSTPONED;
6028 nextchar(pRExC_state);
6030 } /* named and numeric backreferences */
6033 case '?': /* (??...) */
6035 if (*RExC_parse != '{') {
6037 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6040 *flagp |= POSTPONED;
6041 paren = *RExC_parse++;
6043 case '{': /* (?{...}) */
6048 char *s = RExC_parse;
6050 RExC_seen_zerolen++;
6051 RExC_seen |= REG_SEEN_EVAL;
6052 while (count && (c = *RExC_parse)) {
6063 if (*RExC_parse != ')') {
6065 vFAIL("Sequence (?{...}) not terminated or not {}-balanced");
6069 OP_4tree *sop, *rop;
6070 SV * const sv = newSVpvn(s, RExC_parse - 1 - s);
6073 Perl_save_re_context(aTHX);
6074 rop = sv_compile_2op(sv, &sop, "re", &pad);
6075 sop->op_private |= OPpREFCOUNTED;
6076 /* re_dup will OpREFCNT_inc */
6077 OpREFCNT_set(sop, 1);
6080 n = add_data(pRExC_state, 3, "nop");
6081 RExC_rxi->data->data[n] = (void*)rop;
6082 RExC_rxi->data->data[n+1] = (void*)sop;
6083 RExC_rxi->data->data[n+2] = (void*)pad;
6086 else { /* First pass */
6087 if (PL_reginterp_cnt < ++RExC_seen_evals
6089 /* No compiled RE interpolated, has runtime
6090 components ===> unsafe. */
6091 FAIL("Eval-group not allowed at runtime, use re 'eval'");
6092 if (PL_tainting && PL_tainted)
6093 FAIL("Eval-group in insecure regular expression");
6094 #if PERL_VERSION > 8
6095 if (IN_PERL_COMPILETIME)
6100 nextchar(pRExC_state);
6102 ret = reg_node(pRExC_state, LOGICAL);
6105 REGTAIL(pRExC_state, ret, reganode(pRExC_state, EVAL, n));
6106 /* deal with the length of this later - MJD */
6109 ret = reganode(pRExC_state, EVAL, n);
6110 Set_Node_Length(ret, RExC_parse - parse_start + 1);
6111 Set_Node_Offset(ret, parse_start);
6114 case '(': /* (?(?{...})...) and (?(?=...)...) */
6117 if (RExC_parse[0] == '?') { /* (?(?...)) */
6118 if (RExC_parse[1] == '=' || RExC_parse[1] == '!'
6119 || RExC_parse[1] == '<'
6120 || RExC_parse[1] == '{') { /* Lookahead or eval. */
6123 ret = reg_node(pRExC_state, LOGICAL);
6126 REGTAIL(pRExC_state, ret, reg(pRExC_state, 1, &flag,depth+1));
6130 else if ( RExC_parse[0] == '<' /* (?(<NAME>)...) */
6131 || RExC_parse[0] == '\'' ) /* (?('NAME')...) */
6133 char ch = RExC_parse[0] == '<' ? '>' : '\'';
6134 char *name_start= RExC_parse++;
6136 SV *sv_dat=reg_scan_name(pRExC_state,
6137 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
6138 if (RExC_parse == name_start || *RExC_parse != ch)
6139 vFAIL2("Sequence (?(%c... not terminated",
6140 (ch == '>' ? '<' : ch));
6143 num = add_data( pRExC_state, 1, "S" );
6144 RExC_rxi->data->data[num]=(void*)sv_dat;
6145 SvREFCNT_inc_simple_void(sv_dat);
6147 ret = reganode(pRExC_state,NGROUPP,num);
6148 goto insert_if_check_paren;
6150 else if (RExC_parse[0] == 'D' &&
6151 RExC_parse[1] == 'E' &&
6152 RExC_parse[2] == 'F' &&
6153 RExC_parse[3] == 'I' &&
6154 RExC_parse[4] == 'N' &&
6155 RExC_parse[5] == 'E')
6157 ret = reganode(pRExC_state,DEFINEP,0);
6160 goto insert_if_check_paren;
6162 else if (RExC_parse[0] == 'R') {
6165 if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
6166 parno = atoi(RExC_parse++);
6167 while (isDIGIT(*RExC_parse))
6169 } else if (RExC_parse[0] == '&') {
6172 sv_dat = reg_scan_name(pRExC_state,
6173 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
6174 parno = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
6176 ret = reganode(pRExC_state,INSUBP,parno);
6177 goto insert_if_check_paren;
6179 else if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
6182 parno = atoi(RExC_parse++);
6184 while (isDIGIT(*RExC_parse))
6186 ret = reganode(pRExC_state, GROUPP, parno);
6188 insert_if_check_paren:
6189 if ((c = *nextchar(pRExC_state)) != ')')
6190 vFAIL("Switch condition not recognized");
6192 REGTAIL(pRExC_state, ret, reganode(pRExC_state, IFTHEN, 0));
6193 br = regbranch(pRExC_state, &flags, 1,depth+1);
6195 br = reganode(pRExC_state, LONGJMP, 0);
6197 REGTAIL(pRExC_state, br, reganode(pRExC_state, LONGJMP, 0));
6198 c = *nextchar(pRExC_state);
6203 vFAIL("(?(DEFINE)....) does not allow branches");
6204 lastbr = reganode(pRExC_state, IFTHEN, 0); /* Fake one for optimizer. */
6205 regbranch(pRExC_state, &flags, 1,depth+1);
6206 REGTAIL(pRExC_state, ret, lastbr);
6209 c = *nextchar(pRExC_state);
6214 vFAIL("Switch (?(condition)... contains too many branches");
6215 ender = reg_node(pRExC_state, TAIL);
6216 REGTAIL(pRExC_state, br, ender);
6218 REGTAIL(pRExC_state, lastbr, ender);
6219 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender);
6222 REGTAIL(pRExC_state, ret, ender);
6223 RExC_size++; /* XXX WHY do we need this?!!
6224 For large programs it seems to be required
6225 but I can't figure out why. -- dmq*/
6229 vFAIL2("Unknown switch condition (?(%.2s", RExC_parse);
6233 RExC_parse--; /* for vFAIL to print correctly */
6234 vFAIL("Sequence (? incomplete");
6236 case DEFAULT_PAT_MOD: /* Use default flags with the exceptions
6238 has_use_defaults = TRUE;
6239 STD_PMMOD_FLAGS_CLEAR(&RExC_flags);
6240 RExC_flags &= ~(RXf_PMf_LOCALE|RXf_PMf_UNICODE);
6244 parse_flags: /* (?i) */
6246 U32 posflags = 0, negflags = 0;
6247 U32 *flagsp = &posflags;
6248 bool has_charset_modifier = 0;
6250 while (*RExC_parse) {
6251 /* && strchr("iogcmsx", *RExC_parse) */
6252 /* (?g), (?gc) and (?o) are useless here
6253 and must be globally applied -- japhy */
6254 switch (*RExC_parse) {
6255 CASE_STD_PMMOD_FLAGS_PARSE_SET(flagsp);
6256 case LOCALE_PAT_MOD:
6257 if (has_charset_modifier || flagsp == &negflags) {
6258 goto fail_modifiers;
6260 posflags |= RXf_PMf_LOCALE;
6261 negflags |= RXf_PMf_UNICODE;
6262 has_charset_modifier = 1;
6264 case UNICODE_PAT_MOD:
6265 if (has_charset_modifier || flagsp == &negflags) {
6266 goto fail_modifiers;
6268 posflags |= RXf_PMf_UNICODE;
6269 negflags |= RXf_PMf_LOCALE;
6270 has_charset_modifier = 1;
6273 if (has_use_defaults
6274 || has_charset_modifier
6275 || flagsp == &negflags)
6277 goto fail_modifiers;
6279 negflags |= (RXf_PMf_LOCALE|RXf_PMf_UNICODE);
6280 has_charset_modifier = 1;
6282 case ONCE_PAT_MOD: /* 'o' */
6283 case GLOBAL_PAT_MOD: /* 'g' */
6284 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6285 const I32 wflagbit = *RExC_parse == 'o' ? WASTED_O : WASTED_G;
6286 if (! (wastedflags & wflagbit) ) {
6287 wastedflags |= wflagbit;
6290 "Useless (%s%c) - %suse /%c modifier",
6291 flagsp == &negflags ? "?-" : "?",
6293 flagsp == &negflags ? "don't " : "",
6300 case CONTINUE_PAT_MOD: /* 'c' */
6301 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6302 if (! (wastedflags & WASTED_C) ) {
6303 wastedflags |= WASTED_GC;
6306 "Useless (%sc) - %suse /gc modifier",
6307 flagsp == &negflags ? "?-" : "?",
6308 flagsp == &negflags ? "don't " : ""
6313 case KEEPCOPY_PAT_MOD: /* 'p' */
6314 if (flagsp == &negflags) {
6316 ckWARNreg(RExC_parse + 1,"Useless use of (?-p)");
6318 *flagsp |= RXf_PMf_KEEPCOPY;
6322 /* A flag is a default iff it is following a minus, so
6323 * if there is a minus, it means will be trying to
6324 * re-specify a default which is an error */
6325 if (has_use_defaults || flagsp == &negflags) {
6328 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6332 wastedflags = 0; /* reset so (?g-c) warns twice */
6338 RExC_flags |= posflags;
6339 RExC_flags &= ~negflags;
6341 oregflags |= posflags;
6342 oregflags &= ~negflags;
6344 nextchar(pRExC_state);
6355 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6360 }} /* one for the default block, one for the switch */
6367 ret = reganode(pRExC_state, OPEN, parno);
6370 RExC_nestroot = parno;
6371 if (RExC_seen & REG_SEEN_RECURSE
6372 && !RExC_open_parens[parno-1])
6374 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6375 "Setting open paren #%"IVdf" to %d\n",
6376 (IV)parno, REG_NODE_NUM(ret)));
6377 RExC_open_parens[parno-1]= ret;
6380 Set_Node_Length(ret, 1); /* MJD */
6381 Set_Node_Offset(ret, RExC_parse); /* MJD */
6389 /* Pick up the branches, linking them together. */
6390 parse_start = RExC_parse; /* MJD */
6391 br = regbranch(pRExC_state, &flags, 1,depth+1);
6394 if (RExC_npar > after_freeze)
6395 after_freeze = RExC_npar;
6396 RExC_npar = freeze_paren;
6399 /* branch_len = (paren != 0); */
6403 if (*RExC_parse == '|') {
6404 if (!SIZE_ONLY && RExC_extralen) {
6405 reginsert(pRExC_state, BRANCHJ, br, depth+1);
6408 reginsert(pRExC_state, BRANCH, br, depth+1);
6409 Set_Node_Length(br, paren != 0);
6410 Set_Node_Offset_To_R(br-RExC_emit_start, parse_start-RExC_start);
6414 RExC_extralen += 1; /* For BRANCHJ-BRANCH. */
6416 else if (paren == ':') {
6417 *flagp |= flags&SIMPLE;
6419 if (is_open) { /* Starts with OPEN. */
6420 REGTAIL(pRExC_state, ret, br); /* OPEN -> first. */
6422 else if (paren != '?') /* Not Conditional */
6424 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6426 while (*RExC_parse == '|') {
6427 if (!SIZE_ONLY && RExC_extralen) {
6428 ender = reganode(pRExC_state, LONGJMP,0);
6429 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender); /* Append to the previous. */
6432 RExC_extralen += 2; /* Account for LONGJMP. */
6433 nextchar(pRExC_state);
6435 if (RExC_npar > after_freeze)
6436 after_freeze = RExC_npar;
6437 RExC_npar = freeze_paren;
6439 br = regbranch(pRExC_state, &flags, 0, depth+1);
6443 REGTAIL(pRExC_state, lastbr, br); /* BRANCH -> BRANCH. */
6445 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6448 if (have_branch || paren != ':') {
6449 /* Make a closing node, and hook it on the end. */
6452 ender = reg_node(pRExC_state, TAIL);
6455 ender = reganode(pRExC_state, CLOSE, parno);
6456 if (!SIZE_ONLY && RExC_seen & REG_SEEN_RECURSE) {
6457 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6458 "Setting close paren #%"IVdf" to %d\n",
6459 (IV)parno, REG_NODE_NUM(ender)));
6460 RExC_close_parens[parno-1]= ender;
6461 if (RExC_nestroot == parno)
6464 Set_Node_Offset(ender,RExC_parse+1); /* MJD */
6465 Set_Node_Length(ender,1); /* MJD */
6471 *flagp &= ~HASWIDTH;
6474 ender = reg_node(pRExC_state, SUCCEED);
6477 ender = reg_node(pRExC_state, END);
6479 assert(!RExC_opend); /* there can only be one! */
6484 REGTAIL(pRExC_state, lastbr, ender);
6486 if (have_branch && !SIZE_ONLY) {
6488 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
6490 /* Hook the tails of the branches to the closing node. */
6491 for (br = ret; br; br = regnext(br)) {
6492 const U8 op = PL_regkind[OP(br)];
6494 REGTAIL_STUDY(pRExC_state, NEXTOPER(br), ender);
6496 else if (op == BRANCHJ) {
6497 REGTAIL_STUDY(pRExC_state, NEXTOPER(NEXTOPER(br)), ender);
6505 static const char parens[] = "=!<,>";
6507 if (paren && (p = strchr(parens, paren))) {
6508 U8 node = ((p - parens) % 2) ? UNLESSM : IFMATCH;
6509 int flag = (p - parens) > 1;
6512 node = SUSPEND, flag = 0;
6513 reginsert(pRExC_state, node,ret, depth+1);
6514 Set_Node_Cur_Length(ret);
6515 Set_Node_Offset(ret, parse_start + 1);
6517 REGTAIL_STUDY(pRExC_state, ret, reg_node(pRExC_state, TAIL));
6521 /* Check for proper termination. */
6523 RExC_flags = oregflags;
6524 if (RExC_parse >= RExC_end || *nextchar(pRExC_state) != ')') {
6525 RExC_parse = oregcomp_parse;
6526 vFAIL("Unmatched (");
6529 else if (!paren && RExC_parse < RExC_end) {
6530 if (*RExC_parse == ')') {
6532 vFAIL("Unmatched )");
6535 FAIL("Junk on end of regexp"); /* "Can't happen". */
6539 RExC_npar = after_freeze;
6544 - regbranch - one alternative of an | operator
6546 * Implements the concatenation operator.
6549 S_regbranch(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, I32 first, U32 depth)
6552 register regnode *ret;
6553 register regnode *chain = NULL;
6554 register regnode *latest;
6555 I32 flags = 0, c = 0;
6556 GET_RE_DEBUG_FLAGS_DECL;
6558 PERL_ARGS_ASSERT_REGBRANCH;
6560 DEBUG_PARSE("brnc");
6565 if (!SIZE_ONLY && RExC_extralen)
6566 ret = reganode(pRExC_state, BRANCHJ,0);
6568 ret = reg_node(pRExC_state, BRANCH);
6569 Set_Node_Length(ret, 1);
6573 if (!first && SIZE_ONLY)
6574 RExC_extralen += 1; /* BRANCHJ */
6576 *flagp = WORST; /* Tentatively. */
6579 nextchar(pRExC_state);
6580 while (RExC_parse < RExC_end && *RExC_parse != '|' && *RExC_parse != ')') {
6582 latest = regpiece(pRExC_state, &flags,depth+1);
6583 if (latest == NULL) {
6584 if (flags & TRYAGAIN)
6588 else if (ret == NULL)
6590 *flagp |= flags&(HASWIDTH|POSTPONED);
6591 if (chain == NULL) /* First piece. */
6592 *flagp |= flags&SPSTART;
6595 REGTAIL(pRExC_state, chain, latest);
6600 if (chain == NULL) { /* Loop ran zero times. */
6601 chain = reg_node(pRExC_state, NOTHING);
6606 *flagp |= flags&SIMPLE;
6613 - regpiece - something followed by possible [*+?]
6615 * Note that the branching code sequences used for ? and the general cases
6616 * of * and + are somewhat optimized: they use the same NOTHING node as
6617 * both the endmarker for their branch list and the body of the last branch.
6618 * It might seem that this node could be dispensed with entirely, but the
6619 * endmarker role is not redundant.
6622 S_regpiece(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6625 register regnode *ret;
6627 register char *next;
6629 const char * const origparse = RExC_parse;
6631 I32 max = REG_INFTY;
6633 const char *maxpos = NULL;
6634 GET_RE_DEBUG_FLAGS_DECL;
6636 PERL_ARGS_ASSERT_REGPIECE;
6638 DEBUG_PARSE("piec");
6640 ret = regatom(pRExC_state, &flags,depth+1);
6642 if (flags & TRYAGAIN)
6649 if (op == '{' && regcurly(RExC_parse)) {
6651 parse_start = RExC_parse; /* MJD */
6652 next = RExC_parse + 1;
6653 while (isDIGIT(*next) || *next == ',') {
6662 if (*next == '}') { /* got one */
6666 min = atoi(RExC_parse);
6670 maxpos = RExC_parse;
6672 if (!max && *maxpos != '0')
6673 max = REG_INFTY; /* meaning "infinity" */
6674 else if (max >= REG_INFTY)
6675 vFAIL2("Quantifier in {,} bigger than %d", REG_INFTY - 1);
6677 nextchar(pRExC_state);
6680 if ((flags&SIMPLE)) {
6681 RExC_naughty += 2 + RExC_naughty / 2;
6682 reginsert(pRExC_state, CURLY, ret, depth+1);
6683 Set_Node_Offset(ret, parse_start+1); /* MJD */
6684 Set_Node_Cur_Length(ret);
6687 regnode * const w = reg_node(pRExC_state, WHILEM);
6690 REGTAIL(pRExC_state, ret, w);
6691 if (!SIZE_ONLY && RExC_extralen) {
6692 reginsert(pRExC_state, LONGJMP,ret, depth+1);
6693 reginsert(pRExC_state, NOTHING,ret, depth+1);
6694 NEXT_OFF(ret) = 3; /* Go over LONGJMP. */
6696 reginsert(pRExC_state, CURLYX,ret, depth+1);
6698 Set_Node_Offset(ret, parse_start+1);
6699 Set_Node_Length(ret,
6700 op == '{' ? (RExC_parse - parse_start) : 1);
6702 if (!SIZE_ONLY && RExC_extralen)
6703 NEXT_OFF(ret) = 3; /* Go over NOTHING to LONGJMP. */
6704 REGTAIL(pRExC_state, ret, reg_node(pRExC_state, NOTHING));
6706 RExC_whilem_seen++, RExC_extralen += 3;
6707 RExC_naughty += 4 + RExC_naughty; /* compound interest */
6716 vFAIL("Can't do {n,m} with n > m");
6718 ARG1_SET(ret, (U16)min);
6719 ARG2_SET(ret, (U16)max);
6731 #if 0 /* Now runtime fix should be reliable. */
6733 /* if this is reinstated, don't forget to put this back into perldiag:
6735 =item Regexp *+ operand could be empty at {#} in regex m/%s/
6737 (F) The part of the regexp subject to either the * or + quantifier
6738 could match an empty string. The {#} shows in the regular
6739 expression about where the problem was discovered.
6743 if (!(flags&HASWIDTH) && op != '?')
6744 vFAIL("Regexp *+ operand could be empty");
6747 parse_start = RExC_parse;
6748 nextchar(pRExC_state);
6750 *flagp = (op != '+') ? (WORST|SPSTART|HASWIDTH) : (WORST|HASWIDTH);
6752 if (op == '*' && (flags&SIMPLE)) {
6753 reginsert(pRExC_state, STAR, ret, depth+1);
6757 else if (op == '*') {
6761 else if (op == '+' && (flags&SIMPLE)) {
6762 reginsert(pRExC_state, PLUS, ret, depth+1);
6766 else if (op == '+') {
6770 else if (op == '?') {
6775 if (!SIZE_ONLY && !(flags&(HASWIDTH|POSTPONED)) && max > REG_INFTY/3) {
6776 ckWARN3reg(RExC_parse,
6777 "%.*s matches null string many times",
6778 (int)(RExC_parse >= origparse ? RExC_parse - origparse : 0),
6782 if (RExC_parse < RExC_end && *RExC_parse == '?') {
6783 nextchar(pRExC_state);
6784 reginsert(pRExC_state, MINMOD, ret, depth+1);
6785 REGTAIL(pRExC_state, ret, ret + NODE_STEP_REGNODE);
6787 #ifndef REG_ALLOW_MINMOD_SUSPEND
6790 if (RExC_parse < RExC_end && *RExC_parse == '+') {
6792 nextchar(pRExC_state);
6793 ender = reg_node(pRExC_state, SUCCEED);
6794 REGTAIL(pRExC_state, ret, ender);
6795 reginsert(pRExC_state, SUSPEND, ret, depth+1);
6797 ender = reg_node(pRExC_state, TAIL);
6798 REGTAIL(pRExC_state, ret, ender);
6802 if (RExC_parse < RExC_end && ISMULT2(RExC_parse)) {
6804 vFAIL("Nested quantifiers");
6811 /* reg_namedseq(pRExC_state,UVp)
6813 This is expected to be called by a parser routine that has
6814 recognized '\N' and needs to handle the rest. RExC_parse is
6815 expected to point at the first char following the N at the time
6818 The \N may be inside (indicated by valuep not being NULL) or outside a
6821 \N may begin either a named sequence, or if outside a character class, mean
6822 to match a non-newline. For non single-quoted regexes, the tokenizer has
6823 attempted to decide which, and in the case of a named sequence converted it
6824 into one of the forms: \N{} (if the sequence is null), or \N{U+c1.c2...},
6825 where c1... are the characters in the sequence. For single-quoted regexes,
6826 the tokenizer passes the \N sequence through unchanged; this code will not
6827 attempt to determine this nor expand those. The net effect is that if the
6828 beginning of the passed-in pattern isn't '{U+' or there is no '}', it
6829 signals that this \N occurrence means to match a non-newline.
6831 Only the \N{U+...} form should occur in a character class, for the same
6832 reason that '.' inside a character class means to just match a period: it
6833 just doesn't make sense.
6835 If valuep is non-null then it is assumed that we are parsing inside
6836 of a charclass definition and the first codepoint in the resolved
6837 string is returned via *valuep and the routine will return NULL.
6838 In this mode if a multichar string is returned from the charnames
6839 handler, a warning will be issued, and only the first char in the
6840 sequence will be examined. If the string returned is zero length
6841 then the value of *valuep is undefined and NON-NULL will
6842 be returned to indicate failure. (This will NOT be a valid pointer
6845 If valuep is null then it is assumed that we are parsing normal text and a
6846 new EXACT node is inserted into the program containing the resolved string,
6847 and a pointer to the new node is returned. But if the string is zero length
6848 a NOTHING node is emitted instead.
6850 On success RExC_parse is set to the char following the endbrace.
6851 Parsing failures will generate a fatal error via vFAIL(...)
6854 S_reg_namedseq(pTHX_ RExC_state_t *pRExC_state, UV *valuep, I32 *flagp)
6856 char * endbrace; /* '}' following the name */
6857 regnode *ret = NULL;
6859 char* parse_start = RExC_parse - 2; /* points to the '\N' */
6863 GET_RE_DEBUG_FLAGS_DECL;
6865 PERL_ARGS_ASSERT_REG_NAMEDSEQ;
6869 /* The [^\n] meaning of \N ignores spaces and comments under the /x
6870 * modifier. The other meaning does not */
6871 p = (RExC_flags & RXf_PMf_EXTENDED)
6872 ? regwhite( pRExC_state, RExC_parse )
6875 /* Disambiguate between \N meaning a named character versus \N meaning
6876 * [^\n]. The former is assumed when it can't be the latter. */
6877 if (*p != '{' || regcurly(p)) {
6880 /* no bare \N in a charclass */
6881 vFAIL("\\N in a character class must be a named character: \\N{...}");
6883 nextchar(pRExC_state);
6884 ret = reg_node(pRExC_state, REG_ANY);
6885 *flagp |= HASWIDTH|SIMPLE;
6888 Set_Node_Length(ret, 1); /* MJD */
6892 /* Here, we have decided it should be a named sequence */
6894 /* The test above made sure that the next real character is a '{', but
6895 * under the /x modifier, it could be separated by space (or a comment and
6896 * \n) and this is not allowed (for consistency with \x{...} and the
6897 * tokenizer handling of \N{NAME}). */
6898 if (*RExC_parse != '{') {
6899 vFAIL("Missing braces on \\N{}");
6902 RExC_parse++; /* Skip past the '{' */
6904 if (! (endbrace = strchr(RExC_parse, '}')) /* no trailing brace */
6905 || ! (endbrace == RExC_parse /* nothing between the {} */
6906 || (endbrace - RExC_parse >= 2 /* U+ (bad hex is checked below */
6907 && strnEQ(RExC_parse, "U+", 2)))) /* for a better error msg) */
6909 if (endbrace) RExC_parse = endbrace; /* position msg's '<--HERE' */
6910 vFAIL("\\N{NAME} must be resolved by the lexer");
6913 if (endbrace == RExC_parse) { /* empty: \N{} */
6915 RExC_parse = endbrace + 1;
6916 return reg_node(pRExC_state,NOTHING);
6920 ckWARNreg(RExC_parse,
6921 "Ignoring zero length \\N{} in character class"
6923 RExC_parse = endbrace + 1;
6926 return (regnode *) &RExC_parse; /* Invalid regnode pointer */
6929 REQUIRE_UTF8; /* named sequences imply Unicode semantics */
6930 RExC_parse += 2; /* Skip past the 'U+' */
6932 if (valuep) { /* In a bracketed char class */
6933 /* We only pay attention to the first char of
6934 multichar strings being returned. I kinda wonder
6935 if this makes sense as it does change the behaviour
6936 from earlier versions, OTOH that behaviour was broken
6937 as well. XXX Solution is to recharacterize as
6938 [rest-of-class]|multi1|multi2... */
6940 STRLEN length_of_hex;
6941 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
6942 | PERL_SCAN_DISALLOW_PREFIX
6943 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
6945 char * endchar = RExC_parse + strcspn(RExC_parse, ".}");
6946 if (endchar < endbrace) {
6947 ckWARNreg(endchar, "Using just the first character returned by \\N{} in character class");
6950 length_of_hex = (STRLEN)(endchar - RExC_parse);
6951 *valuep = grok_hex(RExC_parse, &length_of_hex, &flags, NULL);
6953 /* The tokenizer should have guaranteed validity, but it's possible to
6954 * bypass it by using single quoting, so check */
6955 if (length_of_hex == 0
6956 || length_of_hex != (STRLEN)(endchar - RExC_parse) )
6958 RExC_parse += length_of_hex; /* Includes all the valid */
6959 RExC_parse += (RExC_orig_utf8) /* point to after 1st invalid */
6960 ? UTF8SKIP(RExC_parse)
6962 /* Guard against malformed utf8 */
6963 if (RExC_parse >= endchar) RExC_parse = endchar;
6964 vFAIL("Invalid hexadecimal number in \\N{U+...}");
6967 RExC_parse = endbrace + 1;
6968 if (endchar == endbrace) return NULL;
6970 ret = (regnode *) &RExC_parse; /* Invalid regnode pointer */
6972 else { /* Not a char class */
6973 char *s; /* String to put in generated EXACT node */
6974 STRLEN len = 0; /* Its current byte length */
6975 char *endchar; /* Points to '.' or '}' ending cur char in the input
6978 ret = reg_node(pRExC_state,
6979 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
6982 /* Exact nodes can hold only a U8 length's of text = 255. Loop through
6983 * the input which is of the form now 'c1.c2.c3...}' until find the
6984 * ending brace or exceed length 255. The characters that exceed this
6985 * limit are dropped. The limit could be relaxed should it become
6986 * desirable by reparsing this as (?:\N{NAME}), so could generate
6987 * multiple EXACT nodes, as is done for just regular input. But this
6988 * is primarily a named character, and not intended to be a huge long
6989 * string, so 255 bytes should be good enough */
6991 STRLEN length_of_hex;
6992 I32 grok_flags = PERL_SCAN_ALLOW_UNDERSCORES
6993 | PERL_SCAN_DISALLOW_PREFIX
6994 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
6995 UV cp; /* Ord of current character */
6997 /* Code points are separated by dots. If none, there is only one
6998 * code point, and is terminated by the brace */
6999 endchar = RExC_parse + strcspn(RExC_parse, ".}");
7001 /* The values are Unicode even on EBCDIC machines */
7002 length_of_hex = (STRLEN)(endchar - RExC_parse);
7003 cp = grok_hex(RExC_parse, &length_of_hex, &grok_flags, NULL);
7004 if ( length_of_hex == 0
7005 || length_of_hex != (STRLEN)(endchar - RExC_parse) )
7007 RExC_parse += length_of_hex; /* Includes all the valid */
7008 RExC_parse += (RExC_orig_utf8) /* point to after 1st invalid */
7009 ? UTF8SKIP(RExC_parse)
7011 /* Guard against malformed utf8 */
7012 if (RExC_parse >= endchar) RExC_parse = endchar;
7013 vFAIL("Invalid hexadecimal number in \\N{U+...}");
7016 if (! FOLD) { /* Not folding, just append to the string */
7019 /* Quit before adding this character if would exceed limit */
7020 if (len + UNISKIP(cp) > U8_MAX) break;
7022 unilen = reguni(pRExC_state, cp, s);
7027 } else { /* Folding, output the folded equivalent */
7028 STRLEN foldlen,numlen;
7029 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
7030 cp = toFOLD_uni(cp, tmpbuf, &foldlen);
7032 /* Quit before exceeding size limit */
7033 if (len + foldlen > U8_MAX) break;
7035 for (foldbuf = tmpbuf;
7039 cp = utf8_to_uvchr(foldbuf, &numlen);
7041 const STRLEN unilen = reguni(pRExC_state, cp, s);
7044 /* In EBCDIC the numlen and unilen can differ. */
7046 if (numlen >= foldlen)
7050 break; /* "Can't happen." */
7054 /* Point to the beginning of the next character in the sequence. */
7055 RExC_parse = endchar + 1;
7057 /* Quit if no more characters */
7058 if (RExC_parse >= endbrace) break;
7063 if (RExC_parse < endbrace) {
7064 ckWARNreg(RExC_parse - 1,
7065 "Using just the first characters returned by \\N{}");
7068 RExC_size += STR_SZ(len);
7071 RExC_emit += STR_SZ(len);
7074 RExC_parse = endbrace + 1;
7076 *flagp |= HASWIDTH; /* Not SIMPLE, as that causes the engine to fail
7077 with malformed in t/re/pat_advanced.t */
7079 Set_Node_Cur_Length(ret); /* MJD */
7080 nextchar(pRExC_state);
7090 * It returns the code point in utf8 for the value in *encp.
7091 * value: a code value in the source encoding
7092 * encp: a pointer to an Encode object
7094 * If the result from Encode is not a single character,
7095 * it returns U+FFFD (Replacement character) and sets *encp to NULL.
7098 S_reg_recode(pTHX_ const char value, SV **encp)
7101 SV * const sv = newSVpvn_flags(&value, numlen, SVs_TEMP);
7102 const char * const s = *encp ? sv_recode_to_utf8(sv, *encp) : SvPVX(sv);
7103 const STRLEN newlen = SvCUR(sv);
7104 UV uv = UNICODE_REPLACEMENT;
7106 PERL_ARGS_ASSERT_REG_RECODE;
7110 ? utf8n_to_uvchr((U8*)s, newlen, &numlen, UTF8_ALLOW_DEFAULT)
7113 if (!newlen || numlen != newlen) {
7114 uv = UNICODE_REPLACEMENT;
7122 - regatom - the lowest level
7124 Try to identify anything special at the start of the pattern. If there
7125 is, then handle it as required. This may involve generating a single regop,
7126 such as for an assertion; or it may involve recursing, such as to
7127 handle a () structure.
7129 If the string doesn't start with something special then we gobble up
7130 as much literal text as we can.
7132 Once we have been able to handle whatever type of thing started the
7133 sequence, we return.
7135 Note: we have to be careful with escapes, as they can be both literal
7136 and special, and in the case of \10 and friends can either, depending
7137 on context. Specifically there are two seperate switches for handling
7138 escape sequences, with the one for handling literal escapes requiring
7139 a dummy entry for all of the special escapes that are actually handled
7144 S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
7147 register regnode *ret = NULL;
7149 char *parse_start = RExC_parse;
7150 GET_RE_DEBUG_FLAGS_DECL;
7151 DEBUG_PARSE("atom");
7152 *flagp = WORST; /* Tentatively. */
7154 PERL_ARGS_ASSERT_REGATOM;
7157 switch ((U8)*RExC_parse) {
7159 RExC_seen_zerolen++;
7160 nextchar(pRExC_state);
7161 if (RExC_flags & RXf_PMf_MULTILINE)
7162 ret = reg_node(pRExC_state, MBOL);
7163 else if (RExC_flags & RXf_PMf_SINGLELINE)
7164 ret = reg_node(pRExC_state, SBOL);
7166 ret = reg_node(pRExC_state, BOL);
7167 Set_Node_Length(ret, 1); /* MJD */
7170 nextchar(pRExC_state);
7172 RExC_seen_zerolen++;
7173 if (RExC_flags & RXf_PMf_MULTILINE)
7174 ret = reg_node(pRExC_state, MEOL);
7175 else if (RExC_flags & RXf_PMf_SINGLELINE)
7176 ret = reg_node(pRExC_state, SEOL);
7178 ret = reg_node(pRExC_state, EOL);
7179 Set_Node_Length(ret, 1); /* MJD */
7182 nextchar(pRExC_state);
7183 if (RExC_flags & RXf_PMf_SINGLELINE)
7184 ret = reg_node(pRExC_state, SANY);
7186 ret = reg_node(pRExC_state, REG_ANY);
7187 *flagp |= HASWIDTH|SIMPLE;
7189 Set_Node_Length(ret, 1); /* MJD */
7193 char * const oregcomp_parse = ++RExC_parse;
7194 ret = regclass(pRExC_state,depth+1);
7195 if (*RExC_parse != ']') {
7196 RExC_parse = oregcomp_parse;
7197 vFAIL("Unmatched [");
7199 nextchar(pRExC_state);
7200 *flagp |= HASWIDTH|SIMPLE;
7201 Set_Node_Length(ret, RExC_parse - oregcomp_parse + 1); /* MJD */
7205 nextchar(pRExC_state);
7206 ret = reg(pRExC_state, 1, &flags,depth+1);
7208 if (flags & TRYAGAIN) {
7209 if (RExC_parse == RExC_end) {
7210 /* Make parent create an empty node if needed. */
7218 *flagp |= flags&(HASWIDTH|SPSTART|SIMPLE|POSTPONED);
7222 if (flags & TRYAGAIN) {
7226 vFAIL("Internal urp");
7227 /* Supposed to be caught earlier. */
7230 if (!regcurly(RExC_parse)) {
7239 vFAIL("Quantifier follows nothing");
7247 len=0; /* silence a spurious compiler warning */
7248 if ((cp = what_len_TRICKYFOLD_safe(RExC_parse,RExC_end,UTF,len))) {
7249 *flagp |= HASWIDTH; /* could be SIMPLE too, but needs a handler in regexec.regrepeat */
7250 RExC_parse+=len-1; /* we get one from nextchar() as well. :-( */
7251 ret = reganode(pRExC_state, FOLDCHAR, cp);
7252 Set_Node_Length(ret, 1); /* MJD */
7253 nextchar(pRExC_state); /* kill whitespace under /x */
7261 This switch handles escape sequences that resolve to some kind
7262 of special regop and not to literal text. Escape sequnces that
7263 resolve to literal text are handled below in the switch marked
7266 Every entry in this switch *must* have a corresponding entry
7267 in the literal escape switch. However, the opposite is not
7268 required, as the default for this switch is to jump to the
7269 literal text handling code.
7271 switch ((U8)*++RExC_parse) {
7276 /* Special Escapes */
7278 RExC_seen_zerolen++;
7279 ret = reg_node(pRExC_state, SBOL);
7281 goto finish_meta_pat;
7283 ret = reg_node(pRExC_state, GPOS);
7284 RExC_seen |= REG_SEEN_GPOS;
7286 goto finish_meta_pat;
7288 RExC_seen_zerolen++;
7289 ret = reg_node(pRExC_state, KEEPS);
7291 /* XXX:dmq : disabling in-place substitution seems to
7292 * be necessary here to avoid cases of memory corruption, as
7293 * with: C<$_="x" x 80; s/x\K/y/> -- rgs
7295 RExC_seen |= REG_SEEN_LOOKBEHIND;
7296 goto finish_meta_pat;
7298 ret = reg_node(pRExC_state, SEOL);
7300 RExC_seen_zerolen++; /* Do not optimize RE away */
7301 goto finish_meta_pat;
7303 ret = reg_node(pRExC_state, EOS);
7305 RExC_seen_zerolen++; /* Do not optimize RE away */
7306 goto finish_meta_pat;
7308 ret = reg_node(pRExC_state, CANY);
7309 RExC_seen |= REG_SEEN_CANY;
7310 *flagp |= HASWIDTH|SIMPLE;
7311 goto finish_meta_pat;
7313 ret = reg_node(pRExC_state, CLUMP);
7315 goto finish_meta_pat;
7318 ret = reg_node(pRExC_state, (U8)(ALNUML));
7320 ret = reg_node(pRExC_state, (U8)(ALNUM));
7321 FLAGS(ret) = (UNI_SEMANTICS) ? USE_UNI : 0;
7323 *flagp |= HASWIDTH|SIMPLE;
7324 goto finish_meta_pat;
7327 ret = reg_node(pRExC_state, (U8)(NALNUML));
7329 ret = reg_node(pRExC_state, (U8)(NALNUM));
7330 FLAGS(ret) = (UNI_SEMANTICS) ? USE_UNI : 0;
7332 *flagp |= HASWIDTH|SIMPLE;
7333 goto finish_meta_pat;
7335 RExC_seen_zerolen++;
7336 RExC_seen |= REG_SEEN_LOOKBEHIND;
7338 ret = reg_node(pRExC_state, (U8)(BOUNDL));
7340 ret = reg_node(pRExC_state, (U8)(BOUND));
7341 FLAGS(ret) = (UNI_SEMANTICS) ? USE_UNI : 0;
7344 goto finish_meta_pat;
7346 RExC_seen_zerolen++;
7347 RExC_seen |= REG_SEEN_LOOKBEHIND;
7349 ret = reg_node(pRExC_state, (U8)(NBOUNDL));
7351 ret = reg_node(pRExC_state, (U8)(NBOUND));
7352 FLAGS(ret) = (UNI_SEMANTICS) ? USE_UNI : 0;
7355 goto finish_meta_pat;
7358 ret = reg_node(pRExC_state, (U8)(SPACEL));
7360 ret = reg_node(pRExC_state, (U8)(SPACE));
7361 FLAGS(ret) = (UNI_SEMANTICS) ? USE_UNI : 0;
7363 *flagp |= HASWIDTH|SIMPLE;
7364 goto finish_meta_pat;
7367 ret = reg_node(pRExC_state, (U8)(NSPACEL));
7369 ret = reg_node(pRExC_state, (U8)(NSPACE));
7370 FLAGS(ret) = (UNI_SEMANTICS) ? USE_UNI : 0;
7372 *flagp |= HASWIDTH|SIMPLE;
7373 goto finish_meta_pat;
7375 ret = reg_node(pRExC_state, DIGIT);
7376 *flagp |= HASWIDTH|SIMPLE;
7377 goto finish_meta_pat;
7379 ret = reg_node(pRExC_state, NDIGIT);
7380 *flagp |= HASWIDTH|SIMPLE;
7381 goto finish_meta_pat;
7383 ret = reg_node(pRExC_state, LNBREAK);
7384 *flagp |= HASWIDTH|SIMPLE;
7385 goto finish_meta_pat;
7387 ret = reg_node(pRExC_state, HORIZWS);
7388 *flagp |= HASWIDTH|SIMPLE;
7389 goto finish_meta_pat;
7391 ret = reg_node(pRExC_state, NHORIZWS);
7392 *flagp |= HASWIDTH|SIMPLE;
7393 goto finish_meta_pat;
7395 ret = reg_node(pRExC_state, VERTWS);
7396 *flagp |= HASWIDTH|SIMPLE;
7397 goto finish_meta_pat;
7399 ret = reg_node(pRExC_state, NVERTWS);
7400 *flagp |= HASWIDTH|SIMPLE;
7402 nextchar(pRExC_state);
7403 Set_Node_Length(ret, 2); /* MJD */
7408 char* const oldregxend = RExC_end;
7410 char* parse_start = RExC_parse - 2;
7413 if (RExC_parse[1] == '{') {
7414 /* a lovely hack--pretend we saw [\pX] instead */
7415 RExC_end = strchr(RExC_parse, '}');
7417 const U8 c = (U8)*RExC_parse;
7419 RExC_end = oldregxend;
7420 vFAIL2("Missing right brace on \\%c{}", c);
7425 RExC_end = RExC_parse + 2;
7426 if (RExC_end > oldregxend)
7427 RExC_end = oldregxend;
7431 ret = regclass(pRExC_state,depth+1);
7433 RExC_end = oldregxend;
7436 Set_Node_Offset(ret, parse_start + 2);
7437 Set_Node_Cur_Length(ret);
7438 nextchar(pRExC_state);
7439 *flagp |= HASWIDTH|SIMPLE;
7443 /* Handle \N and \N{NAME} here and not below because it can be
7444 multicharacter. join_exact() will join them up later on.
7445 Also this makes sure that things like /\N{BLAH}+/ and
7446 \N{BLAH} being multi char Just Happen. dmq*/
7448 ret= reg_namedseq(pRExC_state, NULL, flagp);
7450 case 'k': /* Handle \k<NAME> and \k'NAME' */
7453 char ch= RExC_parse[1];
7454 if (ch != '<' && ch != '\'' && ch != '{') {
7456 vFAIL2("Sequence %.2s... not terminated",parse_start);
7458 /* this pretty much dupes the code for (?P=...) in reg(), if
7459 you change this make sure you change that */
7460 char* name_start = (RExC_parse += 2);
7462 SV *sv_dat = reg_scan_name(pRExC_state,
7463 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7464 ch= (ch == '<') ? '>' : (ch == '{') ? '}' : '\'';
7465 if (RExC_parse == name_start || *RExC_parse != ch)
7466 vFAIL2("Sequence %.3s... not terminated",parse_start);
7469 num = add_data( pRExC_state, 1, "S" );
7470 RExC_rxi->data->data[num]=(void*)sv_dat;
7471 SvREFCNT_inc_simple_void(sv_dat);
7475 ret = reganode(pRExC_state,
7476 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
7480 /* override incorrect value set in reganode MJD */
7481 Set_Node_Offset(ret, parse_start+1);
7482 Set_Node_Cur_Length(ret); /* MJD */
7483 nextchar(pRExC_state);
7489 case '1': case '2': case '3': case '4':
7490 case '5': case '6': case '7': case '8': case '9':
7493 bool isg = *RExC_parse == 'g';
7498 if (*RExC_parse == '{') {
7502 if (*RExC_parse == '-') {
7506 if (hasbrace && !isDIGIT(*RExC_parse)) {
7507 if (isrel) RExC_parse--;
7509 goto parse_named_seq;
7511 num = atoi(RExC_parse);
7512 if (isg && num == 0)
7513 vFAIL("Reference to invalid group 0");
7515 num = RExC_npar - num;
7517 vFAIL("Reference to nonexistent or unclosed group");
7519 if (!isg && num > 9 && num >= RExC_npar)
7522 char * const parse_start = RExC_parse - 1; /* MJD */
7523 while (isDIGIT(*RExC_parse))
7525 if (parse_start == RExC_parse - 1)
7526 vFAIL("Unterminated \\g... pattern");
7528 if (*RExC_parse != '}')
7529 vFAIL("Unterminated \\g{...} pattern");
7533 if (num > (I32)RExC_rx->nparens)
7534 vFAIL("Reference to nonexistent group");
7537 ret = reganode(pRExC_state,
7538 (U8)(FOLD ? (LOC ? REFFL : REFF) : REF),
7542 /* override incorrect value set in reganode MJD */
7543 Set_Node_Offset(ret, parse_start+1);
7544 Set_Node_Cur_Length(ret); /* MJD */
7546 nextchar(pRExC_state);
7551 if (RExC_parse >= RExC_end)
7552 FAIL("Trailing \\");
7555 /* Do not generate "unrecognized" warnings here, we fall
7556 back into the quick-grab loop below */
7563 if (RExC_flags & RXf_PMf_EXTENDED) {
7564 if ( reg_skipcomment( pRExC_state ) )
7571 register STRLEN len;
7576 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
7578 parse_start = RExC_parse - 1;
7584 ret = reg_node(pRExC_state,
7585 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
7587 for (len = 0, p = RExC_parse - 1;
7588 len < 127 && p < RExC_end;
7591 char * const oldp = p;
7593 if (RExC_flags & RXf_PMf_EXTENDED)
7594 p = regwhite( pRExC_state, p );
7599 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7600 goto normal_default;
7610 /* Literal Escapes Switch
7612 This switch is meant to handle escape sequences that
7613 resolve to a literal character.
7615 Every escape sequence that represents something
7616 else, like an assertion or a char class, is handled
7617 in the switch marked 'Special Escapes' above in this
7618 routine, but also has an entry here as anything that
7619 isn't explicitly mentioned here will be treated as
7620 an unescaped equivalent literal.
7624 /* These are all the special escapes. */
7628 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7629 goto normal_default;
7630 case 'A': /* Start assertion */
7631 case 'b': case 'B': /* Word-boundary assertion*/
7632 case 'C': /* Single char !DANGEROUS! */
7633 case 'd': case 'D': /* digit class */
7634 case 'g': case 'G': /* generic-backref, pos assertion */
7635 case 'h': case 'H': /* HORIZWS */
7636 case 'k': case 'K': /* named backref, keep marker */
7637 case 'N': /* named char sequence */
7638 case 'p': case 'P': /* Unicode property */
7639 case 'R': /* LNBREAK */
7640 case 's': case 'S': /* space class */
7641 case 'v': case 'V': /* VERTWS */
7642 case 'w': case 'W': /* word class */
7643 case 'X': /* eXtended Unicode "combining character sequence" */
7644 case 'z': case 'Z': /* End of line/string assertion */
7648 /* Anything after here is an escape that resolves to a
7649 literal. (Except digits, which may or may not)
7668 ender = ASCII_TO_NATIVE('\033');
7672 ender = ASCII_TO_NATIVE('\007');
7677 STRLEN brace_len = len;
7679 const char* error_msg;
7681 bool valid = grok_bslash_o(p,
7688 RExC_parse = p; /* going to die anyway; point
7689 to exact spot of failure */
7696 if (PL_encoding && ender < 0x100) {
7697 goto recode_encoding;
7706 char* const e = strchr(p, '}');
7710 vFAIL("Missing right brace on \\x{}");
7713 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
7714 | PERL_SCAN_DISALLOW_PREFIX;
7715 STRLEN numlen = e - p - 1;
7716 ender = grok_hex(p + 1, &numlen, &flags, NULL);
7723 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
7725 ender = grok_hex(p, &numlen, &flags, NULL);
7728 if (PL_encoding && ender < 0x100)
7729 goto recode_encoding;
7733 ender = grok_bslash_c(*p++, SIZE_ONLY);
7735 case '0': case '1': case '2': case '3':case '4':
7736 case '5': case '6': case '7': case '8':case '9':
7738 (isDIGIT(p[1]) && atoi(p) >= RExC_npar))
7740 I32 flags = PERL_SCAN_SILENT_ILLDIGIT;
7742 ender = grok_oct(p, &numlen, &flags, NULL);
7752 if (PL_encoding && ender < 0x100)
7753 goto recode_encoding;
7757 SV* enc = PL_encoding;
7758 ender = reg_recode((const char)(U8)ender, &enc);
7759 if (!enc && SIZE_ONLY)
7760 ckWARNreg(p, "Invalid escape in the specified encoding");
7766 FAIL("Trailing \\");
7769 if (!SIZE_ONLY&& isALPHA(*p))
7770 ckWARN2reg(p + 1, "Unrecognized escape \\%c passed through", UCHARAT(p));
7771 goto normal_default;
7776 if (UTF8_IS_START(*p) && UTF) {
7778 ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
7779 &numlen, UTF8_ALLOW_DEFAULT);
7786 if ( RExC_flags & RXf_PMf_EXTENDED)
7787 p = regwhite( pRExC_state, p );
7789 /* Prime the casefolded buffer. */
7790 ender = toFOLD_uni(ender, tmpbuf, &foldlen);
7792 if (p < RExC_end && ISMULT2(p)) { /* Back off on ?+*. */
7797 /* Emit all the Unicode characters. */
7799 for (foldbuf = tmpbuf;
7801 foldlen -= numlen) {
7802 ender = utf8_to_uvchr(foldbuf, &numlen);
7804 const STRLEN unilen = reguni(pRExC_state, ender, s);
7807 /* In EBCDIC the numlen
7808 * and unilen can differ. */
7810 if (numlen >= foldlen)
7814 break; /* "Can't happen." */
7818 const STRLEN unilen = reguni(pRExC_state, ender, s);
7827 REGC((char)ender, s++);
7833 /* Emit all the Unicode characters. */
7835 for (foldbuf = tmpbuf;
7837 foldlen -= numlen) {
7838 ender = utf8_to_uvchr(foldbuf, &numlen);
7840 const STRLEN unilen = reguni(pRExC_state, ender, s);
7843 /* In EBCDIC the numlen
7844 * and unilen can differ. */
7846 if (numlen >= foldlen)
7854 const STRLEN unilen = reguni(pRExC_state, ender, s);
7863 REGC((char)ender, s++);
7867 Set_Node_Cur_Length(ret); /* MJD */
7868 nextchar(pRExC_state);
7870 /* len is STRLEN which is unsigned, need to copy to signed */
7873 vFAIL("Internal disaster");
7877 if (len == 1 && UNI_IS_INVARIANT(ender))
7881 RExC_size += STR_SZ(len);
7884 RExC_emit += STR_SZ(len);
7894 S_regwhite( RExC_state_t *pRExC_state, char *p )
7896 const char *e = RExC_end;
7898 PERL_ARGS_ASSERT_REGWHITE;
7903 else if (*p == '#') {
7912 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
7920 /* Parse POSIX character classes: [[:foo:]], [[=foo=]], [[.foo.]].
7921 Character classes ([:foo:]) can also be negated ([:^foo:]).
7922 Returns a named class id (ANYOF_XXX) if successful, -1 otherwise.
7923 Equivalence classes ([=foo=]) and composites ([.foo.]) are parsed,
7924 but trigger failures because they are currently unimplemented. */
7926 #define POSIXCC_DONE(c) ((c) == ':')
7927 #define POSIXCC_NOTYET(c) ((c) == '=' || (c) == '.')
7928 #define POSIXCC(c) (POSIXCC_DONE(c) || POSIXCC_NOTYET(c))
7931 S_regpposixcc(pTHX_ RExC_state_t *pRExC_state, I32 value)
7934 I32 namedclass = OOB_NAMEDCLASS;
7936 PERL_ARGS_ASSERT_REGPPOSIXCC;
7938 if (value == '[' && RExC_parse + 1 < RExC_end &&
7939 /* I smell either [: or [= or [. -- POSIX has been here, right? */
7940 POSIXCC(UCHARAT(RExC_parse))) {
7941 const char c = UCHARAT(RExC_parse);
7942 char* const s = RExC_parse++;
7944 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != c)
7946 if (RExC_parse == RExC_end)
7947 /* Grandfather lone [:, [=, [. */
7950 const char* const t = RExC_parse++; /* skip over the c */
7953 if (UCHARAT(RExC_parse) == ']') {
7954 const char *posixcc = s + 1;
7955 RExC_parse++; /* skip over the ending ] */
7958 const I32 complement = *posixcc == '^' ? *posixcc++ : 0;
7959 const I32 skip = t - posixcc;
7961 /* Initially switch on the length of the name. */
7964 if (memEQ(posixcc, "word", 4)) /* this is not POSIX, this is the Perl \w */
7965 namedclass = complement ? ANYOF_NALNUM : ANYOF_ALNUM;
7968 /* Names all of length 5. */
7969 /* alnum alpha ascii blank cntrl digit graph lower
7970 print punct space upper */
7971 /* Offset 4 gives the best switch position. */
7972 switch (posixcc[4]) {
7974 if (memEQ(posixcc, "alph", 4)) /* alpha */
7975 namedclass = complement ? ANYOF_NALPHA : ANYOF_ALPHA;
7978 if (memEQ(posixcc, "spac", 4)) /* space */
7979 namedclass = complement ? ANYOF_NPSXSPC : ANYOF_PSXSPC;
7982 if (memEQ(posixcc, "grap", 4)) /* graph */
7983 namedclass = complement ? ANYOF_NGRAPH : ANYOF_GRAPH;
7986 if (memEQ(posixcc, "asci", 4)) /* ascii */
7987 namedclass = complement ? ANYOF_NASCII : ANYOF_ASCII;
7990 if (memEQ(posixcc, "blan", 4)) /* blank */
7991 namedclass = complement ? ANYOF_NBLANK : ANYOF_BLANK;
7994 if (memEQ(posixcc, "cntr", 4)) /* cntrl */
7995 namedclass = complement ? ANYOF_NCNTRL : ANYOF_CNTRL;
7998 if (memEQ(posixcc, "alnu", 4)) /* alnum */
7999 namedclass = complement ? ANYOF_NALNUMC : ANYOF_ALNUMC;
8002 if (memEQ(posixcc, "lowe", 4)) /* lower */
8003 namedclass = complement ? ANYOF_NLOWER : ANYOF_LOWER;
8004 else if (memEQ(posixcc, "uppe", 4)) /* upper */
8005 namedclass = complement ? ANYOF_NUPPER : ANYOF_UPPER;
8008 if (memEQ(posixcc, "digi", 4)) /* digit */
8009 namedclass = complement ? ANYOF_NDIGIT : ANYOF_DIGIT;
8010 else if (memEQ(posixcc, "prin", 4)) /* print */
8011 namedclass = complement ? ANYOF_NPRINT : ANYOF_PRINT;
8012 else if (memEQ(posixcc, "punc", 4)) /* punct */
8013 namedclass = complement ? ANYOF_NPUNCT : ANYOF_PUNCT;
8018 if (memEQ(posixcc, "xdigit", 6))
8019 namedclass = complement ? ANYOF_NXDIGIT : ANYOF_XDIGIT;
8023 if (namedclass == OOB_NAMEDCLASS)
8024 Simple_vFAIL3("POSIX class [:%.*s:] unknown",
8026 assert (posixcc[skip] == ':');
8027 assert (posixcc[skip+1] == ']');
8028 } else if (!SIZE_ONLY) {
8029 /* [[=foo=]] and [[.foo.]] are still future. */
8031 /* adjust RExC_parse so the warning shows after
8033 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse) != ']')
8035 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
8038 /* Maternal grandfather:
8039 * "[:" ending in ":" but not in ":]" */
8049 S_checkposixcc(pTHX_ RExC_state_t *pRExC_state)
8053 PERL_ARGS_ASSERT_CHECKPOSIXCC;
8055 if (POSIXCC(UCHARAT(RExC_parse))) {
8056 const char *s = RExC_parse;
8057 const char c = *s++;
8061 if (*s && c == *s && s[1] == ']') {
8063 "POSIX syntax [%c %c] belongs inside character classes",
8066 /* [[=foo=]] and [[.foo.]] are still future. */
8067 if (POSIXCC_NOTYET(c)) {
8068 /* adjust RExC_parse so the error shows after
8070 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse++) != ']')
8072 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
8078 /* No locale test */
8079 #define _C_C_T_NOLOC_(NAME,TEST,WORD) \
8081 for (value = 0; value < 256; value++) \
8083 ANYOF_BITMAP_SET(ret, value); \
8087 case ANYOF_N##NAME: \
8088 for (value = 0; value < 256; value++) \
8090 ANYOF_BITMAP_SET(ret, value); \
8095 /* Like the above, but there are differences if we are in uni-8-bit or not, so
8096 * there are two tests passed in, to use depending on that. There aren't any
8097 * cases where the label is different from the name, so no need for that
8099 #define _C_C_T_(NAME,TEST_8,TEST_7,WORD) \
8101 if (LOC) ANYOF_CLASS_SET(ret, ANYOF_##NAME); \
8102 else if (UNI_SEMANTICS) { \
8103 for (value = 0; value < 256; value++) { \
8104 if (TEST_8) ANYOF_BITMAP_SET(ret, value); \
8108 for (value = 0; value < 256; value++) { \
8109 if (TEST_7) ANYOF_BITMAP_SET(ret, value); \
8115 case ANYOF_N##NAME: \
8116 if (LOC) ANYOF_CLASS_SET(ret, ANYOF_N##NAME); \
8117 else if (UNI_SEMANTICS) { \
8118 for (value = 0; value < 256; value++) { \
8119 if (! TEST_8) ANYOF_BITMAP_SET(ret, value); \
8123 for (value = 0; value < 256; value++) { \
8124 if (! TEST_7) ANYOF_BITMAP_SET(ret, value); \
8132 We dont use PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS as the direct test
8133 so that it is possible to override the option here without having to
8134 rebuild the entire core. as we are required to do if we change regcomp.h
8135 which is where PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS is defined.
8137 #if PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS
8138 #define BROKEN_UNICODE_CHARCLASS_MAPPINGS
8141 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
8142 #define POSIX_CC_UNI_NAME(CCNAME) CCNAME
8144 #define POSIX_CC_UNI_NAME(CCNAME) "Posix" CCNAME
8148 parse a class specification and produce either an ANYOF node that
8149 matches the pattern or if the pattern matches a single char only and
8150 that char is < 256 and we are case insensitive then we produce an
8155 S_regclass(pTHX_ RExC_state_t *pRExC_state, U32 depth)
8158 register UV nextvalue;
8159 register IV prevvalue = OOB_UNICODE;
8160 register IV range = 0;
8161 UV value = 0; /* XXX:dmq: needs to be referenceable (unfortunately) */
8162 register regnode *ret;
8165 char *rangebegin = NULL;
8166 bool need_class = 0;
8169 bool optimize_invert = TRUE;
8170 AV* unicode_alternate = NULL;
8172 UV literal_endpoint = 0;
8174 UV stored = 0; /* 0, 1, or more than 1 chars stored in the class */
8176 regnode * const orig_emit = RExC_emit; /* Save the original RExC_emit in
8177 case we need to change the emitted regop to an EXACT. */
8178 const char * orig_parse = RExC_parse;
8179 GET_RE_DEBUG_FLAGS_DECL;
8181 PERL_ARGS_ASSERT_REGCLASS;
8183 PERL_UNUSED_ARG(depth);
8186 DEBUG_PARSE("clas");
8188 /* Assume we are going to generate an ANYOF node. */
8189 ret = reganode(pRExC_state, ANYOF, 0);
8192 ANYOF_FLAGS(ret) = 0;
8194 if (UCHARAT(RExC_parse) == '^') { /* Complement of range. */
8198 ANYOF_FLAGS(ret) |= ANYOF_INVERT;
8202 RExC_size += ANYOF_SKIP;
8203 listsv = &PL_sv_undef; /* For code scanners: listsv always non-NULL. */
8206 RExC_emit += ANYOF_SKIP;
8208 ANYOF_FLAGS(ret) |= ANYOF_FOLD;
8210 ANYOF_FLAGS(ret) |= ANYOF_LOCALE;
8211 ANYOF_BITMAP_ZERO(ret);
8212 listsv = newSVpvs("# comment\n");
8215 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
8217 if (!SIZE_ONLY && POSIXCC(nextvalue))
8218 checkposixcc(pRExC_state);
8220 /* allow 1st char to be ] (allowing it to be - is dealt with later) */
8221 if (UCHARAT(RExC_parse) == ']')
8225 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != ']') {
8229 namedclass = OOB_NAMEDCLASS; /* initialize as illegal */
8232 rangebegin = RExC_parse;
8234 value = utf8n_to_uvchr((U8*)RExC_parse,
8235 RExC_end - RExC_parse,
8236 &numlen, UTF8_ALLOW_DEFAULT);
8237 RExC_parse += numlen;
8240 value = UCHARAT(RExC_parse++);
8242 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
8243 if (value == '[' && POSIXCC(nextvalue))
8244 namedclass = regpposixcc(pRExC_state, value);
8245 else if (value == '\\') {
8247 value = utf8n_to_uvchr((U8*)RExC_parse,
8248 RExC_end - RExC_parse,
8249 &numlen, UTF8_ALLOW_DEFAULT);
8250 RExC_parse += numlen;
8253 value = UCHARAT(RExC_parse++);
8254 /* Some compilers cannot handle switching on 64-bit integer
8255 * values, therefore value cannot be an UV. Yes, this will
8256 * be a problem later if we want switch on Unicode.
8257 * A similar issue a little bit later when switching on
8258 * namedclass. --jhi */
8259 switch ((I32)value) {
8260 case 'w': namedclass = ANYOF_ALNUM; break;
8261 case 'W': namedclass = ANYOF_NALNUM; break;
8262 case 's': namedclass = ANYOF_SPACE; break;
8263 case 'S': namedclass = ANYOF_NSPACE; break;
8264 case 'd': namedclass = ANYOF_DIGIT; break;
8265 case 'D': namedclass = ANYOF_NDIGIT; break;
8266 case 'v': namedclass = ANYOF_VERTWS; break;
8267 case 'V': namedclass = ANYOF_NVERTWS; break;
8268 case 'h': namedclass = ANYOF_HORIZWS; break;
8269 case 'H': namedclass = ANYOF_NHORIZWS; break;
8270 case 'N': /* Handle \N{NAME} in class */
8272 /* We only pay attention to the first char of
8273 multichar strings being returned. I kinda wonder
8274 if this makes sense as it does change the behaviour
8275 from earlier versions, OTOH that behaviour was broken
8277 UV v; /* value is register so we cant & it /grrr */
8278 if (reg_namedseq(pRExC_state, &v, NULL)) {
8288 if (RExC_parse >= RExC_end)
8289 vFAIL2("Empty \\%c{}", (U8)value);
8290 if (*RExC_parse == '{') {
8291 const U8 c = (U8)value;
8292 e = strchr(RExC_parse++, '}');
8294 vFAIL2("Missing right brace on \\%c{}", c);
8295 while (isSPACE(UCHARAT(RExC_parse)))
8297 if (e == RExC_parse)
8298 vFAIL2("Empty \\%c{}", c);
8300 while (isSPACE(UCHARAT(RExC_parse + n - 1)))
8308 if (UCHARAT(RExC_parse) == '^') {
8311 value = value == 'p' ? 'P' : 'p'; /* toggle */
8312 while (isSPACE(UCHARAT(RExC_parse))) {
8317 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::%.*s\n",
8318 (value=='p' ? '+' : '!'), (int)n, RExC_parse);
8321 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8322 namedclass = ANYOF_MAX; /* no official name, but it's named */
8325 case 'n': value = '\n'; break;
8326 case 'r': value = '\r'; break;
8327 case 't': value = '\t'; break;
8328 case 'f': value = '\f'; break;
8329 case 'b': value = '\b'; break;
8330 case 'e': value = ASCII_TO_NATIVE('\033');break;
8331 case 'a': value = ASCII_TO_NATIVE('\007');break;
8333 RExC_parse--; /* function expects to be pointed at the 'o' */
8335 const char* error_msg;
8336 bool valid = grok_bslash_o(RExC_parse,
8341 RExC_parse += numlen;
8346 if (PL_encoding && value < 0x100) {
8347 goto recode_encoding;
8351 if (*RExC_parse == '{') {
8352 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
8353 | PERL_SCAN_DISALLOW_PREFIX;
8354 char * const e = strchr(RExC_parse++, '}');
8356 vFAIL("Missing right brace on \\x{}");
8358 numlen = e - RExC_parse;
8359 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8363 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
8365 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8366 RExC_parse += numlen;
8368 if (PL_encoding && value < 0x100)
8369 goto recode_encoding;
8372 value = grok_bslash_c(*RExC_parse++, SIZE_ONLY);
8374 case '0': case '1': case '2': case '3': case '4':
8375 case '5': case '6': case '7':
8377 /* Take 1-3 octal digits */
8378 I32 flags = PERL_SCAN_SILENT_ILLDIGIT;
8380 value = grok_oct(--RExC_parse, &numlen, &flags, NULL);
8381 RExC_parse += numlen;
8382 if (PL_encoding && value < 0x100)
8383 goto recode_encoding;
8388 SV* enc = PL_encoding;
8389 value = reg_recode((const char)(U8)value, &enc);
8390 if (!enc && SIZE_ONLY)
8391 ckWARNreg(RExC_parse,
8392 "Invalid escape in the specified encoding");
8396 /* Allow \_ to not give an error */
8397 if (!SIZE_ONLY && isALNUM(value) && value != '_') {
8398 ckWARN2reg(RExC_parse,
8399 "Unrecognized escape \\%c in character class passed through",
8404 } /* end of \blah */
8410 if (namedclass > OOB_NAMEDCLASS) { /* this is a named class \blah */
8412 /* What matches in a locale is not known until runtime, so need to
8413 * (one time per class) allocate extra space to pass to regexec.
8414 * The space will contain a bit for each named class that is to be
8415 * matched against. This isn't needed for \p{} and pseudo-classes,
8416 * as they are not affected by locale, and hence are dealt with
8418 if (LOC && namedclass < ANYOF_MAX && ! need_class) {
8421 RExC_size += ANYOF_CLASS_ADD_SKIP;
8424 RExC_emit += ANYOF_CLASS_ADD_SKIP;
8425 ANYOF_CLASS_ZERO(ret);
8427 ANYOF_FLAGS(ret) |= ANYOF_CLASS|ANYOF_LARGE;
8430 /* a bad range like a-\d, a-[:digit:] ? */
8434 RExC_parse >= rangebegin ?
8435 RExC_parse - rangebegin : 0;
8436 ckWARN4reg(RExC_parse,
8437 "False [] range \"%*.*s\"",
8440 if (prevvalue < 256) {
8441 ANYOF_BITMAP_SET(ret, prevvalue);
8442 ANYOF_BITMAP_SET(ret, '-');
8445 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8446 Perl_sv_catpvf(aTHX_ listsv,
8447 "%04"UVxf"\n%04"UVxf"\n", (UV)prevvalue, (UV) '-');
8451 range = 0; /* this was not a true range */
8457 const char *what = NULL;
8460 if (namedclass > OOB_NAMEDCLASS)
8461 optimize_invert = FALSE;
8462 /* Possible truncation here but in some 64-bit environments
8463 * the compiler gets heartburn about switch on 64-bit values.
8464 * A similar issue a little earlier when switching on value.
8466 switch ((I32)namedclass) {
8468 case _C_C_T_(ALNUMC, isALNUMC_L1(value), isALNUMC(value), "XPosixAlnum");
8469 case _C_C_T_(ALPHA, isALPHA_L1(value), isALPHA(value), "XPosixAlpha");
8470 case _C_C_T_(BLANK, isBLANK_L1(value), isBLANK(value), "XPosixBlank");
8471 case _C_C_T_(CNTRL, isCNTRL_L1(value), isCNTRL(value), "XPosixCntrl");
8472 case _C_C_T_(GRAPH, isGRAPH_L1(value), isGRAPH(value), "XPosixGraph");
8473 case _C_C_T_(LOWER, isLOWER_L1(value), isLOWER(value), "XPosixLower");
8474 case _C_C_T_(PRINT, isPRINT_L1(value), isPRINT(value), "XPosixPrint");
8475 case _C_C_T_(PSXSPC, isPSXSPC_L1(value), isPSXSPC(value), "XPosixSpace");
8476 case _C_C_T_(PUNCT, isPUNCT_L1(value), isPUNCT(value), "XPosixPunct");
8477 case _C_C_T_(UPPER, isUPPER_L1(value), isUPPER(value), "XPosixUpper");
8478 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
8479 /* \s, \w match all unicode if utf8. */
8480 case _C_C_T_(SPACE, isSPACE_L1(value), isSPACE(value), "SpacePerl");
8481 case _C_C_T_(ALNUM, isWORDCHAR_L1(value), isALNUM(value), "Word");
8483 /* \s, \w match ascii and locale only */
8484 case _C_C_T_(SPACE, isSPACE_L1(value), isSPACE(value), "PerlSpace");
8485 case _C_C_T_(ALNUM, isWORDCHAR_L1(value), isALNUM(value), "PerlWord");
8487 case _C_C_T_(XDIGIT, isXDIGIT_L1(value), isXDIGIT(value), "XPosixXDigit");
8488 case _C_C_T_NOLOC_(VERTWS, is_VERTWS_latin1(&value), "VertSpace");
8489 case _C_C_T_NOLOC_(HORIZWS, is_HORIZWS_latin1(&value), "HorizSpace");
8492 ANYOF_CLASS_SET(ret, ANYOF_ASCII);
8495 for (value = 0; value < 128; value++)
8496 ANYOF_BITMAP_SET(ret, value);
8498 for (value = 0; value < 256; value++) {
8500 ANYOF_BITMAP_SET(ret, value);
8509 ANYOF_CLASS_SET(ret, ANYOF_NASCII);
8512 for (value = 128; value < 256; value++)
8513 ANYOF_BITMAP_SET(ret, value);
8515 for (value = 0; value < 256; value++) {
8516 if (!isASCII(value))
8517 ANYOF_BITMAP_SET(ret, value);
8526 ANYOF_CLASS_SET(ret, ANYOF_DIGIT);
8528 /* consecutive digits assumed */
8529 for (value = '0'; value <= '9'; value++)
8530 ANYOF_BITMAP_SET(ret, value);
8533 what = POSIX_CC_UNI_NAME("Digit");
8537 ANYOF_CLASS_SET(ret, ANYOF_NDIGIT);
8539 /* consecutive digits assumed */
8540 for (value = 0; value < '0'; value++)
8541 ANYOF_BITMAP_SET(ret, value);
8542 for (value = '9' + 1; value < 256; value++)
8543 ANYOF_BITMAP_SET(ret, value);
8546 what = POSIX_CC_UNI_NAME("Digit");
8549 /* this is to handle \p and \P */
8552 vFAIL("Invalid [::] class");
8556 /* Strings such as "+utf8::isWord\n" */
8557 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::Is%s\n", yesno, what);
8559 stored+=2; /* can't optimize this class */
8562 } /* end of namedclass \blah */
8565 if (prevvalue > (IV)value) /* b-a */ {
8566 const int w = RExC_parse - rangebegin;
8567 Simple_vFAIL4("Invalid [] range \"%*.*s\"", w, w, rangebegin);
8568 range = 0; /* not a valid range */
8572 prevvalue = value; /* save the beginning of the range */
8573 if (*RExC_parse == '-' && RExC_parse+1 < RExC_end &&
8574 RExC_parse[1] != ']') {
8577 /* a bad range like \w-, [:word:]- ? */
8578 if (namedclass > OOB_NAMEDCLASS) {
8579 if (ckWARN(WARN_REGEXP)) {
8581 RExC_parse >= rangebegin ?
8582 RExC_parse - rangebegin : 0;
8584 "False [] range \"%*.*s\"",
8588 ANYOF_BITMAP_SET(ret, '-');
8590 range = 1; /* yeah, it's a range! */
8591 continue; /* but do it the next time */
8595 /* now is the next time */
8596 /*stored += (value - prevvalue + 1);*/
8598 if (prevvalue < 256) {
8599 const IV ceilvalue = value < 256 ? value : 255;
8602 /* In EBCDIC [\x89-\x91] should include
8603 * the \x8e but [i-j] should not. */
8604 if (literal_endpoint == 2 &&
8605 ((isLOWER(prevvalue) && isLOWER(ceilvalue)) ||
8606 (isUPPER(prevvalue) && isUPPER(ceilvalue))))
8608 if (isLOWER(prevvalue)) {
8609 for (i = prevvalue; i <= ceilvalue; i++)
8610 if (isLOWER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8612 ANYOF_BITMAP_SET(ret, i);
8615 for (i = prevvalue; i <= ceilvalue; i++)
8616 if (isUPPER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8618 ANYOF_BITMAP_SET(ret, i);
8624 for (i = prevvalue; i <= ceilvalue; i++) {
8625 if (!ANYOF_BITMAP_TEST(ret,i)) {
8627 ANYOF_BITMAP_SET(ret, i);
8631 if (value > 255 || UTF) {
8632 const UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
8633 const UV natvalue = NATIVE_TO_UNI(value);
8634 stored+=2; /* can't optimize this class */
8635 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8636 if (prevnatvalue < natvalue) { /* what about > ? */
8637 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\t%04"UVxf"\n",
8638 prevnatvalue, natvalue);
8640 else if (prevnatvalue == natvalue) {
8641 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n", natvalue);
8643 U8 foldbuf[UTF8_MAXBYTES_CASE+1];
8645 const UV f = to_uni_fold(natvalue, foldbuf, &foldlen);
8647 #ifdef EBCDIC /* RD t/uni/fold ff and 6b */
8648 if (RExC_precomp[0] == ':' &&
8649 RExC_precomp[1] == '[' &&
8650 (f == 0xDF || f == 0x92)) {
8651 f = NATIVE_TO_UNI(f);
8654 /* If folding and foldable and a single
8655 * character, insert also the folded version
8656 * to the charclass. */
8658 #ifdef EBCDIC /* RD tunifold ligatures s,t fb05, fb06 */
8659 if ((RExC_precomp[0] == ':' &&
8660 RExC_precomp[1] == '[' &&
8662 (value == 0xFB05 || value == 0xFB06))) ?
8663 foldlen == ((STRLEN)UNISKIP(f) - 1) :
8664 foldlen == (STRLEN)UNISKIP(f) )
8666 if (foldlen == (STRLEN)UNISKIP(f))
8668 Perl_sv_catpvf(aTHX_ listsv,
8671 /* Any multicharacter foldings
8672 * require the following transform:
8673 * [ABCDEF] -> (?:[ABCabcDEFd]|pq|rst)
8674 * where E folds into "pq" and F folds
8675 * into "rst", all other characters
8676 * fold to single characters. We save
8677 * away these multicharacter foldings,
8678 * to be later saved as part of the
8679 * additional "s" data. */
8682 if (!unicode_alternate)
8683 unicode_alternate = newAV();
8684 sv = newSVpvn_utf8((char*)foldbuf, foldlen,
8686 av_push(unicode_alternate, sv);
8690 /* If folding and the value is one of the Greek
8691 * sigmas insert a few more sigmas to make the
8692 * folding rules of the sigmas to work right.
8693 * Note that not all the possible combinations
8694 * are handled here: some of them are handled
8695 * by the standard folding rules, and some of
8696 * them (literal or EXACTF cases) are handled
8697 * during runtime in regexec.c:S_find_byclass(). */
8698 if (value == UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA) {
8699 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8700 (UV)UNICODE_GREEK_CAPITAL_LETTER_SIGMA);
8701 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8702 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8704 else if (value == UNICODE_GREEK_CAPITAL_LETTER_SIGMA)
8705 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8706 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8711 literal_endpoint = 0;
8715 range = 0; /* this range (if it was one) is done now */
8722 /****** !SIZE_ONLY AFTER HERE *********/
8724 if( stored == 1 && (value < 128 || (value < 256 && !UTF))
8725 && !( ANYOF_FLAGS(ret) & ( ANYOF_FLAGS_ALL ^ ANYOF_FOLD ) )
8727 /* optimize single char class to an EXACT node
8728 but *only* when its not a UTF/high char */
8729 const char * cur_parse= RExC_parse;
8730 RExC_emit = (regnode *)orig_emit;
8731 RExC_parse = (char *)orig_parse;
8732 ret = reg_node(pRExC_state,
8733 (U8)((ANYOF_FLAGS(ret) & ANYOF_FOLD) ? EXACTF : EXACT));
8734 RExC_parse = (char *)cur_parse;
8735 *STRING(ret)= (char)value;
8737 RExC_emit += STR_SZ(1);
8738 SvREFCNT_dec(listsv);
8741 /* optimize case-insensitive simple patterns (e.g. /[a-z]/i) */
8742 if ( /* If the only flag is folding (plus possibly inversion). */
8743 ((ANYOF_FLAGS(ret) & (ANYOF_FLAGS_ALL ^ ANYOF_INVERT)) == ANYOF_FOLD)
8745 for (value = 0; value < 256; ++value) {
8746 if (ANYOF_BITMAP_TEST(ret, value)) {
8747 UV fold = PL_fold[value];
8750 ANYOF_BITMAP_SET(ret, fold);
8753 ANYOF_FLAGS(ret) &= ~ANYOF_FOLD;
8756 /* optimize inverted simple patterns (e.g. [^a-z]) */
8757 if (optimize_invert &&
8758 /* If the only flag is inversion. */
8759 (ANYOF_FLAGS(ret) & ANYOF_FLAGS_ALL) == ANYOF_INVERT) {
8760 for (value = 0; value < ANYOF_BITMAP_SIZE; ++value)
8761 ANYOF_BITMAP(ret)[value] ^= ANYOF_FLAGS_ALL;
8762 ANYOF_FLAGS(ret) = ANYOF_UNICODE_ALL;
8765 AV * const av = newAV();
8767 /* The 0th element stores the character class description
8768 * in its textual form: used later (regexec.c:Perl_regclass_swash())
8769 * to initialize the appropriate swash (which gets stored in
8770 * the 1st element), and also useful for dumping the regnode.
8771 * The 2nd element stores the multicharacter foldings,
8772 * used later (regexec.c:S_reginclass()). */
8773 av_store(av, 0, listsv);
8774 av_store(av, 1, NULL);
8775 av_store(av, 2, MUTABLE_SV(unicode_alternate));
8776 rv = newRV_noinc(MUTABLE_SV(av));
8777 n = add_data(pRExC_state, 1, "s");
8778 RExC_rxi->data->data[n] = (void*)rv;
8786 /* reg_skipcomment()
8788 Absorbs an /x style # comments from the input stream.
8789 Returns true if there is more text remaining in the stream.
8790 Will set the REG_SEEN_RUN_ON_COMMENT flag if the comment
8791 terminates the pattern without including a newline.
8793 Note its the callers responsibility to ensure that we are
8799 S_reg_skipcomment(pTHX_ RExC_state_t *pRExC_state)
8803 PERL_ARGS_ASSERT_REG_SKIPCOMMENT;
8805 while (RExC_parse < RExC_end)
8806 if (*RExC_parse++ == '\n') {
8811 /* we ran off the end of the pattern without ending
8812 the comment, so we have to add an \n when wrapping */
8813 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
8821 Advances the parse position, and optionally absorbs
8822 "whitespace" from the inputstream.
8824 Without /x "whitespace" means (?#...) style comments only,
8825 with /x this means (?#...) and # comments and whitespace proper.
8827 Returns the RExC_parse point from BEFORE the scan occurs.
8829 This is the /x friendly way of saying RExC_parse++.
8833 S_nextchar(pTHX_ RExC_state_t *pRExC_state)
8835 char* const retval = RExC_parse++;
8837 PERL_ARGS_ASSERT_NEXTCHAR;
8840 if (*RExC_parse == '(' && RExC_parse[1] == '?' &&
8841 RExC_parse[2] == '#') {
8842 while (*RExC_parse != ')') {
8843 if (RExC_parse == RExC_end)
8844 FAIL("Sequence (?#... not terminated");
8850 if (RExC_flags & RXf_PMf_EXTENDED) {
8851 if (isSPACE(*RExC_parse)) {
8855 else if (*RExC_parse == '#') {
8856 if ( reg_skipcomment( pRExC_state ) )
8865 - reg_node - emit a node
8867 STATIC regnode * /* Location. */
8868 S_reg_node(pTHX_ RExC_state_t *pRExC_state, U8 op)
8871 register regnode *ptr;
8872 regnode * const ret = RExC_emit;
8873 GET_RE_DEBUG_FLAGS_DECL;
8875 PERL_ARGS_ASSERT_REG_NODE;
8878 SIZE_ALIGN(RExC_size);
8882 if (RExC_emit >= RExC_emit_bound)
8883 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8885 NODE_ALIGN_FILL(ret);
8887 FILL_ADVANCE_NODE(ptr, op);
8888 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (ptr) - 1);
8889 #ifdef RE_TRACK_PATTERN_OFFSETS
8890 if (RExC_offsets) { /* MJD */
8891 MJD_OFFSET_DEBUG(("%s:%d: (op %s) %s %"UVuf" (len %"UVuf") (max %"UVuf").\n",
8892 "reg_node", __LINE__,
8894 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0]
8895 ? "Overwriting end of array!\n" : "OK",
8896 (UV)(RExC_emit - RExC_emit_start),
8897 (UV)(RExC_parse - RExC_start),
8898 (UV)RExC_offsets[0]));
8899 Set_Node_Offset(RExC_emit, RExC_parse + (op == END));
8907 - reganode - emit a node with an argument
8909 STATIC regnode * /* Location. */
8910 S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
8913 register regnode *ptr;
8914 regnode * const ret = RExC_emit;
8915 GET_RE_DEBUG_FLAGS_DECL;
8917 PERL_ARGS_ASSERT_REGANODE;
8920 SIZE_ALIGN(RExC_size);
8925 assert(2==regarglen[op]+1);
8927 Anything larger than this has to allocate the extra amount.
8928 If we changed this to be:
8930 RExC_size += (1 + regarglen[op]);
8932 then it wouldn't matter. Its not clear what side effect
8933 might come from that so its not done so far.
8938 if (RExC_emit >= RExC_emit_bound)
8939 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8941 NODE_ALIGN_FILL(ret);
8943 FILL_ADVANCE_NODE_ARG(ptr, op, arg);
8944 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (ptr) - 2);
8945 #ifdef RE_TRACK_PATTERN_OFFSETS
8946 if (RExC_offsets) { /* MJD */
8947 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8951 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0] ?
8952 "Overwriting end of array!\n" : "OK",
8953 (UV)(RExC_emit - RExC_emit_start),
8954 (UV)(RExC_parse - RExC_start),
8955 (UV)RExC_offsets[0]));
8956 Set_Cur_Node_Offset;
8964 - reguni - emit (if appropriate) a Unicode character
8967 S_reguni(pTHX_ const RExC_state_t *pRExC_state, UV uv, char* s)
8971 PERL_ARGS_ASSERT_REGUNI;
8973 return SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
8977 - reginsert - insert an operator in front of already-emitted operand
8979 * Means relocating the operand.
8982 S_reginsert(pTHX_ RExC_state_t *pRExC_state, U8 op, regnode *opnd, U32 depth)
8985 register regnode *src;
8986 register regnode *dst;
8987 register regnode *place;
8988 const int offset = regarglen[(U8)op];
8989 const int size = NODE_STEP_REGNODE + offset;
8990 GET_RE_DEBUG_FLAGS_DECL;
8992 PERL_ARGS_ASSERT_REGINSERT;
8993 PERL_UNUSED_ARG(depth);
8994 /* (PL_regkind[(U8)op] == CURLY ? EXTRA_STEP_2ARGS : 0); */
8995 DEBUG_PARSE_FMT("inst"," - %s",PL_reg_name[op]);
9004 if (RExC_open_parens) {
9006 /*DEBUG_PARSE_FMT("inst"," - %"IVdf, (IV)RExC_npar);*/
9007 for ( paren=0 ; paren < RExC_npar ; paren++ ) {
9008 if ( RExC_open_parens[paren] >= opnd ) {
9009 /*DEBUG_PARSE_FMT("open"," - %d",size);*/
9010 RExC_open_parens[paren] += size;
9012 /*DEBUG_PARSE_FMT("open"," - %s","ok");*/
9014 if ( RExC_close_parens[paren] >= opnd ) {
9015 /*DEBUG_PARSE_FMT("close"," - %d",size);*/
9016 RExC_close_parens[paren] += size;
9018 /*DEBUG_PARSE_FMT("close"," - %s","ok");*/
9023 while (src > opnd) {
9024 StructCopy(--src, --dst, regnode);
9025 #ifdef RE_TRACK_PATTERN_OFFSETS
9026 if (RExC_offsets) { /* MJD 20010112 */
9027 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s copy %"UVuf" -> %"UVuf" (max %"UVuf").\n",
9031 (UV)(dst - RExC_emit_start) > RExC_offsets[0]
9032 ? "Overwriting end of array!\n" : "OK",
9033 (UV)(src - RExC_emit_start),
9034 (UV)(dst - RExC_emit_start),
9035 (UV)RExC_offsets[0]));
9036 Set_Node_Offset_To_R(dst-RExC_emit_start, Node_Offset(src));
9037 Set_Node_Length_To_R(dst-RExC_emit_start, Node_Length(src));
9043 place = opnd; /* Op node, where operand used to be. */
9044 #ifdef RE_TRACK_PATTERN_OFFSETS
9045 if (RExC_offsets) { /* MJD */
9046 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
9050 (UV)(place - RExC_emit_start) > RExC_offsets[0]
9051 ? "Overwriting end of array!\n" : "OK",
9052 (UV)(place - RExC_emit_start),
9053 (UV)(RExC_parse - RExC_start),
9054 (UV)RExC_offsets[0]));
9055 Set_Node_Offset(place, RExC_parse);
9056 Set_Node_Length(place, 1);
9059 src = NEXTOPER(place);
9060 FILL_ADVANCE_NODE(place, op);
9061 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (place) - 1);
9062 Zero(src, offset, regnode);
9066 - regtail - set the next-pointer at the end of a node chain of p to val.
9067 - SEE ALSO: regtail_study
9069 /* TODO: All three parms should be const */
9071 S_regtail(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
9074 register regnode *scan;
9075 GET_RE_DEBUG_FLAGS_DECL;
9077 PERL_ARGS_ASSERT_REGTAIL;
9079 PERL_UNUSED_ARG(depth);
9085 /* Find last node. */
9088 regnode * const temp = regnext(scan);
9090 SV * const mysv=sv_newmortal();
9091 DEBUG_PARSE_MSG((scan==p ? "tail" : ""));
9092 regprop(RExC_rx, mysv, scan);
9093 PerlIO_printf(Perl_debug_log, "~ %s (%d) %s %s\n",
9094 SvPV_nolen_const(mysv), REG_NODE_NUM(scan),
9095 (temp == NULL ? "->" : ""),
9096 (temp == NULL ? PL_reg_name[OP(val)] : "")
9104 if (reg_off_by_arg[OP(scan)]) {
9105 ARG_SET(scan, val - scan);
9108 NEXT_OFF(scan) = val - scan;
9114 - regtail_study - set the next-pointer at the end of a node chain of p to val.
9115 - Look for optimizable sequences at the same time.
9116 - currently only looks for EXACT chains.
9118 This is expermental code. The idea is to use this routine to perform
9119 in place optimizations on branches and groups as they are constructed,
9120 with the long term intention of removing optimization from study_chunk so
9121 that it is purely analytical.
9123 Currently only used when in DEBUG mode. The macro REGTAIL_STUDY() is used
9124 to control which is which.
9127 /* TODO: All four parms should be const */
9130 S_regtail_study(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
9133 register regnode *scan;
9135 #ifdef EXPERIMENTAL_INPLACESCAN
9138 GET_RE_DEBUG_FLAGS_DECL;
9140 PERL_ARGS_ASSERT_REGTAIL_STUDY;
9146 /* Find last node. */
9150 regnode * const temp = regnext(scan);
9151 #ifdef EXPERIMENTAL_INPLACESCAN
9152 if (PL_regkind[OP(scan)] == EXACT)
9153 if (join_exact(pRExC_state,scan,&min,1,val,depth+1))
9161 if( exact == PSEUDO )
9163 else if ( exact != OP(scan) )
9172 SV * const mysv=sv_newmortal();
9173 DEBUG_PARSE_MSG((scan==p ? "tsdy" : ""));
9174 regprop(RExC_rx, mysv, scan);
9175 PerlIO_printf(Perl_debug_log, "~ %s (%d) -> %s\n",
9176 SvPV_nolen_const(mysv),
9178 PL_reg_name[exact]);
9185 SV * const mysv_val=sv_newmortal();
9186 DEBUG_PARSE_MSG("");
9187 regprop(RExC_rx, mysv_val, val);
9188 PerlIO_printf(Perl_debug_log, "~ attach to %s (%"IVdf") offset to %"IVdf"\n",
9189 SvPV_nolen_const(mysv_val),
9190 (IV)REG_NODE_NUM(val),
9194 if (reg_off_by_arg[OP(scan)]) {
9195 ARG_SET(scan, val - scan);
9198 NEXT_OFF(scan) = val - scan;
9206 - regdump - dump a regexp onto Perl_debug_log in vaguely comprehensible form
9210 S_regdump_extflags(pTHX_ const char *lead, const U32 flags)
9215 for (bit=0; bit<32; bit++) {
9216 if (flags & (1<<bit)) {
9218 PerlIO_printf(Perl_debug_log, "%s",lead);
9219 PerlIO_printf(Perl_debug_log, "%s ",PL_reg_extflags_name[bit]);
9224 PerlIO_printf(Perl_debug_log, "\n");
9226 PerlIO_printf(Perl_debug_log, "%s[none-set]\n",lead);
9232 Perl_regdump(pTHX_ const regexp *r)
9236 SV * const sv = sv_newmortal();
9237 SV *dsv= sv_newmortal();
9239 GET_RE_DEBUG_FLAGS_DECL;
9241 PERL_ARGS_ASSERT_REGDUMP;
9243 (void)dumpuntil(r, ri->program, ri->program + 1, NULL, NULL, sv, 0, 0);
9245 /* Header fields of interest. */
9246 if (r->anchored_substr) {
9247 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->anchored_substr),
9248 RE_SV_DUMPLEN(r->anchored_substr), 30);
9249 PerlIO_printf(Perl_debug_log,
9250 "anchored %s%s at %"IVdf" ",
9251 s, RE_SV_TAIL(r->anchored_substr),
9252 (IV)r->anchored_offset);
9253 } else if (r->anchored_utf8) {
9254 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->anchored_utf8),
9255 RE_SV_DUMPLEN(r->anchored_utf8), 30);
9256 PerlIO_printf(Perl_debug_log,
9257 "anchored utf8 %s%s at %"IVdf" ",
9258 s, RE_SV_TAIL(r->anchored_utf8),
9259 (IV)r->anchored_offset);
9261 if (r->float_substr) {
9262 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->float_substr),
9263 RE_SV_DUMPLEN(r->float_substr), 30);
9264 PerlIO_printf(Perl_debug_log,
9265 "floating %s%s at %"IVdf"..%"UVuf" ",
9266 s, RE_SV_TAIL(r->float_substr),
9267 (IV)r->float_min_offset, (UV)r->float_max_offset);
9268 } else if (r->float_utf8) {
9269 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->float_utf8),
9270 RE_SV_DUMPLEN(r->float_utf8), 30);
9271 PerlIO_printf(Perl_debug_log,
9272 "floating utf8 %s%s at %"IVdf"..%"UVuf" ",
9273 s, RE_SV_TAIL(r->float_utf8),
9274 (IV)r->float_min_offset, (UV)r->float_max_offset);
9276 if (r->check_substr || r->check_utf8)
9277 PerlIO_printf(Perl_debug_log,
9279 (r->check_substr == r->float_substr
9280 && r->check_utf8 == r->float_utf8
9281 ? "(checking floating" : "(checking anchored"));
9282 if (r->extflags & RXf_NOSCAN)
9283 PerlIO_printf(Perl_debug_log, " noscan");
9284 if (r->extflags & RXf_CHECK_ALL)
9285 PerlIO_printf(Perl_debug_log, " isall");
9286 if (r->check_substr || r->check_utf8)
9287 PerlIO_printf(Perl_debug_log, ") ");
9289 if (ri->regstclass) {
9290 regprop(r, sv, ri->regstclass);
9291 PerlIO_printf(Perl_debug_log, "stclass %s ", SvPVX_const(sv));
9293 if (r->extflags & RXf_ANCH) {
9294 PerlIO_printf(Perl_debug_log, "anchored");
9295 if (r->extflags & RXf_ANCH_BOL)
9296 PerlIO_printf(Perl_debug_log, "(BOL)");
9297 if (r->extflags & RXf_ANCH_MBOL)
9298 PerlIO_printf(Perl_debug_log, "(MBOL)");
9299 if (r->extflags & RXf_ANCH_SBOL)
9300 PerlIO_printf(Perl_debug_log, "(SBOL)");
9301 if (r->extflags & RXf_ANCH_GPOS)
9302 PerlIO_printf(Perl_debug_log, "(GPOS)");
9303 PerlIO_putc(Perl_debug_log, ' ');
9305 if (r->extflags & RXf_GPOS_SEEN)
9306 PerlIO_printf(Perl_debug_log, "GPOS:%"UVuf" ", (UV)r->gofs);
9307 if (r->intflags & PREGf_SKIP)
9308 PerlIO_printf(Perl_debug_log, "plus ");
9309 if (r->intflags & PREGf_IMPLICIT)
9310 PerlIO_printf(Perl_debug_log, "implicit ");
9311 PerlIO_printf(Perl_debug_log, "minlen %"IVdf" ", (IV)r->minlen);
9312 if (r->extflags & RXf_EVAL_SEEN)
9313 PerlIO_printf(Perl_debug_log, "with eval ");
9314 PerlIO_printf(Perl_debug_log, "\n");
9315 DEBUG_FLAGS_r(regdump_extflags("r->extflags: ",r->extflags));
9317 PERL_ARGS_ASSERT_REGDUMP;
9318 PERL_UNUSED_CONTEXT;
9320 #endif /* DEBUGGING */
9324 - regprop - printable representation of opcode
9326 #define EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags) \
9329 Perl_sv_catpvf(aTHX_ sv,"%s][%s",PL_colors[1],PL_colors[0]); \
9330 if (flags & ANYOF_INVERT) \
9331 /*make sure the invert info is in each */ \
9332 sv_catpvs(sv, "^"); \
9338 Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
9343 RXi_GET_DECL(prog,progi);
9344 GET_RE_DEBUG_FLAGS_DECL;
9346 PERL_ARGS_ASSERT_REGPROP;
9350 if (OP(o) > REGNODE_MAX) /* regnode.type is unsigned */
9351 /* It would be nice to FAIL() here, but this may be called from
9352 regexec.c, and it would be hard to supply pRExC_state. */
9353 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(o), (int)REGNODE_MAX);
9354 sv_catpv(sv, PL_reg_name[OP(o)]); /* Take off const! */
9356 k = PL_regkind[OP(o)];
9360 /* Using is_utf8_string() (via PERL_PV_UNI_DETECT)
9361 * is a crude hack but it may be the best for now since
9362 * we have no flag "this EXACTish node was UTF-8"
9364 pv_pretty(sv, STRING(o), STR_LEN(o), 60, PL_colors[0], PL_colors[1],
9365 PERL_PV_ESCAPE_UNI_DETECT |
9366 PERL_PV_PRETTY_ELLIPSES |
9367 PERL_PV_PRETTY_LTGT |
9368 PERL_PV_PRETTY_NOCLEAR
9370 } else if (k == TRIE) {
9371 /* print the details of the trie in dumpuntil instead, as
9372 * progi->data isn't available here */
9373 const char op = OP(o);
9374 const U32 n = ARG(o);
9375 const reg_ac_data * const ac = IS_TRIE_AC(op) ?
9376 (reg_ac_data *)progi->data->data[n] :
9378 const reg_trie_data * const trie
9379 = (reg_trie_data*)progi->data->data[!IS_TRIE_AC(op) ? n : ac->trie];
9381 Perl_sv_catpvf(aTHX_ sv, "-%s",PL_reg_name[o->flags]);
9382 DEBUG_TRIE_COMPILE_r(
9383 Perl_sv_catpvf(aTHX_ sv,
9384 "<S:%"UVuf"/%"IVdf" W:%"UVuf" L:%"UVuf"/%"UVuf" C:%"UVuf"/%"UVuf">",
9385 (UV)trie->startstate,
9386 (IV)trie->statecount-1, /* -1 because of the unused 0 element */
9387 (UV)trie->wordcount,
9390 (UV)TRIE_CHARCOUNT(trie),
9391 (UV)trie->uniquecharcount
9394 if ( IS_ANYOF_TRIE(op) || trie->bitmap ) {
9396 int rangestart = -1;
9397 U8* bitmap = IS_ANYOF_TRIE(op) ? (U8*)ANYOF_BITMAP(o) : (U8*)TRIE_BITMAP(trie);
9399 for (i = 0; i <= 256; i++) {
9400 if (i < 256 && BITMAP_TEST(bitmap,i)) {
9401 if (rangestart == -1)
9403 } else if (rangestart != -1) {
9404 if (i <= rangestart + 3)
9405 for (; rangestart < i; rangestart++)
9406 put_byte(sv, rangestart);
9408 put_byte(sv, rangestart);
9410 put_byte(sv, i - 1);
9418 } else if (k == CURLY) {
9419 if (OP(o) == CURLYM || OP(o) == CURLYN || OP(o) == CURLYX)
9420 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* Parenth number */
9421 Perl_sv_catpvf(aTHX_ sv, " {%d,%d}", ARG1(o), ARG2(o));
9423 else if (k == WHILEM && o->flags) /* Ordinal/of */
9424 Perl_sv_catpvf(aTHX_ sv, "[%d/%d]", o->flags & 0xf, o->flags>>4);
9425 else if (k == REF || k == OPEN || k == CLOSE || k == GROUPP || OP(o)==ACCEPT) {
9426 Perl_sv_catpvf(aTHX_ sv, "%d", (int)ARG(o)); /* Parenth number */
9427 if ( RXp_PAREN_NAMES(prog) ) {
9428 if ( k != REF || OP(o) < NREF) {
9429 AV *list= MUTABLE_AV(progi->data->data[progi->name_list_idx]);
9430 SV **name= av_fetch(list, ARG(o), 0 );
9432 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9435 AV *list= MUTABLE_AV(progi->data->data[ progi->name_list_idx ]);
9436 SV *sv_dat= MUTABLE_SV(progi->data->data[ ARG( o ) ]);
9437 I32 *nums=(I32*)SvPVX(sv_dat);
9438 SV **name= av_fetch(list, nums[0], 0 );
9441 for ( n=0; n<SvIVX(sv_dat); n++ ) {
9442 Perl_sv_catpvf(aTHX_ sv, "%s%"IVdf,
9443 (n ? "," : ""), (IV)nums[n]);
9445 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9449 } else if (k == GOSUB)
9450 Perl_sv_catpvf(aTHX_ sv, "%d[%+d]", (int)ARG(o),(int)ARG2L(o)); /* Paren and offset */
9451 else if (k == VERB) {
9453 Perl_sv_catpvf(aTHX_ sv, ":%"SVf,
9454 SVfARG((MUTABLE_SV(progi->data->data[ ARG( o ) ]))));
9455 } else if (k == LOGICAL)
9456 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* 2: embedded, otherwise 1 */
9457 else if (k == FOLDCHAR)
9458 Perl_sv_catpvf(aTHX_ sv, "[0x%"UVXf"]", PTR2UV(ARG(o)) );
9459 else if (k == ANYOF) {
9460 int i, rangestart = -1;
9461 const U8 flags = ANYOF_FLAGS(o);
9464 /* Should be synchronized with * ANYOF_ #xdefines in regcomp.h */
9465 static const char * const anyofs[] = {
9498 if (flags & ANYOF_LOCALE)
9499 sv_catpvs(sv, "{loc}");
9500 if (flags & ANYOF_FOLD)
9501 sv_catpvs(sv, "{i}");
9502 Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
9503 if (flags & ANYOF_INVERT)
9506 /* output what the standard cp 0-255 bitmap matches */
9507 for (i = 0; i <= 256; i++) {
9508 if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
9509 if (rangestart == -1)
9511 } else if (rangestart != -1) {
9512 if (i <= rangestart + 3)
9513 for (; rangestart < i; rangestart++)
9514 put_byte(sv, rangestart);
9516 put_byte(sv, rangestart);
9518 put_byte(sv, i - 1);
9525 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9526 /* output any special charclass tests (used mostly under use locale) */
9527 if (o->flags & ANYOF_CLASS)
9528 for (i = 0; i < (int)(sizeof(anyofs)/sizeof(char*)); i++)
9529 if (ANYOF_CLASS_TEST(o,i)) {
9530 sv_catpv(sv, anyofs[i]);
9534 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9536 /* output information about the unicode matching */
9537 if (flags & ANYOF_UNICODE)
9538 sv_catpvs(sv, "{unicode}");
9539 else if (flags & ANYOF_UNICODE_ALL)
9540 sv_catpvs(sv, "{unicode_all}");
9544 SV * const sw = regclass_swash(prog, o, FALSE, &lv, 0);
9548 U8 s[UTF8_MAXBYTES_CASE+1];
9550 for (i = 0; i <= 256; i++) { /* just the first 256 */
9551 uvchr_to_utf8(s, i);
9553 if (i < 256 && swash_fetch(sw, s, TRUE)) {
9554 if (rangestart == -1)
9556 } else if (rangestart != -1) {
9557 if (i <= rangestart + 3)
9558 for (; rangestart < i; rangestart++) {
9559 const U8 * const e = uvchr_to_utf8(s,rangestart);
9561 for(p = s; p < e; p++)
9565 const U8 *e = uvchr_to_utf8(s,rangestart);
9567 for (p = s; p < e; p++)
9570 e = uvchr_to_utf8(s, i-1);
9571 for (p = s; p < e; p++)
9578 sv_catpvs(sv, "..."); /* et cetera */
9582 char *s = savesvpv(lv);
9583 char * const origs = s;
9585 while (*s && *s != '\n')
9589 const char * const t = ++s;
9607 Perl_sv_catpvf(aTHX_ sv, "%s]", PL_colors[1]);
9609 else if (k == BRANCHJ && (OP(o) == UNLESSM || OP(o) == IFMATCH))
9610 Perl_sv_catpvf(aTHX_ sv, "[%d]", -(o->flags));
9612 PERL_UNUSED_CONTEXT;
9613 PERL_UNUSED_ARG(sv);
9615 PERL_UNUSED_ARG(prog);
9616 #endif /* DEBUGGING */
9620 Perl_re_intuit_string(pTHX_ REGEXP * const r)
9621 { /* Assume that RE_INTUIT is set */
9623 struct regexp *const prog = (struct regexp *)SvANY(r);
9624 GET_RE_DEBUG_FLAGS_DECL;
9626 PERL_ARGS_ASSERT_RE_INTUIT_STRING;
9627 PERL_UNUSED_CONTEXT;
9631 const char * const s = SvPV_nolen_const(prog->check_substr
9632 ? prog->check_substr : prog->check_utf8);
9634 if (!PL_colorset) reginitcolors();
9635 PerlIO_printf(Perl_debug_log,
9636 "%sUsing REx %ssubstr:%s \"%s%.60s%s%s\"\n",
9638 prog->check_substr ? "" : "utf8 ",
9639 PL_colors[5],PL_colors[0],
9642 (strlen(s) > 60 ? "..." : ""));
9645 return prog->check_substr ? prog->check_substr : prog->check_utf8;
9651 handles refcounting and freeing the perl core regexp structure. When
9652 it is necessary to actually free the structure the first thing it
9653 does is call the 'free' method of the regexp_engine associated to
9654 the regexp, allowing the handling of the void *pprivate; member
9655 first. (This routine is not overridable by extensions, which is why
9656 the extensions free is called first.)
9658 See regdupe and regdupe_internal if you change anything here.
9660 #ifndef PERL_IN_XSUB_RE
9662 Perl_pregfree(pTHX_ REGEXP *r)
9668 Perl_pregfree2(pTHX_ REGEXP *rx)
9671 struct regexp *const r = (struct regexp *)SvANY(rx);
9672 GET_RE_DEBUG_FLAGS_DECL;
9674 PERL_ARGS_ASSERT_PREGFREE2;
9677 ReREFCNT_dec(r->mother_re);
9679 CALLREGFREE_PVT(rx); /* free the private data */
9680 SvREFCNT_dec(RXp_PAREN_NAMES(r));
9683 SvREFCNT_dec(r->anchored_substr);
9684 SvREFCNT_dec(r->anchored_utf8);
9685 SvREFCNT_dec(r->float_substr);
9686 SvREFCNT_dec(r->float_utf8);
9687 Safefree(r->substrs);
9689 RX_MATCH_COPY_FREE(rx);
9690 #ifdef PERL_OLD_COPY_ON_WRITE
9691 SvREFCNT_dec(r->saved_copy);
9698 This is a hacky workaround to the structural issue of match results
9699 being stored in the regexp structure which is in turn stored in
9700 PL_curpm/PL_reg_curpm. The problem is that due to qr// the pattern
9701 could be PL_curpm in multiple contexts, and could require multiple
9702 result sets being associated with the pattern simultaneously, such
9703 as when doing a recursive match with (??{$qr})
9705 The solution is to make a lightweight copy of the regexp structure
9706 when a qr// is returned from the code executed by (??{$qr}) this
9707 lightweight copy doesnt actually own any of its data except for
9708 the starp/end and the actual regexp structure itself.
9714 Perl_reg_temp_copy (pTHX_ REGEXP *ret_x, REGEXP *rx)
9717 struct regexp *const r = (struct regexp *)SvANY(rx);
9718 register const I32 npar = r->nparens+1;
9720 PERL_ARGS_ASSERT_REG_TEMP_COPY;
9723 ret_x = (REGEXP*) newSV_type(SVt_REGEXP);
9724 ret = (struct regexp *)SvANY(ret_x);
9726 (void)ReREFCNT_inc(rx);
9727 /* We can take advantage of the existing "copied buffer" mechanism in SVs
9728 by pointing directly at the buffer, but flagging that the allocated
9729 space in the copy is zero. As we've just done a struct copy, it's now
9730 a case of zero-ing that, rather than copying the current length. */
9731 SvPV_set(ret_x, RX_WRAPPED(rx));
9732 SvFLAGS(ret_x) |= SvFLAGS(rx) & (SVf_POK|SVp_POK|SVf_UTF8);
9733 memcpy(&(ret->xpv_cur), &(r->xpv_cur),
9734 sizeof(regexp) - STRUCT_OFFSET(regexp, xpv_cur));
9735 SvLEN_set(ret_x, 0);
9736 SvSTASH_set(ret_x, NULL);
9737 SvMAGIC_set(ret_x, NULL);
9738 Newx(ret->offs, npar, regexp_paren_pair);
9739 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9741 Newx(ret->substrs, 1, struct reg_substr_data);
9742 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9744 SvREFCNT_inc_void(ret->anchored_substr);
9745 SvREFCNT_inc_void(ret->anchored_utf8);
9746 SvREFCNT_inc_void(ret->float_substr);
9747 SvREFCNT_inc_void(ret->float_utf8);
9749 /* check_substr and check_utf8, if non-NULL, point to either their
9750 anchored or float namesakes, and don't hold a second reference. */
9752 RX_MATCH_COPIED_off(ret_x);
9753 #ifdef PERL_OLD_COPY_ON_WRITE
9754 ret->saved_copy = NULL;
9756 ret->mother_re = rx;
9762 /* regfree_internal()
9764 Free the private data in a regexp. This is overloadable by
9765 extensions. Perl takes care of the regexp structure in pregfree(),
9766 this covers the *pprivate pointer which technically perl doesn't
9767 know about, however of course we have to handle the
9768 regexp_internal structure when no extension is in use.
9770 Note this is called before freeing anything in the regexp
9775 Perl_regfree_internal(pTHX_ REGEXP * const rx)
9778 struct regexp *const r = (struct regexp *)SvANY(rx);
9780 GET_RE_DEBUG_FLAGS_DECL;
9782 PERL_ARGS_ASSERT_REGFREE_INTERNAL;
9788 SV *dsv= sv_newmortal();
9789 RE_PV_QUOTED_DECL(s, RX_UTF8(rx),
9790 dsv, RX_PRECOMP(rx), RX_PRELEN(rx), 60);
9791 PerlIO_printf(Perl_debug_log,"%sFreeing REx:%s %s\n",
9792 PL_colors[4],PL_colors[5],s);
9795 #ifdef RE_TRACK_PATTERN_OFFSETS
9797 Safefree(ri->u.offsets); /* 20010421 MJD */
9800 int n = ri->data->count;
9801 PAD* new_comppad = NULL;
9806 /* If you add a ->what type here, update the comment in regcomp.h */
9807 switch (ri->data->what[n]) {
9812 SvREFCNT_dec(MUTABLE_SV(ri->data->data[n]));
9815 Safefree(ri->data->data[n]);
9818 new_comppad = MUTABLE_AV(ri->data->data[n]);
9821 if (new_comppad == NULL)
9822 Perl_croak(aTHX_ "panic: pregfree comppad");
9823 PAD_SAVE_LOCAL(old_comppad,
9824 /* Watch out for global destruction's random ordering. */
9825 (SvTYPE(new_comppad) == SVt_PVAV) ? new_comppad : NULL
9828 refcnt = OpREFCNT_dec((OP_4tree*)ri->data->data[n]);
9831 op_free((OP_4tree*)ri->data->data[n]);
9833 PAD_RESTORE_LOCAL(old_comppad);
9834 SvREFCNT_dec(MUTABLE_SV(new_comppad));
9840 { /* Aho Corasick add-on structure for a trie node.
9841 Used in stclass optimization only */
9843 reg_ac_data *aho=(reg_ac_data*)ri->data->data[n];
9845 refcount = --aho->refcount;
9848 PerlMemShared_free(aho->states);
9849 PerlMemShared_free(aho->fail);
9850 /* do this last!!!! */
9851 PerlMemShared_free(ri->data->data[n]);
9852 PerlMemShared_free(ri->regstclass);
9858 /* trie structure. */
9860 reg_trie_data *trie=(reg_trie_data*)ri->data->data[n];
9862 refcount = --trie->refcount;
9865 PerlMemShared_free(trie->charmap);
9866 PerlMemShared_free(trie->states);
9867 PerlMemShared_free(trie->trans);
9869 PerlMemShared_free(trie->bitmap);
9871 PerlMemShared_free(trie->jump);
9872 PerlMemShared_free(trie->wordinfo);
9873 /* do this last!!!! */
9874 PerlMemShared_free(ri->data->data[n]);
9879 Perl_croak(aTHX_ "panic: regfree data code '%c'", ri->data->what[n]);
9882 Safefree(ri->data->what);
9889 #define av_dup_inc(s,t) MUTABLE_AV(sv_dup_inc((const SV *)s,t))
9890 #define hv_dup_inc(s,t) MUTABLE_HV(sv_dup_inc((const SV *)s,t))
9891 #define SAVEPVN(p,n) ((p) ? savepvn(p,n) : NULL)
9894 re_dup - duplicate a regexp.
9896 This routine is expected to clone a given regexp structure. It is only
9897 compiled under USE_ITHREADS.
9899 After all of the core data stored in struct regexp is duplicated
9900 the regexp_engine.dupe method is used to copy any private data
9901 stored in the *pprivate pointer. This allows extensions to handle
9902 any duplication it needs to do.
9904 See pregfree() and regfree_internal() if you change anything here.
9906 #if defined(USE_ITHREADS)
9907 #ifndef PERL_IN_XSUB_RE
9909 Perl_re_dup_guts(pTHX_ const REGEXP *sstr, REGEXP *dstr, CLONE_PARAMS *param)
9913 const struct regexp *r = (const struct regexp *)SvANY(sstr);
9914 struct regexp *ret = (struct regexp *)SvANY(dstr);
9916 PERL_ARGS_ASSERT_RE_DUP_GUTS;
9918 npar = r->nparens+1;
9919 Newx(ret->offs, npar, regexp_paren_pair);
9920 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9922 /* no need to copy these */
9923 Newx(ret->swap, npar, regexp_paren_pair);
9927 /* Do it this way to avoid reading from *r after the StructCopy().
9928 That way, if any of the sv_dup_inc()s dislodge *r from the L1
9929 cache, it doesn't matter. */
9930 const bool anchored = r->check_substr
9931 ? r->check_substr == r->anchored_substr
9932 : r->check_utf8 == r->anchored_utf8;
9933 Newx(ret->substrs, 1, struct reg_substr_data);
9934 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9936 ret->anchored_substr = sv_dup_inc(ret->anchored_substr, param);
9937 ret->anchored_utf8 = sv_dup_inc(ret->anchored_utf8, param);
9938 ret->float_substr = sv_dup_inc(ret->float_substr, param);
9939 ret->float_utf8 = sv_dup_inc(ret->float_utf8, param);
9941 /* check_substr and check_utf8, if non-NULL, point to either their
9942 anchored or float namesakes, and don't hold a second reference. */
9944 if (ret->check_substr) {
9946 assert(r->check_utf8 == r->anchored_utf8);
9947 ret->check_substr = ret->anchored_substr;
9948 ret->check_utf8 = ret->anchored_utf8;
9950 assert(r->check_substr == r->float_substr);
9951 assert(r->check_utf8 == r->float_utf8);
9952 ret->check_substr = ret->float_substr;
9953 ret->check_utf8 = ret->float_utf8;
9955 } else if (ret->check_utf8) {
9957 ret->check_utf8 = ret->anchored_utf8;
9959 ret->check_utf8 = ret->float_utf8;
9964 RXp_PAREN_NAMES(ret) = hv_dup_inc(RXp_PAREN_NAMES(ret), param);
9967 RXi_SET(ret,CALLREGDUPE_PVT(dstr,param));
9969 if (RX_MATCH_COPIED(dstr))
9970 ret->subbeg = SAVEPVN(ret->subbeg, ret->sublen);
9973 #ifdef PERL_OLD_COPY_ON_WRITE
9974 ret->saved_copy = NULL;
9977 if (ret->mother_re) {
9978 if (SvPVX_const(dstr) == SvPVX_const(ret->mother_re)) {
9979 /* Our storage points directly to our mother regexp, but that's
9980 1: a buffer in a different thread
9981 2: something we no longer hold a reference on
9982 so we need to copy it locally. */
9983 /* Note we need to sue SvCUR() on our mother_re, because it, in
9984 turn, may well be pointing to its own mother_re. */
9985 SvPV_set(dstr, SAVEPVN(SvPVX_const(ret->mother_re),
9986 SvCUR(ret->mother_re)+1));
9987 SvLEN_set(dstr, SvCUR(ret->mother_re)+1);
9989 ret->mother_re = NULL;
9993 #endif /* PERL_IN_XSUB_RE */
9998 This is the internal complement to regdupe() which is used to copy
9999 the structure pointed to by the *pprivate pointer in the regexp.
10000 This is the core version of the extension overridable cloning hook.
10001 The regexp structure being duplicated will be copied by perl prior
10002 to this and will be provided as the regexp *r argument, however
10003 with the /old/ structures pprivate pointer value. Thus this routine
10004 may override any copying normally done by perl.
10006 It returns a pointer to the new regexp_internal structure.
10010 Perl_regdupe_internal(pTHX_ REGEXP * const rx, CLONE_PARAMS *param)
10013 struct regexp *const r = (struct regexp *)SvANY(rx);
10014 regexp_internal *reti;
10016 RXi_GET_DECL(r,ri);
10018 PERL_ARGS_ASSERT_REGDUPE_INTERNAL;
10020 npar = r->nparens+1;
10023 Newxc(reti, sizeof(regexp_internal) + len*sizeof(regnode), char, regexp_internal);
10024 Copy(ri->program, reti->program, len+1, regnode);
10027 reti->regstclass = NULL;
10030 struct reg_data *d;
10031 const int count = ri->data->count;
10034 Newxc(d, sizeof(struct reg_data) + count*sizeof(void *),
10035 char, struct reg_data);
10036 Newx(d->what, count, U8);
10039 for (i = 0; i < count; i++) {
10040 d->what[i] = ri->data->what[i];
10041 switch (d->what[i]) {
10042 /* legal options are one of: sSfpontTua
10043 see also regcomp.h and pregfree() */
10044 case 'a': /* actually an AV, but the dup function is identical. */
10047 case 'p': /* actually an AV, but the dup function is identical. */
10048 case 'u': /* actually an HV, but the dup function is identical. */
10049 d->data[i] = sv_dup_inc((const SV *)ri->data->data[i], param);
10052 /* This is cheating. */
10053 Newx(d->data[i], 1, struct regnode_charclass_class);
10054 StructCopy(ri->data->data[i], d->data[i],
10055 struct regnode_charclass_class);
10056 reti->regstclass = (regnode*)d->data[i];
10059 /* Compiled op trees are readonly and in shared memory,
10060 and can thus be shared without duplication. */
10062 d->data[i] = (void*)OpREFCNT_inc((OP*)ri->data->data[i]);
10066 /* Trie stclasses are readonly and can thus be shared
10067 * without duplication. We free the stclass in pregfree
10068 * when the corresponding reg_ac_data struct is freed.
10070 reti->regstclass= ri->regstclass;
10074 ((reg_trie_data*)ri->data->data[i])->refcount++;
10078 d->data[i] = ri->data->data[i];
10081 Perl_croak(aTHX_ "panic: re_dup unknown data code '%c'", ri->data->what[i]);
10090 reti->name_list_idx = ri->name_list_idx;
10092 #ifdef RE_TRACK_PATTERN_OFFSETS
10093 if (ri->u.offsets) {
10094 Newx(reti->u.offsets, 2*len+1, U32);
10095 Copy(ri->u.offsets, reti->u.offsets, 2*len+1, U32);
10098 SetProgLen(reti,len);
10101 return (void*)reti;
10104 #endif /* USE_ITHREADS */
10106 #ifndef PERL_IN_XSUB_RE
10109 - regnext - dig the "next" pointer out of a node
10112 Perl_regnext(pTHX_ register regnode *p)
10115 register I32 offset;
10120 if (OP(p) > REGNODE_MAX) { /* regnode.type is unsigned */
10121 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(p), (int)REGNODE_MAX);
10124 offset = (reg_off_by_arg[OP(p)] ? ARG(p) : NEXT_OFF(p));
10133 S_re_croak2(pTHX_ const char* pat1,const char* pat2,...)
10136 STRLEN l1 = strlen(pat1);
10137 STRLEN l2 = strlen(pat2);
10140 const char *message;
10142 PERL_ARGS_ASSERT_RE_CROAK2;
10148 Copy(pat1, buf, l1 , char);
10149 Copy(pat2, buf + l1, l2 , char);
10150 buf[l1 + l2] = '\n';
10151 buf[l1 + l2 + 1] = '\0';
10153 /* ANSI variant takes additional second argument */
10154 va_start(args, pat2);
10158 msv = vmess(buf, &args);
10160 message = SvPV_const(msv,l1);
10163 Copy(message, buf, l1 , char);
10164 buf[l1-1] = '\0'; /* Overwrite \n */
10165 Perl_croak(aTHX_ "%s", buf);
10168 /* XXX Here's a total kludge. But we need to re-enter for swash routines. */
10170 #ifndef PERL_IN_XSUB_RE
10172 Perl_save_re_context(pTHX)
10176 struct re_save_state *state;
10178 SAVEVPTR(PL_curcop);
10179 SSGROW(SAVESTACK_ALLOC_FOR_RE_SAVE_STATE + 1);
10181 state = (struct re_save_state *)(PL_savestack + PL_savestack_ix);
10182 PL_savestack_ix += SAVESTACK_ALLOC_FOR_RE_SAVE_STATE;
10183 SSPUSHUV(SAVEt_RE_STATE);
10185 Copy(&PL_reg_state, state, 1, struct re_save_state);
10187 PL_reg_start_tmp = 0;
10188 PL_reg_start_tmpl = 0;
10189 PL_reg_oldsaved = NULL;
10190 PL_reg_oldsavedlen = 0;
10191 PL_reg_maxiter = 0;
10192 PL_reg_leftiter = 0;
10193 PL_reg_poscache = NULL;
10194 PL_reg_poscache_size = 0;
10195 #ifdef PERL_OLD_COPY_ON_WRITE
10199 /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
10201 const REGEXP * const rx = PM_GETRE(PL_curpm);
10204 for (i = 1; i <= RX_NPARENS(rx); i++) {
10205 char digits[TYPE_CHARS(long)];
10206 const STRLEN len = my_snprintf(digits, sizeof(digits), "%lu", (long)i);
10207 GV *const *const gvp
10208 = (GV**)hv_fetch(PL_defstash, digits, len, 0);
10211 GV * const gv = *gvp;
10212 if (SvTYPE(gv) == SVt_PVGV && GvSV(gv))
10222 clear_re(pTHX_ void *r)
10225 ReREFCNT_dec((REGEXP *)r);
10231 S_put_byte(pTHX_ SV *sv, int c)
10233 PERL_ARGS_ASSERT_PUT_BYTE;
10235 /* Our definition of isPRINT() ignores locales, so only bytes that are
10236 not part of UTF-8 are considered printable. I assume that the same
10237 holds for UTF-EBCDIC.
10238 Also, code point 255 is not printable in either (it's E0 in EBCDIC,
10239 which Wikipedia says:
10241 EO, or Eight Ones, is an 8-bit EBCDIC character code represented as all
10242 ones (binary 1111 1111, hexadecimal FF). It is similar, but not
10243 identical, to the ASCII delete (DEL) or rubout control character.
10244 ) So the old condition can be simplified to !isPRINT(c) */
10246 Perl_sv_catpvf(aTHX_ sv, "\\%o", c);
10248 const char string = c;
10249 if (c == '-' || c == ']' || c == '\\' || c == '^')
10250 sv_catpvs(sv, "\\");
10251 sv_catpvn(sv, &string, 1);
10256 #define CLEAR_OPTSTART \
10257 if (optstart) STMT_START { \
10258 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log, " (%"IVdf" nodes)\n", (IV)(node - optstart))); \
10262 #define DUMPUNTIL(b,e) CLEAR_OPTSTART; node=dumpuntil(r,start,(b),(e),last,sv,indent+1,depth+1);
10264 STATIC const regnode *
10265 S_dumpuntil(pTHX_ const regexp *r, const regnode *start, const regnode *node,
10266 const regnode *last, const regnode *plast,
10267 SV* sv, I32 indent, U32 depth)
10270 register U8 op = PSEUDO; /* Arbitrary non-END op. */
10271 register const regnode *next;
10272 const regnode *optstart= NULL;
10274 RXi_GET_DECL(r,ri);
10275 GET_RE_DEBUG_FLAGS_DECL;
10277 PERL_ARGS_ASSERT_DUMPUNTIL;
10279 #ifdef DEBUG_DUMPUNTIL
10280 PerlIO_printf(Perl_debug_log, "--- %d : %d - %d - %d\n",indent,node-start,
10281 last ? last-start : 0,plast ? plast-start : 0);
10284 if (plast && plast < last)
10287 while (PL_regkind[op] != END && (!last || node < last)) {
10288 /* While that wasn't END last time... */
10291 if (op == CLOSE || op == WHILEM)
10293 next = regnext((regnode *)node);
10296 if (OP(node) == OPTIMIZED) {
10297 if (!optstart && RE_DEBUG_FLAG(RE_DEBUG_COMPILE_OPTIMISE))
10304 regprop(r, sv, node);
10305 PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
10306 (int)(2*indent + 1), "", SvPVX_const(sv));
10308 if (OP(node) != OPTIMIZED) {
10309 if (next == NULL) /* Next ptr. */
10310 PerlIO_printf(Perl_debug_log, " (0)");
10311 else if (PL_regkind[(U8)op] == BRANCH && PL_regkind[OP(next)] != BRANCH )
10312 PerlIO_printf(Perl_debug_log, " (FAIL)");
10314 PerlIO_printf(Perl_debug_log, " (%"IVdf")", (IV)(next - start));
10315 (void)PerlIO_putc(Perl_debug_log, '\n');
10319 if (PL_regkind[(U8)op] == BRANCHJ) {
10322 register const regnode *nnode = (OP(next) == LONGJMP
10323 ? regnext((regnode *)next)
10325 if (last && nnode > last)
10327 DUMPUNTIL(NEXTOPER(NEXTOPER(node)), nnode);
10330 else if (PL_regkind[(U8)op] == BRANCH) {
10332 DUMPUNTIL(NEXTOPER(node), next);
10334 else if ( PL_regkind[(U8)op] == TRIE ) {
10335 const regnode *this_trie = node;
10336 const char op = OP(node);
10337 const U32 n = ARG(node);
10338 const reg_ac_data * const ac = op>=AHOCORASICK ?
10339 (reg_ac_data *)ri->data->data[n] :
10341 const reg_trie_data * const trie =
10342 (reg_trie_data*)ri->data->data[op<AHOCORASICK ? n : ac->trie];
10344 AV *const trie_words = MUTABLE_AV(ri->data->data[n + TRIE_WORDS_OFFSET]);
10346 const regnode *nextbranch= NULL;
10349 for (word_idx= 0; word_idx < (I32)trie->wordcount; word_idx++) {
10350 SV ** const elem_ptr = av_fetch(trie_words,word_idx,0);
10352 PerlIO_printf(Perl_debug_log, "%*s%s ",
10353 (int)(2*(indent+3)), "",
10354 elem_ptr ? pv_pretty(sv, SvPV_nolen_const(*elem_ptr), SvCUR(*elem_ptr), 60,
10355 PL_colors[0], PL_colors[1],
10356 (SvUTF8(*elem_ptr) ? PERL_PV_ESCAPE_UNI : 0) |
10357 PERL_PV_PRETTY_ELLIPSES |
10358 PERL_PV_PRETTY_LTGT
10363 U16 dist= trie->jump[word_idx+1];
10364 PerlIO_printf(Perl_debug_log, "(%"UVuf")\n",
10365 (UV)((dist ? this_trie + dist : next) - start));
10368 nextbranch= this_trie + trie->jump[0];
10369 DUMPUNTIL(this_trie + dist, nextbranch);
10371 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
10372 nextbranch= regnext((regnode *)nextbranch);
10374 PerlIO_printf(Perl_debug_log, "\n");
10377 if (last && next > last)
10382 else if ( op == CURLY ) { /* "next" might be very big: optimizer */
10383 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS,
10384 NEXTOPER(node) + EXTRA_STEP_2ARGS + 1);
10386 else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
10388 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS, next);
10390 else if ( op == PLUS || op == STAR) {
10391 DUMPUNTIL(NEXTOPER(node), NEXTOPER(node) + 1);
10393 else if (op == ANYOF) {
10394 /* arglen 1 + class block */
10395 node += 1 + ((ANYOF_FLAGS(node) & ANYOF_LARGE)
10396 ? ANYOF_CLASS_SKIP : ANYOF_SKIP);
10397 node = NEXTOPER(node);
10399 else if (PL_regkind[(U8)op] == EXACT) {
10400 /* Literal string, where present. */
10401 node += NODE_SZ_STR(node) - 1;
10402 node = NEXTOPER(node);
10405 node = NEXTOPER(node);
10406 node += regarglen[(U8)op];
10408 if (op == CURLYX || op == OPEN)
10412 #ifdef DEBUG_DUMPUNTIL
10413 PerlIO_printf(Perl_debug_log, "--- %d\n", (int)indent);
10418 #endif /* DEBUGGING */
10422 * c-indentation-style: bsd
10423 * c-basic-offset: 4
10424 * indent-tabs-mode: t
10427 * ex: set ts=8 sts=4 sw=4 noet:
projects / perl / modules / re-engine-Hooks.git / blob