5 * 'A fair jaw-cracker dwarf-language must be.' --Samwise Gamgee
7 * [p.285 of _The Lord of the Rings_, II/iii: "The Ring Goes South"]
10 /* This file contains functions for compiling a regular expression. See
11 * also regexec.c which funnily enough, contains functions for executing
12 * a regular expression.
14 * This file is also copied at build time to ext/re/re_comp.c, where
15 * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
16 * This causes the main functions to be compiled under new names and with
17 * debugging support added, which makes "use re 'debug'" work.
20 /* NOTE: this is derived from Henry Spencer's regexp code, and should not
21 * confused with the original package (see point 3 below). Thanks, Henry!
24 /* Additional note: this code is very heavily munged from Henry's version
25 * in places. In some spots I've traded clarity for efficiency, so don't
26 * blame Henry for some of the lack of readability.
29 /* The names of the functions have been changed from regcomp and
30 * regexec to pregcomp and pregexec in order to avoid conflicts
31 * with the POSIX routines of the same names.
34 #ifdef PERL_EXT_RE_BUILD
39 * pregcomp and pregexec -- regsub and regerror are not used in perl
41 * Copyright (c) 1986 by University of Toronto.
42 * Written by Henry Spencer. Not derived from licensed software.
44 * Permission is granted to anyone to use this software for any
45 * purpose on any computer system, and to redistribute it freely,
46 * subject to the following restrictions:
48 * 1. The author is not responsible for the consequences of use of
49 * this software, no matter how awful, even if they arise
52 * 2. The origin of this software must not be misrepresented, either
53 * by explicit claim or by omission.
55 * 3. Altered versions must be plainly marked as such, and must not
56 * be misrepresented as being the original software.
59 **** Alterations to Henry's code are...
61 **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
62 **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
63 **** by Larry Wall and others
65 **** You may distribute under the terms of either the GNU General Public
66 **** License or the Artistic License, as specified in the README file.
69 * Beware that some of this code is subtly aware of the way operator
70 * precedence is structured in regular expressions. Serious changes in
71 * regular-expression syntax might require a total rethink.
74 #define PERL_IN_REGCOMP_C
77 #ifndef PERL_IN_XSUB_RE
82 #ifdef PERL_IN_XSUB_RE
93 # if defined(BUGGY_MSC6)
94 /* MSC 6.00A breaks on op/regexp.t test 85 unless we turn this off */
95 # pragma optimize("a",off)
96 /* But MSC 6.00A is happy with 'w', for aliases only across function calls*/
97 # pragma optimize("w",on )
98 # endif /* BUGGY_MSC6 */
102 #define STATIC static
105 typedef struct RExC_state_t {
106 U32 flags; /* are we folding, multilining? */
107 char *precomp; /* uncompiled string. */
108 REGEXP *rx_sv; /* The SV that is the regexp. */
109 regexp *rx; /* perl core regexp structure */
110 regexp_internal *rxi; /* internal data for regexp object pprivate field */
111 char *start; /* Start of input for compile */
112 char *end; /* End of input for compile */
113 char *parse; /* Input-scan pointer. */
114 I32 whilem_seen; /* number of WHILEM in this expr */
115 regnode *emit_start; /* Start of emitted-code area */
116 regnode *emit_bound; /* First regnode outside of the allocated space */
117 regnode *emit; /* Code-emit pointer; ®dummy = don't = compiling */
118 I32 naughty; /* How bad is this pattern? */
119 I32 sawback; /* Did we see \1, ...? */
121 I32 size; /* Code size. */
122 I32 npar; /* Capture buffer count, (OPEN). */
123 I32 cpar; /* Capture buffer count, (CLOSE). */
124 I32 nestroot; /* root parens we are in - used by accept */
128 regnode **open_parens; /* pointers to open parens */
129 regnode **close_parens; /* pointers to close parens */
130 regnode *opend; /* END node in program */
131 I32 utf8; /* whether the pattern is utf8 or not */
132 I32 orig_utf8; /* whether the pattern was originally in utf8 */
133 /* XXX use this for future optimisation of case
134 * where pattern must be upgraded to utf8. */
135 HV *charnames; /* cache of named sequences */
136 HV *paren_names; /* Paren names */
138 regnode **recurse; /* Recurse regops */
139 I32 recurse_count; /* Number of recurse regops */
141 char *starttry; /* -Dr: where regtry was called. */
142 #define RExC_starttry (pRExC_state->starttry)
145 const char *lastparse;
147 AV *paren_name_list; /* idx -> name */
148 #define RExC_lastparse (pRExC_state->lastparse)
149 #define RExC_lastnum (pRExC_state->lastnum)
150 #define RExC_paren_name_list (pRExC_state->paren_name_list)
154 #define RExC_flags (pRExC_state->flags)
155 #define RExC_precomp (pRExC_state->precomp)
156 #define RExC_rx_sv (pRExC_state->rx_sv)
157 #define RExC_rx (pRExC_state->rx)
158 #define RExC_rxi (pRExC_state->rxi)
159 #define RExC_start (pRExC_state->start)
160 #define RExC_end (pRExC_state->end)
161 #define RExC_parse (pRExC_state->parse)
162 #define RExC_whilem_seen (pRExC_state->whilem_seen)
163 #ifdef RE_TRACK_PATTERN_OFFSETS
164 #define RExC_offsets (pRExC_state->rxi->u.offsets) /* I am not like the others */
166 #define RExC_emit (pRExC_state->emit)
167 #define RExC_emit_start (pRExC_state->emit_start)
168 #define RExC_emit_bound (pRExC_state->emit_bound)
169 #define RExC_naughty (pRExC_state->naughty)
170 #define RExC_sawback (pRExC_state->sawback)
171 #define RExC_seen (pRExC_state->seen)
172 #define RExC_size (pRExC_state->size)
173 #define RExC_npar (pRExC_state->npar)
174 #define RExC_nestroot (pRExC_state->nestroot)
175 #define RExC_extralen (pRExC_state->extralen)
176 #define RExC_seen_zerolen (pRExC_state->seen_zerolen)
177 #define RExC_seen_evals (pRExC_state->seen_evals)
178 #define RExC_utf8 (pRExC_state->utf8)
179 #define RExC_orig_utf8 (pRExC_state->orig_utf8)
180 #define RExC_charnames (pRExC_state->charnames)
181 #define RExC_open_parens (pRExC_state->open_parens)
182 #define RExC_close_parens (pRExC_state->close_parens)
183 #define RExC_opend (pRExC_state->opend)
184 #define RExC_paren_names (pRExC_state->paren_names)
185 #define RExC_recurse (pRExC_state->recurse)
186 #define RExC_recurse_count (pRExC_state->recurse_count)
189 #define ISMULT1(c) ((c) == '*' || (c) == '+' || (c) == '?')
190 #define ISMULT2(s) ((*s) == '*' || (*s) == '+' || (*s) == '?' || \
191 ((*s) == '{' && regcurly(s)))
194 #undef SPSTART /* dratted cpp namespace... */
197 * Flags to be passed up and down.
199 #define WORST 0 /* Worst case. */
200 #define HASWIDTH 0x01 /* Known to match non-null strings. */
201 #define SIMPLE 0x02 /* Simple enough to be STAR/PLUS operand. */
202 #define SPSTART 0x04 /* Starts with * or +. */
203 #define TRYAGAIN 0x08 /* Weeded out a declaration. */
204 #define POSTPONED 0x10 /* (?1),(?&name), (??{...}) or similar */
206 #define REG_NODE_NUM(x) ((x) ? (int)((x)-RExC_emit_start) : -1)
208 /* whether trie related optimizations are enabled */
209 #if PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION
210 #define TRIE_STUDY_OPT
211 #define FULL_TRIE_STUDY
217 #define PBYTE(u8str,paren) ((U8*)(u8str))[(paren) >> 3]
218 #define PBITVAL(paren) (1 << ((paren) & 7))
219 #define PAREN_TEST(u8str,paren) ( PBYTE(u8str,paren) & PBITVAL(paren))
220 #define PAREN_SET(u8str,paren) PBYTE(u8str,paren) |= PBITVAL(paren)
221 #define PAREN_UNSET(u8str,paren) PBYTE(u8str,paren) &= (~PBITVAL(paren))
224 /* About scan_data_t.
226 During optimisation we recurse through the regexp program performing
227 various inplace (keyhole style) optimisations. In addition study_chunk
228 and scan_commit populate this data structure with information about
229 what strings MUST appear in the pattern. We look for the longest
230 string that must appear for at a fixed location, and we look for the
231 longest string that may appear at a floating location. So for instance
236 Both 'FOO' and 'A' are fixed strings. Both 'B' and 'BAR' are floating
237 strings (because they follow a .* construct). study_chunk will identify
238 both FOO and BAR as being the longest fixed and floating strings respectively.
240 The strings can be composites, for instance
244 will result in a composite fixed substring 'foo'.
246 For each string some basic information is maintained:
248 - offset or min_offset
249 This is the position the string must appear at, or not before.
250 It also implicitly (when combined with minlenp) tells us how many
251 character must match before the string we are searching.
252 Likewise when combined with minlenp and the length of the string
253 tells us how many characters must appear after the string we have
257 Only used for floating strings. This is the rightmost point that
258 the string can appear at. Ifset to I32 max it indicates that the
259 string can occur infinitely far to the right.
262 A pointer to the minimum length of the pattern that the string
263 was found inside. This is important as in the case of positive
264 lookahead or positive lookbehind we can have multiple patterns
269 The minimum length of the pattern overall is 3, the minimum length
270 of the lookahead part is 3, but the minimum length of the part that
271 will actually match is 1. So 'FOO's minimum length is 3, but the
272 minimum length for the F is 1. This is important as the minimum length
273 is used to determine offsets in front of and behind the string being
274 looked for. Since strings can be composites this is the length of the
275 pattern at the time it was commited with a scan_commit. Note that
276 the length is calculated by study_chunk, so that the minimum lengths
277 are not known until the full pattern has been compiled, thus the
278 pointer to the value.
282 In the case of lookbehind the string being searched for can be
283 offset past the start point of the final matching string.
284 If this value was just blithely removed from the min_offset it would
285 invalidate some of the calculations for how many chars must match
286 before or after (as they are derived from min_offset and minlen and
287 the length of the string being searched for).
288 When the final pattern is compiled and the data is moved from the
289 scan_data_t structure into the regexp structure the information
290 about lookbehind is factored in, with the information that would
291 have been lost precalculated in the end_shift field for the
294 The fields pos_min and pos_delta are used to store the minimum offset
295 and the delta to the maximum offset at the current point in the pattern.
299 typedef struct scan_data_t {
300 /*I32 len_min; unused */
301 /*I32 len_delta; unused */
305 I32 last_end; /* min value, <0 unless valid. */
308 SV **longest; /* Either &l_fixed, or &l_float. */
309 SV *longest_fixed; /* longest fixed string found in pattern */
310 I32 offset_fixed; /* offset where it starts */
311 I32 *minlen_fixed; /* pointer to the minlen relevent to the string */
312 I32 lookbehind_fixed; /* is the position of the string modfied by LB */
313 SV *longest_float; /* longest floating string found in pattern */
314 I32 offset_float_min; /* earliest point in string it can appear */
315 I32 offset_float_max; /* latest point in string it can appear */
316 I32 *minlen_float; /* pointer to the minlen relevent to the string */
317 I32 lookbehind_float; /* is the position of the string modified by LB */
321 struct regnode_charclass_class *start_class;
325 * Forward declarations for pregcomp()'s friends.
328 static const scan_data_t zero_scan_data =
329 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ,0};
331 #define SF_BEFORE_EOL (SF_BEFORE_SEOL|SF_BEFORE_MEOL)
332 #define SF_BEFORE_SEOL 0x0001
333 #define SF_BEFORE_MEOL 0x0002
334 #define SF_FIX_BEFORE_EOL (SF_FIX_BEFORE_SEOL|SF_FIX_BEFORE_MEOL)
335 #define SF_FL_BEFORE_EOL (SF_FL_BEFORE_SEOL|SF_FL_BEFORE_MEOL)
338 # define SF_FIX_SHIFT_EOL (0+2)
339 # define SF_FL_SHIFT_EOL (0+4)
341 # define SF_FIX_SHIFT_EOL (+2)
342 # define SF_FL_SHIFT_EOL (+4)
345 #define SF_FIX_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FIX_SHIFT_EOL)
346 #define SF_FIX_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FIX_SHIFT_EOL)
348 #define SF_FL_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FL_SHIFT_EOL)
349 #define SF_FL_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FL_SHIFT_EOL) /* 0x20 */
350 #define SF_IS_INF 0x0040
351 #define SF_HAS_PAR 0x0080
352 #define SF_IN_PAR 0x0100
353 #define SF_HAS_EVAL 0x0200
354 #define SCF_DO_SUBSTR 0x0400
355 #define SCF_DO_STCLASS_AND 0x0800
356 #define SCF_DO_STCLASS_OR 0x1000
357 #define SCF_DO_STCLASS (SCF_DO_STCLASS_AND|SCF_DO_STCLASS_OR)
358 #define SCF_WHILEM_VISITED_POS 0x2000
360 #define SCF_TRIE_RESTUDY 0x4000 /* Do restudy? */
361 #define SCF_SEEN_ACCEPT 0x8000
363 #define UTF (RExC_utf8 != 0)
364 #define LOC ((RExC_flags & RXf_PMf_LOCALE) != 0)
365 #define FOLD ((RExC_flags & RXf_PMf_FOLD) != 0)
367 #define OOB_UNICODE 12345678
368 #define OOB_NAMEDCLASS -1
370 #define CHR_SVLEN(sv) (UTF ? sv_len_utf8(sv) : SvCUR(sv))
371 #define CHR_DIST(a,b) (UTF ? utf8_distance(a,b) : a - b)
374 /* length of regex to show in messages that don't mark a position within */
375 #define RegexLengthToShowInErrorMessages 127
378 * If MARKER[12] are adjusted, be sure to adjust the constants at the top
379 * of t/op/regmesg.t, the tests in t/op/re_tests, and those in
380 * op/pragma/warn/regcomp.
382 #define MARKER1 "<-- HERE" /* marker as it appears in the description */
383 #define MARKER2 " <-- HERE " /* marker as it appears within the regex */
385 #define REPORT_LOCATION " in regex; marked by " MARKER1 " in m/%.*s" MARKER2 "%s/"
388 * Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
389 * arg. Show regex, up to a maximum length. If it's too long, chop and add
392 #define _FAIL(code) STMT_START { \
393 const char *ellipses = ""; \
394 IV len = RExC_end - RExC_precomp; \
397 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
398 if (len > RegexLengthToShowInErrorMessages) { \
399 /* chop 10 shorter than the max, to ensure meaning of "..." */ \
400 len = RegexLengthToShowInErrorMessages - 10; \
406 #define FAIL(msg) _FAIL( \
407 Perl_croak(aTHX_ "%s in regex m/%.*s%s/", \
408 msg, (int)len, RExC_precomp, ellipses))
410 #define FAIL2(msg,arg) _FAIL( \
411 Perl_croak(aTHX_ msg " in regex m/%.*s%s/", \
412 arg, (int)len, RExC_precomp, ellipses))
415 * Simple_vFAIL -- like FAIL, but marks the current location in the scan
417 #define Simple_vFAIL(m) STMT_START { \
418 const IV offset = RExC_parse - RExC_precomp; \
419 Perl_croak(aTHX_ "%s" REPORT_LOCATION, \
420 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
424 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL()
426 #define vFAIL(m) STMT_START { \
428 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
433 * Like Simple_vFAIL(), but accepts two arguments.
435 #define Simple_vFAIL2(m,a1) STMT_START { \
436 const IV offset = RExC_parse - RExC_precomp; \
437 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, \
438 (int)offset, RExC_precomp, RExC_precomp + offset); \
442 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL2().
444 #define vFAIL2(m,a1) STMT_START { \
446 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
447 Simple_vFAIL2(m, a1); \
452 * Like Simple_vFAIL(), but accepts three arguments.
454 #define Simple_vFAIL3(m, a1, a2) STMT_START { \
455 const IV offset = RExC_parse - RExC_precomp; \
456 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, \
457 (int)offset, RExC_precomp, RExC_precomp + offset); \
461 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL3().
463 #define vFAIL3(m,a1,a2) STMT_START { \
465 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
466 Simple_vFAIL3(m, a1, a2); \
470 * Like Simple_vFAIL(), but accepts four arguments.
472 #define Simple_vFAIL4(m, a1, a2, a3) STMT_START { \
473 const IV offset = RExC_parse - RExC_precomp; \
474 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, \
475 (int)offset, RExC_precomp, RExC_precomp + offset); \
478 #define ckWARNreg(loc,m) STMT_START { \
479 const IV offset = loc - RExC_precomp; \
480 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
481 (int)offset, RExC_precomp, RExC_precomp + offset); \
484 #define ckWARNregdep(loc,m) STMT_START { \
485 const IV offset = loc - RExC_precomp; \
486 Perl_ck_warner_d(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
488 (int)offset, RExC_precomp, RExC_precomp + offset); \
491 #define ckWARN2reg(loc, m, a1) STMT_START { \
492 const IV offset = loc - RExC_precomp; \
493 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
494 a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
497 #define vWARN3(loc, m, a1, a2) STMT_START { \
498 const IV offset = loc - RExC_precomp; \
499 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
500 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
503 #define ckWARN3reg(loc, m, a1, a2) STMT_START { \
504 const IV offset = loc - RExC_precomp; \
505 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
506 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
509 #define vWARN4(loc, m, a1, a2, a3) STMT_START { \
510 const IV offset = loc - RExC_precomp; \
511 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
512 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
515 #define ckWARN4reg(loc, m, a1, a2, a3) STMT_START { \
516 const IV offset = loc - RExC_precomp; \
517 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
518 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
521 #define vWARN5(loc, m, a1, a2, a3, a4) STMT_START { \
522 const IV offset = loc - RExC_precomp; \
523 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
524 a1, a2, a3, a4, (int)offset, RExC_precomp, RExC_precomp + offset); \
528 /* Allow for side effects in s */
529 #define REGC(c,s) STMT_START { \
530 if (!SIZE_ONLY) *(s) = (c); else (void)(s); \
533 /* Macros for recording node offsets. 20001227 mjd@plover.com
534 * Nodes are numbered 1, 2, 3, 4. Node #n's position is recorded in
535 * element 2*n-1 of the array. Element #2n holds the byte length node #n.
536 * Element 0 holds the number n.
537 * Position is 1 indexed.
539 #ifndef RE_TRACK_PATTERN_OFFSETS
540 #define Set_Node_Offset_To_R(node,byte)
541 #define Set_Node_Offset(node,byte)
542 #define Set_Cur_Node_Offset
543 #define Set_Node_Length_To_R(node,len)
544 #define Set_Node_Length(node,len)
545 #define Set_Node_Cur_Length(node)
546 #define Node_Offset(n)
547 #define Node_Length(n)
548 #define Set_Node_Offset_Length(node,offset,len)
549 #define ProgLen(ri) ri->u.proglen
550 #define SetProgLen(ri,x) ri->u.proglen = x
552 #define ProgLen(ri) ri->u.offsets[0]
553 #define SetProgLen(ri,x) ri->u.offsets[0] = x
554 #define Set_Node_Offset_To_R(node,byte) STMT_START { \
556 MJD_OFFSET_DEBUG(("** (%d) offset of node %d is %d.\n", \
557 __LINE__, (int)(node), (int)(byte))); \
559 Perl_croak(aTHX_ "value of node is %d in Offset macro", (int)(node)); \
561 RExC_offsets[2*(node)-1] = (byte); \
566 #define Set_Node_Offset(node,byte) \
567 Set_Node_Offset_To_R((node)-RExC_emit_start, (byte)-RExC_start)
568 #define Set_Cur_Node_Offset Set_Node_Offset(RExC_emit, RExC_parse)
570 #define Set_Node_Length_To_R(node,len) STMT_START { \
572 MJD_OFFSET_DEBUG(("** (%d) size of node %d is %d.\n", \
573 __LINE__, (int)(node), (int)(len))); \
575 Perl_croak(aTHX_ "value of node is %d in Length macro", (int)(node)); \
577 RExC_offsets[2*(node)] = (len); \
582 #define Set_Node_Length(node,len) \
583 Set_Node_Length_To_R((node)-RExC_emit_start, len)
584 #define Set_Cur_Node_Length(len) Set_Node_Length(RExC_emit, len)
585 #define Set_Node_Cur_Length(node) \
586 Set_Node_Length(node, RExC_parse - parse_start)
588 /* Get offsets and lengths */
589 #define Node_Offset(n) (RExC_offsets[2*((n)-RExC_emit_start)-1])
590 #define Node_Length(n) (RExC_offsets[2*((n)-RExC_emit_start)])
592 #define Set_Node_Offset_Length(node,offset,len) STMT_START { \
593 Set_Node_Offset_To_R((node)-RExC_emit_start, (offset)); \
594 Set_Node_Length_To_R((node)-RExC_emit_start, (len)); \
598 #if PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS
599 #define EXPERIMENTAL_INPLACESCAN
600 #endif /*RE_TRACK_PATTERN_OFFSETS*/
602 #define DEBUG_STUDYDATA(str,data,depth) \
603 DEBUG_OPTIMISE_MORE_r(if(data){ \
604 PerlIO_printf(Perl_debug_log, \
605 "%*s" str "Pos:%"IVdf"/%"IVdf \
606 " Flags: 0x%"UVXf" Whilem_c: %"IVdf" Lcp: %"IVdf" %s", \
607 (int)(depth)*2, "", \
608 (IV)((data)->pos_min), \
609 (IV)((data)->pos_delta), \
610 (UV)((data)->flags), \
611 (IV)((data)->whilem_c), \
612 (IV)((data)->last_closep ? *((data)->last_closep) : -1), \
613 is_inf ? "INF " : "" \
615 if ((data)->last_found) \
616 PerlIO_printf(Perl_debug_log, \
617 "Last:'%s' %"IVdf":%"IVdf"/%"IVdf" %sFixed:'%s' @ %"IVdf \
618 " %sFloat: '%s' @ %"IVdf"/%"IVdf"", \
619 SvPVX_const((data)->last_found), \
620 (IV)((data)->last_end), \
621 (IV)((data)->last_start_min), \
622 (IV)((data)->last_start_max), \
623 ((data)->longest && \
624 (data)->longest==&((data)->longest_fixed)) ? "*" : "", \
625 SvPVX_const((data)->longest_fixed), \
626 (IV)((data)->offset_fixed), \
627 ((data)->longest && \
628 (data)->longest==&((data)->longest_float)) ? "*" : "", \
629 SvPVX_const((data)->longest_float), \
630 (IV)((data)->offset_float_min), \
631 (IV)((data)->offset_float_max) \
633 PerlIO_printf(Perl_debug_log,"\n"); \
636 static void clear_re(pTHX_ void *r);
638 /* Mark that we cannot extend a found fixed substring at this point.
639 Update the longest found anchored substring and the longest found
640 floating substrings if needed. */
643 S_scan_commit(pTHX_ const RExC_state_t *pRExC_state, scan_data_t *data, I32 *minlenp, int is_inf)
645 const STRLEN l = CHR_SVLEN(data->last_found);
646 const STRLEN old_l = CHR_SVLEN(*data->longest);
647 GET_RE_DEBUG_FLAGS_DECL;
649 PERL_ARGS_ASSERT_SCAN_COMMIT;
651 if ((l >= old_l) && ((l > old_l) || (data->flags & SF_BEFORE_EOL))) {
652 SvSetMagicSV(*data->longest, data->last_found);
653 if (*data->longest == data->longest_fixed) {
654 data->offset_fixed = l ? data->last_start_min : data->pos_min;
655 if (data->flags & SF_BEFORE_EOL)
657 |= ((data->flags & SF_BEFORE_EOL) << SF_FIX_SHIFT_EOL);
659 data->flags &= ~SF_FIX_BEFORE_EOL;
660 data->minlen_fixed=minlenp;
661 data->lookbehind_fixed=0;
663 else { /* *data->longest == data->longest_float */
664 data->offset_float_min = l ? data->last_start_min : data->pos_min;
665 data->offset_float_max = (l
666 ? data->last_start_max
667 : data->pos_min + data->pos_delta);
668 if (is_inf || (U32)data->offset_float_max > (U32)I32_MAX)
669 data->offset_float_max = I32_MAX;
670 if (data->flags & SF_BEFORE_EOL)
672 |= ((data->flags & SF_BEFORE_EOL) << SF_FL_SHIFT_EOL);
674 data->flags &= ~SF_FL_BEFORE_EOL;
675 data->minlen_float=minlenp;
676 data->lookbehind_float=0;
679 SvCUR_set(data->last_found, 0);
681 SV * const sv = data->last_found;
682 if (SvUTF8(sv) && SvMAGICAL(sv)) {
683 MAGIC * const mg = mg_find(sv, PERL_MAGIC_utf8);
689 data->flags &= ~SF_BEFORE_EOL;
690 DEBUG_STUDYDATA("commit: ",data,0);
693 /* Can match anything (initialization) */
695 S_cl_anything(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
697 PERL_ARGS_ASSERT_CL_ANYTHING;
699 ANYOF_CLASS_ZERO(cl);
700 ANYOF_BITMAP_SETALL(cl);
701 cl->flags = ANYOF_EOS|ANYOF_UNICODE_ALL;
703 cl->flags |= ANYOF_LOCALE;
706 /* Can match anything (initialization) */
708 S_cl_is_anything(const struct regnode_charclass_class *cl)
712 PERL_ARGS_ASSERT_CL_IS_ANYTHING;
714 for (value = 0; value <= ANYOF_MAX; value += 2)
715 if (ANYOF_CLASS_TEST(cl, value) && ANYOF_CLASS_TEST(cl, value + 1))
717 if (!(cl->flags & ANYOF_UNICODE_ALL))
719 if (!ANYOF_BITMAP_TESTALLSET((const void*)cl))
724 /* Can match anything (initialization) */
726 S_cl_init(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
728 PERL_ARGS_ASSERT_CL_INIT;
730 Zero(cl, 1, struct regnode_charclass_class);
732 cl_anything(pRExC_state, cl);
736 S_cl_init_zero(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
738 PERL_ARGS_ASSERT_CL_INIT_ZERO;
740 Zero(cl, 1, struct regnode_charclass_class);
742 cl_anything(pRExC_state, cl);
744 cl->flags |= ANYOF_LOCALE;
747 /* 'And' a given class with another one. Can create false positives */
748 /* We assume that cl is not inverted */
750 S_cl_and(struct regnode_charclass_class *cl,
751 const struct regnode_charclass_class *and_with)
753 PERL_ARGS_ASSERT_CL_AND;
755 assert(and_with->type == ANYOF);
756 if (!(and_with->flags & ANYOF_CLASS)
757 && !(cl->flags & ANYOF_CLASS)
758 && (and_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
759 && !(and_with->flags & ANYOF_FOLD)
760 && !(cl->flags & ANYOF_FOLD)) {
763 if (and_with->flags & ANYOF_INVERT)
764 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
765 cl->bitmap[i] &= ~and_with->bitmap[i];
767 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
768 cl->bitmap[i] &= and_with->bitmap[i];
769 } /* XXXX: logic is complicated otherwise, leave it along for a moment. */
770 if (!(and_with->flags & ANYOF_EOS))
771 cl->flags &= ~ANYOF_EOS;
773 if (cl->flags & ANYOF_UNICODE_ALL && and_with->flags & ANYOF_UNICODE &&
774 !(and_with->flags & ANYOF_INVERT)) {
775 cl->flags &= ~ANYOF_UNICODE_ALL;
776 cl->flags |= ANYOF_UNICODE;
777 ARG_SET(cl, ARG(and_with));
779 if (!(and_with->flags & ANYOF_UNICODE_ALL) &&
780 !(and_with->flags & ANYOF_INVERT))
781 cl->flags &= ~ANYOF_UNICODE_ALL;
782 if (!(and_with->flags & (ANYOF_UNICODE|ANYOF_UNICODE_ALL)) &&
783 !(and_with->flags & ANYOF_INVERT))
784 cl->flags &= ~ANYOF_UNICODE;
787 /* 'OR' a given class with another one. Can create false positives */
788 /* We assume that cl is not inverted */
790 S_cl_or(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, const struct regnode_charclass_class *or_with)
792 PERL_ARGS_ASSERT_CL_OR;
794 if (or_with->flags & ANYOF_INVERT) {
796 * (B1 | CL1) | (!B2 & !CL2) = (B1 | !B2 & !CL2) | (CL1 | (!B2 & !CL2))
797 * <= (B1 | !B2) | (CL1 | !CL2)
798 * which is wasteful if CL2 is small, but we ignore CL2:
799 * (B1 | CL1) | (!B2 & !CL2) <= (B1 | CL1) | !B2 = (B1 | !B2) | CL1
800 * XXXX Can we handle case-fold? Unclear:
801 * (OK1(i) | OK1(i')) | !(OK1(i) | OK1(i')) =
802 * (OK1(i) | OK1(i')) | (!OK1(i) & !OK1(i'))
804 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
805 && !(or_with->flags & ANYOF_FOLD)
806 && !(cl->flags & ANYOF_FOLD) ) {
809 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
810 cl->bitmap[i] |= ~or_with->bitmap[i];
811 } /* XXXX: logic is complicated otherwise */
813 cl_anything(pRExC_state, cl);
816 /* (B1 | CL1) | (B2 | CL2) = (B1 | B2) | (CL1 | CL2)) */
817 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
818 && (!(or_with->flags & ANYOF_FOLD)
819 || (cl->flags & ANYOF_FOLD)) ) {
822 /* OR char bitmap and class bitmap separately */
823 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
824 cl->bitmap[i] |= or_with->bitmap[i];
825 if (or_with->flags & ANYOF_CLASS) {
826 for (i = 0; i < ANYOF_CLASSBITMAP_SIZE; i++)
827 cl->classflags[i] |= or_with->classflags[i];
828 cl->flags |= ANYOF_CLASS;
831 else { /* XXXX: logic is complicated, leave it along for a moment. */
832 cl_anything(pRExC_state, cl);
835 if (or_with->flags & ANYOF_EOS)
836 cl->flags |= ANYOF_EOS;
838 if (cl->flags & ANYOF_UNICODE && or_with->flags & ANYOF_UNICODE &&
839 ARG(cl) != ARG(or_with)) {
840 cl->flags |= ANYOF_UNICODE_ALL;
841 cl->flags &= ~ANYOF_UNICODE;
843 if (or_with->flags & ANYOF_UNICODE_ALL) {
844 cl->flags |= ANYOF_UNICODE_ALL;
845 cl->flags &= ~ANYOF_UNICODE;
849 #define TRIE_LIST_ITEM(state,idx) (trie->states[state].trans.list)[ idx ]
850 #define TRIE_LIST_CUR(state) ( TRIE_LIST_ITEM( state, 0 ).forid )
851 #define TRIE_LIST_LEN(state) ( TRIE_LIST_ITEM( state, 0 ).newstate )
852 #define TRIE_LIST_USED(idx) ( trie->states[state].trans.list ? (TRIE_LIST_CUR( idx ) - 1) : 0 )
857 dump_trie(trie,widecharmap,revcharmap)
858 dump_trie_interim_list(trie,widecharmap,revcharmap,next_alloc)
859 dump_trie_interim_table(trie,widecharmap,revcharmap,next_alloc)
861 These routines dump out a trie in a somewhat readable format.
862 The _interim_ variants are used for debugging the interim
863 tables that are used to generate the final compressed
864 representation which is what dump_trie expects.
866 Part of the reason for their existance is to provide a form
867 of documentation as to how the different representations function.
872 Dumps the final compressed table form of the trie to Perl_debug_log.
873 Used for debugging make_trie().
877 S_dump_trie(pTHX_ const struct _reg_trie_data *trie, HV *widecharmap,
878 AV *revcharmap, U32 depth)
881 SV *sv=sv_newmortal();
882 int colwidth= widecharmap ? 6 : 4;
883 GET_RE_DEBUG_FLAGS_DECL;
885 PERL_ARGS_ASSERT_DUMP_TRIE;
887 PerlIO_printf( Perl_debug_log, "%*sChar : %-6s%-6s%-4s ",
888 (int)depth * 2 + 2,"",
889 "Match","Base","Ofs" );
891 for( state = 0 ; state < trie->uniquecharcount ; state++ ) {
892 SV ** const tmp = av_fetch( revcharmap, state, 0);
894 PerlIO_printf( Perl_debug_log, "%*s",
896 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
897 PL_colors[0], PL_colors[1],
898 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
899 PERL_PV_ESCAPE_FIRSTCHAR
904 PerlIO_printf( Perl_debug_log, "\n%*sState|-----------------------",
905 (int)depth * 2 + 2,"");
907 for( state = 0 ; state < trie->uniquecharcount ; state++ )
908 PerlIO_printf( Perl_debug_log, "%.*s", colwidth, "--------");
909 PerlIO_printf( Perl_debug_log, "\n");
911 for( state = 1 ; state < trie->statecount ; state++ ) {
912 const U32 base = trie->states[ state ].trans.base;
914 PerlIO_printf( Perl_debug_log, "%*s#%4"UVXf"|", (int)depth * 2 + 2,"", (UV)state);
916 if ( trie->states[ state ].wordnum ) {
917 PerlIO_printf( Perl_debug_log, " W%4X", trie->states[ state ].wordnum );
919 PerlIO_printf( Perl_debug_log, "%6s", "" );
922 PerlIO_printf( Perl_debug_log, " @%4"UVXf" ", (UV)base );
927 while( ( base + ofs < trie->uniquecharcount ) ||
928 ( base + ofs - trie->uniquecharcount < trie->lasttrans
929 && trie->trans[ base + ofs - trie->uniquecharcount ].check != state))
932 PerlIO_printf( Perl_debug_log, "+%2"UVXf"[ ", (UV)ofs);
934 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
935 if ( ( base + ofs >= trie->uniquecharcount ) &&
936 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
937 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
939 PerlIO_printf( Perl_debug_log, "%*"UVXf,
941 (UV)trie->trans[ base + ofs - trie->uniquecharcount ].next );
943 PerlIO_printf( Perl_debug_log, "%*s",colwidth," ." );
947 PerlIO_printf( Perl_debug_log, "]");
950 PerlIO_printf( Perl_debug_log, "\n" );
954 Dumps a fully constructed but uncompressed trie in list form.
955 List tries normally only are used for construction when the number of
956 possible chars (trie->uniquecharcount) is very high.
957 Used for debugging make_trie().
960 S_dump_trie_interim_list(pTHX_ const struct _reg_trie_data *trie,
961 HV *widecharmap, AV *revcharmap, U32 next_alloc,
965 SV *sv=sv_newmortal();
966 int colwidth= widecharmap ? 6 : 4;
967 GET_RE_DEBUG_FLAGS_DECL;
969 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_LIST;
971 /* print out the table precompression. */
972 PerlIO_printf( Perl_debug_log, "%*sState :Word | Transition Data\n%*s%s",
973 (int)depth * 2 + 2,"", (int)depth * 2 + 2,"",
974 "------:-----+-----------------\n" );
976 for( state=1 ; state < next_alloc ; state ++ ) {
979 PerlIO_printf( Perl_debug_log, "%*s %4"UVXf" :",
980 (int)depth * 2 + 2,"", (UV)state );
981 if ( ! trie->states[ state ].wordnum ) {
982 PerlIO_printf( Perl_debug_log, "%5s| ","");
984 PerlIO_printf( Perl_debug_log, "W%4x| ",
985 trie->states[ state ].wordnum
988 for( charid = 1 ; charid <= TRIE_LIST_USED( state ) ; charid++ ) {
989 SV ** const tmp = av_fetch( revcharmap, TRIE_LIST_ITEM(state,charid).forid, 0);
991 PerlIO_printf( Perl_debug_log, "%*s:%3X=%4"UVXf" | ",
993 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
994 PL_colors[0], PL_colors[1],
995 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
996 PERL_PV_ESCAPE_FIRSTCHAR
998 TRIE_LIST_ITEM(state,charid).forid,
999 (UV)TRIE_LIST_ITEM(state,charid).newstate
1002 PerlIO_printf(Perl_debug_log, "\n%*s| ",
1003 (int)((depth * 2) + 14), "");
1006 PerlIO_printf( Perl_debug_log, "\n");
1011 Dumps a fully constructed but uncompressed trie in table form.
1012 This is the normal DFA style state transition table, with a few
1013 twists to facilitate compression later.
1014 Used for debugging make_trie().
1017 S_dump_trie_interim_table(pTHX_ const struct _reg_trie_data *trie,
1018 HV *widecharmap, AV *revcharmap, U32 next_alloc,
1023 SV *sv=sv_newmortal();
1024 int colwidth= widecharmap ? 6 : 4;
1025 GET_RE_DEBUG_FLAGS_DECL;
1027 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_TABLE;
1030 print out the table precompression so that we can do a visual check
1031 that they are identical.
1034 PerlIO_printf( Perl_debug_log, "%*sChar : ",(int)depth * 2 + 2,"" );
1036 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1037 SV ** const tmp = av_fetch( revcharmap, charid, 0);
1039 PerlIO_printf( Perl_debug_log, "%*s",
1041 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1042 PL_colors[0], PL_colors[1],
1043 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1044 PERL_PV_ESCAPE_FIRSTCHAR
1050 PerlIO_printf( Perl_debug_log, "\n%*sState+-",(int)depth * 2 + 2,"" );
1052 for( charid=0 ; charid < trie->uniquecharcount ; charid++ ) {
1053 PerlIO_printf( Perl_debug_log, "%.*s", colwidth,"--------");
1056 PerlIO_printf( Perl_debug_log, "\n" );
1058 for( state=1 ; state < next_alloc ; state += trie->uniquecharcount ) {
1060 PerlIO_printf( Perl_debug_log, "%*s%4"UVXf" : ",
1061 (int)depth * 2 + 2,"",
1062 (UV)TRIE_NODENUM( state ) );
1064 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1065 UV v=(UV)SAFE_TRIE_NODENUM( trie->trans[ state + charid ].next );
1067 PerlIO_printf( Perl_debug_log, "%*"UVXf, colwidth, v );
1069 PerlIO_printf( Perl_debug_log, "%*s", colwidth, "." );
1071 if ( ! trie->states[ TRIE_NODENUM( state ) ].wordnum ) {
1072 PerlIO_printf( Perl_debug_log, " (%4"UVXf")\n", (UV)trie->trans[ state ].check );
1074 PerlIO_printf( Perl_debug_log, " (%4"UVXf") W%4X\n", (UV)trie->trans[ state ].check,
1075 trie->states[ TRIE_NODENUM( state ) ].wordnum );
1082 /* make_trie(startbranch,first,last,tail,word_count,flags,depth)
1083 startbranch: the first branch in the whole branch sequence
1084 first : start branch of sequence of branch-exact nodes.
1085 May be the same as startbranch
1086 last : Thing following the last branch.
1087 May be the same as tail.
1088 tail : item following the branch sequence
1089 count : words in the sequence
1090 flags : currently the OP() type we will be building one of /EXACT(|F|Fl)/
1091 depth : indent depth
1093 Inplace optimizes a sequence of 2 or more Branch-Exact nodes into a TRIE node.
1095 A trie is an N'ary tree where the branches are determined by digital
1096 decomposition of the key. IE, at the root node you look up the 1st character and
1097 follow that branch repeat until you find the end of the branches. Nodes can be
1098 marked as "accepting" meaning they represent a complete word. Eg:
1102 would convert into the following structure. Numbers represent states, letters
1103 following numbers represent valid transitions on the letter from that state, if
1104 the number is in square brackets it represents an accepting state, otherwise it
1105 will be in parenthesis.
1107 +-h->+-e->[3]-+-r->(8)-+-s->[9]
1111 (1) +-i->(6)-+-s->[7]
1113 +-s->(3)-+-h->(4)-+-e->[5]
1115 Accept Word Mapping: 3=>1 (he),5=>2 (she), 7=>3 (his), 9=>4 (hers)
1117 This shows that when matching against the string 'hers' we will begin at state 1
1118 read 'h' and move to state 2, read 'e' and move to state 3 which is accepting,
1119 then read 'r' and go to state 8 followed by 's' which takes us to state 9 which
1120 is also accepting. Thus we know that we can match both 'he' and 'hers' with a
1121 single traverse. We store a mapping from accepting to state to which word was
1122 matched, and then when we have multiple possibilities we try to complete the
1123 rest of the regex in the order in which they occured in the alternation.
1125 The only prior NFA like behaviour that would be changed by the TRIE support is
1126 the silent ignoring of duplicate alternations which are of the form:
1128 / (DUPE|DUPE) X? (?{ ... }) Y /x
1130 Thus EVAL blocks follwing a trie may be called a different number of times with
1131 and without the optimisation. With the optimisations dupes will be silently
1132 ignored. This inconsistant behaviour of EVAL type nodes is well established as
1133 the following demonstrates:
1135 'words'=~/(word|word|word)(?{ print $1 })[xyz]/
1137 which prints out 'word' three times, but
1139 'words'=~/(word|word|word)(?{ print $1 })S/
1141 which doesnt print it out at all. This is due to other optimisations kicking in.
1143 Example of what happens on a structural level:
1145 The regexp /(ac|ad|ab)+/ will produce the folowing debug output:
1147 1: CURLYM[1] {1,32767}(18)
1158 This would be optimizable with startbranch=5, first=5, last=16, tail=16
1159 and should turn into:
1161 1: CURLYM[1] {1,32767}(18)
1163 [Words:3 Chars Stored:6 Unique Chars:4 States:5 NCP:1]
1171 Cases where tail != last would be like /(?foo|bar)baz/:
1181 which would be optimizable with startbranch=1, first=1, last=7, tail=8
1182 and would end up looking like:
1185 [Words:2 Chars Stored:6 Unique Chars:5 States:7 NCP:1]
1192 d = uvuni_to_utf8_flags(d, uv, 0);
1194 is the recommended Unicode-aware way of saying
1199 #define TRIE_STORE_REVCHAR \
1202 SV *zlopp = newSV(2); \
1203 unsigned char *flrbbbbb = (unsigned char *) SvPVX(zlopp); \
1204 unsigned const char *const kapow = uvuni_to_utf8(flrbbbbb, uvc & 0xFF); \
1205 SvCUR_set(zlopp, kapow - flrbbbbb); \
1208 av_push(revcharmap, zlopp); \
1210 char ooooff = (char)uvc; \
1211 av_push(revcharmap, newSVpvn(&ooooff, 1)); \
1215 #define TRIE_READ_CHAR STMT_START { \
1219 if ( foldlen > 0 ) { \
1220 uvc = utf8n_to_uvuni( scan, UTF8_MAXLEN, &len, uniflags ); \
1225 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1226 uvc = to_uni_fold( uvc, foldbuf, &foldlen ); \
1227 foldlen -= UNISKIP( uvc ); \
1228 scan = foldbuf + UNISKIP( uvc ); \
1231 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1241 #define TRIE_LIST_PUSH(state,fid,ns) STMT_START { \
1242 if ( TRIE_LIST_CUR( state ) >=TRIE_LIST_LEN( state ) ) { \
1243 U32 ging = TRIE_LIST_LEN( state ) *= 2; \
1244 Renew( trie->states[ state ].trans.list, ging, reg_trie_trans_le ); \
1246 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).forid = fid; \
1247 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).newstate = ns; \
1248 TRIE_LIST_CUR( state )++; \
1251 #define TRIE_LIST_NEW(state) STMT_START { \
1252 Newxz( trie->states[ state ].trans.list, \
1253 4, reg_trie_trans_le ); \
1254 TRIE_LIST_CUR( state ) = 1; \
1255 TRIE_LIST_LEN( state ) = 4; \
1258 #define TRIE_HANDLE_WORD(state) STMT_START { \
1259 U16 dupe= trie->states[ state ].wordnum; \
1260 regnode * const noper_next = regnext( noper ); \
1262 if (trie->wordlen) \
1263 trie->wordlen[ curword ] = wordlen; \
1265 /* store the word for dumping */ \
1267 if (OP(noper) != NOTHING) \
1268 tmp = newSVpvn_utf8(STRING(noper), STR_LEN(noper), UTF); \
1270 tmp = newSVpvn_utf8( "", 0, UTF ); \
1271 av_push( trie_words, tmp ); \
1276 if ( noper_next < tail ) { \
1278 trie->jump = (U16 *) PerlMemShared_calloc( word_count + 1, sizeof(U16) ); \
1279 trie->jump[curword] = (U16)(noper_next - convert); \
1281 jumper = noper_next; \
1283 nextbranch= regnext(cur); \
1287 /* So it's a dupe. This means we need to maintain a */\
1288 /* linked-list from the first to the next. */\
1289 /* we only allocate the nextword buffer when there */\
1290 /* a dupe, so first time we have to do the allocation */\
1291 if (!trie->nextword) \
1292 trie->nextword = (U16 *) \
1293 PerlMemShared_calloc( word_count + 1, sizeof(U16)); \
1294 while ( trie->nextword[dupe] ) \
1295 dupe= trie->nextword[dupe]; \
1296 trie->nextword[dupe]= curword; \
1298 /* we haven't inserted this word yet. */ \
1299 trie->states[ state ].wordnum = curword; \
1304 #define TRIE_TRANS_STATE(state,base,ucharcount,charid,special) \
1305 ( ( base + charid >= ucharcount \
1306 && base + charid < ubound \
1307 && state == trie->trans[ base - ucharcount + charid ].check \
1308 && trie->trans[ base - ucharcount + charid ].next ) \
1309 ? trie->trans[ base - ucharcount + charid ].next \
1310 : ( state==1 ? special : 0 ) \
1314 #define MADE_JUMP_TRIE 2
1315 #define MADE_EXACT_TRIE 4
1318 S_make_trie(pTHX_ RExC_state_t *pRExC_state, regnode *startbranch, regnode *first, regnode *last, regnode *tail, U32 word_count, U32 flags, U32 depth)
1321 /* first pass, loop through and scan words */
1322 reg_trie_data *trie;
1323 HV *widecharmap = NULL;
1324 AV *revcharmap = newAV();
1326 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1331 regnode *jumper = NULL;
1332 regnode *nextbranch = NULL;
1333 regnode *convert = NULL;
1334 /* we just use folder as a flag in utf8 */
1335 const U8 * const folder = ( flags == EXACTF
1337 : ( flags == EXACTFL
1344 const U32 data_slot = add_data( pRExC_state, 4, "tuuu" );
1345 AV *trie_words = NULL;
1346 /* along with revcharmap, this only used during construction but both are
1347 * useful during debugging so we store them in the struct when debugging.
1350 const U32 data_slot = add_data( pRExC_state, 2, "tu" );
1351 STRLEN trie_charcount=0;
1353 SV *re_trie_maxbuff;
1354 GET_RE_DEBUG_FLAGS_DECL;
1356 PERL_ARGS_ASSERT_MAKE_TRIE;
1358 PERL_UNUSED_ARG(depth);
1361 trie = (reg_trie_data *) PerlMemShared_calloc( 1, sizeof(reg_trie_data) );
1363 trie->startstate = 1;
1364 trie->wordcount = word_count;
1365 RExC_rxi->data->data[ data_slot ] = (void*)trie;
1366 trie->charmap = (U16 *) PerlMemShared_calloc( 256, sizeof(U16) );
1367 if (!(UTF && folder))
1368 trie->bitmap = (char *) PerlMemShared_calloc( ANYOF_BITMAP_SIZE, 1 );
1370 trie_words = newAV();
1373 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
1374 if (!SvIOK(re_trie_maxbuff)) {
1375 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
1378 PerlIO_printf( Perl_debug_log,
1379 "%*smake_trie start==%d, first==%d, last==%d, tail==%d depth=%d\n",
1380 (int)depth * 2 + 2, "",
1381 REG_NODE_NUM(startbranch),REG_NODE_NUM(first),
1382 REG_NODE_NUM(last), REG_NODE_NUM(tail),
1386 /* Find the node we are going to overwrite */
1387 if ( first == startbranch && OP( last ) != BRANCH ) {
1388 /* whole branch chain */
1391 /* branch sub-chain */
1392 convert = NEXTOPER( first );
1395 /* -- First loop and Setup --
1397 We first traverse the branches and scan each word to determine if it
1398 contains widechars, and how many unique chars there are, this is
1399 important as we have to build a table with at least as many columns as we
1402 We use an array of integers to represent the character codes 0..255
1403 (trie->charmap) and we use a an HV* to store Unicode characters. We use the
1404 native representation of the character value as the key and IV's for the
1407 *TODO* If we keep track of how many times each character is used we can
1408 remap the columns so that the table compression later on is more
1409 efficient in terms of memory by ensuring most common value is in the
1410 middle and the least common are on the outside. IMO this would be better
1411 than a most to least common mapping as theres a decent chance the most
1412 common letter will share a node with the least common, meaning the node
1413 will not be compressable. With a middle is most common approach the worst
1414 case is when we have the least common nodes twice.
1418 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1419 regnode * const noper = NEXTOPER( cur );
1420 const U8 *uc = (U8*)STRING( noper );
1421 const U8 * const e = uc + STR_LEN( noper );
1423 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1424 const U8 *scan = (U8*)NULL;
1425 U32 wordlen = 0; /* required init */
1427 bool set_bit = trie->bitmap ? 1 : 0; /*store the first char in the bitmap?*/
1429 if (OP(noper) == NOTHING) {
1433 if ( set_bit ) /* bitmap only alloced when !(UTF&&Folding) */
1434 TRIE_BITMAP_SET(trie,*uc); /* store the raw first byte
1435 regardless of encoding */
1437 for ( ; uc < e ; uc += len ) {
1438 TRIE_CHARCOUNT(trie)++;
1442 if ( !trie->charmap[ uvc ] ) {
1443 trie->charmap[ uvc ]=( ++trie->uniquecharcount );
1445 trie->charmap[ folder[ uvc ] ] = trie->charmap[ uvc ];
1449 /* store the codepoint in the bitmap, and if its ascii
1450 also store its folded equivelent. */
1451 TRIE_BITMAP_SET(trie,uvc);
1453 /* store the folded codepoint */
1454 if ( folder ) TRIE_BITMAP_SET(trie,folder[ uvc ]);
1457 /* store first byte of utf8 representation of
1458 codepoints in the 127 < uvc < 256 range */
1459 if (127 < uvc && uvc < 192) {
1460 TRIE_BITMAP_SET(trie,194);
1461 } else if (191 < uvc ) {
1462 TRIE_BITMAP_SET(trie,195);
1463 /* && uvc < 256 -- we know uvc is < 256 already */
1466 set_bit = 0; /* We've done our bit :-) */
1471 widecharmap = newHV();
1473 svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 1 );
1476 Perl_croak( aTHX_ "error creating/fetching widecharmap entry for 0x%"UVXf, uvc );
1478 if ( !SvTRUE( *svpp ) ) {
1479 sv_setiv( *svpp, ++trie->uniquecharcount );
1484 if( cur == first ) {
1487 } else if (chars < trie->minlen) {
1489 } else if (chars > trie->maxlen) {
1493 } /* end first pass */
1494 DEBUG_TRIE_COMPILE_r(
1495 PerlIO_printf( Perl_debug_log, "%*sTRIE(%s): W:%d C:%d Uq:%d Min:%d Max:%d\n",
1496 (int)depth * 2 + 2,"",
1497 ( widecharmap ? "UTF8" : "NATIVE" ), (int)word_count,
1498 (int)TRIE_CHARCOUNT(trie), trie->uniquecharcount,
1499 (int)trie->minlen, (int)trie->maxlen )
1501 trie->wordlen = (U32 *) PerlMemShared_calloc( word_count, sizeof(U32) );
1504 We now know what we are dealing with in terms of unique chars and
1505 string sizes so we can calculate how much memory a naive
1506 representation using a flat table will take. If it's over a reasonable
1507 limit (as specified by ${^RE_TRIE_MAXBUF}) we use a more memory
1508 conservative but potentially much slower representation using an array
1511 At the end we convert both representations into the same compressed
1512 form that will be used in regexec.c for matching with. The latter
1513 is a form that cannot be used to construct with but has memory
1514 properties similar to the list form and access properties similar
1515 to the table form making it both suitable for fast searches and
1516 small enough that its feasable to store for the duration of a program.
1518 See the comment in the code where the compressed table is produced
1519 inplace from the flat tabe representation for an explanation of how
1520 the compression works.
1525 if ( (IV)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1) > SvIV(re_trie_maxbuff) ) {
1527 Second Pass -- Array Of Lists Representation
1529 Each state will be represented by a list of charid:state records
1530 (reg_trie_trans_le) the first such element holds the CUR and LEN
1531 points of the allocated array. (See defines above).
1533 We build the initial structure using the lists, and then convert
1534 it into the compressed table form which allows faster lookups
1535 (but cant be modified once converted).
1538 STRLEN transcount = 1;
1540 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1541 "%*sCompiling trie using list compiler\n",
1542 (int)depth * 2 + 2, ""));
1544 trie->states = (reg_trie_state *)
1545 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1546 sizeof(reg_trie_state) );
1550 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1552 regnode * const noper = NEXTOPER( cur );
1553 U8 *uc = (U8*)STRING( noper );
1554 const U8 * const e = uc + STR_LEN( noper );
1555 U32 state = 1; /* required init */
1556 U16 charid = 0; /* sanity init */
1557 U8 *scan = (U8*)NULL; /* sanity init */
1558 STRLEN foldlen = 0; /* required init */
1559 U32 wordlen = 0; /* required init */
1560 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1562 if (OP(noper) != NOTHING) {
1563 for ( ; uc < e ; uc += len ) {
1568 charid = trie->charmap[ uvc ];
1570 SV** const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1574 charid=(U16)SvIV( *svpp );
1577 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1584 if ( !trie->states[ state ].trans.list ) {
1585 TRIE_LIST_NEW( state );
1587 for ( check = 1; check <= TRIE_LIST_USED( state ); check++ ) {
1588 if ( TRIE_LIST_ITEM( state, check ).forid == charid ) {
1589 newstate = TRIE_LIST_ITEM( state, check ).newstate;
1594 newstate = next_alloc++;
1595 TRIE_LIST_PUSH( state, charid, newstate );
1600 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1604 TRIE_HANDLE_WORD(state);
1606 } /* end second pass */
1608 /* next alloc is the NEXT state to be allocated */
1609 trie->statecount = next_alloc;
1610 trie->states = (reg_trie_state *)
1611 PerlMemShared_realloc( trie->states,
1613 * sizeof(reg_trie_state) );
1615 /* and now dump it out before we compress it */
1616 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_list(trie, widecharmap,
1617 revcharmap, next_alloc,
1621 trie->trans = (reg_trie_trans *)
1622 PerlMemShared_calloc( transcount, sizeof(reg_trie_trans) );
1629 for( state=1 ; state < next_alloc ; state ++ ) {
1633 DEBUG_TRIE_COMPILE_MORE_r(
1634 PerlIO_printf( Perl_debug_log, "tp: %d zp: %d ",tp,zp)
1638 if (trie->states[state].trans.list) {
1639 U16 minid=TRIE_LIST_ITEM( state, 1).forid;
1643 for( idx = 2 ; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1644 const U16 forid = TRIE_LIST_ITEM( state, idx).forid;
1645 if ( forid < minid ) {
1647 } else if ( forid > maxid ) {
1651 if ( transcount < tp + maxid - minid + 1) {
1653 trie->trans = (reg_trie_trans *)
1654 PerlMemShared_realloc( trie->trans,
1656 * sizeof(reg_trie_trans) );
1657 Zero( trie->trans + (transcount / 2), transcount / 2 , reg_trie_trans );
1659 base = trie->uniquecharcount + tp - minid;
1660 if ( maxid == minid ) {
1662 for ( ; zp < tp ; zp++ ) {
1663 if ( ! trie->trans[ zp ].next ) {
1664 base = trie->uniquecharcount + zp - minid;
1665 trie->trans[ zp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1666 trie->trans[ zp ].check = state;
1672 trie->trans[ tp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1673 trie->trans[ tp ].check = state;
1678 for ( idx=1; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1679 const U32 tid = base - trie->uniquecharcount + TRIE_LIST_ITEM( state, idx ).forid;
1680 trie->trans[ tid ].next = TRIE_LIST_ITEM( state, idx ).newstate;
1681 trie->trans[ tid ].check = state;
1683 tp += ( maxid - minid + 1 );
1685 Safefree(trie->states[ state ].trans.list);
1688 DEBUG_TRIE_COMPILE_MORE_r(
1689 PerlIO_printf( Perl_debug_log, " base: %d\n",base);
1692 trie->states[ state ].trans.base=base;
1694 trie->lasttrans = tp + 1;
1698 Second Pass -- Flat Table Representation.
1700 we dont use the 0 slot of either trans[] or states[] so we add 1 to each.
1701 We know that we will need Charcount+1 trans at most to store the data
1702 (one row per char at worst case) So we preallocate both structures
1703 assuming worst case.
1705 We then construct the trie using only the .next slots of the entry
1708 We use the .check field of the first entry of the node temporarily to
1709 make compression both faster and easier by keeping track of how many non
1710 zero fields are in the node.
1712 Since trans are numbered from 1 any 0 pointer in the table is a FAIL
1715 There are two terms at use here: state as a TRIE_NODEIDX() which is a
1716 number representing the first entry of the node, and state as a
1717 TRIE_NODENUM() which is the trans number. state 1 is TRIE_NODEIDX(1) and
1718 TRIE_NODENUM(1), state 2 is TRIE_NODEIDX(2) and TRIE_NODENUM(3) if there
1719 are 2 entrys per node. eg:
1727 The table is internally in the right hand, idx form. However as we also
1728 have to deal with the states array which is indexed by nodenum we have to
1729 use TRIE_NODENUM() to convert.
1732 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1733 "%*sCompiling trie using table compiler\n",
1734 (int)depth * 2 + 2, ""));
1736 trie->trans = (reg_trie_trans *)
1737 PerlMemShared_calloc( ( TRIE_CHARCOUNT(trie) + 1 )
1738 * trie->uniquecharcount + 1,
1739 sizeof(reg_trie_trans) );
1740 trie->states = (reg_trie_state *)
1741 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1742 sizeof(reg_trie_state) );
1743 next_alloc = trie->uniquecharcount + 1;
1746 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1748 regnode * const noper = NEXTOPER( cur );
1749 const U8 *uc = (U8*)STRING( noper );
1750 const U8 * const e = uc + STR_LEN( noper );
1752 U32 state = 1; /* required init */
1754 U16 charid = 0; /* sanity init */
1755 U32 accept_state = 0; /* sanity init */
1756 U8 *scan = (U8*)NULL; /* sanity init */
1758 STRLEN foldlen = 0; /* required init */
1759 U32 wordlen = 0; /* required init */
1760 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1762 if ( OP(noper) != NOTHING ) {
1763 for ( ; uc < e ; uc += len ) {
1768 charid = trie->charmap[ uvc ];
1770 SV* const * const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1771 charid = svpp ? (U16)SvIV(*svpp) : 0;
1775 if ( !trie->trans[ state + charid ].next ) {
1776 trie->trans[ state + charid ].next = next_alloc;
1777 trie->trans[ state ].check++;
1778 next_alloc += trie->uniquecharcount;
1780 state = trie->trans[ state + charid ].next;
1782 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1784 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1787 accept_state = TRIE_NODENUM( state );
1788 TRIE_HANDLE_WORD(accept_state);
1790 } /* end second pass */
1792 /* and now dump it out before we compress it */
1793 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_table(trie, widecharmap,
1795 next_alloc, depth+1));
1799 * Inplace compress the table.*
1801 For sparse data sets the table constructed by the trie algorithm will
1802 be mostly 0/FAIL transitions or to put it another way mostly empty.
1803 (Note that leaf nodes will not contain any transitions.)
1805 This algorithm compresses the tables by eliminating most such
1806 transitions, at the cost of a modest bit of extra work during lookup:
1808 - Each states[] entry contains a .base field which indicates the
1809 index in the state[] array wheres its transition data is stored.
1811 - If .base is 0 there are no valid transitions from that node.
1813 - If .base is nonzero then charid is added to it to find an entry in
1816 -If trans[states[state].base+charid].check!=state then the
1817 transition is taken to be a 0/Fail transition. Thus if there are fail
1818 transitions at the front of the node then the .base offset will point
1819 somewhere inside the previous nodes data (or maybe even into a node
1820 even earlier), but the .check field determines if the transition is
1824 The following process inplace converts the table to the compressed
1825 table: We first do not compress the root node 1,and mark its all its
1826 .check pointers as 1 and set its .base pointer as 1 as well. This
1827 allows to do a DFA construction from the compressed table later, and
1828 ensures that any .base pointers we calculate later are greater than
1831 - We set 'pos' to indicate the first entry of the second node.
1833 - We then iterate over the columns of the node, finding the first and
1834 last used entry at l and m. We then copy l..m into pos..(pos+m-l),
1835 and set the .check pointers accordingly, and advance pos
1836 appropriately and repreat for the next node. Note that when we copy
1837 the next pointers we have to convert them from the original
1838 NODEIDX form to NODENUM form as the former is not valid post
1841 - If a node has no transitions used we mark its base as 0 and do not
1842 advance the pos pointer.
1844 - If a node only has one transition we use a second pointer into the
1845 structure to fill in allocated fail transitions from other states.
1846 This pointer is independent of the main pointer and scans forward
1847 looking for null transitions that are allocated to a state. When it
1848 finds one it writes the single transition into the "hole". If the
1849 pointer doesnt find one the single transition is appended as normal.
1851 - Once compressed we can Renew/realloc the structures to release the
1854 See "Table-Compression Methods" in sec 3.9 of the Red Dragon,
1855 specifically Fig 3.47 and the associated pseudocode.
1859 const U32 laststate = TRIE_NODENUM( next_alloc );
1862 trie->statecount = laststate;
1864 for ( state = 1 ; state < laststate ; state++ ) {
1866 const U32 stateidx = TRIE_NODEIDX( state );
1867 const U32 o_used = trie->trans[ stateidx ].check;
1868 U32 used = trie->trans[ stateidx ].check;
1869 trie->trans[ stateidx ].check = 0;
1871 for ( charid = 0 ; used && charid < trie->uniquecharcount ; charid++ ) {
1872 if ( flag || trie->trans[ stateidx + charid ].next ) {
1873 if ( trie->trans[ stateidx + charid ].next ) {
1875 for ( ; zp < pos ; zp++ ) {
1876 if ( ! trie->trans[ zp ].next ) {
1880 trie->states[ state ].trans.base = zp + trie->uniquecharcount - charid ;
1881 trie->trans[ zp ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1882 trie->trans[ zp ].check = state;
1883 if ( ++zp > pos ) pos = zp;
1890 trie->states[ state ].trans.base = pos + trie->uniquecharcount - charid ;
1892 trie->trans[ pos ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
1893 trie->trans[ pos ].check = state;
1898 trie->lasttrans = pos + 1;
1899 trie->states = (reg_trie_state *)
1900 PerlMemShared_realloc( trie->states, laststate
1901 * sizeof(reg_trie_state) );
1902 DEBUG_TRIE_COMPILE_MORE_r(
1903 PerlIO_printf( Perl_debug_log,
1904 "%*sAlloc: %d Orig: %"IVdf" elements, Final:%"IVdf". Savings of %%%5.2f\n",
1905 (int)depth * 2 + 2,"",
1906 (int)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1 ),
1909 ( ( next_alloc - pos ) * 100 ) / (double)next_alloc );
1912 } /* end table compress */
1914 DEBUG_TRIE_COMPILE_MORE_r(
1915 PerlIO_printf(Perl_debug_log, "%*sStatecount:%"UVxf" Lasttrans:%"UVxf"\n",
1916 (int)depth * 2 + 2, "",
1917 (UV)trie->statecount,
1918 (UV)trie->lasttrans)
1920 /* resize the trans array to remove unused space */
1921 trie->trans = (reg_trie_trans *)
1922 PerlMemShared_realloc( trie->trans, trie->lasttrans
1923 * sizeof(reg_trie_trans) );
1925 /* and now dump out the compressed format */
1926 DEBUG_TRIE_COMPILE_r(dump_trie(trie, widecharmap, revcharmap, depth+1));
1928 { /* Modify the program and insert the new TRIE node*/
1929 U8 nodetype =(U8)(flags & 0xFF);
1933 regnode *optimize = NULL;
1934 #ifdef RE_TRACK_PATTERN_OFFSETS
1937 U32 mjd_nodelen = 0;
1938 #endif /* RE_TRACK_PATTERN_OFFSETS */
1939 #endif /* DEBUGGING */
1941 This means we convert either the first branch or the first Exact,
1942 depending on whether the thing following (in 'last') is a branch
1943 or not and whther first is the startbranch (ie is it a sub part of
1944 the alternation or is it the whole thing.)
1945 Assuming its a sub part we conver the EXACT otherwise we convert
1946 the whole branch sequence, including the first.
1948 /* Find the node we are going to overwrite */
1949 if ( first != startbranch || OP( last ) == BRANCH ) {
1950 /* branch sub-chain */
1951 NEXT_OFF( first ) = (U16)(last - first);
1952 #ifdef RE_TRACK_PATTERN_OFFSETS
1954 mjd_offset= Node_Offset((convert));
1955 mjd_nodelen= Node_Length((convert));
1958 /* whole branch chain */
1960 #ifdef RE_TRACK_PATTERN_OFFSETS
1963 const regnode *nop = NEXTOPER( convert );
1964 mjd_offset= Node_Offset((nop));
1965 mjd_nodelen= Node_Length((nop));
1969 PerlIO_printf(Perl_debug_log, "%*sMJD offset:%"UVuf" MJD length:%"UVuf"\n",
1970 (int)depth * 2 + 2, "",
1971 (UV)mjd_offset, (UV)mjd_nodelen)
1974 /* But first we check to see if there is a common prefix we can
1975 split out as an EXACT and put in front of the TRIE node. */
1976 trie->startstate= 1;
1977 if ( trie->bitmap && !widecharmap && !trie->jump ) {
1979 for ( state = 1 ; state < trie->statecount-1 ; state++ ) {
1983 const U32 base = trie->states[ state ].trans.base;
1985 if ( trie->states[state].wordnum )
1988 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
1989 if ( ( base + ofs >= trie->uniquecharcount ) &&
1990 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
1991 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
1993 if ( ++count > 1 ) {
1994 SV **tmp = av_fetch( revcharmap, ofs, 0);
1995 const U8 *ch = (U8*)SvPV_nolen_const( *tmp );
1996 if ( state == 1 ) break;
1998 Zero(trie->bitmap, ANYOF_BITMAP_SIZE, char);
2000 PerlIO_printf(Perl_debug_log,
2001 "%*sNew Start State=%"UVuf" Class: [",
2002 (int)depth * 2 + 2, "",
2005 SV ** const tmp = av_fetch( revcharmap, idx, 0);
2006 const U8 * const ch = (U8*)SvPV_nolen_const( *tmp );
2008 TRIE_BITMAP_SET(trie,*ch);
2010 TRIE_BITMAP_SET(trie, folder[ *ch ]);
2012 PerlIO_printf(Perl_debug_log, "%s", (char*)ch)
2016 TRIE_BITMAP_SET(trie,*ch);
2018 TRIE_BITMAP_SET(trie,folder[ *ch ]);
2019 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"%s", ch));
2025 SV **tmp = av_fetch( revcharmap, idx, 0);
2027 char *ch = SvPV( *tmp, len );
2029 SV *sv=sv_newmortal();
2030 PerlIO_printf( Perl_debug_log,
2031 "%*sPrefix State: %"UVuf" Idx:%"UVuf" Char='%s'\n",
2032 (int)depth * 2 + 2, "",
2034 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 6,
2035 PL_colors[0], PL_colors[1],
2036 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
2037 PERL_PV_ESCAPE_FIRSTCHAR
2042 OP( convert ) = nodetype;
2043 str=STRING(convert);
2046 STR_LEN(convert) += len;
2052 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"]\n"));
2058 regnode *n = convert+NODE_SZ_STR(convert);
2059 NEXT_OFF(convert) = NODE_SZ_STR(convert);
2060 trie->startstate = state;
2061 trie->minlen -= (state - 1);
2062 trie->maxlen -= (state - 1);
2064 /* At least the UNICOS C compiler choked on this
2065 * being argument to DEBUG_r(), so let's just have
2068 #ifdef PERL_EXT_RE_BUILD
2074 regnode *fix = convert;
2075 U32 word = trie->wordcount;
2077 Set_Node_Offset_Length(convert, mjd_offset, state - 1);
2078 while( ++fix < n ) {
2079 Set_Node_Offset_Length(fix, 0, 0);
2082 SV ** const tmp = av_fetch( trie_words, word, 0 );
2084 if ( STR_LEN(convert) <= SvCUR(*tmp) )
2085 sv_chop(*tmp, SvPV_nolen(*tmp) + STR_LEN(convert));
2087 sv_chop(*tmp, SvPV_nolen(*tmp) + SvCUR(*tmp));
2095 NEXT_OFF(convert) = (U16)(tail - convert);
2096 DEBUG_r(optimize= n);
2102 if ( trie->maxlen ) {
2103 NEXT_OFF( convert ) = (U16)(tail - convert);
2104 ARG_SET( convert, data_slot );
2105 /* Store the offset to the first unabsorbed branch in
2106 jump[0], which is otherwise unused by the jump logic.
2107 We use this when dumping a trie and during optimisation. */
2109 trie->jump[0] = (U16)(nextbranch - convert);
2112 if ( !trie->states[trie->startstate].wordnum && trie->bitmap &&
2113 ( (char *)jumper - (char *)convert) >= (int)sizeof(struct regnode_charclass) )
2115 OP( convert ) = TRIEC;
2116 Copy(trie->bitmap, ((struct regnode_charclass *)convert)->bitmap, ANYOF_BITMAP_SIZE, char);
2117 PerlMemShared_free(trie->bitmap);
2120 OP( convert ) = TRIE;
2122 /* store the type in the flags */
2123 convert->flags = nodetype;
2127 + regarglen[ OP( convert ) ];
2129 /* XXX We really should free up the resource in trie now,
2130 as we won't use them - (which resources?) dmq */
2132 /* needed for dumping*/
2133 DEBUG_r(if (optimize) {
2134 regnode *opt = convert;
2136 while ( ++opt < optimize) {
2137 Set_Node_Offset_Length(opt,0,0);
2140 Try to clean up some of the debris left after the
2143 while( optimize < jumper ) {
2144 mjd_nodelen += Node_Length((optimize));
2145 OP( optimize ) = OPTIMIZED;
2146 Set_Node_Offset_Length(optimize,0,0);
2149 Set_Node_Offset_Length(convert,mjd_offset,mjd_nodelen);
2151 } /* end node insert */
2152 RExC_rxi->data->data[ data_slot + 1 ] = (void*)widecharmap;
2154 RExC_rxi->data->data[ data_slot + TRIE_WORDS_OFFSET ] = (void*)trie_words;
2155 RExC_rxi->data->data[ data_slot + 3 ] = (void*)revcharmap;
2157 SvREFCNT_dec(revcharmap);
2161 : trie->startstate>1
2167 S_make_trie_failtable(pTHX_ RExC_state_t *pRExC_state, regnode *source, regnode *stclass, U32 depth)
2169 /* The Trie is constructed and compressed now so we can build a fail array now if its needed
2171 This is basically the Aho-Corasick algorithm. Its from exercise 3.31 and 3.32 in the
2172 "Red Dragon" -- Compilers, principles, techniques, and tools. Aho, Sethi, Ullman 1985/88
2175 We find the fail state for each state in the trie, this state is the longest proper
2176 suffix of the current states 'word' that is also a proper prefix of another word in our
2177 trie. State 1 represents the word '' and is the thus the default fail state. This allows
2178 the DFA not to have to restart after its tried and failed a word at a given point, it
2179 simply continues as though it had been matching the other word in the first place.
2181 'abcdgu'=~/abcdefg|cdgu/
2182 When we get to 'd' we are still matching the first word, we would encounter 'g' which would
2183 fail, which would bring use to the state representing 'd' in the second word where we would
2184 try 'g' and succeed, prodceding to match 'cdgu'.
2186 /* add a fail transition */
2187 const U32 trie_offset = ARG(source);
2188 reg_trie_data *trie=(reg_trie_data *)RExC_rxi->data->data[trie_offset];
2190 const U32 ucharcount = trie->uniquecharcount;
2191 const U32 numstates = trie->statecount;
2192 const U32 ubound = trie->lasttrans + ucharcount;
2196 U32 base = trie->states[ 1 ].trans.base;
2199 const U32 data_slot = add_data( pRExC_state, 1, "T" );
2200 GET_RE_DEBUG_FLAGS_DECL;
2202 PERL_ARGS_ASSERT_MAKE_TRIE_FAILTABLE;
2204 PERL_UNUSED_ARG(depth);
2208 ARG_SET( stclass, data_slot );
2209 aho = (reg_ac_data *) PerlMemShared_calloc( 1, sizeof(reg_ac_data) );
2210 RExC_rxi->data->data[ data_slot ] = (void*)aho;
2211 aho->trie=trie_offset;
2212 aho->states=(reg_trie_state *)PerlMemShared_malloc( numstates * sizeof(reg_trie_state) );
2213 Copy( trie->states, aho->states, numstates, reg_trie_state );
2214 Newxz( q, numstates, U32);
2215 aho->fail = (U32 *) PerlMemShared_calloc( numstates, sizeof(U32) );
2218 /* initialize fail[0..1] to be 1 so that we always have
2219 a valid final fail state */
2220 fail[ 0 ] = fail[ 1 ] = 1;
2222 for ( charid = 0; charid < ucharcount ; charid++ ) {
2223 const U32 newstate = TRIE_TRANS_STATE( 1, base, ucharcount, charid, 0 );
2225 q[ q_write ] = newstate;
2226 /* set to point at the root */
2227 fail[ q[ q_write++ ] ]=1;
2230 while ( q_read < q_write) {
2231 const U32 cur = q[ q_read++ % numstates ];
2232 base = trie->states[ cur ].trans.base;
2234 for ( charid = 0 ; charid < ucharcount ; charid++ ) {
2235 const U32 ch_state = TRIE_TRANS_STATE( cur, base, ucharcount, charid, 1 );
2237 U32 fail_state = cur;
2240 fail_state = fail[ fail_state ];
2241 fail_base = aho->states[ fail_state ].trans.base;
2242 } while ( !TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 ) );
2244 fail_state = TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 );
2245 fail[ ch_state ] = fail_state;
2246 if ( !aho->states[ ch_state ].wordnum && aho->states[ fail_state ].wordnum )
2248 aho->states[ ch_state ].wordnum = aho->states[ fail_state ].wordnum;
2250 q[ q_write++ % numstates] = ch_state;
2254 /* restore fail[0..1] to 0 so that we "fall out" of the AC loop
2255 when we fail in state 1, this allows us to use the
2256 charclass scan to find a valid start char. This is based on the principle
2257 that theres a good chance the string being searched contains lots of stuff
2258 that cant be a start char.
2260 fail[ 0 ] = fail[ 1 ] = 0;
2261 DEBUG_TRIE_COMPILE_r({
2262 PerlIO_printf(Perl_debug_log,
2263 "%*sStclass Failtable (%"UVuf" states): 0",
2264 (int)(depth * 2), "", (UV)numstates
2266 for( q_read=1; q_read<numstates; q_read++ ) {
2267 PerlIO_printf(Perl_debug_log, ", %"UVuf, (UV)fail[q_read]);
2269 PerlIO_printf(Perl_debug_log, "\n");
2272 /*RExC_seen |= REG_SEEN_TRIEDFA;*/
2277 * There are strange code-generation bugs caused on sparc64 by gcc-2.95.2.
2278 * These need to be revisited when a newer toolchain becomes available.
2280 #if defined(__sparc64__) && defined(__GNUC__)
2281 # if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
2282 # undef SPARC64_GCC_WORKAROUND
2283 # define SPARC64_GCC_WORKAROUND 1
2287 #define DEBUG_PEEP(str,scan,depth) \
2288 DEBUG_OPTIMISE_r({if (scan){ \
2289 SV * const mysv=sv_newmortal(); \
2290 regnode *Next = regnext(scan); \
2291 regprop(RExC_rx, mysv, scan); \
2292 PerlIO_printf(Perl_debug_log, "%*s" str ">%3d: %s (%d)\n", \
2293 (int)depth*2, "", REG_NODE_NUM(scan), SvPV_nolen_const(mysv),\
2294 Next ? (REG_NODE_NUM(Next)) : 0 ); \
2301 #define JOIN_EXACT(scan,min,flags) \
2302 if (PL_regkind[OP(scan)] == EXACT) \
2303 join_exact(pRExC_state,(scan),(min),(flags),NULL,depth+1)
2306 S_join_exact(pTHX_ RExC_state_t *pRExC_state, regnode *scan, I32 *min, U32 flags,regnode *val, U32 depth) {
2307 /* Merge several consecutive EXACTish nodes into one. */
2308 regnode *n = regnext(scan);
2310 regnode *next = scan + NODE_SZ_STR(scan);
2314 regnode *stop = scan;
2315 GET_RE_DEBUG_FLAGS_DECL;
2317 PERL_UNUSED_ARG(depth);
2320 PERL_ARGS_ASSERT_JOIN_EXACT;
2321 #ifndef EXPERIMENTAL_INPLACESCAN
2322 PERL_UNUSED_ARG(flags);
2323 PERL_UNUSED_ARG(val);
2325 DEBUG_PEEP("join",scan,depth);
2327 /* Skip NOTHING, merge EXACT*. */
2329 ( PL_regkind[OP(n)] == NOTHING ||
2330 (stringok && (OP(n) == OP(scan))))
2332 && NEXT_OFF(scan) + NEXT_OFF(n) < I16_MAX) {
2334 if (OP(n) == TAIL || n > next)
2336 if (PL_regkind[OP(n)] == NOTHING) {
2337 DEBUG_PEEP("skip:",n,depth);
2338 NEXT_OFF(scan) += NEXT_OFF(n);
2339 next = n + NODE_STEP_REGNODE;
2346 else if (stringok) {
2347 const unsigned int oldl = STR_LEN(scan);
2348 regnode * const nnext = regnext(n);
2350 DEBUG_PEEP("merg",n,depth);
2353 if (oldl + STR_LEN(n) > U8_MAX)
2355 NEXT_OFF(scan) += NEXT_OFF(n);
2356 STR_LEN(scan) += STR_LEN(n);
2357 next = n + NODE_SZ_STR(n);
2358 /* Now we can overwrite *n : */
2359 Move(STRING(n), STRING(scan) + oldl, STR_LEN(n), char);
2367 #ifdef EXPERIMENTAL_INPLACESCAN
2368 if (flags && !NEXT_OFF(n)) {
2369 DEBUG_PEEP("atch", val, depth);
2370 if (reg_off_by_arg[OP(n)]) {
2371 ARG_SET(n, val - n);
2374 NEXT_OFF(n) = val - n;
2381 if (UTF && ( OP(scan) == EXACTF ) && ( STR_LEN(scan) >= 6 ) ) {
2383 Two problematic code points in Unicode casefolding of EXACT nodes:
2385 U+0390 - GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
2386 U+03B0 - GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
2392 U+03B9 U+0308 U+0301 0xCE 0xB9 0xCC 0x88 0xCC 0x81
2393 U+03C5 U+0308 U+0301 0xCF 0x85 0xCC 0x88 0xCC 0x81
2395 This means that in case-insensitive matching (or "loose matching",
2396 as Unicode calls it), an EXACTF of length six (the UTF-8 encoded byte
2397 length of the above casefolded versions) can match a target string
2398 of length two (the byte length of UTF-8 encoded U+0390 or U+03B0).
2399 This would rather mess up the minimum length computation.
2401 What we'll do is to look for the tail four bytes, and then peek
2402 at the preceding two bytes to see whether we need to decrease
2403 the minimum length by four (six minus two).
2405 Thanks to the design of UTF-8, there cannot be false matches:
2406 A sequence of valid UTF-8 bytes cannot be a subsequence of
2407 another valid sequence of UTF-8 bytes.
2410 char * const s0 = STRING(scan), *s, *t;
2411 char * const s1 = s0 + STR_LEN(scan) - 1;
2412 char * const s2 = s1 - 4;
2413 #ifdef EBCDIC /* RD tunifold greek 0390 and 03B0 */
2414 const char t0[] = "\xaf\x49\xaf\x42";
2416 const char t0[] = "\xcc\x88\xcc\x81";
2418 const char * const t1 = t0 + 3;
2421 s < s2 && (t = ninstr(s, s1, t0, t1));
2424 if (((U8)t[-1] == 0x68 && (U8)t[-2] == 0xB4) ||
2425 ((U8)t[-1] == 0x46 && (U8)t[-2] == 0xB5))
2427 if (((U8)t[-1] == 0xB9 && (U8)t[-2] == 0xCE) ||
2428 ((U8)t[-1] == 0x85 && (U8)t[-2] == 0xCF))
2436 n = scan + NODE_SZ_STR(scan);
2438 if (PL_regkind[OP(n)] != NOTHING || OP(n) == NOTHING) {
2445 DEBUG_OPTIMISE_r(if (merged){DEBUG_PEEP("finl",scan,depth)});
2449 /* REx optimizer. Converts nodes into quickier variants "in place".
2450 Finds fixed substrings. */
2452 /* Stops at toplevel WHILEM as well as at "last". At end *scanp is set
2453 to the position after last scanned or to NULL. */
2455 #define INIT_AND_WITHP \
2456 assert(!and_withp); \
2457 Newx(and_withp,1,struct regnode_charclass_class); \
2458 SAVEFREEPV(and_withp)
2460 /* this is a chain of data about sub patterns we are processing that
2461 need to be handled seperately/specially in study_chunk. Its so
2462 we can simulate recursion without losing state. */
2464 typedef struct scan_frame {
2465 regnode *last; /* last node to process in this frame */
2466 regnode *next; /* next node to process when last is reached */
2467 struct scan_frame *prev; /*previous frame*/
2468 I32 stop; /* what stopparen do we use */
2472 #define SCAN_COMMIT(s, data, m) scan_commit(s, data, m, is_inf)
2474 #define CASE_SYNST_FNC(nAmE) \
2476 if (flags & SCF_DO_STCLASS_AND) { \
2477 for (value = 0; value < 256; value++) \
2478 if (!is_ ## nAmE ## _cp(value)) \
2479 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2482 for (value = 0; value < 256; value++) \
2483 if (is_ ## nAmE ## _cp(value)) \
2484 ANYOF_BITMAP_SET(data->start_class, value); \
2488 if (flags & SCF_DO_STCLASS_AND) { \
2489 for (value = 0; value < 256; value++) \
2490 if (is_ ## nAmE ## _cp(value)) \
2491 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2494 for (value = 0; value < 256; value++) \
2495 if (!is_ ## nAmE ## _cp(value)) \
2496 ANYOF_BITMAP_SET(data->start_class, value); \
2503 S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
2504 I32 *minlenp, I32 *deltap,
2509 struct regnode_charclass_class *and_withp,
2510 U32 flags, U32 depth)
2511 /* scanp: Start here (read-write). */
2512 /* deltap: Write maxlen-minlen here. */
2513 /* last: Stop before this one. */
2514 /* data: string data about the pattern */
2515 /* stopparen: treat close N as END */
2516 /* recursed: which subroutines have we recursed into */
2517 /* and_withp: Valid if flags & SCF_DO_STCLASS_OR */
2520 I32 min = 0, pars = 0, code;
2521 regnode *scan = *scanp, *next;
2523 int is_inf = (flags & SCF_DO_SUBSTR) && (data->flags & SF_IS_INF);
2524 int is_inf_internal = 0; /* The studied chunk is infinite */
2525 I32 is_par = OP(scan) == OPEN ? ARG(scan) : 0;
2526 scan_data_t data_fake;
2527 SV *re_trie_maxbuff = NULL;
2528 regnode *first_non_open = scan;
2529 I32 stopmin = I32_MAX;
2530 scan_frame *frame = NULL;
2531 GET_RE_DEBUG_FLAGS_DECL;
2533 PERL_ARGS_ASSERT_STUDY_CHUNK;
2536 StructCopy(&zero_scan_data, &data_fake, scan_data_t);
2540 while (first_non_open && OP(first_non_open) == OPEN)
2541 first_non_open=regnext(first_non_open);
2546 while ( scan && OP(scan) != END && scan < last ){
2547 /* Peephole optimizer: */
2548 DEBUG_STUDYDATA("Peep:", data,depth);
2549 DEBUG_PEEP("Peep",scan,depth);
2550 JOIN_EXACT(scan,&min,0);
2552 /* Follow the next-chain of the current node and optimize
2553 away all the NOTHINGs from it. */
2554 if (OP(scan) != CURLYX) {
2555 const int max = (reg_off_by_arg[OP(scan)]
2557 /* I32 may be smaller than U16 on CRAYs! */
2558 : (I32_MAX < U16_MAX ? I32_MAX : U16_MAX));
2559 int off = (reg_off_by_arg[OP(scan)] ? ARG(scan) : NEXT_OFF(scan));
2563 /* Skip NOTHING and LONGJMP. */
2564 while ((n = regnext(n))
2565 && ((PL_regkind[OP(n)] == NOTHING && (noff = NEXT_OFF(n)))
2566 || ((OP(n) == LONGJMP) && (noff = ARG(n))))
2567 && off + noff < max)
2569 if (reg_off_by_arg[OP(scan)])
2572 NEXT_OFF(scan) = off;
2577 /* The principal pseudo-switch. Cannot be a switch, since we
2578 look into several different things. */
2579 if (OP(scan) == BRANCH || OP(scan) == BRANCHJ
2580 || OP(scan) == IFTHEN) {
2581 next = regnext(scan);
2583 /* demq: the op(next)==code check is to see if we have "branch-branch" AFAICT */
2585 if (OP(next) == code || code == IFTHEN) {
2586 /* NOTE - There is similar code to this block below for handling
2587 TRIE nodes on a re-study. If you change stuff here check there
2589 I32 max1 = 0, min1 = I32_MAX, num = 0;
2590 struct regnode_charclass_class accum;
2591 regnode * const startbranch=scan;
2593 if (flags & SCF_DO_SUBSTR)
2594 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot merge strings after this. */
2595 if (flags & SCF_DO_STCLASS)
2596 cl_init_zero(pRExC_state, &accum);
2598 while (OP(scan) == code) {
2599 I32 deltanext, minnext, f = 0, fake;
2600 struct regnode_charclass_class this_class;
2603 data_fake.flags = 0;
2605 data_fake.whilem_c = data->whilem_c;
2606 data_fake.last_closep = data->last_closep;
2609 data_fake.last_closep = &fake;
2611 data_fake.pos_delta = delta;
2612 next = regnext(scan);
2613 scan = NEXTOPER(scan);
2615 scan = NEXTOPER(scan);
2616 if (flags & SCF_DO_STCLASS) {
2617 cl_init(pRExC_state, &this_class);
2618 data_fake.start_class = &this_class;
2619 f = SCF_DO_STCLASS_AND;
2621 if (flags & SCF_WHILEM_VISITED_POS)
2622 f |= SCF_WHILEM_VISITED_POS;
2624 /* we suppose the run is continuous, last=next...*/
2625 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
2627 stopparen, recursed, NULL, f,depth+1);
2630 if (max1 < minnext + deltanext)
2631 max1 = minnext + deltanext;
2632 if (deltanext == I32_MAX)
2633 is_inf = is_inf_internal = 1;
2635 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
2637 if (data_fake.flags & SCF_SEEN_ACCEPT) {
2638 if ( stopmin > minnext)
2639 stopmin = min + min1;
2640 flags &= ~SCF_DO_SUBSTR;
2642 data->flags |= SCF_SEEN_ACCEPT;
2645 if (data_fake.flags & SF_HAS_EVAL)
2646 data->flags |= SF_HAS_EVAL;
2647 data->whilem_c = data_fake.whilem_c;
2649 if (flags & SCF_DO_STCLASS)
2650 cl_or(pRExC_state, &accum, &this_class);
2652 if (code == IFTHEN && num < 2) /* Empty ELSE branch */
2654 if (flags & SCF_DO_SUBSTR) {
2655 data->pos_min += min1;
2656 data->pos_delta += max1 - min1;
2657 if (max1 != min1 || is_inf)
2658 data->longest = &(data->longest_float);
2661 delta += max1 - min1;
2662 if (flags & SCF_DO_STCLASS_OR) {
2663 cl_or(pRExC_state, data->start_class, &accum);
2665 cl_and(data->start_class, and_withp);
2666 flags &= ~SCF_DO_STCLASS;
2669 else if (flags & SCF_DO_STCLASS_AND) {
2671 cl_and(data->start_class, &accum);
2672 flags &= ~SCF_DO_STCLASS;
2675 /* Switch to OR mode: cache the old value of
2676 * data->start_class */
2678 StructCopy(data->start_class, and_withp,
2679 struct regnode_charclass_class);
2680 flags &= ~SCF_DO_STCLASS_AND;
2681 StructCopy(&accum, data->start_class,
2682 struct regnode_charclass_class);
2683 flags |= SCF_DO_STCLASS_OR;
2684 data->start_class->flags |= ANYOF_EOS;
2688 if (PERL_ENABLE_TRIE_OPTIMISATION && OP( startbranch ) == BRANCH ) {
2691 Assuming this was/is a branch we are dealing with: 'scan' now
2692 points at the item that follows the branch sequence, whatever
2693 it is. We now start at the beginning of the sequence and look
2700 which would be constructed from a pattern like /A|LIST|OF|WORDS/
2702 If we can find such a subseqence we need to turn the first
2703 element into a trie and then add the subsequent branch exact
2704 strings to the trie.
2708 1. patterns where the whole set of branch can be converted.
2710 2. patterns where only a subset can be converted.
2712 In case 1 we can replace the whole set with a single regop
2713 for the trie. In case 2 we need to keep the start and end
2716 'BRANCH EXACT; BRANCH EXACT; BRANCH X'
2717 becomes BRANCH TRIE; BRANCH X;
2719 There is an additional case, that being where there is a
2720 common prefix, which gets split out into an EXACT like node
2721 preceding the TRIE node.
2723 If x(1..n)==tail then we can do a simple trie, if not we make
2724 a "jump" trie, such that when we match the appropriate word
2725 we "jump" to the appopriate tail node. Essentailly we turn
2726 a nested if into a case structure of sorts.
2731 if (!re_trie_maxbuff) {
2732 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
2733 if (!SvIOK(re_trie_maxbuff))
2734 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
2736 if ( SvIV(re_trie_maxbuff)>=0 ) {
2738 regnode *first = (regnode *)NULL;
2739 regnode *last = (regnode *)NULL;
2740 regnode *tail = scan;
2745 SV * const mysv = sv_newmortal(); /* for dumping */
2747 /* var tail is used because there may be a TAIL
2748 regop in the way. Ie, the exacts will point to the
2749 thing following the TAIL, but the last branch will
2750 point at the TAIL. So we advance tail. If we
2751 have nested (?:) we may have to move through several
2755 while ( OP( tail ) == TAIL ) {
2756 /* this is the TAIL generated by (?:) */
2757 tail = regnext( tail );
2762 regprop(RExC_rx, mysv, tail );
2763 PerlIO_printf( Perl_debug_log, "%*s%s%s\n",
2764 (int)depth * 2 + 2, "",
2765 "Looking for TRIE'able sequences. Tail node is: ",
2766 SvPV_nolen_const( mysv )
2772 step through the branches, cur represents each
2773 branch, noper is the first thing to be matched
2774 as part of that branch and noper_next is the
2775 regnext() of that node. if noper is an EXACT
2776 and noper_next is the same as scan (our current
2777 position in the regex) then the EXACT branch is
2778 a possible optimization target. Once we have
2779 two or more consequetive such branches we can
2780 create a trie of the EXACT's contents and stich
2781 it in place. If the sequence represents all of
2782 the branches we eliminate the whole thing and
2783 replace it with a single TRIE. If it is a
2784 subsequence then we need to stitch it in. This
2785 means the first branch has to remain, and needs
2786 to be repointed at the item on the branch chain
2787 following the last branch optimized. This could
2788 be either a BRANCH, in which case the
2789 subsequence is internal, or it could be the
2790 item following the branch sequence in which
2791 case the subsequence is at the end.
2795 /* dont use tail as the end marker for this traverse */
2796 for ( cur = startbranch ; cur != scan ; cur = regnext( cur ) ) {
2797 regnode * const noper = NEXTOPER( cur );
2798 #if defined(DEBUGGING) || defined(NOJUMPTRIE)
2799 regnode * const noper_next = regnext( noper );
2803 regprop(RExC_rx, mysv, cur);
2804 PerlIO_printf( Perl_debug_log, "%*s- %s (%d)",
2805 (int)depth * 2 + 2,"", SvPV_nolen_const( mysv ), REG_NODE_NUM(cur) );
2807 regprop(RExC_rx, mysv, noper);
2808 PerlIO_printf( Perl_debug_log, " -> %s",
2809 SvPV_nolen_const(mysv));
2812 regprop(RExC_rx, mysv, noper_next );
2813 PerlIO_printf( Perl_debug_log,"\t=> %s\t",
2814 SvPV_nolen_const(mysv));
2816 PerlIO_printf( Perl_debug_log, "(First==%d,Last==%d,Cur==%d)\n",
2817 REG_NODE_NUM(first), REG_NODE_NUM(last), REG_NODE_NUM(cur) );
2819 if ( (((first && optype!=NOTHING) ? OP( noper ) == optype
2820 : PL_regkind[ OP( noper ) ] == EXACT )
2821 || OP(noper) == NOTHING )
2823 && noper_next == tail
2828 if ( !first || optype == NOTHING ) {
2829 if (!first) first = cur;
2830 optype = OP( noper );
2836 Currently we do not believe that the trie logic can
2837 handle case insensitive matching properly when the
2838 pattern is not unicode (thus forcing unicode semantics).
2840 If/when this is fixed the following define can be swapped
2841 in below to fully enable trie logic.
2843 #define TRIE_TYPE_IS_SAFE 1
2846 #define TRIE_TYPE_IS_SAFE (UTF || optype==EXACT)
2848 if ( last && TRIE_TYPE_IS_SAFE ) {
2849 make_trie( pRExC_state,
2850 startbranch, first, cur, tail, count,
2853 if ( PL_regkind[ OP( noper ) ] == EXACT
2855 && noper_next == tail
2860 optype = OP( noper );
2870 regprop(RExC_rx, mysv, cur);
2871 PerlIO_printf( Perl_debug_log,
2872 "%*s- %s (%d) <SCAN FINISHED>\n", (int)depth * 2 + 2,
2873 "", SvPV_nolen_const( mysv ),REG_NODE_NUM(cur));
2877 if ( last && TRIE_TYPE_IS_SAFE ) {
2878 made= make_trie( pRExC_state, startbranch, first, scan, tail, count, optype, depth+1 );
2879 #ifdef TRIE_STUDY_OPT
2880 if ( ((made == MADE_EXACT_TRIE &&
2881 startbranch == first)
2882 || ( first_non_open == first )) &&
2884 flags |= SCF_TRIE_RESTUDY;
2885 if ( startbranch == first
2888 RExC_seen &=~REG_TOP_LEVEL_BRANCHES;
2898 else if ( code == BRANCHJ ) { /* single branch is optimized. */
2899 scan = NEXTOPER(NEXTOPER(scan));
2900 } else /* single branch is optimized. */
2901 scan = NEXTOPER(scan);
2903 } else if (OP(scan) == SUSPEND || OP(scan) == GOSUB || OP(scan) == GOSTART) {
2904 scan_frame *newframe = NULL;
2909 if (OP(scan) != SUSPEND) {
2910 /* set the pointer */
2911 if (OP(scan) == GOSUB) {
2913 RExC_recurse[ARG2L(scan)] = scan;
2914 start = RExC_open_parens[paren-1];
2915 end = RExC_close_parens[paren-1];
2918 start = RExC_rxi->program + 1;
2922 Newxz(recursed, (((RExC_npar)>>3) +1), U8);
2923 SAVEFREEPV(recursed);
2925 if (!PAREN_TEST(recursed,paren+1)) {
2926 PAREN_SET(recursed,paren+1);
2927 Newx(newframe,1,scan_frame);
2929 if (flags & SCF_DO_SUBSTR) {
2930 SCAN_COMMIT(pRExC_state,data,minlenp);
2931 data->longest = &(data->longest_float);
2933 is_inf = is_inf_internal = 1;
2934 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
2935 cl_anything(pRExC_state, data->start_class);
2936 flags &= ~SCF_DO_STCLASS;
2939 Newx(newframe,1,scan_frame);
2942 end = regnext(scan);
2947 SAVEFREEPV(newframe);
2948 newframe->next = regnext(scan);
2949 newframe->last = last;
2950 newframe->stop = stopparen;
2951 newframe->prev = frame;
2961 else if (OP(scan) == EXACT) {
2962 I32 l = STR_LEN(scan);
2965 const U8 * const s = (U8*)STRING(scan);
2966 l = utf8_length(s, s + l);
2967 uc = utf8_to_uvchr(s, NULL);
2969 uc = *((U8*)STRING(scan));
2972 if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
2973 /* The code below prefers earlier match for fixed
2974 offset, later match for variable offset. */
2975 if (data->last_end == -1) { /* Update the start info. */
2976 data->last_start_min = data->pos_min;
2977 data->last_start_max = is_inf
2978 ? I32_MAX : data->pos_min + data->pos_delta;
2980 sv_catpvn(data->last_found, STRING(scan), STR_LEN(scan));
2982 SvUTF8_on(data->last_found);
2984 SV * const sv = data->last_found;
2985 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
2986 mg_find(sv, PERL_MAGIC_utf8) : NULL;
2987 if (mg && mg->mg_len >= 0)
2988 mg->mg_len += utf8_length((U8*)STRING(scan),
2989 (U8*)STRING(scan)+STR_LEN(scan));
2991 data->last_end = data->pos_min + l;
2992 data->pos_min += l; /* As in the first entry. */
2993 data->flags &= ~SF_BEFORE_EOL;
2995 if (flags & SCF_DO_STCLASS_AND) {
2996 /* Check whether it is compatible with what we know already! */
3000 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3001 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3002 && (!(data->start_class->flags & ANYOF_FOLD)
3003 || !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
3006 ANYOF_CLASS_ZERO(data->start_class);
3007 ANYOF_BITMAP_ZERO(data->start_class);
3009 ANYOF_BITMAP_SET(data->start_class, uc);
3010 data->start_class->flags &= ~ANYOF_EOS;
3012 data->start_class->flags &= ~ANYOF_UNICODE_ALL;
3014 else if (flags & SCF_DO_STCLASS_OR) {
3015 /* false positive possible if the class is case-folded */
3017 ANYOF_BITMAP_SET(data->start_class, uc);
3019 data->start_class->flags |= ANYOF_UNICODE_ALL;
3020 data->start_class->flags &= ~ANYOF_EOS;
3021 cl_and(data->start_class, and_withp);
3023 flags &= ~SCF_DO_STCLASS;
3025 else if (PL_regkind[OP(scan)] == EXACT) { /* But OP != EXACT! */
3026 I32 l = STR_LEN(scan);
3027 UV uc = *((U8*)STRING(scan));
3029 /* Search for fixed substrings supports EXACT only. */
3030 if (flags & SCF_DO_SUBSTR) {
3032 SCAN_COMMIT(pRExC_state, data, minlenp);
3035 const U8 * const s = (U8 *)STRING(scan);
3036 l = utf8_length(s, s + l);
3037 uc = utf8_to_uvchr(s, NULL);
3040 if (flags & SCF_DO_SUBSTR)
3042 if (flags & SCF_DO_STCLASS_AND) {
3043 /* Check whether it is compatible with what we know already! */
3047 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3048 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3049 && !ANYOF_BITMAP_TEST(data->start_class, PL_fold[uc])))
3051 ANYOF_CLASS_ZERO(data->start_class);
3052 ANYOF_BITMAP_ZERO(data->start_class);
3054 ANYOF_BITMAP_SET(data->start_class, uc);
3055 data->start_class->flags &= ~ANYOF_EOS;
3056 data->start_class->flags |= ANYOF_FOLD;
3057 if (OP(scan) == EXACTFL)
3058 data->start_class->flags |= ANYOF_LOCALE;
3061 else if (flags & SCF_DO_STCLASS_OR) {
3062 if (data->start_class->flags & ANYOF_FOLD) {
3063 /* false positive possible if the class is case-folded.
3064 Assume that the locale settings are the same... */
3066 ANYOF_BITMAP_SET(data->start_class, uc);
3067 data->start_class->flags &= ~ANYOF_EOS;
3069 cl_and(data->start_class, and_withp);
3071 flags &= ~SCF_DO_STCLASS;
3073 else if (strchr((const char*)PL_varies,OP(scan))) {
3074 I32 mincount, maxcount, minnext, deltanext, fl = 0;
3075 I32 f = flags, pos_before = 0;
3076 regnode * const oscan = scan;
3077 struct regnode_charclass_class this_class;
3078 struct regnode_charclass_class *oclass = NULL;
3079 I32 next_is_eval = 0;
3081 switch (PL_regkind[OP(scan)]) {
3082 case WHILEM: /* End of (?:...)* . */
3083 scan = NEXTOPER(scan);
3086 if (flags & (SCF_DO_SUBSTR | SCF_DO_STCLASS)) {
3087 next = NEXTOPER(scan);
3088 if (OP(next) == EXACT || (flags & SCF_DO_STCLASS)) {
3090 maxcount = REG_INFTY;
3091 next = regnext(scan);
3092 scan = NEXTOPER(scan);
3096 if (flags & SCF_DO_SUBSTR)
3101 if (flags & SCF_DO_STCLASS) {
3103 maxcount = REG_INFTY;
3104 next = regnext(scan);
3105 scan = NEXTOPER(scan);
3108 is_inf = is_inf_internal = 1;
3109 scan = regnext(scan);
3110 if (flags & SCF_DO_SUBSTR) {
3111 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot extend fixed substrings */
3112 data->longest = &(data->longest_float);
3114 goto optimize_curly_tail;
3116 if (stopparen>0 && (OP(scan)==CURLYN || OP(scan)==CURLYM)
3117 && (scan->flags == stopparen))
3122 mincount = ARG1(scan);
3123 maxcount = ARG2(scan);
3125 next = regnext(scan);
3126 if (OP(scan) == CURLYX) {
3127 I32 lp = (data ? *(data->last_closep) : 0);
3128 scan->flags = ((lp <= (I32)U8_MAX) ? (U8)lp : U8_MAX);
3130 scan = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
3131 next_is_eval = (OP(scan) == EVAL);
3133 if (flags & SCF_DO_SUBSTR) {
3134 if (mincount == 0) SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot extend fixed substrings */
3135 pos_before = data->pos_min;
3139 data->flags &= ~(SF_HAS_PAR|SF_IN_PAR|SF_HAS_EVAL);
3141 data->flags |= SF_IS_INF;
3143 if (flags & SCF_DO_STCLASS) {
3144 cl_init(pRExC_state, &this_class);
3145 oclass = data->start_class;
3146 data->start_class = &this_class;
3147 f |= SCF_DO_STCLASS_AND;
3148 f &= ~SCF_DO_STCLASS_OR;
3150 /* These are the cases when once a subexpression
3151 fails at a particular position, it cannot succeed
3152 even after backtracking at the enclosing scope.
3154 XXXX what if minimal match and we are at the
3155 initial run of {n,m}? */
3156 if ((mincount != maxcount - 1) && (maxcount != REG_INFTY))
3157 f &= ~SCF_WHILEM_VISITED_POS;
3159 /* This will finish on WHILEM, setting scan, or on NULL: */
3160 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
3161 last, data, stopparen, recursed, NULL,
3163 ? (f & ~SCF_DO_SUBSTR) : f),depth+1);
3165 if (flags & SCF_DO_STCLASS)
3166 data->start_class = oclass;
3167 if (mincount == 0 || minnext == 0) {
3168 if (flags & SCF_DO_STCLASS_OR) {
3169 cl_or(pRExC_state, data->start_class, &this_class);
3171 else if (flags & SCF_DO_STCLASS_AND) {
3172 /* Switch to OR mode: cache the old value of
3173 * data->start_class */
3175 StructCopy(data->start_class, and_withp,
3176 struct regnode_charclass_class);
3177 flags &= ~SCF_DO_STCLASS_AND;
3178 StructCopy(&this_class, data->start_class,
3179 struct regnode_charclass_class);
3180 flags |= SCF_DO_STCLASS_OR;
3181 data->start_class->flags |= ANYOF_EOS;
3183 } else { /* Non-zero len */
3184 if (flags & SCF_DO_STCLASS_OR) {
3185 cl_or(pRExC_state, data->start_class, &this_class);
3186 cl_and(data->start_class, and_withp);
3188 else if (flags & SCF_DO_STCLASS_AND)
3189 cl_and(data->start_class, &this_class);
3190 flags &= ~SCF_DO_STCLASS;
3192 if (!scan) /* It was not CURLYX, but CURLY. */
3194 if ( /* ? quantifier ok, except for (?{ ... }) */
3195 (next_is_eval || !(mincount == 0 && maxcount == 1))
3196 && (minnext == 0) && (deltanext == 0)
3197 && data && !(data->flags & (SF_HAS_PAR|SF_IN_PAR))
3198 && maxcount <= REG_INFTY/3) /* Complement check for big count */
3200 ckWARNreg(RExC_parse,
3201 "Quantifier unexpected on zero-length expression");
3204 min += minnext * mincount;
3205 is_inf_internal |= ((maxcount == REG_INFTY
3206 && (minnext + deltanext) > 0)
3207 || deltanext == I32_MAX);
3208 is_inf |= is_inf_internal;
3209 delta += (minnext + deltanext) * maxcount - minnext * mincount;
3211 /* Try powerful optimization CURLYX => CURLYN. */
3212 if ( OP(oscan) == CURLYX && data
3213 && data->flags & SF_IN_PAR
3214 && !(data->flags & SF_HAS_EVAL)
3215 && !deltanext && minnext == 1 ) {
3216 /* Try to optimize to CURLYN. */
3217 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS;
3218 regnode * const nxt1 = nxt;
3225 if (!strchr((const char*)PL_simple,OP(nxt))
3226 && !(PL_regkind[OP(nxt)] == EXACT
3227 && STR_LEN(nxt) == 1))
3233 if (OP(nxt) != CLOSE)
3235 if (RExC_open_parens) {
3236 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3237 RExC_close_parens[ARG(nxt1)-1]=nxt+2; /*close->while*/
3239 /* Now we know that nxt2 is the only contents: */
3240 oscan->flags = (U8)ARG(nxt);
3242 OP(nxt1) = NOTHING; /* was OPEN. */
3245 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3246 NEXT_OFF(nxt1+ 1) = 0; /* just for consistancy. */
3247 NEXT_OFF(nxt2) = 0; /* just for consistancy with CURLY. */
3248 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3249 OP(nxt + 1) = OPTIMIZED; /* was count. */
3250 NEXT_OFF(nxt+ 1) = 0; /* just for consistancy. */
3255 /* Try optimization CURLYX => CURLYM. */
3256 if ( OP(oscan) == CURLYX && data
3257 && !(data->flags & SF_HAS_PAR)
3258 && !(data->flags & SF_HAS_EVAL)
3259 && !deltanext /* atom is fixed width */
3260 && minnext != 0 /* CURLYM can't handle zero width */
3262 /* XXXX How to optimize if data == 0? */
3263 /* Optimize to a simpler form. */
3264 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN */
3268 while ( (nxt2 = regnext(nxt)) /* skip over embedded stuff*/
3269 && (OP(nxt2) != WHILEM))
3271 OP(nxt2) = SUCCEED; /* Whas WHILEM */
3272 /* Need to optimize away parenths. */
3273 if (data->flags & SF_IN_PAR) {
3274 /* Set the parenth number. */
3275 regnode *nxt1 = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN*/
3277 if (OP(nxt) != CLOSE)
3278 FAIL("Panic opt close");
3279 oscan->flags = (U8)ARG(nxt);
3280 if (RExC_open_parens) {
3281 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3282 RExC_close_parens[ARG(nxt1)-1]=nxt2+1; /*close->NOTHING*/
3284 OP(nxt1) = OPTIMIZED; /* was OPEN. */
3285 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3288 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3289 OP(nxt + 1) = OPTIMIZED; /* was count. */
3290 NEXT_OFF(nxt1 + 1) = 0; /* just for consistancy. */
3291 NEXT_OFF(nxt + 1) = 0; /* just for consistancy. */
3294 while ( nxt1 && (OP(nxt1) != WHILEM)) {
3295 regnode *nnxt = regnext(nxt1);
3298 if (reg_off_by_arg[OP(nxt1)])
3299 ARG_SET(nxt1, nxt2 - nxt1);
3300 else if (nxt2 - nxt1 < U16_MAX)
3301 NEXT_OFF(nxt1) = nxt2 - nxt1;
3303 OP(nxt) = NOTHING; /* Cannot beautify */
3308 /* Optimize again: */
3309 study_chunk(pRExC_state, &nxt1, minlenp, &deltanext, nxt,
3310 NULL, stopparen, recursed, NULL, 0,depth+1);
3315 else if ((OP(oscan) == CURLYX)
3316 && (flags & SCF_WHILEM_VISITED_POS)
3317 /* See the comment on a similar expression above.
3318 However, this time it not a subexpression
3319 we care about, but the expression itself. */
3320 && (maxcount == REG_INFTY)
3321 && data && ++data->whilem_c < 16) {
3322 /* This stays as CURLYX, we can put the count/of pair. */
3323 /* Find WHILEM (as in regexec.c) */
3324 regnode *nxt = oscan + NEXT_OFF(oscan);
3326 if (OP(PREVOPER(nxt)) == NOTHING) /* LONGJMP */
3328 PREVOPER(nxt)->flags = (U8)(data->whilem_c
3329 | (RExC_whilem_seen << 4)); /* On WHILEM */
3331 if (data && fl & (SF_HAS_PAR|SF_IN_PAR))
3333 if (flags & SCF_DO_SUBSTR) {
3334 SV *last_str = NULL;
3335 int counted = mincount != 0;
3337 if (data->last_end > 0 && mincount != 0) { /* Ends with a string. */
3338 #if defined(SPARC64_GCC_WORKAROUND)
3341 const char *s = NULL;
3344 if (pos_before >= data->last_start_min)
3347 b = data->last_start_min;
3350 s = SvPV_const(data->last_found, l);
3351 old = b - data->last_start_min;
3354 I32 b = pos_before >= data->last_start_min
3355 ? pos_before : data->last_start_min;
3357 const char * const s = SvPV_const(data->last_found, l);
3358 I32 old = b - data->last_start_min;
3362 old = utf8_hop((U8*)s, old) - (U8*)s;
3365 /* Get the added string: */
3366 last_str = newSVpvn_utf8(s + old, l, UTF);
3367 if (deltanext == 0 && pos_before == b) {
3368 /* What was added is a constant string */
3370 SvGROW(last_str, (mincount * l) + 1);
3371 repeatcpy(SvPVX(last_str) + l,
3372 SvPVX_const(last_str), l, mincount - 1);
3373 SvCUR_set(last_str, SvCUR(last_str) * mincount);
3374 /* Add additional parts. */
3375 SvCUR_set(data->last_found,
3376 SvCUR(data->last_found) - l);
3377 sv_catsv(data->last_found, last_str);
3379 SV * sv = data->last_found;
3381 SvUTF8(sv) && SvMAGICAL(sv) ?
3382 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3383 if (mg && mg->mg_len >= 0)
3384 mg->mg_len += CHR_SVLEN(last_str) - l;
3386 data->last_end += l * (mincount - 1);
3389 /* start offset must point into the last copy */
3390 data->last_start_min += minnext * (mincount - 1);
3391 data->last_start_max += is_inf ? I32_MAX
3392 : (maxcount - 1) * (minnext + data->pos_delta);
3395 /* It is counted once already... */
3396 data->pos_min += minnext * (mincount - counted);
3397 data->pos_delta += - counted * deltanext +
3398 (minnext + deltanext) * maxcount - minnext * mincount;
3399 if (mincount != maxcount) {
3400 /* Cannot extend fixed substrings found inside
3402 SCAN_COMMIT(pRExC_state,data,minlenp);
3403 if (mincount && last_str) {
3404 SV * const sv = data->last_found;
3405 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3406 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3410 sv_setsv(sv, last_str);
3411 data->last_end = data->pos_min;
3412 data->last_start_min =
3413 data->pos_min - CHR_SVLEN(last_str);
3414 data->last_start_max = is_inf
3416 : data->pos_min + data->pos_delta
3417 - CHR_SVLEN(last_str);
3419 data->longest = &(data->longest_float);
3421 SvREFCNT_dec(last_str);
3423 if (data && (fl & SF_HAS_EVAL))
3424 data->flags |= SF_HAS_EVAL;
3425 optimize_curly_tail:
3426 if (OP(oscan) != CURLYX) {
3427 while (PL_regkind[OP(next = regnext(oscan))] == NOTHING
3429 NEXT_OFF(oscan) += NEXT_OFF(next);
3432 default: /* REF and CLUMP only? */
3433 if (flags & SCF_DO_SUBSTR) {
3434 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3435 data->longest = &(data->longest_float);
3437 is_inf = is_inf_internal = 1;
3438 if (flags & SCF_DO_STCLASS_OR)
3439 cl_anything(pRExC_state, data->start_class);
3440 flags &= ~SCF_DO_STCLASS;
3444 else if (OP(scan) == LNBREAK) {
3445 if (flags & SCF_DO_STCLASS) {
3447 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3448 if (flags & SCF_DO_STCLASS_AND) {
3449 for (value = 0; value < 256; value++)
3450 if (!is_VERTWS_cp(value))
3451 ANYOF_BITMAP_CLEAR(data->start_class, value);
3454 for (value = 0; value < 256; value++)
3455 if (is_VERTWS_cp(value))
3456 ANYOF_BITMAP_SET(data->start_class, value);
3458 if (flags & SCF_DO_STCLASS_OR)
3459 cl_and(data->start_class, and_withp);
3460 flags &= ~SCF_DO_STCLASS;
3464 if (flags & SCF_DO_SUBSTR) {
3465 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3467 data->pos_delta += 1;
3468 data->longest = &(data->longest_float);
3472 else if (OP(scan) == FOLDCHAR) {
3473 int d = ARG(scan)==0xDF ? 1 : 2;
3474 flags &= ~SCF_DO_STCLASS;
3477 if (flags & SCF_DO_SUBSTR) {
3478 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3480 data->pos_delta += d;
3481 data->longest = &(data->longest_float);
3484 else if (strchr((const char*)PL_simple,OP(scan))) {
3487 if (flags & SCF_DO_SUBSTR) {
3488 SCAN_COMMIT(pRExC_state,data,minlenp);
3492 if (flags & SCF_DO_STCLASS) {
3493 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3495 /* Some of the logic below assumes that switching
3496 locale on will only add false positives. */
3497 switch (PL_regkind[OP(scan)]) {
3501 /* Perl_croak(aTHX_ "panic: unexpected simple REx opcode %d", OP(scan)); */
3502 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3503 cl_anything(pRExC_state, data->start_class);
3506 if (OP(scan) == SANY)
3508 if (flags & SCF_DO_STCLASS_OR) { /* Everything but \n */
3509 value = (ANYOF_BITMAP_TEST(data->start_class,'\n')
3510 || (data->start_class->flags & ANYOF_CLASS));
3511 cl_anything(pRExC_state, data->start_class);
3513 if (flags & SCF_DO_STCLASS_AND || !value)
3514 ANYOF_BITMAP_CLEAR(data->start_class,'\n');
3517 if (flags & SCF_DO_STCLASS_AND)
3518 cl_and(data->start_class,
3519 (struct regnode_charclass_class*)scan);
3521 cl_or(pRExC_state, data->start_class,
3522 (struct regnode_charclass_class*)scan);
3525 if (flags & SCF_DO_STCLASS_AND) {
3526 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3527 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3528 for (value = 0; value < 256; value++)
3529 if (!isALNUM(value))
3530 ANYOF_BITMAP_CLEAR(data->start_class, value);
3534 if (data->start_class->flags & ANYOF_LOCALE)
3535 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3537 for (value = 0; value < 256; value++)
3539 ANYOF_BITMAP_SET(data->start_class, value);
3544 if (flags & SCF_DO_STCLASS_AND) {
3545 if (data->start_class->flags & ANYOF_LOCALE)
3546 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
3549 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
3550 data->start_class->flags |= ANYOF_LOCALE;
3554 if (flags & SCF_DO_STCLASS_AND) {
3555 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3556 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3557 for (value = 0; value < 256; value++)
3559 ANYOF_BITMAP_CLEAR(data->start_class, value);
3563 if (data->start_class->flags & ANYOF_LOCALE)
3564 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3566 for (value = 0; value < 256; value++)
3567 if (!isALNUM(value))
3568 ANYOF_BITMAP_SET(data->start_class, value);
3573 if (flags & SCF_DO_STCLASS_AND) {
3574 if (data->start_class->flags & ANYOF_LOCALE)
3575 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
3578 data->start_class->flags |= ANYOF_LOCALE;
3579 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
3583 if (flags & SCF_DO_STCLASS_AND) {
3584 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3585 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3586 for (value = 0; value < 256; value++)
3587 if (!isSPACE(value))
3588 ANYOF_BITMAP_CLEAR(data->start_class, value);
3592 if (data->start_class->flags & ANYOF_LOCALE)
3593 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3595 for (value = 0; value < 256; value++)
3597 ANYOF_BITMAP_SET(data->start_class, value);
3602 if (flags & SCF_DO_STCLASS_AND) {
3603 if (data->start_class->flags & ANYOF_LOCALE)
3604 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
3607 data->start_class->flags |= ANYOF_LOCALE;
3608 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
3612 if (flags & SCF_DO_STCLASS_AND) {
3613 if (!(data->start_class->flags & ANYOF_LOCALE)) {
3614 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3615 for (value = 0; value < 256; value++)
3617 ANYOF_BITMAP_CLEAR(data->start_class, value);
3621 if (data->start_class->flags & ANYOF_LOCALE)
3622 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3624 for (value = 0; value < 256; value++)
3625 if (!isSPACE(value))
3626 ANYOF_BITMAP_SET(data->start_class, value);
3631 if (flags & SCF_DO_STCLASS_AND) {
3632 if (data->start_class->flags & ANYOF_LOCALE) {
3633 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
3634 for (value = 0; value < 256; value++)
3635 if (!isSPACE(value))
3636 ANYOF_BITMAP_CLEAR(data->start_class, value);
3640 data->start_class->flags |= ANYOF_LOCALE;
3641 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
3645 if (flags & SCF_DO_STCLASS_AND) {
3646 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NDIGIT);
3647 for (value = 0; value < 256; value++)
3648 if (!isDIGIT(value))
3649 ANYOF_BITMAP_CLEAR(data->start_class, value);
3652 if (data->start_class->flags & ANYOF_LOCALE)
3653 ANYOF_CLASS_SET(data->start_class,ANYOF_DIGIT);
3655 for (value = 0; value < 256; value++)
3657 ANYOF_BITMAP_SET(data->start_class, value);
3662 if (flags & SCF_DO_STCLASS_AND) {
3663 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_DIGIT);
3664 for (value = 0; value < 256; value++)
3666 ANYOF_BITMAP_CLEAR(data->start_class, value);
3669 if (data->start_class->flags & ANYOF_LOCALE)
3670 ANYOF_CLASS_SET(data->start_class,ANYOF_NDIGIT);
3672 for (value = 0; value < 256; value++)
3673 if (!isDIGIT(value))
3674 ANYOF_BITMAP_SET(data->start_class, value);
3678 CASE_SYNST_FNC(VERTWS);
3679 CASE_SYNST_FNC(HORIZWS);
3682 if (flags & SCF_DO_STCLASS_OR)
3683 cl_and(data->start_class, and_withp);
3684 flags &= ~SCF_DO_STCLASS;
3687 else if (PL_regkind[OP(scan)] == EOL && flags & SCF_DO_SUBSTR) {
3688 data->flags |= (OP(scan) == MEOL
3692 else if ( PL_regkind[OP(scan)] == BRANCHJ
3693 /* Lookbehind, or need to calculate parens/evals/stclass: */
3694 && (scan->flags || data || (flags & SCF_DO_STCLASS))
3695 && (OP(scan) == IFMATCH || OP(scan) == UNLESSM)) {
3696 if ( !PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3697 || OP(scan) == UNLESSM )
3699 /* Negative Lookahead/lookbehind
3700 In this case we can't do fixed string optimisation.
3703 I32 deltanext, minnext, fake = 0;
3705 struct regnode_charclass_class intrnl;
3708 data_fake.flags = 0;
3710 data_fake.whilem_c = data->whilem_c;
3711 data_fake.last_closep = data->last_closep;
3714 data_fake.last_closep = &fake;
3715 data_fake.pos_delta = delta;
3716 if ( flags & SCF_DO_STCLASS && !scan->flags
3717 && OP(scan) == IFMATCH ) { /* Lookahead */
3718 cl_init(pRExC_state, &intrnl);
3719 data_fake.start_class = &intrnl;
3720 f |= SCF_DO_STCLASS_AND;
3722 if (flags & SCF_WHILEM_VISITED_POS)
3723 f |= SCF_WHILEM_VISITED_POS;
3724 next = regnext(scan);
3725 nscan = NEXTOPER(NEXTOPER(scan));
3726 minnext = study_chunk(pRExC_state, &nscan, minlenp, &deltanext,
3727 last, &data_fake, stopparen, recursed, NULL, f, depth+1);
3730 FAIL("Variable length lookbehind not implemented");
3732 else if (minnext > (I32)U8_MAX) {
3733 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3735 scan->flags = (U8)minnext;
3738 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3740 if (data_fake.flags & SF_HAS_EVAL)
3741 data->flags |= SF_HAS_EVAL;
3742 data->whilem_c = data_fake.whilem_c;
3744 if (f & SCF_DO_STCLASS_AND) {
3745 if (flags & SCF_DO_STCLASS_OR) {
3746 /* OR before, AND after: ideally we would recurse with
3747 * data_fake to get the AND applied by study of the
3748 * remainder of the pattern, and then derecurse;
3749 * *** HACK *** for now just treat as "no information".
3750 * See [perl #56690].
3752 cl_init(pRExC_state, data->start_class);
3754 /* AND before and after: combine and continue */
3755 const int was = (data->start_class->flags & ANYOF_EOS);
3757 cl_and(data->start_class, &intrnl);
3759 data->start_class->flags |= ANYOF_EOS;
3763 #if PERL_ENABLE_POSITIVE_ASSERTION_STUDY
3765 /* Positive Lookahead/lookbehind
3766 In this case we can do fixed string optimisation,
3767 but we must be careful about it. Note in the case of
3768 lookbehind the positions will be offset by the minimum
3769 length of the pattern, something we won't know about
3770 until after the recurse.
3772 I32 deltanext, fake = 0;
3774 struct regnode_charclass_class intrnl;
3776 /* We use SAVEFREEPV so that when the full compile
3777 is finished perl will clean up the allocated
3778 minlens when its all done. This was we don't
3779 have to worry about freeing them when we know
3780 they wont be used, which would be a pain.
3783 Newx( minnextp, 1, I32 );
3784 SAVEFREEPV(minnextp);
3787 StructCopy(data, &data_fake, scan_data_t);
3788 if ((flags & SCF_DO_SUBSTR) && data->last_found) {
3791 SCAN_COMMIT(pRExC_state, &data_fake,minlenp);
3792 data_fake.last_found=newSVsv(data->last_found);
3796 data_fake.last_closep = &fake;
3797 data_fake.flags = 0;
3798 data_fake.pos_delta = delta;
3800 data_fake.flags |= SF_IS_INF;
3801 if ( flags & SCF_DO_STCLASS && !scan->flags
3802 && OP(scan) == IFMATCH ) { /* Lookahead */
3803 cl_init(pRExC_state, &intrnl);
3804 data_fake.start_class = &intrnl;
3805 f |= SCF_DO_STCLASS_AND;
3807 if (flags & SCF_WHILEM_VISITED_POS)
3808 f |= SCF_WHILEM_VISITED_POS;
3809 next = regnext(scan);
3810 nscan = NEXTOPER(NEXTOPER(scan));
3812 *minnextp = study_chunk(pRExC_state, &nscan, minnextp, &deltanext,
3813 last, &data_fake, stopparen, recursed, NULL, f,depth+1);
3816 FAIL("Variable length lookbehind not implemented");
3818 else if (*minnextp > (I32)U8_MAX) {
3819 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
3821 scan->flags = (U8)*minnextp;
3826 if (f & SCF_DO_STCLASS_AND) {
3827 const int was = (data->start_class->flags & ANYOF_EOS);
3829 cl_and(data->start_class, &intrnl);
3831 data->start_class->flags |= ANYOF_EOS;
3834 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3836 if (data_fake.flags & SF_HAS_EVAL)
3837 data->flags |= SF_HAS_EVAL;
3838 data->whilem_c = data_fake.whilem_c;
3839 if ((flags & SCF_DO_SUBSTR) && data_fake.last_found) {
3840 if (RExC_rx->minlen<*minnextp)
3841 RExC_rx->minlen=*minnextp;
3842 SCAN_COMMIT(pRExC_state, &data_fake, minnextp);
3843 SvREFCNT_dec(data_fake.last_found);
3845 if ( data_fake.minlen_fixed != minlenp )
3847 data->offset_fixed= data_fake.offset_fixed;
3848 data->minlen_fixed= data_fake.minlen_fixed;
3849 data->lookbehind_fixed+= scan->flags;
3851 if ( data_fake.minlen_float != minlenp )
3853 data->minlen_float= data_fake.minlen_float;
3854 data->offset_float_min=data_fake.offset_float_min;
3855 data->offset_float_max=data_fake.offset_float_max;
3856 data->lookbehind_float+= scan->flags;
3865 else if (OP(scan) == OPEN) {
3866 if (stopparen != (I32)ARG(scan))
3869 else if (OP(scan) == CLOSE) {
3870 if (stopparen == (I32)ARG(scan)) {
3873 if ((I32)ARG(scan) == is_par) {
3874 next = regnext(scan);
3876 if ( next && (OP(next) != WHILEM) && next < last)
3877 is_par = 0; /* Disable optimization */
3880 *(data->last_closep) = ARG(scan);
3882 else if (OP(scan) == EVAL) {
3884 data->flags |= SF_HAS_EVAL;
3886 else if ( PL_regkind[OP(scan)] == ENDLIKE ) {
3887 if (flags & SCF_DO_SUBSTR) {
3888 SCAN_COMMIT(pRExC_state,data,minlenp);
3889 flags &= ~SCF_DO_SUBSTR;
3891 if (data && OP(scan)==ACCEPT) {
3892 data->flags |= SCF_SEEN_ACCEPT;
3897 else if (OP(scan) == LOGICAL && scan->flags == 2) /* Embedded follows */
3899 if (flags & SCF_DO_SUBSTR) {
3900 SCAN_COMMIT(pRExC_state,data,minlenp);
3901 data->longest = &(data->longest_float);
3903 is_inf = is_inf_internal = 1;
3904 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3905 cl_anything(pRExC_state, data->start_class);
3906 flags &= ~SCF_DO_STCLASS;
3908 else if (OP(scan) == GPOS) {
3909 if (!(RExC_rx->extflags & RXf_GPOS_FLOAT) &&
3910 !(delta || is_inf || (data && data->pos_delta)))
3912 if (!(RExC_rx->extflags & RXf_ANCH) && (flags & SCF_DO_SUBSTR))
3913 RExC_rx->extflags |= RXf_ANCH_GPOS;
3914 if (RExC_rx->gofs < (U32)min)
3915 RExC_rx->gofs = min;
3917 RExC_rx->extflags |= RXf_GPOS_FLOAT;
3921 #ifdef TRIE_STUDY_OPT
3922 #ifdef FULL_TRIE_STUDY
3923 else if (PL_regkind[OP(scan)] == TRIE) {
3924 /* NOTE - There is similar code to this block above for handling
3925 BRANCH nodes on the initial study. If you change stuff here
3927 regnode *trie_node= scan;
3928 regnode *tail= regnext(scan);
3929 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
3930 I32 max1 = 0, min1 = I32_MAX;
3931 struct regnode_charclass_class accum;
3933 if (flags & SCF_DO_SUBSTR) /* XXXX Add !SUSPEND? */
3934 SCAN_COMMIT(pRExC_state, data,minlenp); /* Cannot merge strings after this. */
3935 if (flags & SCF_DO_STCLASS)
3936 cl_init_zero(pRExC_state, &accum);
3942 const regnode *nextbranch= NULL;
3945 for ( word=1 ; word <= trie->wordcount ; word++)
3947 I32 deltanext=0, minnext=0, f = 0, fake;
3948 struct regnode_charclass_class this_class;
3950 data_fake.flags = 0;
3952 data_fake.whilem_c = data->whilem_c;
3953 data_fake.last_closep = data->last_closep;
3956 data_fake.last_closep = &fake;
3957 data_fake.pos_delta = delta;
3958 if (flags & SCF_DO_STCLASS) {
3959 cl_init(pRExC_state, &this_class);
3960 data_fake.start_class = &this_class;
3961 f = SCF_DO_STCLASS_AND;
3963 if (flags & SCF_WHILEM_VISITED_POS)
3964 f |= SCF_WHILEM_VISITED_POS;
3966 if (trie->jump[word]) {
3968 nextbranch = trie_node + trie->jump[0];
3969 scan= trie_node + trie->jump[word];
3970 /* We go from the jump point to the branch that follows
3971 it. Note this means we need the vestigal unused branches
3972 even though they arent otherwise used.
3974 minnext = study_chunk(pRExC_state, &scan, minlenp,
3975 &deltanext, (regnode *)nextbranch, &data_fake,
3976 stopparen, recursed, NULL, f,depth+1);
3978 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
3979 nextbranch= regnext((regnode*)nextbranch);
3981 if (min1 > (I32)(minnext + trie->minlen))
3982 min1 = minnext + trie->minlen;
3983 if (max1 < (I32)(minnext + deltanext + trie->maxlen))
3984 max1 = minnext + deltanext + trie->maxlen;
3985 if (deltanext == I32_MAX)
3986 is_inf = is_inf_internal = 1;
3988 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3990 if (data_fake.flags & SCF_SEEN_ACCEPT) {
3991 if ( stopmin > min + min1)
3992 stopmin = min + min1;
3993 flags &= ~SCF_DO_SUBSTR;
3995 data->flags |= SCF_SEEN_ACCEPT;
3998 if (data_fake.flags & SF_HAS_EVAL)
3999 data->flags |= SF_HAS_EVAL;
4000 data->whilem_c = data_fake.whilem_c;
4002 if (flags & SCF_DO_STCLASS)
4003 cl_or(pRExC_state, &accum, &this_class);
4006 if (flags & SCF_DO_SUBSTR) {
4007 data->pos_min += min1;
4008 data->pos_delta += max1 - min1;
4009 if (max1 != min1 || is_inf)
4010 data->longest = &(data->longest_float);
4013 delta += max1 - min1;
4014 if (flags & SCF_DO_STCLASS_OR) {
4015 cl_or(pRExC_state, data->start_class, &accum);
4017 cl_and(data->start_class, and_withp);
4018 flags &= ~SCF_DO_STCLASS;
4021 else if (flags & SCF_DO_STCLASS_AND) {
4023 cl_and(data->start_class, &accum);
4024 flags &= ~SCF_DO_STCLASS;
4027 /* Switch to OR mode: cache the old value of
4028 * data->start_class */
4030 StructCopy(data->start_class, and_withp,
4031 struct regnode_charclass_class);
4032 flags &= ~SCF_DO_STCLASS_AND;
4033 StructCopy(&accum, data->start_class,
4034 struct regnode_charclass_class);
4035 flags |= SCF_DO_STCLASS_OR;
4036 data->start_class->flags |= ANYOF_EOS;
4043 else if (PL_regkind[OP(scan)] == TRIE) {
4044 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
4047 min += trie->minlen;
4048 delta += (trie->maxlen - trie->minlen);
4049 flags &= ~SCF_DO_STCLASS; /* xxx */
4050 if (flags & SCF_DO_SUBSTR) {
4051 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
4052 data->pos_min += trie->minlen;
4053 data->pos_delta += (trie->maxlen - trie->minlen);
4054 if (trie->maxlen != trie->minlen)
4055 data->longest = &(data->longest_float);
4057 if (trie->jump) /* no more substrings -- for now /grr*/
4058 flags &= ~SCF_DO_SUBSTR;
4060 #endif /* old or new */
4061 #endif /* TRIE_STUDY_OPT */
4063 /* Else: zero-length, ignore. */
4064 scan = regnext(scan);
4069 stopparen = frame->stop;
4070 frame = frame->prev;
4071 goto fake_study_recurse;
4076 DEBUG_STUDYDATA("pre-fin:",data,depth);
4079 *deltap = is_inf_internal ? I32_MAX : delta;
4080 if (flags & SCF_DO_SUBSTR && is_inf)
4081 data->pos_delta = I32_MAX - data->pos_min;
4082 if (is_par > (I32)U8_MAX)
4084 if (is_par && pars==1 && data) {
4085 data->flags |= SF_IN_PAR;
4086 data->flags &= ~SF_HAS_PAR;
4088 else if (pars && data) {
4089 data->flags |= SF_HAS_PAR;
4090 data->flags &= ~SF_IN_PAR;
4092 if (flags & SCF_DO_STCLASS_OR)
4093 cl_and(data->start_class, and_withp);
4094 if (flags & SCF_TRIE_RESTUDY)
4095 data->flags |= SCF_TRIE_RESTUDY;
4097 DEBUG_STUDYDATA("post-fin:",data,depth);
4099 return min < stopmin ? min : stopmin;
4103 S_add_data(RExC_state_t *pRExC_state, U32 n, const char *s)
4105 U32 count = RExC_rxi->data ? RExC_rxi->data->count : 0;
4107 PERL_ARGS_ASSERT_ADD_DATA;
4109 Renewc(RExC_rxi->data,
4110 sizeof(*RExC_rxi->data) + sizeof(void*) * (count + n - 1),
4111 char, struct reg_data);
4113 Renew(RExC_rxi->data->what, count + n, U8);
4115 Newx(RExC_rxi->data->what, n, U8);
4116 RExC_rxi->data->count = count + n;
4117 Copy(s, RExC_rxi->data->what + count, n, U8);
4121 /*XXX: todo make this not included in a non debugging perl */
4122 #ifndef PERL_IN_XSUB_RE
4124 Perl_reginitcolors(pTHX)
4127 const char * const s = PerlEnv_getenv("PERL_RE_COLORS");
4129 char *t = savepv(s);
4133 t = strchr(t, '\t');
4139 PL_colors[i] = t = (char *)"";
4144 PL_colors[i++] = (char *)"";
4151 #ifdef TRIE_STUDY_OPT
4152 #define CHECK_RESTUDY_GOTO \
4154 (data.flags & SCF_TRIE_RESTUDY) \
4158 #define CHECK_RESTUDY_GOTO
4162 - pregcomp - compile a regular expression into internal code
4164 * We can't allocate space until we know how big the compiled form will be,
4165 * but we can't compile it (and thus know how big it is) until we've got a
4166 * place to put the code. So we cheat: we compile it twice, once with code
4167 * generation turned off and size counting turned on, and once "for real".
4168 * This also means that we don't allocate space until we are sure that the
4169 * thing really will compile successfully, and we never have to move the
4170 * code and thus invalidate pointers into it. (Note that it has to be in
4171 * one piece because free() must be able to free it all.) [NB: not true in perl]
4173 * Beware that the optimization-preparation code in here knows about some
4174 * of the structure of the compiled regexp. [I'll say.]
4179 #ifndef PERL_IN_XSUB_RE
4180 #define RE_ENGINE_PTR &PL_core_reg_engine
4182 extern const struct regexp_engine my_reg_engine;
4183 #define RE_ENGINE_PTR &my_reg_engine
4186 #ifndef PERL_IN_XSUB_RE
4188 Perl_pregcomp(pTHX_ SV * const pattern, const U32 flags)
4191 HV * const table = GvHV(PL_hintgv);
4193 PERL_ARGS_ASSERT_PREGCOMP;
4195 /* Dispatch a request to compile a regexp to correct
4198 SV **ptr= hv_fetchs(table, "regcomp", FALSE);
4199 GET_RE_DEBUG_FLAGS_DECL;
4200 if (ptr && SvIOK(*ptr) && SvIV(*ptr)) {
4201 const regexp_engine *eng=INT2PTR(regexp_engine*,SvIV(*ptr));
4203 PerlIO_printf(Perl_debug_log, "Using engine %"UVxf"\n",
4206 return CALLREGCOMP_ENG(eng, pattern, flags);
4209 return Perl_re_compile(aTHX_ pattern, flags);
4214 Perl_re_compile(pTHX_ SV * const pattern, U32 pm_flags)
4219 register regexp_internal *ri;
4221 char *exp = SvPV(pattern, plen);
4222 char* xend = exp + plen;
4229 RExC_state_t RExC_state;
4230 RExC_state_t * const pRExC_state = &RExC_state;
4231 #ifdef TRIE_STUDY_OPT
4233 RExC_state_t copyRExC_state;
4235 GET_RE_DEBUG_FLAGS_DECL;
4237 PERL_ARGS_ASSERT_RE_COMPILE;
4239 DEBUG_r(if (!PL_colorset) reginitcolors());
4241 RExC_utf8 = RExC_orig_utf8 = SvUTF8(pattern);
4244 SV *dsv= sv_newmortal();
4245 RE_PV_QUOTED_DECL(s, RExC_utf8,
4246 dsv, exp, plen, 60);
4247 PerlIO_printf(Perl_debug_log, "%sCompiling REx%s %s\n",
4248 PL_colors[4],PL_colors[5],s);
4253 RExC_flags = pm_flags;
4257 RExC_seen_zerolen = *exp == '^' ? -1 : 0;
4258 RExC_seen_evals = 0;
4261 /* First pass: determine size, legality. */
4269 RExC_emit = &PL_regdummy;
4270 RExC_whilem_seen = 0;
4271 RExC_charnames = NULL;
4272 RExC_open_parens = NULL;
4273 RExC_close_parens = NULL;
4275 RExC_paren_names = NULL;
4277 RExC_paren_name_list = NULL;
4279 RExC_recurse = NULL;
4280 RExC_recurse_count = 0;
4282 #if 0 /* REGC() is (currently) a NOP at the first pass.
4283 * Clever compilers notice this and complain. --jhi */
4284 REGC((U8)REG_MAGIC, (char*)RExC_emit);
4286 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "Starting first pass (sizing)\n"));
4287 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4288 RExC_precomp = NULL;
4291 if (RExC_utf8 && !RExC_orig_utf8) {
4292 /* It's possible to write a regexp in ascii that represents Unicode
4293 codepoints outside of the byte range, such as via \x{100}. If we
4294 detect such a sequence we have to convert the entire pattern to utf8
4295 and then recompile, as our sizing calculation will have been based
4296 on 1 byte == 1 character, but we will need to use utf8 to encode
4297 at least some part of the pattern, and therefore must convert the whole
4299 XXX: somehow figure out how to make this less expensive...
4302 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log,
4303 "UTF8 mismatch! Converting to utf8 for resizing and compile\n"));
4304 exp = (char*)Perl_bytes_to_utf8(aTHX_ (U8*)exp, &len);
4306 RExC_orig_utf8 = RExC_utf8;
4308 goto redo_first_pass;
4311 PerlIO_printf(Perl_debug_log,
4312 "Required size %"IVdf" nodes\n"
4313 "Starting second pass (creation)\n",
4316 RExC_lastparse=NULL;
4318 /* Small enough for pointer-storage convention?
4319 If extralen==0, this means that we will not need long jumps. */
4320 if (RExC_size >= 0x10000L && RExC_extralen)
4321 RExC_size += RExC_extralen;
4324 if (RExC_whilem_seen > 15)
4325 RExC_whilem_seen = 15;
4327 /* Allocate space and zero-initialize. Note, the two step process
4328 of zeroing when in debug mode, thus anything assigned has to
4329 happen after that */
4330 rx = (REGEXP*) newSV_type(SVt_REGEXP);
4331 r = (struct regexp*)SvANY(rx);
4332 Newxc(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode),
4333 char, regexp_internal);
4334 if ( r == NULL || ri == NULL )
4335 FAIL("Regexp out of space");
4337 /* avoid reading uninitialized memory in DEBUGGING code in study_chunk() */
4338 Zero(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode), char);
4340 /* bulk initialize base fields with 0. */
4341 Zero(ri, sizeof(regexp_internal), char);
4344 /* non-zero initialization begins here */
4346 r->engine= RE_ENGINE_PTR;
4347 r->extflags = pm_flags;
4349 bool has_p = ((r->extflags & RXf_PMf_KEEPCOPY) == RXf_PMf_KEEPCOPY);
4350 bool has_minus = ((r->extflags & RXf_PMf_STD_PMMOD) != RXf_PMf_STD_PMMOD);
4351 bool has_runon = ((RExC_seen & REG_SEEN_RUN_ON_COMMENT)==REG_SEEN_RUN_ON_COMMENT);
4352 U16 reganch = (U16)((r->extflags & RXf_PMf_STD_PMMOD)
4353 >> RXf_PMf_STD_PMMOD_SHIFT);
4354 const char *fptr = STD_PAT_MODS; /*"msix"*/
4356 const STRLEN wraplen = plen + has_minus + has_p + has_runon
4357 + (sizeof(STD_PAT_MODS) - 1)
4358 + (sizeof("(?:)") - 1);
4360 p = sv_grow(MUTABLE_SV(rx), wraplen + 1);
4361 SvCUR_set(rx, wraplen);
4363 SvFLAGS(rx) |= SvUTF8(pattern);
4366 *p++ = KEEPCOPY_PAT_MOD; /*'p'*/
4368 char *r = p + (sizeof(STD_PAT_MODS) - 1) + has_minus - 1;
4369 char *colon = r + 1;
4372 while((ch = *fptr++)) {
4386 Copy(RExC_precomp, p, plen, char);
4387 assert ((RX_WRAPPED(rx) - p) < 16);
4388 r->pre_prefix = p - RX_WRAPPED(rx);
4397 r->nparens = RExC_npar - 1; /* set early to validate backrefs */
4399 if (RExC_seen & REG_SEEN_RECURSE) {
4400 Newxz(RExC_open_parens, RExC_npar,regnode *);
4401 SAVEFREEPV(RExC_open_parens);
4402 Newxz(RExC_close_parens,RExC_npar,regnode *);
4403 SAVEFREEPV(RExC_close_parens);
4406 /* Useful during FAIL. */
4407 #ifdef RE_TRACK_PATTERN_OFFSETS
4408 Newxz(ri->u.offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
4409 DEBUG_OFFSETS_r(PerlIO_printf(Perl_debug_log,
4410 "%s %"UVuf" bytes for offset annotations.\n",
4411 ri->u.offsets ? "Got" : "Couldn't get",
4412 (UV)((2*RExC_size+1) * sizeof(U32))));
4414 SetProgLen(ri,RExC_size);
4418 REH_CALL_COMP_BEGIN_HOOK(pRExC_state->rx);
4420 /* Second pass: emit code. */
4421 RExC_flags = pm_flags; /* don't let top level (?i) bleed */
4426 RExC_emit_start = ri->program;
4427 RExC_emit = ri->program;
4428 RExC_emit_bound = ri->program + RExC_size + 1;
4430 /* Store the count of eval-groups for security checks: */
4431 RExC_rx->seen_evals = RExC_seen_evals;
4432 REGC((U8)REG_MAGIC, (char*) RExC_emit++);
4433 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4437 /* XXXX To minimize changes to RE engine we always allocate
4438 3-units-long substrs field. */
4439 Newx(r->substrs, 1, struct reg_substr_data);
4440 if (RExC_recurse_count) {
4441 Newxz(RExC_recurse,RExC_recurse_count,regnode *);
4442 SAVEFREEPV(RExC_recurse);
4446 r->minlen = minlen = sawplus = sawopen = 0;
4447 Zero(r->substrs, 1, struct reg_substr_data);
4449 #ifdef TRIE_STUDY_OPT
4451 StructCopy(&zero_scan_data, &data, scan_data_t);
4452 copyRExC_state = RExC_state;
4455 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log,"Restudying\n"));
4457 RExC_state = copyRExC_state;
4458 if (seen & REG_TOP_LEVEL_BRANCHES)
4459 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
4461 RExC_seen &= ~REG_TOP_LEVEL_BRANCHES;
4462 if (data.last_found) {
4463 SvREFCNT_dec(data.longest_fixed);
4464 SvREFCNT_dec(data.longest_float);
4465 SvREFCNT_dec(data.last_found);
4467 StructCopy(&zero_scan_data, &data, scan_data_t);
4470 StructCopy(&zero_scan_data, &data, scan_data_t);
4473 /* Dig out information for optimizations. */
4474 r->extflags = RExC_flags; /* was pm_op */
4475 /*dmq: removed as part of de-PMOP: pm->op_pmflags = RExC_flags; */
4478 SvUTF8_on(rx); /* Unicode in it? */
4479 ri->regstclass = NULL;
4480 if (RExC_naughty >= 10) /* Probably an expensive pattern. */
4481 r->intflags |= PREGf_NAUGHTY;
4482 scan = ri->program + 1; /* First BRANCH. */
4484 /* testing for BRANCH here tells us whether there is "must appear"
4485 data in the pattern. If there is then we can use it for optimisations */
4486 if (!(RExC_seen & REG_TOP_LEVEL_BRANCHES)) { /* Only one top-level choice. */
4488 STRLEN longest_float_length, longest_fixed_length;
4489 struct regnode_charclass_class ch_class; /* pointed to by data */
4491 I32 last_close = 0; /* pointed to by data */
4492 regnode *first= scan;
4493 regnode *first_next= regnext(first);
4496 * Skip introductions and multiplicators >= 1
4497 * so that we can extract the 'meat' of the pattern that must
4498 * match in the large if() sequence following.
4499 * NOTE that EXACT is NOT covered here, as it is normally
4500 * picked up by the optimiser separately.
4502 * This is unfortunate as the optimiser isnt handling lookahead
4503 * properly currently.
4506 while ((OP(first) == OPEN && (sawopen = 1)) ||
4507 /* An OR of *one* alternative - should not happen now. */
4508 (OP(first) == BRANCH && OP(first_next) != BRANCH) ||
4509 /* for now we can't handle lookbehind IFMATCH*/
4510 (OP(first) == IFMATCH && !first->flags) ||
4511 (OP(first) == PLUS) ||
4512 (OP(first) == MINMOD) ||
4513 /* An {n,m} with n>0 */
4514 (PL_regkind[OP(first)] == CURLY && ARG1(first) > 0) ||
4515 (OP(first) == NOTHING && PL_regkind[OP(first_next)] != END ))
4518 * the only op that could be a regnode is PLUS, all the rest
4519 * will be regnode_1 or regnode_2.
4522 if (OP(first) == PLUS)
4525 first += regarglen[OP(first)];
4527 first = NEXTOPER(first);
4528 first_next= regnext(first);
4531 /* Starting-point info. */
4533 DEBUG_PEEP("first:",first,0);
4534 /* Ignore EXACT as we deal with it later. */
4535 if (PL_regkind[OP(first)] == EXACT) {
4536 if (OP(first) == EXACT)
4537 NOOP; /* Empty, get anchored substr later. */
4538 else if ((OP(first) == EXACTF || OP(first) == EXACTFL))
4539 ri->regstclass = first;
4542 else if (PL_regkind[OP(first)] == TRIE &&
4543 ((reg_trie_data *)ri->data->data[ ARG(first) ])->minlen>0)
4546 /* this can happen only on restudy */
4547 if ( OP(first) == TRIE ) {
4548 struct regnode_1 *trieop = (struct regnode_1 *)
4549 PerlMemShared_calloc(1, sizeof(struct regnode_1));
4550 StructCopy(first,trieop,struct regnode_1);
4551 trie_op=(regnode *)trieop;
4553 struct regnode_charclass *trieop = (struct regnode_charclass *)
4554 PerlMemShared_calloc(1, sizeof(struct regnode_charclass));
4555 StructCopy(first,trieop,struct regnode_charclass);
4556 trie_op=(regnode *)trieop;
4559 make_trie_failtable(pRExC_state, (regnode *)first, trie_op, 0);
4560 ri->regstclass = trie_op;
4563 else if (strchr((const char*)PL_simple,OP(first)))
4564 ri->regstclass = first;
4565 else if (PL_regkind[OP(first)] == BOUND ||
4566 PL_regkind[OP(first)] == NBOUND)
4567 ri->regstclass = first;
4568 else if (PL_regkind[OP(first)] == BOL) {
4569 r->extflags |= (OP(first) == MBOL
4571 : (OP(first) == SBOL
4574 first = NEXTOPER(first);
4577 else if (OP(first) == GPOS) {
4578 r->extflags |= RXf_ANCH_GPOS;
4579 first = NEXTOPER(first);
4582 else if ((!sawopen || !RExC_sawback) &&
4583 (OP(first) == STAR &&
4584 PL_regkind[OP(NEXTOPER(first))] == REG_ANY) &&
4585 !(r->extflags & RXf_ANCH) && !(RExC_seen & REG_SEEN_EVAL))
4587 /* turn .* into ^.* with an implied $*=1 */
4589 (OP(NEXTOPER(first)) == REG_ANY)
4592 r->extflags |= type;
4593 r->intflags |= PREGf_IMPLICIT;
4594 first = NEXTOPER(first);
4597 if (sawplus && (!sawopen || !RExC_sawback)
4598 && !(RExC_seen & REG_SEEN_EVAL)) /* May examine pos and $& */
4599 /* x+ must match at the 1st pos of run of x's */
4600 r->intflags |= PREGf_SKIP;
4602 /* Scan is after the zeroth branch, first is atomic matcher. */
4603 #ifdef TRIE_STUDY_OPT
4606 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4607 (IV)(first - scan + 1))
4611 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
4612 (IV)(first - scan + 1))
4618 * If there's something expensive in the r.e., find the
4619 * longest literal string that must appear and make it the
4620 * regmust. Resolve ties in favor of later strings, since
4621 * the regstart check works with the beginning of the r.e.
4622 * and avoiding duplication strengthens checking. Not a
4623 * strong reason, but sufficient in the absence of others.
4624 * [Now we resolve ties in favor of the earlier string if
4625 * it happens that c_offset_min has been invalidated, since the
4626 * earlier string may buy us something the later one won't.]
4629 data.longest_fixed = newSVpvs("");
4630 data.longest_float = newSVpvs("");
4631 data.last_found = newSVpvs("");
4632 data.longest = &(data.longest_fixed);
4634 if (!ri->regstclass) {
4635 cl_init(pRExC_state, &ch_class);
4636 data.start_class = &ch_class;
4637 stclass_flag = SCF_DO_STCLASS_AND;
4638 } else /* XXXX Check for BOUND? */
4640 data.last_closep = &last_close;
4642 minlen = study_chunk(pRExC_state, &first, &minlen, &fake, scan + RExC_size, /* Up to end */
4643 &data, -1, NULL, NULL,
4644 SCF_DO_SUBSTR | SCF_WHILEM_VISITED_POS | stclass_flag,0);
4650 if ( RExC_npar == 1 && data.longest == &(data.longest_fixed)
4651 && data.last_start_min == 0 && data.last_end > 0
4652 && !RExC_seen_zerolen
4653 && !(RExC_seen & REG_SEEN_VERBARG)
4654 && (!(RExC_seen & REG_SEEN_GPOS) || (r->extflags & RXf_ANCH_GPOS)))
4655 r->extflags |= RXf_CHECK_ALL;
4656 scan_commit(pRExC_state, &data,&minlen,0);
4657 SvREFCNT_dec(data.last_found);
4659 /* Note that code very similar to this but for anchored string
4660 follows immediately below, changes may need to be made to both.
4663 longest_float_length = CHR_SVLEN(data.longest_float);
4664 if (longest_float_length
4665 || (data.flags & SF_FL_BEFORE_EOL
4666 && (!(data.flags & SF_FL_BEFORE_MEOL)
4667 || (RExC_flags & RXf_PMf_MULTILINE))))
4671 if (SvCUR(data.longest_fixed) /* ok to leave SvCUR */
4672 && data.offset_fixed == data.offset_float_min
4673 && SvCUR(data.longest_fixed) == SvCUR(data.longest_float))
4674 goto remove_float; /* As in (a)+. */
4676 /* copy the information about the longest float from the reg_scan_data
4677 over to the program. */
4678 if (SvUTF8(data.longest_float)) {
4679 r->float_utf8 = data.longest_float;
4680 r->float_substr = NULL;
4682 r->float_substr = data.longest_float;
4683 r->float_utf8 = NULL;
4685 /* float_end_shift is how many chars that must be matched that
4686 follow this item. We calculate it ahead of time as once the
4687 lookbehind offset is added in we lose the ability to correctly
4689 ml = data.minlen_float ? *(data.minlen_float)
4690 : (I32)longest_float_length;
4691 r->float_end_shift = ml - data.offset_float_min
4692 - longest_float_length + (SvTAIL(data.longest_float) != 0)
4693 + data.lookbehind_float;
4694 r->float_min_offset = data.offset_float_min - data.lookbehind_float;
4695 r->float_max_offset = data.offset_float_max;
4696 if (data.offset_float_max < I32_MAX) /* Don't offset infinity */
4697 r->float_max_offset -= data.lookbehind_float;
4699 t = (data.flags & SF_FL_BEFORE_EOL /* Can't have SEOL and MULTI */
4700 && (!(data.flags & SF_FL_BEFORE_MEOL)
4701 || (RExC_flags & RXf_PMf_MULTILINE)));
4702 fbm_compile(data.longest_float, t ? FBMcf_TAIL : 0);
4706 r->float_substr = r->float_utf8 = NULL;
4707 SvREFCNT_dec(data.longest_float);
4708 longest_float_length = 0;
4711 /* Note that code very similar to this but for floating string
4712 is immediately above, changes may need to be made to both.
4715 longest_fixed_length = CHR_SVLEN(data.longest_fixed);
4716 if (longest_fixed_length
4717 || (data.flags & SF_FIX_BEFORE_EOL /* Cannot have SEOL and MULTI */
4718 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4719 || (RExC_flags & RXf_PMf_MULTILINE))))
4723 /* copy the information about the longest fixed
4724 from the reg_scan_data over to the program. */
4725 if (SvUTF8(data.longest_fixed)) {
4726 r->anchored_utf8 = data.longest_fixed;
4727 r->anchored_substr = NULL;
4729 r->anchored_substr = data.longest_fixed;
4730 r->anchored_utf8 = NULL;
4732 /* fixed_end_shift is how many chars that must be matched that
4733 follow this item. We calculate it ahead of time as once the
4734 lookbehind offset is added in we lose the ability to correctly
4736 ml = data.minlen_fixed ? *(data.minlen_fixed)
4737 : (I32)longest_fixed_length;
4738 r->anchored_end_shift = ml - data.offset_fixed
4739 - longest_fixed_length + (SvTAIL(data.longest_fixed) != 0)
4740 + data.lookbehind_fixed;
4741 r->anchored_offset = data.offset_fixed - data.lookbehind_fixed;
4743 t = (data.flags & SF_FIX_BEFORE_EOL /* Can't have SEOL and MULTI */
4744 && (!(data.flags & SF_FIX_BEFORE_MEOL)
4745 || (RExC_flags & RXf_PMf_MULTILINE)));
4746 fbm_compile(data.longest_fixed, t ? FBMcf_TAIL : 0);
4749 r->anchored_substr = r->anchored_utf8 = NULL;
4750 SvREFCNT_dec(data.longest_fixed);
4751 longest_fixed_length = 0;
4754 && (OP(ri->regstclass) == REG_ANY || OP(ri->regstclass) == SANY))
4755 ri->regstclass = NULL;
4756 if ((!(r->anchored_substr || r->anchored_utf8) || r->anchored_offset)
4758 && !(data.start_class->flags & ANYOF_EOS)
4759 && !cl_is_anything(data.start_class))
4761 const U32 n = add_data(pRExC_state, 1, "f");
4763 Newx(RExC_rxi->data->data[n], 1,
4764 struct regnode_charclass_class);
4765 StructCopy(data.start_class,
4766 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4767 struct regnode_charclass_class);
4768 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4769 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4770 DEBUG_COMPILE_r({ SV *sv = sv_newmortal();
4771 regprop(r, sv, (regnode*)data.start_class);
4772 PerlIO_printf(Perl_debug_log,
4773 "synthetic stclass \"%s\".\n",
4774 SvPVX_const(sv));});
4777 /* A temporary algorithm prefers floated substr to fixed one to dig more info. */
4778 if (longest_fixed_length > longest_float_length) {
4779 r->check_end_shift = r->anchored_end_shift;
4780 r->check_substr = r->anchored_substr;
4781 r->check_utf8 = r->anchored_utf8;
4782 r->check_offset_min = r->check_offset_max = r->anchored_offset;
4783 if (r->extflags & RXf_ANCH_SINGLE)
4784 r->extflags |= RXf_NOSCAN;
4787 r->check_end_shift = r->float_end_shift;
4788 r->check_substr = r->float_substr;
4789 r->check_utf8 = r->float_utf8;
4790 r->check_offset_min = r->float_min_offset;
4791 r->check_offset_max = r->float_max_offset;
4793 /* XXXX Currently intuiting is not compatible with ANCH_GPOS.
4794 This should be changed ASAP! */
4795 if ((r->check_substr || r->check_utf8) && !(r->extflags & RXf_ANCH_GPOS)) {
4796 r->extflags |= RXf_USE_INTUIT;
4797 if (SvTAIL(r->check_substr ? r->check_substr : r->check_utf8))
4798 r->extflags |= RXf_INTUIT_TAIL;
4800 /* XXX Unneeded? dmq (shouldn't as this is handled elsewhere)
4801 if ( (STRLEN)minlen < longest_float_length )
4802 minlen= longest_float_length;
4803 if ( (STRLEN)minlen < longest_fixed_length )
4804 minlen= longest_fixed_length;
4808 /* Several toplevels. Best we can is to set minlen. */
4810 struct regnode_charclass_class ch_class;
4813 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "\nMulti Top Level\n"));
4815 scan = ri->program + 1;
4816 cl_init(pRExC_state, &ch_class);
4817 data.start_class = &ch_class;
4818 data.last_closep = &last_close;
4821 minlen = study_chunk(pRExC_state, &scan, &minlen, &fake, scan + RExC_size,
4822 &data, -1, NULL, NULL, SCF_DO_STCLASS_AND|SCF_WHILEM_VISITED_POS,0);
4826 r->check_substr = r->check_utf8 = r->anchored_substr = r->anchored_utf8
4827 = r->float_substr = r->float_utf8 = NULL;
4828 if (!(data.start_class->flags & ANYOF_EOS)
4829 && !cl_is_anything(data.start_class))
4831 const U32 n = add_data(pRExC_state, 1, "f");
4833 Newx(RExC_rxi->data->data[n], 1,
4834 struct regnode_charclass_class);
4835 StructCopy(data.start_class,
4836 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
4837 struct regnode_charclass_class);
4838 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
4839 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
4840 DEBUG_COMPILE_r({ SV* sv = sv_newmortal();
4841 regprop(r, sv, (regnode*)data.start_class);
4842 PerlIO_printf(Perl_debug_log,
4843 "synthetic stclass \"%s\".\n",
4844 SvPVX_const(sv));});
4848 /* Guard against an embedded (?=) or (?<=) with a longer minlen than
4849 the "real" pattern. */
4851 PerlIO_printf(Perl_debug_log,"minlen: %"IVdf" r->minlen:%"IVdf"\n",
4852 (IV)minlen, (IV)r->minlen);
4854 r->minlenret = minlen;
4855 if (r->minlen < minlen)
4858 if (RExC_seen & REG_SEEN_GPOS)
4859 r->extflags |= RXf_GPOS_SEEN;
4860 if (RExC_seen & REG_SEEN_LOOKBEHIND)
4861 r->extflags |= RXf_LOOKBEHIND_SEEN;
4862 if (RExC_seen & REG_SEEN_EVAL)
4863 r->extflags |= RXf_EVAL_SEEN;
4864 if (RExC_seen & REG_SEEN_CANY)
4865 r->extflags |= RXf_CANY_SEEN;
4866 if (RExC_seen & REG_SEEN_VERBARG)
4867 r->intflags |= PREGf_VERBARG_SEEN;
4868 if (RExC_seen & REG_SEEN_CUTGROUP)
4869 r->intflags |= PREGf_CUTGROUP_SEEN;
4870 if (RExC_paren_names)
4871 RXp_PAREN_NAMES(r) = MUTABLE_HV(SvREFCNT_inc(RExC_paren_names));
4873 RXp_PAREN_NAMES(r) = NULL;
4875 #ifdef STUPID_PATTERN_CHECKS
4876 if (RX_PRELEN(rx) == 0)
4877 r->extflags |= RXf_NULL;
4878 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
4879 /* XXX: this should happen BEFORE we compile */
4880 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
4881 else if (RX_PRELEN(rx) == 3 && memEQ("\\s+", RX_PRECOMP(rx), 3))
4882 r->extflags |= RXf_WHITE;
4883 else if (RX_PRELEN(rx) == 1 && RXp_PRECOMP(rx)[0] == '^')
4884 r->extflags |= RXf_START_ONLY;
4886 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
4887 /* XXX: this should happen BEFORE we compile */
4888 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
4890 regnode *first = ri->program + 1;
4892 U8 nop = OP(NEXTOPER(first));
4894 if (PL_regkind[fop] == NOTHING && nop == END)
4895 r->extflags |= RXf_NULL;
4896 else if (PL_regkind[fop] == BOL && nop == END)
4897 r->extflags |= RXf_START_ONLY;
4898 else if (fop == PLUS && nop ==SPACE && OP(regnext(first))==END)
4899 r->extflags |= RXf_WHITE;
4903 if (RExC_paren_names) {
4904 ri->name_list_idx = add_data( pRExC_state, 1, "p" );
4905 ri->data->data[ri->name_list_idx] = (void*)SvREFCNT_inc(RExC_paren_name_list);
4908 ri->name_list_idx = 0;
4910 if (RExC_recurse_count) {
4911 for ( ; RExC_recurse_count ; RExC_recurse_count-- ) {
4912 const regnode *scan = RExC_recurse[RExC_recurse_count-1];
4913 ARG2L_SET( scan, RExC_open_parens[ARG(scan)-1] - scan );
4916 Newxz(r->offs, RExC_npar, regexp_paren_pair);
4917 /* assume we don't need to swap parens around before we match */
4920 PerlIO_printf(Perl_debug_log,"Final program:\n");
4923 #ifdef RE_TRACK_PATTERN_OFFSETS
4924 DEBUG_OFFSETS_r(if (ri->u.offsets) {
4925 const U32 len = ri->u.offsets[0];
4927 GET_RE_DEBUG_FLAGS_DECL;
4928 PerlIO_printf(Perl_debug_log, "Offsets: [%"UVuf"]\n\t", (UV)ri->u.offsets[0]);
4929 for (i = 1; i <= len; i++) {
4930 if (ri->u.offsets[i*2-1] || ri->u.offsets[i*2])
4931 PerlIO_printf(Perl_debug_log, "%"UVuf":%"UVuf"[%"UVuf"] ",
4932 (UV)i, (UV)ri->u.offsets[i*2-1], (UV)ri->u.offsets[i*2]);
4934 PerlIO_printf(Perl_debug_log, "\n");
4940 #undef RE_ENGINE_PTR
4944 Perl_reg_named_buff(pTHX_ REGEXP * const rx, SV * const key, SV * const value,
4947 PERL_ARGS_ASSERT_REG_NAMED_BUFF;
4949 PERL_UNUSED_ARG(value);
4951 if (flags & RXapif_FETCH) {
4952 return reg_named_buff_fetch(rx, key, flags);
4953 } else if (flags & (RXapif_STORE | RXapif_DELETE | RXapif_CLEAR)) {
4954 Perl_croak(aTHX_ "%s", PL_no_modify);
4956 } else if (flags & RXapif_EXISTS) {
4957 return reg_named_buff_exists(rx, key, flags)
4960 } else if (flags & RXapif_REGNAMES) {
4961 return reg_named_buff_all(rx, flags);
4962 } else if (flags & (RXapif_SCALAR | RXapif_REGNAMES_COUNT)) {
4963 return reg_named_buff_scalar(rx, flags);
4965 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff", (int)flags);
4971 Perl_reg_named_buff_iter(pTHX_ REGEXP * const rx, const SV * const lastkey,
4974 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ITER;
4975 PERL_UNUSED_ARG(lastkey);
4977 if (flags & RXapif_FIRSTKEY)
4978 return reg_named_buff_firstkey(rx, flags);
4979 else if (flags & RXapif_NEXTKEY)
4980 return reg_named_buff_nextkey(rx, flags);
4982 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_iter", (int)flags);
4988 Perl_reg_named_buff_fetch(pTHX_ REGEXP * const r, SV * const namesv,
4991 AV *retarray = NULL;
4993 struct regexp *const rx = (struct regexp *)SvANY(r);
4995 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FETCH;
4997 if (flags & RXapif_ALL)
5000 if (rx && RXp_PAREN_NAMES(rx)) {
5001 HE *he_str = hv_fetch_ent( RXp_PAREN_NAMES(rx), namesv, 0, 0 );
5004 SV* sv_dat=HeVAL(he_str);
5005 I32 *nums=(I32*)SvPVX(sv_dat);
5006 for ( i=0; i<SvIVX(sv_dat); i++ ) {
5007 if ((I32)(rx->nparens) >= nums[i]
5008 && rx->offs[nums[i]].start != -1
5009 && rx->offs[nums[i]].end != -1)
5012 CALLREG_NUMBUF_FETCH(r,nums[i],ret);
5016 ret = newSVsv(&PL_sv_undef);
5019 av_push(retarray, ret);
5022 return newRV_noinc(MUTABLE_SV(retarray));
5029 Perl_reg_named_buff_exists(pTHX_ REGEXP * const r, SV * const key,
5032 struct regexp *const rx = (struct regexp *)SvANY(r);
5034 PERL_ARGS_ASSERT_REG_NAMED_BUFF_EXISTS;
5036 if (rx && RXp_PAREN_NAMES(rx)) {
5037 if (flags & RXapif_ALL) {
5038 return hv_exists_ent(RXp_PAREN_NAMES(rx), key, 0);
5040 SV *sv = CALLREG_NAMED_BUFF_FETCH(r, key, flags);
5054 Perl_reg_named_buff_firstkey(pTHX_ REGEXP * const r, const U32 flags)
5056 struct regexp *const rx = (struct regexp *)SvANY(r);
5058 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FIRSTKEY;
5060 if ( rx && RXp_PAREN_NAMES(rx) ) {
5061 (void)hv_iterinit(RXp_PAREN_NAMES(rx));
5063 return CALLREG_NAMED_BUFF_NEXTKEY(r, NULL, flags & ~RXapif_FIRSTKEY);
5070 Perl_reg_named_buff_nextkey(pTHX_ REGEXP * const r, const U32 flags)
5072 struct regexp *const rx = (struct regexp *)SvANY(r);
5073 GET_RE_DEBUG_FLAGS_DECL;
5075 PERL_ARGS_ASSERT_REG_NAMED_BUFF_NEXTKEY;
5077 if (rx && RXp_PAREN_NAMES(rx)) {
5078 HV *hv = RXp_PAREN_NAMES(rx);
5080 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5083 SV* sv_dat = HeVAL(temphe);
5084 I32 *nums = (I32*)SvPVX(sv_dat);
5085 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5086 if ((I32)(rx->lastparen) >= nums[i] &&
5087 rx->offs[nums[i]].start != -1 &&
5088 rx->offs[nums[i]].end != -1)
5094 if (parno || flags & RXapif_ALL) {
5095 return newSVhek(HeKEY_hek(temphe));
5103 Perl_reg_named_buff_scalar(pTHX_ REGEXP * const r, const U32 flags)
5108 struct regexp *const rx = (struct regexp *)SvANY(r);
5110 PERL_ARGS_ASSERT_REG_NAMED_BUFF_SCALAR;
5112 if (rx && RXp_PAREN_NAMES(rx)) {
5113 if (flags & (RXapif_ALL | RXapif_REGNAMES_COUNT)) {
5114 return newSViv(HvTOTALKEYS(RXp_PAREN_NAMES(rx)));
5115 } else if (flags & RXapif_ONE) {
5116 ret = CALLREG_NAMED_BUFF_ALL(r, (flags | RXapif_REGNAMES));
5117 av = MUTABLE_AV(SvRV(ret));
5118 length = av_len(av);
5120 return newSViv(length + 1);
5122 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_scalar", (int)flags);
5126 return &PL_sv_undef;
5130 Perl_reg_named_buff_all(pTHX_ REGEXP * const r, const U32 flags)
5132 struct regexp *const rx = (struct regexp *)SvANY(r);
5135 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ALL;
5137 if (rx && RXp_PAREN_NAMES(rx)) {
5138 HV *hv= RXp_PAREN_NAMES(rx);
5140 (void)hv_iterinit(hv);
5141 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5144 SV* sv_dat = HeVAL(temphe);
5145 I32 *nums = (I32*)SvPVX(sv_dat);
5146 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5147 if ((I32)(rx->lastparen) >= nums[i] &&
5148 rx->offs[nums[i]].start != -1 &&
5149 rx->offs[nums[i]].end != -1)
5155 if (parno || flags & RXapif_ALL) {
5156 av_push(av, newSVhek(HeKEY_hek(temphe)));
5161 return newRV_noinc(MUTABLE_SV(av));
5165 Perl_reg_numbered_buff_fetch(pTHX_ REGEXP * const r, const I32 paren,
5168 struct regexp *const rx = (struct regexp *)SvANY(r);
5173 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_FETCH;
5176 sv_setsv(sv,&PL_sv_undef);
5180 if (paren == RX_BUFF_IDX_PREMATCH && rx->offs[0].start != -1) {
5182 i = rx->offs[0].start;
5186 if (paren == RX_BUFF_IDX_POSTMATCH && rx->offs[0].end != -1) {
5188 s = rx->subbeg + rx->offs[0].end;
5189 i = rx->sublen - rx->offs[0].end;
5192 if ( 0 <= paren && paren <= (I32)rx->nparens &&
5193 (s1 = rx->offs[paren].start) != -1 &&
5194 (t1 = rx->offs[paren].end) != -1)
5198 s = rx->subbeg + s1;
5200 sv_setsv(sv,&PL_sv_undef);
5203 assert(rx->sublen >= (s - rx->subbeg) + i );
5205 const int oldtainted = PL_tainted;
5207 sv_setpvn(sv, s, i);
5208 PL_tainted = oldtainted;
5209 if ( (rx->extflags & RXf_CANY_SEEN)
5210 ? (RXp_MATCH_UTF8(rx)
5211 && (!i || is_utf8_string((U8*)s, i)))
5212 : (RXp_MATCH_UTF8(rx)) )
5219 if (RXp_MATCH_TAINTED(rx)) {
5220 if (SvTYPE(sv) >= SVt_PVMG) {
5221 MAGIC* const mg = SvMAGIC(sv);
5224 SvMAGIC_set(sv, mg->mg_moremagic);
5226 if ((mgt = SvMAGIC(sv))) {
5227 mg->mg_moremagic = mgt;
5228 SvMAGIC_set(sv, mg);
5238 sv_setsv(sv,&PL_sv_undef);
5244 Perl_reg_numbered_buff_store(pTHX_ REGEXP * const rx, const I32 paren,
5245 SV const * const value)
5247 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_STORE;
5249 PERL_UNUSED_ARG(rx);
5250 PERL_UNUSED_ARG(paren);
5251 PERL_UNUSED_ARG(value);
5254 Perl_croak(aTHX_ "%s", PL_no_modify);
5258 Perl_reg_numbered_buff_length(pTHX_ REGEXP * const r, const SV * const sv,
5261 struct regexp *const rx = (struct regexp *)SvANY(r);
5265 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_LENGTH;
5267 /* Some of this code was originally in C<Perl_magic_len> in F<mg.c> */
5269 /* $` / ${^PREMATCH} */
5270 case RX_BUFF_IDX_PREMATCH:
5271 if (rx->offs[0].start != -1) {
5272 i = rx->offs[0].start;
5280 /* $' / ${^POSTMATCH} */
5281 case RX_BUFF_IDX_POSTMATCH:
5282 if (rx->offs[0].end != -1) {
5283 i = rx->sublen - rx->offs[0].end;
5285 s1 = rx->offs[0].end;
5291 /* $& / ${^MATCH}, $1, $2, ... */
5293 if (paren <= (I32)rx->nparens &&
5294 (s1 = rx->offs[paren].start) != -1 &&
5295 (t1 = rx->offs[paren].end) != -1)
5300 if (ckWARN(WARN_UNINITIALIZED))
5301 report_uninit((const SV *)sv);
5306 if (i > 0 && RXp_MATCH_UTF8(rx)) {
5307 const char * const s = rx->subbeg + s1;
5312 if (is_utf8_string_loclen((U8*)s, i, &ep, &el))
5319 Perl_reg_qr_package(pTHX_ REGEXP * const rx)
5321 PERL_ARGS_ASSERT_REG_QR_PACKAGE;
5322 PERL_UNUSED_ARG(rx);
5326 return newSVpvs("Regexp");
5329 /* Scans the name of a named buffer from the pattern.
5330 * If flags is REG_RSN_RETURN_NULL returns null.
5331 * If flags is REG_RSN_RETURN_NAME returns an SV* containing the name
5332 * If flags is REG_RSN_RETURN_DATA returns the data SV* corresponding
5333 * to the parsed name as looked up in the RExC_paren_names hash.
5334 * If there is an error throws a vFAIL().. type exception.
5337 #define REG_RSN_RETURN_NULL 0
5338 #define REG_RSN_RETURN_NAME 1
5339 #define REG_RSN_RETURN_DATA 2
5342 S_reg_scan_name(pTHX_ RExC_state_t *pRExC_state, U32 flags)
5344 char *name_start = RExC_parse;
5346 PERL_ARGS_ASSERT_REG_SCAN_NAME;
5348 if (isIDFIRST_lazy_if(RExC_parse, UTF)) {
5349 /* skip IDFIRST by using do...while */
5352 RExC_parse += UTF8SKIP(RExC_parse);
5353 } while (isALNUM_utf8((U8*)RExC_parse));
5357 } while (isALNUM(*RExC_parse));
5362 = newSVpvn_flags(name_start, (int)(RExC_parse - name_start),
5363 SVs_TEMP | (UTF ? SVf_UTF8 : 0));
5364 if ( flags == REG_RSN_RETURN_NAME)
5366 else if (flags==REG_RSN_RETURN_DATA) {
5369 if ( ! sv_name ) /* should not happen*/
5370 Perl_croak(aTHX_ "panic: no svname in reg_scan_name");
5371 if (RExC_paren_names)
5372 he_str = hv_fetch_ent( RExC_paren_names, sv_name, 0, 0 );
5374 sv_dat = HeVAL(he_str);
5376 vFAIL("Reference to nonexistent named group");
5380 Perl_croak(aTHX_ "panic: bad flag in reg_scan_name");
5387 #define DEBUG_PARSE_MSG(funcname) DEBUG_PARSE_r({ \
5388 int rem=(int)(RExC_end - RExC_parse); \
5397 if (RExC_lastparse!=RExC_parse) \
5398 PerlIO_printf(Perl_debug_log," >%.*s%-*s", \
5401 iscut ? "..." : "<" \
5404 PerlIO_printf(Perl_debug_log,"%16s",""); \
5407 num = RExC_size + 1; \
5409 num=REG_NODE_NUM(RExC_emit); \
5410 if (RExC_lastnum!=num) \
5411 PerlIO_printf(Perl_debug_log,"|%4d",num); \
5413 PerlIO_printf(Perl_debug_log,"|%4s",""); \
5414 PerlIO_printf(Perl_debug_log,"|%*s%-4s", \
5415 (int)((depth*2)), "", \
5419 RExC_lastparse=RExC_parse; \
5424 #define DEBUG_PARSE(funcname) DEBUG_PARSE_r({ \
5425 DEBUG_PARSE_MSG((funcname)); \
5426 PerlIO_printf(Perl_debug_log,"%4s","\n"); \
5428 #define DEBUG_PARSE_FMT(funcname,fmt,args) DEBUG_PARSE_r({ \
5429 DEBUG_PARSE_MSG((funcname)); \
5430 PerlIO_printf(Perl_debug_log,fmt "\n",args); \
5433 - reg - regular expression, i.e. main body or parenthesized thing
5435 * Caller must absorb opening parenthesis.
5437 * Combining parenthesis handling with the base level of regular expression
5438 * is a trifle forced, but the need to tie the tails of the branches to what
5439 * follows makes it hard to avoid.
5441 #define REGTAIL(x,y,z) regtail((x),(y),(z),depth+1)
5443 #define REGTAIL_STUDY(x,y,z) regtail_study((x),(y),(z),depth+1)
5445 #define REGTAIL_STUDY(x,y,z) regtail((x),(y),(z),depth+1)
5449 S_reg(pTHX_ RExC_state_t *pRExC_state, I32 paren, I32 *flagp,U32 depth)
5450 /* paren: Parenthesized? 0=top, 1=(, inside: changed to letter. */
5453 register regnode *ret; /* Will be the head of the group. */
5454 register regnode *br;
5455 register regnode *lastbr;
5456 register regnode *ender = NULL;
5457 register I32 parno = 0;
5459 U32 oregflags = RExC_flags;
5460 bool have_branch = 0;
5462 I32 freeze_paren = 0;
5463 I32 after_freeze = 0;
5465 /* for (?g), (?gc), and (?o) warnings; warning
5466 about (?c) will warn about (?g) -- japhy */
5468 #define WASTED_O 0x01
5469 #define WASTED_G 0x02
5470 #define WASTED_C 0x04
5471 #define WASTED_GC (0x02|0x04)
5472 I32 wastedflags = 0x00;
5474 char * parse_start = RExC_parse; /* MJD */
5475 char * const oregcomp_parse = RExC_parse;
5477 GET_RE_DEBUG_FLAGS_DECL;
5479 PERL_ARGS_ASSERT_REG;
5480 DEBUG_PARSE("reg ");
5482 *flagp = 0; /* Tentatively. */
5485 /* Make an OPEN node, if parenthesized. */
5487 if ( *RExC_parse == '*') { /* (*VERB:ARG) */
5488 char *start_verb = RExC_parse;
5489 STRLEN verb_len = 0;
5490 char *start_arg = NULL;
5491 unsigned char op = 0;
5493 int internal_argval = 0; /* internal_argval is only useful if !argok */
5494 while ( *RExC_parse && *RExC_parse != ')' ) {
5495 if ( *RExC_parse == ':' ) {
5496 start_arg = RExC_parse + 1;
5502 verb_len = RExC_parse - start_verb;
5505 while ( *RExC_parse && *RExC_parse != ')' )
5507 if ( *RExC_parse != ')' )
5508 vFAIL("Unterminated verb pattern argument");
5509 if ( RExC_parse == start_arg )
5512 if ( *RExC_parse != ')' )
5513 vFAIL("Unterminated verb pattern");
5516 switch ( *start_verb ) {
5517 case 'A': /* (*ACCEPT) */
5518 if ( memEQs(start_verb,verb_len,"ACCEPT") ) {
5520 internal_argval = RExC_nestroot;
5523 case 'C': /* (*COMMIT) */
5524 if ( memEQs(start_verb,verb_len,"COMMIT") )
5527 case 'F': /* (*FAIL) */
5528 if ( verb_len==1 || memEQs(start_verb,verb_len,"FAIL") ) {
5533 case ':': /* (*:NAME) */
5534 case 'M': /* (*MARK:NAME) */
5535 if ( verb_len==0 || memEQs(start_verb,verb_len,"MARK") ) {
5540 case 'P': /* (*PRUNE) */
5541 if ( memEQs(start_verb,verb_len,"PRUNE") )
5544 case 'S': /* (*SKIP) */
5545 if ( memEQs(start_verb,verb_len,"SKIP") )
5548 case 'T': /* (*THEN) */
5549 /* [19:06] <TimToady> :: is then */
5550 if ( memEQs(start_verb,verb_len,"THEN") ) {
5552 RExC_seen |= REG_SEEN_CUTGROUP;
5558 vFAIL3("Unknown verb pattern '%.*s'",
5559 verb_len, start_verb);
5562 if ( start_arg && internal_argval ) {
5563 vFAIL3("Verb pattern '%.*s' may not have an argument",
5564 verb_len, start_verb);
5565 } else if ( argok < 0 && !start_arg ) {
5566 vFAIL3("Verb pattern '%.*s' has a mandatory argument",
5567 verb_len, start_verb);
5569 ret = reganode(pRExC_state, op, internal_argval);
5570 if ( ! internal_argval && ! SIZE_ONLY ) {
5572 SV *sv = newSVpvn( start_arg, RExC_parse - start_arg);
5573 ARG(ret) = add_data( pRExC_state, 1, "S" );
5574 RExC_rxi->data->data[ARG(ret)]=(void*)sv;
5581 if (!internal_argval)
5582 RExC_seen |= REG_SEEN_VERBARG;
5583 } else if ( start_arg ) {
5584 vFAIL3("Verb pattern '%.*s' may not have an argument",
5585 verb_len, start_verb);
5587 ret = reg_node(pRExC_state, op);
5589 nextchar(pRExC_state);
5592 if (*RExC_parse == '?') { /* (?...) */
5593 bool is_logical = 0;
5594 const char * const seqstart = RExC_parse;
5597 paren = *RExC_parse++;
5598 ret = NULL; /* For look-ahead/behind. */
5601 case 'P': /* (?P...) variants for those used to PCRE/Python */
5602 paren = *RExC_parse++;
5603 if ( paren == '<') /* (?P<...>) named capture */
5605 else if (paren == '>') { /* (?P>name) named recursion */
5606 goto named_recursion;
5608 else if (paren == '=') { /* (?P=...) named backref */
5609 /* this pretty much dupes the code for \k<NAME> in regatom(), if
5610 you change this make sure you change that */
5611 char* name_start = RExC_parse;
5613 SV *sv_dat = reg_scan_name(pRExC_state,
5614 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5615 if (RExC_parse == name_start || *RExC_parse != ')')
5616 vFAIL2("Sequence %.3s... not terminated",parse_start);
5619 num = add_data( pRExC_state, 1, "S" );
5620 RExC_rxi->data->data[num]=(void*)sv_dat;
5621 SvREFCNT_inc_simple_void(sv_dat);
5624 ret = reganode(pRExC_state,
5625 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
5629 Set_Node_Offset(ret, parse_start+1);
5630 Set_Node_Cur_Length(ret); /* MJD */
5632 nextchar(pRExC_state);
5636 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5638 case '<': /* (?<...) */
5639 if (*RExC_parse == '!')
5641 else if (*RExC_parse != '=')
5647 case '\'': /* (?'...') */
5648 name_start= RExC_parse;
5649 svname = reg_scan_name(pRExC_state,
5650 SIZE_ONLY ? /* reverse test from the others */
5651 REG_RSN_RETURN_NAME :
5652 REG_RSN_RETURN_NULL);
5653 if (RExC_parse == name_start) {
5655 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5658 if (*RExC_parse != paren)
5659 vFAIL2("Sequence (?%c... not terminated",
5660 paren=='>' ? '<' : paren);
5664 if (!svname) /* shouldnt happen */
5666 "panic: reg_scan_name returned NULL");
5667 if (!RExC_paren_names) {
5668 RExC_paren_names= newHV();
5669 sv_2mortal(MUTABLE_SV(RExC_paren_names));
5671 RExC_paren_name_list= newAV();
5672 sv_2mortal(MUTABLE_SV(RExC_paren_name_list));
5675 he_str = hv_fetch_ent( RExC_paren_names, svname, 1, 0 );
5677 sv_dat = HeVAL(he_str);
5679 /* croak baby croak */
5681 "panic: paren_name hash element allocation failed");
5682 } else if ( SvPOK(sv_dat) ) {
5683 /* (?|...) can mean we have dupes so scan to check
5684 its already been stored. Maybe a flag indicating
5685 we are inside such a construct would be useful,
5686 but the arrays are likely to be quite small, so
5687 for now we punt -- dmq */
5688 IV count = SvIV(sv_dat);
5689 I32 *pv = (I32*)SvPVX(sv_dat);
5691 for ( i = 0 ; i < count ; i++ ) {
5692 if ( pv[i] == RExC_npar ) {
5698 pv = (I32*)SvGROW(sv_dat, SvCUR(sv_dat) + sizeof(I32)+1);
5699 SvCUR_set(sv_dat, SvCUR(sv_dat) + sizeof(I32));
5700 pv[count] = RExC_npar;
5701 SvIV_set(sv_dat, SvIVX(sv_dat) + 1);
5704 (void)SvUPGRADE(sv_dat,SVt_PVNV);
5705 sv_setpvn(sv_dat, (char *)&(RExC_npar), sizeof(I32));
5707 SvIV_set(sv_dat, 1);
5710 if (!av_store(RExC_paren_name_list, RExC_npar, SvREFCNT_inc(svname)))
5711 SvREFCNT_dec(svname);
5714 /*sv_dump(sv_dat);*/
5716 nextchar(pRExC_state);
5718 goto capturing_parens;
5720 RExC_seen |= REG_SEEN_LOOKBEHIND;
5722 case '=': /* (?=...) */
5723 RExC_seen_zerolen++;
5725 case '!': /* (?!...) */
5726 RExC_seen_zerolen++;
5727 if (*RExC_parse == ')') {
5728 ret=reg_node(pRExC_state, OPFAIL);
5729 nextchar(pRExC_state);
5733 case '|': /* (?|...) */
5734 /* branch reset, behave like a (?:...) except that
5735 buffers in alternations share the same numbers */
5737 after_freeze = freeze_paren = RExC_npar;
5739 case ':': /* (?:...) */
5740 case '>': /* (?>...) */
5742 case '$': /* (?$...) */
5743 case '@': /* (?@...) */
5744 vFAIL2("Sequence (?%c...) not implemented", (int)paren);
5746 case '#': /* (?#...) */
5747 while (*RExC_parse && *RExC_parse != ')')
5749 if (*RExC_parse != ')')
5750 FAIL("Sequence (?#... not terminated");
5751 nextchar(pRExC_state);
5754 case '0' : /* (?0) */
5755 case 'R' : /* (?R) */
5756 if (*RExC_parse != ')')
5757 FAIL("Sequence (?R) not terminated");
5758 ret = reg_node(pRExC_state, GOSTART);
5759 *flagp |= POSTPONED;
5760 nextchar(pRExC_state);
5763 { /* named and numeric backreferences */
5765 case '&': /* (?&NAME) */
5766 parse_start = RExC_parse - 1;
5769 SV *sv_dat = reg_scan_name(pRExC_state,
5770 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5771 num = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5773 goto gen_recurse_regop;
5776 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5778 vFAIL("Illegal pattern");
5780 goto parse_recursion;
5782 case '-': /* (?-1) */
5783 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
5784 RExC_parse--; /* rewind to let it be handled later */
5788 case '1': case '2': case '3': case '4': /* (?1) */
5789 case '5': case '6': case '7': case '8': case '9':
5792 num = atoi(RExC_parse);
5793 parse_start = RExC_parse - 1; /* MJD */
5794 if (*RExC_parse == '-')
5796 while (isDIGIT(*RExC_parse))
5798 if (*RExC_parse!=')')
5799 vFAIL("Expecting close bracket");
5802 if ( paren == '-' ) {
5804 Diagram of capture buffer numbering.
5805 Top line is the normal capture buffer numbers
5806 Botton line is the negative indexing as from
5810 /(a(x)y)(a(b(c(?-2)d)e)f)(g(h))/
5814 num = RExC_npar + num;
5817 vFAIL("Reference to nonexistent group");
5819 } else if ( paren == '+' ) {
5820 num = RExC_npar + num - 1;
5823 ret = reganode(pRExC_state, GOSUB, num);
5825 if (num > (I32)RExC_rx->nparens) {
5827 vFAIL("Reference to nonexistent group");
5829 ARG2L_SET( ret, RExC_recurse_count++);
5831 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
5832 "Recurse #%"UVuf" to %"IVdf"\n", (UV)ARG(ret), (IV)ARG2L(ret)));
5836 RExC_seen |= REG_SEEN_RECURSE;
5837 Set_Node_Length(ret, 1 + regarglen[OP(ret)]); /* MJD */
5838 Set_Node_Offset(ret, parse_start); /* MJD */
5840 *flagp |= POSTPONED;
5841 nextchar(pRExC_state);
5843 } /* named and numeric backreferences */
5846 case '?': /* (??...) */
5848 if (*RExC_parse != '{') {
5850 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
5853 *flagp |= POSTPONED;
5854 paren = *RExC_parse++;
5856 case '{': /* (?{...}) */
5861 char *s = RExC_parse;
5863 RExC_seen_zerolen++;
5864 RExC_seen |= REG_SEEN_EVAL;
5865 while (count && (c = *RExC_parse)) {
5876 if (*RExC_parse != ')') {
5878 vFAIL("Sequence (?{...}) not terminated or not {}-balanced");
5882 OP_4tree *sop, *rop;
5883 SV * const sv = newSVpvn(s, RExC_parse - 1 - s);
5886 Perl_save_re_context(aTHX);
5887 rop = sv_compile_2op(sv, &sop, "re", &pad);
5888 sop->op_private |= OPpREFCOUNTED;
5889 /* re_dup will OpREFCNT_inc */
5890 OpREFCNT_set(sop, 1);
5893 n = add_data(pRExC_state, 3, "nop");
5894 RExC_rxi->data->data[n] = (void*)rop;
5895 RExC_rxi->data->data[n+1] = (void*)sop;
5896 RExC_rxi->data->data[n+2] = (void*)pad;
5899 else { /* First pass */
5900 if (PL_reginterp_cnt < ++RExC_seen_evals
5902 /* No compiled RE interpolated, has runtime
5903 components ===> unsafe. */
5904 FAIL("Eval-group not allowed at runtime, use re 'eval'");
5905 if (PL_tainting && PL_tainted)
5906 FAIL("Eval-group in insecure regular expression");
5907 #if PERL_VERSION > 8
5908 if (IN_PERL_COMPILETIME)
5913 nextchar(pRExC_state);
5915 ret = reg_node(pRExC_state, LOGICAL);
5918 REGTAIL(pRExC_state, ret, reganode(pRExC_state, EVAL, n));
5919 /* deal with the length of this later - MJD */
5922 ret = reganode(pRExC_state, EVAL, n);
5923 Set_Node_Length(ret, RExC_parse - parse_start + 1);
5924 Set_Node_Offset(ret, parse_start);
5927 case '(': /* (?(?{...})...) and (?(?=...)...) */
5930 if (RExC_parse[0] == '?') { /* (?(?...)) */
5931 if (RExC_parse[1] == '=' || RExC_parse[1] == '!'
5932 || RExC_parse[1] == '<'
5933 || RExC_parse[1] == '{') { /* Lookahead or eval. */
5936 ret = reg_node(pRExC_state, LOGICAL);
5939 REGTAIL(pRExC_state, ret, reg(pRExC_state, 1, &flag,depth+1));
5943 else if ( RExC_parse[0] == '<' /* (?(<NAME>)...) */
5944 || RExC_parse[0] == '\'' ) /* (?('NAME')...) */
5946 char ch = RExC_parse[0] == '<' ? '>' : '\'';
5947 char *name_start= RExC_parse++;
5949 SV *sv_dat=reg_scan_name(pRExC_state,
5950 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5951 if (RExC_parse == name_start || *RExC_parse != ch)
5952 vFAIL2("Sequence (?(%c... not terminated",
5953 (ch == '>' ? '<' : ch));
5956 num = add_data( pRExC_state, 1, "S" );
5957 RExC_rxi->data->data[num]=(void*)sv_dat;
5958 SvREFCNT_inc_simple_void(sv_dat);
5960 ret = reganode(pRExC_state,NGROUPP,num);
5961 goto insert_if_check_paren;
5963 else if (RExC_parse[0] == 'D' &&
5964 RExC_parse[1] == 'E' &&
5965 RExC_parse[2] == 'F' &&
5966 RExC_parse[3] == 'I' &&
5967 RExC_parse[4] == 'N' &&
5968 RExC_parse[5] == 'E')
5970 ret = reganode(pRExC_state,DEFINEP,0);
5973 goto insert_if_check_paren;
5975 else if (RExC_parse[0] == 'R') {
5978 if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
5979 parno = atoi(RExC_parse++);
5980 while (isDIGIT(*RExC_parse))
5982 } else if (RExC_parse[0] == '&') {
5985 sv_dat = reg_scan_name(pRExC_state,
5986 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
5987 parno = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
5989 ret = reganode(pRExC_state,INSUBP,parno);
5990 goto insert_if_check_paren;
5992 else if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
5995 parno = atoi(RExC_parse++);
5997 while (isDIGIT(*RExC_parse))
5999 ret = reganode(pRExC_state, GROUPP, parno);
6001 insert_if_check_paren:
6002 if ((c = *nextchar(pRExC_state)) != ')')
6003 vFAIL("Switch condition not recognized");
6005 REGTAIL(pRExC_state, ret, reganode(pRExC_state, IFTHEN, 0));
6006 br = regbranch(pRExC_state, &flags, 1,depth+1);
6008 br = reganode(pRExC_state, LONGJMP, 0);
6010 REGTAIL(pRExC_state, br, reganode(pRExC_state, LONGJMP, 0));
6011 c = *nextchar(pRExC_state);
6016 vFAIL("(?(DEFINE)....) does not allow branches");
6017 lastbr = reganode(pRExC_state, IFTHEN, 0); /* Fake one for optimizer. */
6018 regbranch(pRExC_state, &flags, 1,depth+1);
6019 REGTAIL(pRExC_state, ret, lastbr);
6022 c = *nextchar(pRExC_state);
6027 vFAIL("Switch (?(condition)... contains too many branches");
6028 ender = reg_node(pRExC_state, TAIL);
6029 REGTAIL(pRExC_state, br, ender);
6031 REGTAIL(pRExC_state, lastbr, ender);
6032 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender);
6035 REGTAIL(pRExC_state, ret, ender);
6036 RExC_size++; /* XXX WHY do we need this?!!
6037 For large programs it seems to be required
6038 but I can't figure out why. -- dmq*/
6042 vFAIL2("Unknown switch condition (?(%.2s", RExC_parse);
6046 RExC_parse--; /* for vFAIL to print correctly */
6047 vFAIL("Sequence (? incomplete");
6051 parse_flags: /* (?i) */
6053 U32 posflags = 0, negflags = 0;
6054 U32 *flagsp = &posflags;
6056 while (*RExC_parse) {
6057 /* && strchr("iogcmsx", *RExC_parse) */
6058 /* (?g), (?gc) and (?o) are useless here
6059 and must be globally applied -- japhy */
6060 switch (*RExC_parse) {
6061 CASE_STD_PMMOD_FLAGS_PARSE_SET(flagsp);
6062 case ONCE_PAT_MOD: /* 'o' */
6063 case GLOBAL_PAT_MOD: /* 'g' */
6064 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6065 const I32 wflagbit = *RExC_parse == 'o' ? WASTED_O : WASTED_G;
6066 if (! (wastedflags & wflagbit) ) {
6067 wastedflags |= wflagbit;
6070 "Useless (%s%c) - %suse /%c modifier",
6071 flagsp == &negflags ? "?-" : "?",
6073 flagsp == &negflags ? "don't " : "",
6080 case CONTINUE_PAT_MOD: /* 'c' */
6081 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
6082 if (! (wastedflags & WASTED_C) ) {
6083 wastedflags |= WASTED_GC;
6086 "Useless (%sc) - %suse /gc modifier",
6087 flagsp == &negflags ? "?-" : "?",
6088 flagsp == &negflags ? "don't " : ""
6093 case KEEPCOPY_PAT_MOD: /* 'p' */
6094 if (flagsp == &negflags) {
6096 ckWARNreg(RExC_parse + 1,"Useless use of (?-p)");
6098 *flagsp |= RXf_PMf_KEEPCOPY;
6102 if (flagsp == &negflags) {
6104 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6108 wastedflags = 0; /* reset so (?g-c) warns twice */
6114 RExC_flags |= posflags;
6115 RExC_flags &= ~negflags;
6117 oregflags |= posflags;
6118 oregflags &= ~negflags;
6120 nextchar(pRExC_state);
6131 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
6136 }} /* one for the default block, one for the switch */
6143 ret = reganode(pRExC_state, OPEN, parno);
6146 RExC_nestroot = parno;
6147 if (RExC_seen & REG_SEEN_RECURSE
6148 && !RExC_open_parens[parno-1])
6150 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6151 "Setting open paren #%"IVdf" to %d\n",
6152 (IV)parno, REG_NODE_NUM(ret)));
6153 RExC_open_parens[parno-1]= ret;
6156 Set_Node_Length(ret, 1); /* MJD */
6157 Set_Node_Offset(ret, RExC_parse); /* MJD */
6165 /* Pick up the branches, linking them together. */
6166 parse_start = RExC_parse; /* MJD */
6167 br = regbranch(pRExC_state, &flags, 1,depth+1);
6170 if (RExC_npar > after_freeze)
6171 after_freeze = RExC_npar;
6172 RExC_npar = freeze_paren;
6175 /* branch_len = (paren != 0); */
6179 if (*RExC_parse == '|') {
6180 if (!SIZE_ONLY && RExC_extralen) {
6181 reginsert(pRExC_state, BRANCHJ, br, depth+1);
6184 reginsert(pRExC_state, BRANCH, br, depth+1);
6185 Set_Node_Length(br, paren != 0);
6186 Set_Node_Offset_To_R(br-RExC_emit_start, parse_start-RExC_start);
6190 RExC_extralen += 1; /* For BRANCHJ-BRANCH. */
6192 else if (paren == ':') {
6193 *flagp |= flags&SIMPLE;
6195 if (is_open) { /* Starts with OPEN. */
6196 REGTAIL(pRExC_state, ret, br); /* OPEN -> first. */
6198 else if (paren != '?') /* Not Conditional */
6200 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6202 while (*RExC_parse == '|') {
6203 if (!SIZE_ONLY && RExC_extralen) {
6204 ender = reganode(pRExC_state, LONGJMP,0);
6205 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender); /* Append to the previous. */
6208 RExC_extralen += 2; /* Account for LONGJMP. */
6209 nextchar(pRExC_state);
6211 if (RExC_npar > after_freeze)
6212 after_freeze = RExC_npar;
6213 RExC_npar = freeze_paren;
6215 br = regbranch(pRExC_state, &flags, 0, depth+1);
6219 REGTAIL(pRExC_state, lastbr, br); /* BRANCH -> BRANCH. */
6221 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
6224 if (have_branch || paren != ':') {
6225 /* Make a closing node, and hook it on the end. */
6228 ender = reg_node(pRExC_state, TAIL);
6231 ender = reganode(pRExC_state, CLOSE, parno);
6232 if (!SIZE_ONLY && RExC_seen & REG_SEEN_RECURSE) {
6233 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
6234 "Setting close paren #%"IVdf" to %d\n",
6235 (IV)parno, REG_NODE_NUM(ender)));
6236 RExC_close_parens[parno-1]= ender;
6237 if (RExC_nestroot == parno)
6240 Set_Node_Offset(ender,RExC_parse+1); /* MJD */
6241 Set_Node_Length(ender,1); /* MJD */
6247 *flagp &= ~HASWIDTH;
6250 ender = reg_node(pRExC_state, SUCCEED);
6253 ender = reg_node(pRExC_state, END);
6255 assert(!RExC_opend); /* there can only be one! */
6260 REGTAIL(pRExC_state, lastbr, ender);
6262 if (have_branch && !SIZE_ONLY) {
6264 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
6266 /* Hook the tails of the branches to the closing node. */
6267 for (br = ret; br; br = regnext(br)) {
6268 const U8 op = PL_regkind[OP(br)];
6270 REGTAIL_STUDY(pRExC_state, NEXTOPER(br), ender);
6272 else if (op == BRANCHJ) {
6273 REGTAIL_STUDY(pRExC_state, NEXTOPER(NEXTOPER(br)), ender);
6281 static const char parens[] = "=!<,>";
6283 if (paren && (p = strchr(parens, paren))) {
6284 U8 node = ((p - parens) % 2) ? UNLESSM : IFMATCH;
6285 int flag = (p - parens) > 1;
6288 node = SUSPEND, flag = 0;
6289 reginsert(pRExC_state, node,ret, depth+1);
6290 Set_Node_Cur_Length(ret);
6291 Set_Node_Offset(ret, parse_start + 1);
6293 REGTAIL_STUDY(pRExC_state, ret, reg_node(pRExC_state, TAIL));
6297 /* Check for proper termination. */
6299 RExC_flags = oregflags;
6300 if (RExC_parse >= RExC_end || *nextchar(pRExC_state) != ')') {
6301 RExC_parse = oregcomp_parse;
6302 vFAIL("Unmatched (");
6305 else if (!paren && RExC_parse < RExC_end) {
6306 if (*RExC_parse == ')') {
6308 vFAIL("Unmatched )");
6311 FAIL("Junk on end of regexp"); /* "Can't happen". */
6315 RExC_npar = after_freeze;
6320 - regbranch - one alternative of an | operator
6322 * Implements the concatenation operator.
6325 S_regbranch(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, I32 first, U32 depth)
6328 register regnode *ret;
6329 register regnode *chain = NULL;
6330 register regnode *latest;
6331 I32 flags = 0, c = 0;
6332 GET_RE_DEBUG_FLAGS_DECL;
6334 PERL_ARGS_ASSERT_REGBRANCH;
6336 DEBUG_PARSE("brnc");
6341 if (!SIZE_ONLY && RExC_extralen)
6342 ret = reganode(pRExC_state, BRANCHJ,0);
6344 ret = reg_node(pRExC_state, BRANCH);
6345 Set_Node_Length(ret, 1);
6349 if (!first && SIZE_ONLY)
6350 RExC_extralen += 1; /* BRANCHJ */
6352 *flagp = WORST; /* Tentatively. */
6355 nextchar(pRExC_state);
6356 while (RExC_parse < RExC_end && *RExC_parse != '|' && *RExC_parse != ')') {
6358 latest = regpiece(pRExC_state, &flags,depth+1);
6359 if (latest == NULL) {
6360 if (flags & TRYAGAIN)
6364 else if (ret == NULL)
6366 *flagp |= flags&(HASWIDTH|POSTPONED);
6367 if (chain == NULL) /* First piece. */
6368 *flagp |= flags&SPSTART;
6371 REGTAIL(pRExC_state, chain, latest);
6376 if (chain == NULL) { /* Loop ran zero times. */
6377 chain = reg_node(pRExC_state, NOTHING);
6382 *flagp |= flags&SIMPLE;
6389 - regpiece - something followed by possible [*+?]
6391 * Note that the branching code sequences used for ? and the general cases
6392 * of * and + are somewhat optimized: they use the same NOTHING node as
6393 * both the endmarker for their branch list and the body of the last branch.
6394 * It might seem that this node could be dispensed with entirely, but the
6395 * endmarker role is not redundant.
6398 S_regpiece(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6401 register regnode *ret;
6403 register char *next;
6405 const char * const origparse = RExC_parse;
6407 I32 max = REG_INFTY;
6409 const char *maxpos = NULL;
6410 GET_RE_DEBUG_FLAGS_DECL;
6412 PERL_ARGS_ASSERT_REGPIECE;
6414 DEBUG_PARSE("piec");
6416 ret = regatom(pRExC_state, &flags,depth+1);
6418 if (flags & TRYAGAIN)
6425 if (op == '{' && regcurly(RExC_parse)) {
6427 parse_start = RExC_parse; /* MJD */
6428 next = RExC_parse + 1;
6429 while (isDIGIT(*next) || *next == ',') {
6438 if (*next == '}') { /* got one */
6442 min = atoi(RExC_parse);
6446 maxpos = RExC_parse;
6448 if (!max && *maxpos != '0')
6449 max = REG_INFTY; /* meaning "infinity" */
6450 else if (max >= REG_INFTY)
6451 vFAIL2("Quantifier in {,} bigger than %d", REG_INFTY - 1);
6453 nextchar(pRExC_state);
6456 if ((flags&SIMPLE)) {
6457 RExC_naughty += 2 + RExC_naughty / 2;
6458 reginsert(pRExC_state, CURLY, ret, depth+1);
6459 Set_Node_Offset(ret, parse_start+1); /* MJD */
6460 Set_Node_Cur_Length(ret);
6463 regnode * const w = reg_node(pRExC_state, WHILEM);
6466 REGTAIL(pRExC_state, ret, w);
6467 if (!SIZE_ONLY && RExC_extralen) {
6468 reginsert(pRExC_state, LONGJMP,ret, depth+1);
6469 reginsert(pRExC_state, NOTHING,ret, depth+1);
6470 NEXT_OFF(ret) = 3; /* Go over LONGJMP. */
6472 reginsert(pRExC_state, CURLYX,ret, depth+1);
6474 Set_Node_Offset(ret, parse_start+1);
6475 Set_Node_Length(ret,
6476 op == '{' ? (RExC_parse - parse_start) : 1);
6478 if (!SIZE_ONLY && RExC_extralen)
6479 NEXT_OFF(ret) = 3; /* Go over NOTHING to LONGJMP. */
6480 REGTAIL(pRExC_state, ret, reg_node(pRExC_state, NOTHING));
6482 RExC_whilem_seen++, RExC_extralen += 3;
6483 RExC_naughty += 4 + RExC_naughty; /* compound interest */
6492 vFAIL("Can't do {n,m} with n > m");
6494 ARG1_SET(ret, (U16)min);
6495 ARG2_SET(ret, (U16)max);
6507 #if 0 /* Now runtime fix should be reliable. */
6509 /* if this is reinstated, don't forget to put this back into perldiag:
6511 =item Regexp *+ operand could be empty at {#} in regex m/%s/
6513 (F) The part of the regexp subject to either the * or + quantifier
6514 could match an empty string. The {#} shows in the regular
6515 expression about where the problem was discovered.
6519 if (!(flags&HASWIDTH) && op != '?')
6520 vFAIL("Regexp *+ operand could be empty");
6523 parse_start = RExC_parse;
6524 nextchar(pRExC_state);
6526 *flagp = (op != '+') ? (WORST|SPSTART|HASWIDTH) : (WORST|HASWIDTH);
6528 if (op == '*' && (flags&SIMPLE)) {
6529 reginsert(pRExC_state, STAR, ret, depth+1);
6533 else if (op == '*') {
6537 else if (op == '+' && (flags&SIMPLE)) {
6538 reginsert(pRExC_state, PLUS, ret, depth+1);
6542 else if (op == '+') {
6546 else if (op == '?') {
6551 if (!SIZE_ONLY && !(flags&(HASWIDTH|POSTPONED)) && max > REG_INFTY/3) {
6552 ckWARN3reg(RExC_parse,
6553 "%.*s matches null string many times",
6554 (int)(RExC_parse >= origparse ? RExC_parse - origparse : 0),
6558 if (RExC_parse < RExC_end && *RExC_parse == '?') {
6559 nextchar(pRExC_state);
6560 reginsert(pRExC_state, MINMOD, ret, depth+1);
6561 REGTAIL(pRExC_state, ret, ret + NODE_STEP_REGNODE);
6563 #ifndef REG_ALLOW_MINMOD_SUSPEND
6566 if (RExC_parse < RExC_end && *RExC_parse == '+') {
6568 nextchar(pRExC_state);
6569 ender = reg_node(pRExC_state, SUCCEED);
6570 REGTAIL(pRExC_state, ret, ender);
6571 reginsert(pRExC_state, SUSPEND, ret, depth+1);
6573 ender = reg_node(pRExC_state, TAIL);
6574 REGTAIL(pRExC_state, ret, ender);
6578 if (RExC_parse < RExC_end && ISMULT2(RExC_parse)) {
6580 vFAIL("Nested quantifiers");
6587 /* reg_namedseq(pRExC_state,UVp)
6589 This is expected to be called by a parser routine that has
6590 recognized '\N' and needs to handle the rest. RExC_parse is
6591 expected to point at the first char following the N at the time
6594 If valuep is non-null then it is assumed that we are parsing inside
6595 of a charclass definition and the first codepoint in the resolved
6596 string is returned via *valuep and the routine will return NULL.
6597 In this mode if a multichar string is returned from the charnames
6598 handler a warning will be issued, and only the first char in the
6599 sequence will be examined. If the string returned is zero length
6600 then the value of *valuep is undefined and NON-NULL will
6601 be returned to indicate failure. (This will NOT be a valid pointer
6604 If valuep is null then it is assumed that we are parsing normal text
6605 and inserts a new EXACT node into the program containing the resolved
6606 string and returns a pointer to the new node. If the string is
6607 zerolength a NOTHING node is emitted.
6609 On success RExC_parse is set to the char following the endbrace.
6610 Parsing failures will generate a fatal errorvia vFAIL(...)
6612 NOTE: We cache all results from the charnames handler locally in
6613 the RExC_charnames hash (created on first use) to prevent a charnames
6614 handler from playing silly-buggers and returning a short string and
6615 then a long string for a given pattern. Since the regexp program
6616 size is calculated during an initial parse this would result
6617 in a buffer overrun so we cache to prevent the charname result from
6618 changing during the course of the parse.
6622 S_reg_namedseq(pTHX_ RExC_state_t *pRExC_state, UV *valuep, I32 *flagp)
6624 char * name; /* start of the content of the name */
6625 char * endbrace; /* endbrace following the name */
6628 STRLEN len; /* this has various purposes throughout the code */
6629 bool cached = 0; /* if this is true then we shouldn't refcount dev sv_str */
6630 regnode *ret = NULL;
6632 PERL_ARGS_ASSERT_REG_NAMEDSEQ;
6634 if (*RExC_parse != '{' ||
6635 (*RExC_parse == '{' && RExC_parse[1]
6636 && strchr("0123456789", RExC_parse[1])))
6638 GET_RE_DEBUG_FLAGS_DECL;
6640 /* no bare \N in a charclass */
6641 vFAIL("Missing braces on \\N{}");
6643 nextchar(pRExC_state);
6644 ret = reg_node(pRExC_state, REG_ANY);
6645 *flagp |= HASWIDTH|SIMPLE;
6648 Set_Node_Length(ret, 1); /* MJD */
6651 name = RExC_parse+1;
6652 endbrace = strchr(RExC_parse, '}');
6655 vFAIL("Missing right brace on \\N{}");
6657 RExC_parse = endbrace + 1;
6660 /* RExC_parse points at the beginning brace,
6661 endbrace points at the last */
6662 if ( name[0]=='U' && name[1]=='+' ) {
6663 /* its a "Unicode hex" notation {U+89AB} */
6664 I32 fl = PERL_SCAN_ALLOW_UNDERSCORES
6665 | PERL_SCAN_DISALLOW_PREFIX
6666 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
6668 len = (STRLEN)(endbrace - name - 2);
6669 cp = grok_hex(name + 2, &len, &fl, NULL);
6670 if ( len != (STRLEN)(endbrace - name - 2) ) {
6674 if (cp > 0xff) RExC_utf8 = 1;
6679 /* Need to convert to utf8 if either: won't fit into a byte, or the re
6680 * is going to be in utf8 and the representation changes under utf8. */
6681 if (cp > 0xff || (RExC_utf8 && ! UNI_IS_INVARIANT(cp))) {
6682 U8 string[UTF8_MAXBYTES+1];
6685 tmps = uvuni_to_utf8(string, cp);
6686 sv_str = newSVpvn_utf8((char*)string, tmps - string, TRUE);
6687 } else { /* Otherwise, no need for utf8, can skip that step */
6690 sv_str= newSVpvn(&string, 1);
6693 /* fetch the charnames handler for this scope */
6694 HV * const table = GvHV(PL_hintgv);
6696 hv_fetchs(table, "charnames", FALSE) :
6698 SV *cv= cvp ? *cvp : NULL;
6701 /* create an SV with the name as argument */
6702 sv_name = newSVpvn(name, endbrace - name);
6704 if (!table || !(PL_hints & HINT_LOCALIZE_HH)) {
6705 vFAIL2("Constant(\\N{%" SVf "}) unknown: "
6706 "(possibly a missing \"use charnames ...\")",
6709 if (!cvp || !SvOK(*cvp)) { /* when $^H{charnames} = undef; */
6710 vFAIL2("Constant(\\N{%" SVf "}): "
6711 "$^H{charnames} is not defined", SVfARG(sv_name));
6716 if (!RExC_charnames) {
6717 /* make sure our cache is allocated */
6718 RExC_charnames = newHV();
6719 sv_2mortal(MUTABLE_SV(RExC_charnames));
6721 /* see if we have looked this one up before */
6722 he_str = hv_fetch_ent( RExC_charnames, sv_name, 0, 0 );
6724 sv_str = HeVAL(he_str);
6737 count= call_sv(cv, G_SCALAR);
6739 if (count == 1) { /* XXXX is this right? dmq */
6741 SvREFCNT_inc_simple_void(sv_str);
6749 if ( !sv_str || !SvOK(sv_str) ) {
6750 vFAIL2("Constant(\\N{%" SVf "}): Call to &{$^H{charnames}} "
6751 "did not return a defined value", SVfARG(sv_name));
6753 if (hv_store_ent( RExC_charnames, sv_name, sv_str, 0))
6758 char *p = SvPV(sv_str, len);
6761 if ( SvUTF8(sv_str) ) {
6762 *valuep = utf8_to_uvchr((U8*)p, &numlen);
6766 We have to turn on utf8 for high bit chars otherwise
6767 we get failures with
6769 "ss" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i
6770 "SS" =~ /[\N{LATIN SMALL LETTER SHARP S}]/i
6772 This is different from what \x{} would do with the same
6773 codepoint, where the condition is > 0xFF.
6780 /* warn if we havent used the whole string? */
6782 if (numlen<len && SIZE_ONLY) {
6783 ckWARN2reg(RExC_parse,
6784 "Ignoring excess chars from \\N{%" SVf "} in character class",
6788 } else if (SIZE_ONLY) {
6789 ckWARN2reg(RExC_parse,
6790 "Ignoring zero length \\N{%" SVf "} in character class",
6794 SvREFCNT_dec(sv_name);
6796 SvREFCNT_dec(sv_str);
6797 return len ? NULL : (regnode *)&len;
6798 } else if(SvCUR(sv_str)) {
6804 char * parse_start = name-3; /* needed for the offsets */
6806 GET_RE_DEBUG_FLAGS_DECL; /* needed for the offsets */
6808 ret = reg_node(pRExC_state,
6809 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
6812 if ( RExC_utf8 && !SvUTF8(sv_str) ) {
6813 sv_utf8_upgrade(sv_str);
6814 } else if ( !RExC_utf8 && SvUTF8(sv_str) ) {
6818 p = SvPV(sv_str, len);
6820 /* len is the length written, charlen is the size the char read */
6821 for ( len = 0; p < pend; p += charlen ) {
6823 UV uvc = utf8_to_uvchr((U8*)p, &charlen);
6825 STRLEN foldlen,numlen;
6826 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
6827 uvc = toFOLD_uni(uvc, tmpbuf, &foldlen);
6828 /* Emit all the Unicode characters. */
6830 for (foldbuf = tmpbuf;
6834 uvc = utf8_to_uvchr(foldbuf, &numlen);
6836 const STRLEN unilen = reguni(pRExC_state, uvc, s);
6839 /* In EBCDIC the numlen
6840 * and unilen can differ. */
6842 if (numlen >= foldlen)
6846 break; /* "Can't happen." */
6849 const STRLEN unilen = reguni(pRExC_state, uvc, s);
6861 RExC_size += STR_SZ(len);
6864 RExC_emit += STR_SZ(len);
6866 Set_Node_Cur_Length(ret); /* MJD */
6868 nextchar(pRExC_state);
6869 } else { /* zero length */
6870 ret = reg_node(pRExC_state,NOTHING);
6872 SvREFCNT_dec(sv_name);
6874 SvREFCNT_dec(sv_str);
6883 * It returns the code point in utf8 for the value in *encp.
6884 * value: a code value in the source encoding
6885 * encp: a pointer to an Encode object
6887 * If the result from Encode is not a single character,
6888 * it returns U+FFFD (Replacement character) and sets *encp to NULL.
6891 S_reg_recode(pTHX_ const char value, SV **encp)
6894 SV * const sv = newSVpvn_flags(&value, numlen, SVs_TEMP);
6895 const char * const s = *encp ? sv_recode_to_utf8(sv, *encp) : SvPVX(sv);
6896 const STRLEN newlen = SvCUR(sv);
6897 UV uv = UNICODE_REPLACEMENT;
6899 PERL_ARGS_ASSERT_REG_RECODE;
6903 ? utf8n_to_uvchr((U8*)s, newlen, &numlen, UTF8_ALLOW_DEFAULT)
6906 if (!newlen || numlen != newlen) {
6907 uv = UNICODE_REPLACEMENT;
6915 - regatom - the lowest level
6917 Try to identify anything special at the start of the pattern. If there
6918 is, then handle it as required. This may involve generating a single regop,
6919 such as for an assertion; or it may involve recursing, such as to
6920 handle a () structure.
6922 If the string doesn't start with something special then we gobble up
6923 as much literal text as we can.
6925 Once we have been able to handle whatever type of thing started the
6926 sequence, we return.
6928 Note: we have to be careful with escapes, as they can be both literal
6929 and special, and in the case of \10 and friends can either, depending
6930 on context. Specifically there are two seperate switches for handling
6931 escape sequences, with the one for handling literal escapes requiring
6932 a dummy entry for all of the special escapes that are actually handled
6937 S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
6940 register regnode *ret = NULL;
6942 char *parse_start = RExC_parse;
6943 GET_RE_DEBUG_FLAGS_DECL;
6944 DEBUG_PARSE("atom");
6945 *flagp = WORST; /* Tentatively. */
6947 PERL_ARGS_ASSERT_REGATOM;
6950 switch ((U8)*RExC_parse) {
6952 RExC_seen_zerolen++;
6953 nextchar(pRExC_state);
6954 if (RExC_flags & RXf_PMf_MULTILINE)
6955 ret = reg_node(pRExC_state, MBOL);
6956 else if (RExC_flags & RXf_PMf_SINGLELINE)
6957 ret = reg_node(pRExC_state, SBOL);
6959 ret = reg_node(pRExC_state, BOL);
6960 Set_Node_Length(ret, 1); /* MJD */
6963 nextchar(pRExC_state);
6965 RExC_seen_zerolen++;
6966 if (RExC_flags & RXf_PMf_MULTILINE)
6967 ret = reg_node(pRExC_state, MEOL);
6968 else if (RExC_flags & RXf_PMf_SINGLELINE)
6969 ret = reg_node(pRExC_state, SEOL);
6971 ret = reg_node(pRExC_state, EOL);
6972 Set_Node_Length(ret, 1); /* MJD */
6975 nextchar(pRExC_state);
6976 if (RExC_flags & RXf_PMf_SINGLELINE)
6977 ret = reg_node(pRExC_state, SANY);
6979 ret = reg_node(pRExC_state, REG_ANY);
6980 *flagp |= HASWIDTH|SIMPLE;
6982 Set_Node_Length(ret, 1); /* MJD */
6986 char * const oregcomp_parse = ++RExC_parse;
6987 ret = regclass(pRExC_state,depth+1);
6988 if (*RExC_parse != ']') {
6989 RExC_parse = oregcomp_parse;
6990 vFAIL("Unmatched [");
6992 nextchar(pRExC_state);
6993 *flagp |= HASWIDTH|SIMPLE;
6994 Set_Node_Length(ret, RExC_parse - oregcomp_parse + 1); /* MJD */
6998 nextchar(pRExC_state);
6999 ret = reg(pRExC_state, 1, &flags,depth+1);
7001 if (flags & TRYAGAIN) {
7002 if (RExC_parse == RExC_end) {
7003 /* Make parent create an empty node if needed. */
7011 *flagp |= flags&(HASWIDTH|SPSTART|SIMPLE|POSTPONED);
7015 if (flags & TRYAGAIN) {
7019 vFAIL("Internal urp");
7020 /* Supposed to be caught earlier. */
7023 if (!regcurly(RExC_parse)) {
7032 vFAIL("Quantifier follows nothing");
7040 len=0; /* silence a spurious compiler warning */
7041 if ((cp = what_len_TRICKYFOLD_safe(RExC_parse,RExC_end,UTF,len))) {
7042 *flagp |= HASWIDTH; /* could be SIMPLE too, but needs a handler in regexec.regrepeat */
7043 RExC_parse+=len-1; /* we get one from nextchar() as well. :-( */
7044 ret = reganode(pRExC_state, FOLDCHAR, cp);
7045 Set_Node_Length(ret, 1); /* MJD */
7046 nextchar(pRExC_state); /* kill whitespace under /x */
7054 This switch handles escape sequences that resolve to some kind
7055 of special regop and not to literal text. Escape sequnces that
7056 resolve to literal text are handled below in the switch marked
7059 Every entry in this switch *must* have a corresponding entry
7060 in the literal escape switch. However, the opposite is not
7061 required, as the default for this switch is to jump to the
7062 literal text handling code.
7064 switch ((U8)*++RExC_parse) {
7069 /* Special Escapes */
7071 RExC_seen_zerolen++;
7072 ret = reg_node(pRExC_state, SBOL);
7074 goto finish_meta_pat;
7076 ret = reg_node(pRExC_state, GPOS);
7077 RExC_seen |= REG_SEEN_GPOS;
7079 goto finish_meta_pat;
7081 RExC_seen_zerolen++;
7082 ret = reg_node(pRExC_state, KEEPS);
7084 /* XXX:dmq : disabling in-place substitution seems to
7085 * be necessary here to avoid cases of memory corruption, as
7086 * with: C<$_="x" x 80; s/x\K/y/> -- rgs
7088 RExC_seen |= REG_SEEN_LOOKBEHIND;
7089 goto finish_meta_pat;
7091 ret = reg_node(pRExC_state, SEOL);
7093 RExC_seen_zerolen++; /* Do not optimize RE away */
7094 goto finish_meta_pat;
7096 ret = reg_node(pRExC_state, EOS);
7098 RExC_seen_zerolen++; /* Do not optimize RE away */
7099 goto finish_meta_pat;
7101 ret = reg_node(pRExC_state, CANY);
7102 RExC_seen |= REG_SEEN_CANY;
7103 *flagp |= HASWIDTH|SIMPLE;
7104 goto finish_meta_pat;
7106 ret = reg_node(pRExC_state, CLUMP);
7108 goto finish_meta_pat;
7110 ret = reg_node(pRExC_state, (U8)(LOC ? ALNUML : ALNUM));
7111 *flagp |= HASWIDTH|SIMPLE;
7112 goto finish_meta_pat;
7114 ret = reg_node(pRExC_state, (U8)(LOC ? NALNUML : NALNUM));
7115 *flagp |= HASWIDTH|SIMPLE;
7116 goto finish_meta_pat;
7118 RExC_seen_zerolen++;
7119 RExC_seen |= REG_SEEN_LOOKBEHIND;
7120 ret = reg_node(pRExC_state, (U8)(LOC ? BOUNDL : BOUND));
7122 goto finish_meta_pat;
7124 RExC_seen_zerolen++;
7125 RExC_seen |= REG_SEEN_LOOKBEHIND;
7126 ret = reg_node(pRExC_state, (U8)(LOC ? NBOUNDL : NBOUND));
7128 goto finish_meta_pat;
7130 ret = reg_node(pRExC_state, (U8)(LOC ? SPACEL : SPACE));
7131 *flagp |= HASWIDTH|SIMPLE;
7132 goto finish_meta_pat;
7134 ret = reg_node(pRExC_state, (U8)(LOC ? NSPACEL : NSPACE));
7135 *flagp |= HASWIDTH|SIMPLE;
7136 goto finish_meta_pat;
7138 ret = reg_node(pRExC_state, DIGIT);
7139 *flagp |= HASWIDTH|SIMPLE;
7140 goto finish_meta_pat;
7142 ret = reg_node(pRExC_state, NDIGIT);
7143 *flagp |= HASWIDTH|SIMPLE;
7144 goto finish_meta_pat;
7146 ret = reg_node(pRExC_state, LNBREAK);
7147 *flagp |= HASWIDTH|SIMPLE;
7148 goto finish_meta_pat;
7150 ret = reg_node(pRExC_state, HORIZWS);
7151 *flagp |= HASWIDTH|SIMPLE;
7152 goto finish_meta_pat;
7154 ret = reg_node(pRExC_state, NHORIZWS);
7155 *flagp |= HASWIDTH|SIMPLE;
7156 goto finish_meta_pat;
7158 ret = reg_node(pRExC_state, VERTWS);
7159 *flagp |= HASWIDTH|SIMPLE;
7160 goto finish_meta_pat;
7162 ret = reg_node(pRExC_state, NVERTWS);
7163 *flagp |= HASWIDTH|SIMPLE;
7165 nextchar(pRExC_state);
7166 Set_Node_Length(ret, 2); /* MJD */
7171 char* const oldregxend = RExC_end;
7173 char* parse_start = RExC_parse - 2;
7176 if (RExC_parse[1] == '{') {
7177 /* a lovely hack--pretend we saw [\pX] instead */
7178 RExC_end = strchr(RExC_parse, '}');
7180 const U8 c = (U8)*RExC_parse;
7182 RExC_end = oldregxend;
7183 vFAIL2("Missing right brace on \\%c{}", c);
7188 RExC_end = RExC_parse + 2;
7189 if (RExC_end > oldregxend)
7190 RExC_end = oldregxend;
7194 ret = regclass(pRExC_state,depth+1);
7196 RExC_end = oldregxend;
7199 Set_Node_Offset(ret, parse_start + 2);
7200 Set_Node_Cur_Length(ret);
7201 nextchar(pRExC_state);
7202 *flagp |= HASWIDTH|SIMPLE;
7206 /* Handle \N and \N{NAME} here and not below because it can be
7207 multicharacter. join_exact() will join them up later on.
7208 Also this makes sure that things like /\N{BLAH}+/ and
7209 \N{BLAH} being multi char Just Happen. dmq*/
7211 ret= reg_namedseq(pRExC_state, NULL, flagp);
7213 case 'k': /* Handle \k<NAME> and \k'NAME' */
7216 char ch= RExC_parse[1];
7217 if (ch != '<' && ch != '\'' && ch != '{') {
7219 vFAIL2("Sequence %.2s... not terminated",parse_start);
7221 /* this pretty much dupes the code for (?P=...) in reg(), if
7222 you change this make sure you change that */
7223 char* name_start = (RExC_parse += 2);
7225 SV *sv_dat = reg_scan_name(pRExC_state,
7226 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7227 ch= (ch == '<') ? '>' : (ch == '{') ? '}' : '\'';
7228 if (RExC_parse == name_start || *RExC_parse != ch)
7229 vFAIL2("Sequence %.3s... not terminated",parse_start);
7232 num = add_data( pRExC_state, 1, "S" );
7233 RExC_rxi->data->data[num]=(void*)sv_dat;
7234 SvREFCNT_inc_simple_void(sv_dat);
7238 ret = reganode(pRExC_state,
7239 (U8)(FOLD ? (LOC ? NREFFL : NREFF) : NREF),
7243 /* override incorrect value set in reganode MJD */
7244 Set_Node_Offset(ret, parse_start+1);
7245 Set_Node_Cur_Length(ret); /* MJD */
7246 nextchar(pRExC_state);
7252 case '1': case '2': case '3': case '4':
7253 case '5': case '6': case '7': case '8': case '9':
7256 bool isg = *RExC_parse == 'g';
7261 if (*RExC_parse == '{') {
7265 if (*RExC_parse == '-') {
7269 if (hasbrace && !isDIGIT(*RExC_parse)) {
7270 if (isrel) RExC_parse--;
7272 goto parse_named_seq;
7274 num = atoi(RExC_parse);
7275 if (isg && num == 0)
7276 vFAIL("Reference to invalid group 0");
7278 num = RExC_npar - num;
7280 vFAIL("Reference to nonexistent or unclosed group");
7282 if (!isg && num > 9 && num >= RExC_npar)
7285 char * const parse_start = RExC_parse - 1; /* MJD */
7286 while (isDIGIT(*RExC_parse))
7288 if (parse_start == RExC_parse - 1)
7289 vFAIL("Unterminated \\g... pattern");
7291 if (*RExC_parse != '}')
7292 vFAIL("Unterminated \\g{...} pattern");
7296 if (num > (I32)RExC_rx->nparens)
7297 vFAIL("Reference to nonexistent group");
7300 ret = reganode(pRExC_state,
7301 (U8)(FOLD ? (LOC ? REFFL : REFF) : REF),
7305 /* override incorrect value set in reganode MJD */
7306 Set_Node_Offset(ret, parse_start+1);
7307 Set_Node_Cur_Length(ret); /* MJD */
7309 nextchar(pRExC_state);
7314 if (RExC_parse >= RExC_end)
7315 FAIL("Trailing \\");
7318 /* Do not generate "unrecognized" warnings here, we fall
7319 back into the quick-grab loop below */
7326 if (RExC_flags & RXf_PMf_EXTENDED) {
7327 if ( reg_skipcomment( pRExC_state ) )
7334 register STRLEN len;
7339 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
7341 parse_start = RExC_parse - 1;
7347 ret = reg_node(pRExC_state,
7348 (U8)(FOLD ? (LOC ? EXACTFL : EXACTF) : EXACT));
7350 for (len = 0, p = RExC_parse - 1;
7351 len < 127 && p < RExC_end;
7354 char * const oldp = p;
7356 if (RExC_flags & RXf_PMf_EXTENDED)
7357 p = regwhite( pRExC_state, p );
7362 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7363 goto normal_default;
7373 /* Literal Escapes Switch
7375 This switch is meant to handle escape sequences that
7376 resolve to a literal character.
7378 Every escape sequence that represents something
7379 else, like an assertion or a char class, is handled
7380 in the switch marked 'Special Escapes' above in this
7381 routine, but also has an entry here as anything that
7382 isn't explicitly mentioned here will be treated as
7383 an unescaped equivalent literal.
7387 /* These are all the special escapes. */
7391 if (LOC || !FOLD || !is_TRICKYFOLD_safe(p,RExC_end,UTF))
7392 goto normal_default;
7393 case 'A': /* Start assertion */
7394 case 'b': case 'B': /* Word-boundary assertion*/
7395 case 'C': /* Single char !DANGEROUS! */
7396 case 'd': case 'D': /* digit class */
7397 case 'g': case 'G': /* generic-backref, pos assertion */
7398 case 'h': case 'H': /* HORIZWS */
7399 case 'k': case 'K': /* named backref, keep marker */
7400 case 'N': /* named char sequence */
7401 case 'p': case 'P': /* Unicode property */
7402 case 'R': /* LNBREAK */
7403 case 's': case 'S': /* space class */
7404 case 'v': case 'V': /* VERTWS */
7405 case 'w': case 'W': /* word class */
7406 case 'X': /* eXtended Unicode "combining character sequence" */
7407 case 'z': case 'Z': /* End of line/string assertion */
7411 /* Anything after here is an escape that resolves to a
7412 literal. (Except digits, which may or may not)
7431 ender = ASCII_TO_NATIVE('\033');
7435 ender = ASCII_TO_NATIVE('\007');
7440 char* const e = strchr(p, '}');
7444 vFAIL("Missing right brace on \\x{}");
7447 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
7448 | PERL_SCAN_DISALLOW_PREFIX;
7449 STRLEN numlen = e - p - 1;
7450 ender = grok_hex(p + 1, &numlen, &flags, NULL);
7457 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
7459 ender = grok_hex(p, &numlen, &flags, NULL);
7462 if (PL_encoding && ender < 0x100)
7463 goto recode_encoding;
7467 ender = UCHARAT(p++);
7468 ender = toCTRL(ender);
7470 case '0': case '1': case '2': case '3':case '4':
7471 case '5': case '6': case '7': case '8':case '9':
7473 (isDIGIT(p[1]) && atoi(p) >= RExC_npar) ) {
7476 ender = grok_oct(p, &numlen, &flags, NULL);
7478 /* An octal above 0xff is interpreted differently
7479 * depending on if the re is in utf8 or not. If it
7480 * is in utf8, the value will be itself, otherwise
7481 * it is interpreted as modulo 0x100. It has been
7482 * decided to discourage the use of octal above the
7483 * single-byte range. For now, warn only when
7484 * it ends up modulo */
7485 if (SIZE_ONLY && ender >= 0x100
7486 && ! UTF && ! PL_encoding) {
7487 ckWARNregdep(p, "Use of octal value above 377 is deprecated");
7495 if (PL_encoding && ender < 0x100)
7496 goto recode_encoding;
7500 SV* enc = PL_encoding;
7501 ender = reg_recode((const char)(U8)ender, &enc);
7502 if (!enc && SIZE_ONLY)
7503 ckWARNreg(p, "Invalid escape in the specified encoding");
7509 FAIL("Trailing \\");
7512 if (!SIZE_ONLY&& isALPHA(*p))
7513 ckWARN2reg(p + 1, "Unrecognized escape \\%c passed through", UCHARAT(p));
7514 goto normal_default;
7519 if (UTF8_IS_START(*p) && UTF) {
7521 ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
7522 &numlen, UTF8_ALLOW_DEFAULT);
7529 if ( RExC_flags & RXf_PMf_EXTENDED)
7530 p = regwhite( pRExC_state, p );
7532 /* Prime the casefolded buffer. */
7533 ender = toFOLD_uni(ender, tmpbuf, &foldlen);
7535 if (p < RExC_end && ISMULT2(p)) { /* Back off on ?+*. */
7540 /* Emit all the Unicode characters. */
7542 for (foldbuf = tmpbuf;
7544 foldlen -= numlen) {
7545 ender = utf8_to_uvchr(foldbuf, &numlen);
7547 const STRLEN unilen = reguni(pRExC_state, ender, s);
7550 /* In EBCDIC the numlen
7551 * and unilen can differ. */
7553 if (numlen >= foldlen)
7557 break; /* "Can't happen." */
7561 const STRLEN unilen = reguni(pRExC_state, ender, s);
7570 REGC((char)ender, s++);
7576 /* Emit all the Unicode characters. */
7578 for (foldbuf = tmpbuf;
7580 foldlen -= numlen) {
7581 ender = utf8_to_uvchr(foldbuf, &numlen);
7583 const STRLEN unilen = reguni(pRExC_state, ender, s);
7586 /* In EBCDIC the numlen
7587 * and unilen can differ. */
7589 if (numlen >= foldlen)
7597 const STRLEN unilen = reguni(pRExC_state, ender, s);
7606 REGC((char)ender, s++);
7610 Set_Node_Cur_Length(ret); /* MJD */
7611 nextchar(pRExC_state);
7613 /* len is STRLEN which is unsigned, need to copy to signed */
7616 vFAIL("Internal disaster");
7620 if (len == 1 && UNI_IS_INVARIANT(ender))
7624 RExC_size += STR_SZ(len);
7627 RExC_emit += STR_SZ(len);
7637 S_regwhite( RExC_state_t *pRExC_state, char *p )
7639 const char *e = RExC_end;
7641 PERL_ARGS_ASSERT_REGWHITE;
7646 else if (*p == '#') {
7655 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
7663 /* Parse POSIX character classes: [[:foo:]], [[=foo=]], [[.foo.]].
7664 Character classes ([:foo:]) can also be negated ([:^foo:]).
7665 Returns a named class id (ANYOF_XXX) if successful, -1 otherwise.
7666 Equivalence classes ([=foo=]) and composites ([.foo.]) are parsed,
7667 but trigger failures because they are currently unimplemented. */
7669 #define POSIXCC_DONE(c) ((c) == ':')
7670 #define POSIXCC_NOTYET(c) ((c) == '=' || (c) == '.')
7671 #define POSIXCC(c) (POSIXCC_DONE(c) || POSIXCC_NOTYET(c))
7674 S_regpposixcc(pTHX_ RExC_state_t *pRExC_state, I32 value)
7677 I32 namedclass = OOB_NAMEDCLASS;
7679 PERL_ARGS_ASSERT_REGPPOSIXCC;
7681 if (value == '[' && RExC_parse + 1 < RExC_end &&
7682 /* I smell either [: or [= or [. -- POSIX has been here, right? */
7683 POSIXCC(UCHARAT(RExC_parse))) {
7684 const char c = UCHARAT(RExC_parse);
7685 char* const s = RExC_parse++;
7687 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != c)
7689 if (RExC_parse == RExC_end)
7690 /* Grandfather lone [:, [=, [. */
7693 const char* const t = RExC_parse++; /* skip over the c */
7696 if (UCHARAT(RExC_parse) == ']') {
7697 const char *posixcc = s + 1;
7698 RExC_parse++; /* skip over the ending ] */
7701 const I32 complement = *posixcc == '^' ? *posixcc++ : 0;
7702 const I32 skip = t - posixcc;
7704 /* Initially switch on the length of the name. */
7707 if (memEQ(posixcc, "word", 4)) /* this is not POSIX, this is the Perl \w */
7708 namedclass = complement ? ANYOF_NALNUM : ANYOF_ALNUM;
7711 /* Names all of length 5. */
7712 /* alnum alpha ascii blank cntrl digit graph lower
7713 print punct space upper */
7714 /* Offset 4 gives the best switch position. */
7715 switch (posixcc[4]) {
7717 if (memEQ(posixcc, "alph", 4)) /* alpha */
7718 namedclass = complement ? ANYOF_NALPHA : ANYOF_ALPHA;
7721 if (memEQ(posixcc, "spac", 4)) /* space */
7722 namedclass = complement ? ANYOF_NPSXSPC : ANYOF_PSXSPC;
7725 if (memEQ(posixcc, "grap", 4)) /* graph */
7726 namedclass = complement ? ANYOF_NGRAPH : ANYOF_GRAPH;
7729 if (memEQ(posixcc, "asci", 4)) /* ascii */
7730 namedclass = complement ? ANYOF_NASCII : ANYOF_ASCII;
7733 if (memEQ(posixcc, "blan", 4)) /* blank */
7734 namedclass = complement ? ANYOF_NBLANK : ANYOF_BLANK;
7737 if (memEQ(posixcc, "cntr", 4)) /* cntrl */
7738 namedclass = complement ? ANYOF_NCNTRL : ANYOF_CNTRL;
7741 if (memEQ(posixcc, "alnu", 4)) /* alnum */
7742 namedclass = complement ? ANYOF_NALNUMC : ANYOF_ALNUMC;
7745 if (memEQ(posixcc, "lowe", 4)) /* lower */
7746 namedclass = complement ? ANYOF_NLOWER : ANYOF_LOWER;
7747 else if (memEQ(posixcc, "uppe", 4)) /* upper */
7748 namedclass = complement ? ANYOF_NUPPER : ANYOF_UPPER;
7751 if (memEQ(posixcc, "digi", 4)) /* digit */
7752 namedclass = complement ? ANYOF_NDIGIT : ANYOF_DIGIT;
7753 else if (memEQ(posixcc, "prin", 4)) /* print */
7754 namedclass = complement ? ANYOF_NPRINT : ANYOF_PRINT;
7755 else if (memEQ(posixcc, "punc", 4)) /* punct */
7756 namedclass = complement ? ANYOF_NPUNCT : ANYOF_PUNCT;
7761 if (memEQ(posixcc, "xdigit", 6))
7762 namedclass = complement ? ANYOF_NXDIGIT : ANYOF_XDIGIT;
7766 if (namedclass == OOB_NAMEDCLASS)
7767 Simple_vFAIL3("POSIX class [:%.*s:] unknown",
7769 assert (posixcc[skip] == ':');
7770 assert (posixcc[skip+1] == ']');
7771 } else if (!SIZE_ONLY) {
7772 /* [[=foo=]] and [[.foo.]] are still future. */
7774 /* adjust RExC_parse so the warning shows after
7776 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse) != ']')
7778 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7781 /* Maternal grandfather:
7782 * "[:" ending in ":" but not in ":]" */
7792 S_checkposixcc(pTHX_ RExC_state_t *pRExC_state)
7796 PERL_ARGS_ASSERT_CHECKPOSIXCC;
7798 if (POSIXCC(UCHARAT(RExC_parse))) {
7799 const char *s = RExC_parse;
7800 const char c = *s++;
7804 if (*s && c == *s && s[1] == ']') {
7806 "POSIX syntax [%c %c] belongs inside character classes",
7809 /* [[=foo=]] and [[.foo.]] are still future. */
7810 if (POSIXCC_NOTYET(c)) {
7811 /* adjust RExC_parse so the error shows after
7813 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse++) != ']')
7815 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
7822 #define _C_C_T_(NAME,TEST,WORD) \
7825 ANYOF_CLASS_SET(ret, ANYOF_##NAME); \
7827 for (value = 0; value < 256; value++) \
7829 ANYOF_BITMAP_SET(ret, value); \
7834 case ANYOF_N##NAME: \
7836 ANYOF_CLASS_SET(ret, ANYOF_N##NAME); \
7838 for (value = 0; value < 256; value++) \
7840 ANYOF_BITMAP_SET(ret, value); \
7846 #define _C_C_T_NOLOC_(NAME,TEST,WORD) \
7848 for (value = 0; value < 256; value++) \
7850 ANYOF_BITMAP_SET(ret, value); \
7854 case ANYOF_N##NAME: \
7855 for (value = 0; value < 256; value++) \
7857 ANYOF_BITMAP_SET(ret, value); \
7863 We dont use PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS as the direct test
7864 so that it is possible to override the option here without having to
7865 rebuild the entire core. as we are required to do if we change regcomp.h
7866 which is where PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS is defined.
7868 #if PERL_LEGACY_UNICODE_CHARCLASS_MAPPINGS
7869 #define BROKEN_UNICODE_CHARCLASS_MAPPINGS
7872 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
7873 #define POSIX_CC_UNI_NAME(CCNAME) CCNAME
7875 #define POSIX_CC_UNI_NAME(CCNAME) "Posix" CCNAME
7879 parse a class specification and produce either an ANYOF node that
7880 matches the pattern or if the pattern matches a single char only and
7881 that char is < 256 and we are case insensitive then we produce an
7886 S_regclass(pTHX_ RExC_state_t *pRExC_state, U32 depth)
7889 register UV nextvalue;
7890 register IV prevvalue = OOB_UNICODE;
7891 register IV range = 0;
7892 UV value = 0; /* XXX:dmq: needs to be referenceable (unfortunately) */
7893 register regnode *ret;
7896 char *rangebegin = NULL;
7897 bool need_class = 0;
7900 bool optimize_invert = TRUE;
7901 AV* unicode_alternate = NULL;
7903 UV literal_endpoint = 0;
7905 UV stored = 0; /* number of chars stored in the class */
7907 regnode * const orig_emit = RExC_emit; /* Save the original RExC_emit in
7908 case we need to change the emitted regop to an EXACT. */
7909 const char * orig_parse = RExC_parse;
7910 GET_RE_DEBUG_FLAGS_DECL;
7912 PERL_ARGS_ASSERT_REGCLASS;
7914 PERL_UNUSED_ARG(depth);
7917 DEBUG_PARSE("clas");
7919 /* Assume we are going to generate an ANYOF node. */
7920 ret = reganode(pRExC_state, ANYOF, 0);
7923 ANYOF_FLAGS(ret) = 0;
7925 if (UCHARAT(RExC_parse) == '^') { /* Complement of range. */
7929 ANYOF_FLAGS(ret) |= ANYOF_INVERT;
7933 RExC_size += ANYOF_SKIP;
7934 listsv = &PL_sv_undef; /* For code scanners: listsv always non-NULL. */
7937 RExC_emit += ANYOF_SKIP;
7939 ANYOF_FLAGS(ret) |= ANYOF_FOLD;
7941 ANYOF_FLAGS(ret) |= ANYOF_LOCALE;
7942 ANYOF_BITMAP_ZERO(ret);
7943 listsv = newSVpvs("# comment\n");
7946 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
7948 if (!SIZE_ONLY && POSIXCC(nextvalue))
7949 checkposixcc(pRExC_state);
7951 /* allow 1st char to be ] (allowing it to be - is dealt with later) */
7952 if (UCHARAT(RExC_parse) == ']')
7956 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != ']') {
7960 namedclass = OOB_NAMEDCLASS; /* initialize as illegal */
7963 rangebegin = RExC_parse;
7965 value = utf8n_to_uvchr((U8*)RExC_parse,
7966 RExC_end - RExC_parse,
7967 &numlen, UTF8_ALLOW_DEFAULT);
7968 RExC_parse += numlen;
7971 value = UCHARAT(RExC_parse++);
7973 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
7974 if (value == '[' && POSIXCC(nextvalue))
7975 namedclass = regpposixcc(pRExC_state, value);
7976 else if (value == '\\') {
7978 value = utf8n_to_uvchr((U8*)RExC_parse,
7979 RExC_end - RExC_parse,
7980 &numlen, UTF8_ALLOW_DEFAULT);
7981 RExC_parse += numlen;
7984 value = UCHARAT(RExC_parse++);
7985 /* Some compilers cannot handle switching on 64-bit integer
7986 * values, therefore value cannot be an UV. Yes, this will
7987 * be a problem later if we want switch on Unicode.
7988 * A similar issue a little bit later when switching on
7989 * namedclass. --jhi */
7990 switch ((I32)value) {
7991 case 'w': namedclass = ANYOF_ALNUM; break;
7992 case 'W': namedclass = ANYOF_NALNUM; break;
7993 case 's': namedclass = ANYOF_SPACE; break;
7994 case 'S': namedclass = ANYOF_NSPACE; break;
7995 case 'd': namedclass = ANYOF_DIGIT; break;
7996 case 'D': namedclass = ANYOF_NDIGIT; break;
7997 case 'v': namedclass = ANYOF_VERTWS; break;
7998 case 'V': namedclass = ANYOF_NVERTWS; break;
7999 case 'h': namedclass = ANYOF_HORIZWS; break;
8000 case 'H': namedclass = ANYOF_NHORIZWS; break;
8001 case 'N': /* Handle \N{NAME} in class */
8003 /* We only pay attention to the first char of
8004 multichar strings being returned. I kinda wonder
8005 if this makes sense as it does change the behaviour
8006 from earlier versions, OTOH that behaviour was broken
8008 UV v; /* value is register so we cant & it /grrr */
8009 if (reg_namedseq(pRExC_state, &v, NULL)) {
8019 if (RExC_parse >= RExC_end)
8020 vFAIL2("Empty \\%c{}", (U8)value);
8021 if (*RExC_parse == '{') {
8022 const U8 c = (U8)value;
8023 e = strchr(RExC_parse++, '}');
8025 vFAIL2("Missing right brace on \\%c{}", c);
8026 while (isSPACE(UCHARAT(RExC_parse)))
8028 if (e == RExC_parse)
8029 vFAIL2("Empty \\%c{}", c);
8031 while (isSPACE(UCHARAT(RExC_parse + n - 1)))
8039 if (UCHARAT(RExC_parse) == '^') {
8042 value = value == 'p' ? 'P' : 'p'; /* toggle */
8043 while (isSPACE(UCHARAT(RExC_parse))) {
8048 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::%.*s\n",
8049 (value=='p' ? '+' : '!'), (int)n, RExC_parse);
8052 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8053 namedclass = ANYOF_MAX; /* no official name, but it's named */
8056 case 'n': value = '\n'; break;
8057 case 'r': value = '\r'; break;
8058 case 't': value = '\t'; break;
8059 case 'f': value = '\f'; break;
8060 case 'b': value = '\b'; break;
8061 case 'e': value = ASCII_TO_NATIVE('\033');break;
8062 case 'a': value = ASCII_TO_NATIVE('\007');break;
8064 if (*RExC_parse == '{') {
8065 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
8066 | PERL_SCAN_DISALLOW_PREFIX;
8067 char * const e = strchr(RExC_parse++, '}');
8069 vFAIL("Missing right brace on \\x{}");
8071 numlen = e - RExC_parse;
8072 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8076 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
8078 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
8079 RExC_parse += numlen;
8081 if (PL_encoding && value < 0x100)
8082 goto recode_encoding;
8085 value = UCHARAT(RExC_parse++);
8086 value = toCTRL(value);
8088 case '0': case '1': case '2': case '3': case '4':
8089 case '5': case '6': case '7': case '8': case '9':
8093 value = grok_oct(--RExC_parse, &numlen, &flags, NULL);
8094 RExC_parse += numlen;
8095 if (PL_encoding && value < 0x100)
8096 goto recode_encoding;
8101 SV* enc = PL_encoding;
8102 value = reg_recode((const char)(U8)value, &enc);
8103 if (!enc && SIZE_ONLY)
8104 ckWARNreg(RExC_parse,
8105 "Invalid escape in the specified encoding");
8109 if (!SIZE_ONLY && isALPHA(value))
8110 ckWARN2reg(RExC_parse,
8111 "Unrecognized escape \\%c in character class passed through",
8115 } /* end of \blah */
8121 if (namedclass > OOB_NAMEDCLASS) { /* this is a named class \blah */
8123 if (!SIZE_ONLY && !need_class)
8124 ANYOF_CLASS_ZERO(ret);
8128 /* a bad range like a-\d, a-[:digit:] ? */
8132 RExC_parse >= rangebegin ?
8133 RExC_parse - rangebegin : 0;
8134 ckWARN4reg(RExC_parse,
8135 "False [] range \"%*.*s\"",
8138 if (prevvalue < 256) {
8139 ANYOF_BITMAP_SET(ret, prevvalue);
8140 ANYOF_BITMAP_SET(ret, '-');
8143 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8144 Perl_sv_catpvf(aTHX_ listsv,
8145 "%04"UVxf"\n%04"UVxf"\n", (UV)prevvalue, (UV) '-');
8149 range = 0; /* this was not a true range */
8155 const char *what = NULL;
8158 if (namedclass > OOB_NAMEDCLASS)
8159 optimize_invert = FALSE;
8160 /* Possible truncation here but in some 64-bit environments
8161 * the compiler gets heartburn about switch on 64-bit values.
8162 * A similar issue a little earlier when switching on value.
8164 switch ((I32)namedclass) {
8166 case _C_C_T_(ALNUMC, isALNUMC(value), POSIX_CC_UNI_NAME("Alnum"));
8167 case _C_C_T_(ALPHA, isALPHA(value), POSIX_CC_UNI_NAME("Alpha"));
8168 case _C_C_T_(BLANK, isBLANK(value), POSIX_CC_UNI_NAME("Blank"));
8169 case _C_C_T_(CNTRL, isCNTRL(value), POSIX_CC_UNI_NAME("Cntrl"));
8170 case _C_C_T_(GRAPH, isGRAPH(value), POSIX_CC_UNI_NAME("Graph"));
8171 case _C_C_T_(LOWER, isLOWER(value), POSIX_CC_UNI_NAME("Lower"));
8172 case _C_C_T_(PRINT, isPRINT(value), POSIX_CC_UNI_NAME("Print"));
8173 case _C_C_T_(PSXSPC, isPSXSPC(value), POSIX_CC_UNI_NAME("Space"));
8174 case _C_C_T_(PUNCT, isPUNCT(value), POSIX_CC_UNI_NAME("Punct"));
8175 case _C_C_T_(UPPER, isUPPER(value), POSIX_CC_UNI_NAME("Upper"));
8176 #ifdef BROKEN_UNICODE_CHARCLASS_MAPPINGS
8177 case _C_C_T_(ALNUM, isALNUM(value), "Word");
8178 case _C_C_T_(SPACE, isSPACE(value), "SpacePerl");
8180 case _C_C_T_(SPACE, isSPACE(value), "PerlSpace");
8181 case _C_C_T_(ALNUM, isALNUM(value), "PerlWord");
8183 case _C_C_T_(XDIGIT, isXDIGIT(value), "XDigit");
8184 case _C_C_T_NOLOC_(VERTWS, is_VERTWS_latin1(&value), "VertSpace");
8185 case _C_C_T_NOLOC_(HORIZWS, is_HORIZWS_latin1(&value), "HorizSpace");
8188 ANYOF_CLASS_SET(ret, ANYOF_ASCII);
8191 for (value = 0; value < 128; value++)
8192 ANYOF_BITMAP_SET(ret, value);
8194 for (value = 0; value < 256; value++) {
8196 ANYOF_BITMAP_SET(ret, value);
8205 ANYOF_CLASS_SET(ret, ANYOF_NASCII);
8208 for (value = 128; value < 256; value++)
8209 ANYOF_BITMAP_SET(ret, value);
8211 for (value = 0; value < 256; value++) {
8212 if (!isASCII(value))
8213 ANYOF_BITMAP_SET(ret, value);
8222 ANYOF_CLASS_SET(ret, ANYOF_DIGIT);
8224 /* consecutive digits assumed */
8225 for (value = '0'; value <= '9'; value++)
8226 ANYOF_BITMAP_SET(ret, value);
8229 what = POSIX_CC_UNI_NAME("Digit");
8233 ANYOF_CLASS_SET(ret, ANYOF_NDIGIT);
8235 /* consecutive digits assumed */
8236 for (value = 0; value < '0'; value++)
8237 ANYOF_BITMAP_SET(ret, value);
8238 for (value = '9' + 1; value < 256; value++)
8239 ANYOF_BITMAP_SET(ret, value);
8242 what = POSIX_CC_UNI_NAME("Digit");
8245 /* this is to handle \p and \P */
8248 vFAIL("Invalid [::] class");
8252 /* Strings such as "+utf8::isWord\n" */
8253 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::Is%s\n", yesno, what);
8256 ANYOF_FLAGS(ret) |= ANYOF_CLASS;
8259 } /* end of namedclass \blah */
8262 if (prevvalue > (IV)value) /* b-a */ {
8263 const int w = RExC_parse - rangebegin;
8264 Simple_vFAIL4("Invalid [] range \"%*.*s\"", w, w, rangebegin);
8265 range = 0; /* not a valid range */
8269 prevvalue = value; /* save the beginning of the range */
8270 if (*RExC_parse == '-' && RExC_parse+1 < RExC_end &&
8271 RExC_parse[1] != ']') {
8274 /* a bad range like \w-, [:word:]- ? */
8275 if (namedclass > OOB_NAMEDCLASS) {
8276 if (ckWARN(WARN_REGEXP)) {
8278 RExC_parse >= rangebegin ?
8279 RExC_parse - rangebegin : 0;
8281 "False [] range \"%*.*s\"",
8285 ANYOF_BITMAP_SET(ret, '-');
8287 range = 1; /* yeah, it's a range! */
8288 continue; /* but do it the next time */
8292 /* now is the next time */
8293 /*stored += (value - prevvalue + 1);*/
8295 if (prevvalue < 256) {
8296 const IV ceilvalue = value < 256 ? value : 255;
8299 /* In EBCDIC [\x89-\x91] should include
8300 * the \x8e but [i-j] should not. */
8301 if (literal_endpoint == 2 &&
8302 ((isLOWER(prevvalue) && isLOWER(ceilvalue)) ||
8303 (isUPPER(prevvalue) && isUPPER(ceilvalue))))
8305 if (isLOWER(prevvalue)) {
8306 for (i = prevvalue; i <= ceilvalue; i++)
8307 if (isLOWER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8309 ANYOF_BITMAP_SET(ret, i);
8312 for (i = prevvalue; i <= ceilvalue; i++)
8313 if (isUPPER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
8315 ANYOF_BITMAP_SET(ret, i);
8321 for (i = prevvalue; i <= ceilvalue; i++) {
8322 if (!ANYOF_BITMAP_TEST(ret,i)) {
8324 ANYOF_BITMAP_SET(ret, i);
8328 if (value > 255 || UTF) {
8329 const UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
8330 const UV natvalue = NATIVE_TO_UNI(value);
8331 stored+=2; /* can't optimize this class */
8332 ANYOF_FLAGS(ret) |= ANYOF_UNICODE;
8333 if (prevnatvalue < natvalue) { /* what about > ? */
8334 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\t%04"UVxf"\n",
8335 prevnatvalue, natvalue);
8337 else if (prevnatvalue == natvalue) {
8338 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n", natvalue);
8340 U8 foldbuf[UTF8_MAXBYTES_CASE+1];
8342 const UV f = to_uni_fold(natvalue, foldbuf, &foldlen);
8344 #ifdef EBCDIC /* RD t/uni/fold ff and 6b */
8345 if (RExC_precomp[0] == ':' &&
8346 RExC_precomp[1] == '[' &&
8347 (f == 0xDF || f == 0x92)) {
8348 f = NATIVE_TO_UNI(f);
8351 /* If folding and foldable and a single
8352 * character, insert also the folded version
8353 * to the charclass. */
8355 #ifdef EBCDIC /* RD tunifold ligatures s,t fb05, fb06 */
8356 if ((RExC_precomp[0] == ':' &&
8357 RExC_precomp[1] == '[' &&
8359 (value == 0xFB05 || value == 0xFB06))) ?
8360 foldlen == ((STRLEN)UNISKIP(f) - 1) :
8361 foldlen == (STRLEN)UNISKIP(f) )
8363 if (foldlen == (STRLEN)UNISKIP(f))
8365 Perl_sv_catpvf(aTHX_ listsv,
8368 /* Any multicharacter foldings
8369 * require the following transform:
8370 * [ABCDEF] -> (?:[ABCabcDEFd]|pq|rst)
8371 * where E folds into "pq" and F folds
8372 * into "rst", all other characters
8373 * fold to single characters. We save
8374 * away these multicharacter foldings,
8375 * to be later saved as part of the
8376 * additional "s" data. */
8379 if (!unicode_alternate)
8380 unicode_alternate = newAV();
8381 sv = newSVpvn_utf8((char*)foldbuf, foldlen,
8383 av_push(unicode_alternate, sv);
8387 /* If folding and the value is one of the Greek
8388 * sigmas insert a few more sigmas to make the
8389 * folding rules of the sigmas to work right.
8390 * Note that not all the possible combinations
8391 * are handled here: some of them are handled
8392 * by the standard folding rules, and some of
8393 * them (literal or EXACTF cases) are handled
8394 * during runtime in regexec.c:S_find_byclass(). */
8395 if (value == UNICODE_GREEK_SMALL_LETTER_FINAL_SIGMA) {
8396 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8397 (UV)UNICODE_GREEK_CAPITAL_LETTER_SIGMA);
8398 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8399 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8401 else if (value == UNICODE_GREEK_CAPITAL_LETTER_SIGMA)
8402 Perl_sv_catpvf(aTHX_ listsv, "%04"UVxf"\n",
8403 (UV)UNICODE_GREEK_SMALL_LETTER_SIGMA);
8408 literal_endpoint = 0;
8412 range = 0; /* this range (if it was one) is done now */
8416 ANYOF_FLAGS(ret) |= ANYOF_LARGE;
8418 RExC_size += ANYOF_CLASS_ADD_SKIP;
8420 RExC_emit += ANYOF_CLASS_ADD_SKIP;
8426 /****** !SIZE_ONLY AFTER HERE *********/
8428 if( stored == 1 && (value < 128 || (value < 256 && !UTF))
8429 && !( ANYOF_FLAGS(ret) & ( ANYOF_FLAGS_ALL ^ ANYOF_FOLD ) )
8431 /* optimize single char class to an EXACT node
8432 but *only* when its not a UTF/high char */
8433 const char * cur_parse= RExC_parse;
8434 RExC_emit = (regnode *)orig_emit;
8435 RExC_parse = (char *)orig_parse;
8436 ret = reg_node(pRExC_state,
8437 (U8)((ANYOF_FLAGS(ret) & ANYOF_FOLD) ? EXACTF : EXACT));
8438 RExC_parse = (char *)cur_parse;
8439 *STRING(ret)= (char)value;
8441 RExC_emit += STR_SZ(1);
8442 SvREFCNT_dec(listsv);
8445 /* optimize case-insensitive simple patterns (e.g. /[a-z]/i) */
8446 if ( /* If the only flag is folding (plus possibly inversion). */
8447 ((ANYOF_FLAGS(ret) & (ANYOF_FLAGS_ALL ^ ANYOF_INVERT)) == ANYOF_FOLD)
8449 for (value = 0; value < 256; ++value) {
8450 if (ANYOF_BITMAP_TEST(ret, value)) {
8451 UV fold = PL_fold[value];
8454 ANYOF_BITMAP_SET(ret, fold);
8457 ANYOF_FLAGS(ret) &= ~ANYOF_FOLD;
8460 /* optimize inverted simple patterns (e.g. [^a-z]) */
8461 if (optimize_invert &&
8462 /* If the only flag is inversion. */
8463 (ANYOF_FLAGS(ret) & ANYOF_FLAGS_ALL) == ANYOF_INVERT) {
8464 for (value = 0; value < ANYOF_BITMAP_SIZE; ++value)
8465 ANYOF_BITMAP(ret)[value] ^= ANYOF_FLAGS_ALL;
8466 ANYOF_FLAGS(ret) = ANYOF_UNICODE_ALL;
8469 AV * const av = newAV();
8471 /* The 0th element stores the character class description
8472 * in its textual form: used later (regexec.c:Perl_regclass_swash())
8473 * to initialize the appropriate swash (which gets stored in
8474 * the 1st element), and also useful for dumping the regnode.
8475 * The 2nd element stores the multicharacter foldings,
8476 * used later (regexec.c:S_reginclass()). */
8477 av_store(av, 0, listsv);
8478 av_store(av, 1, NULL);
8479 av_store(av, 2, MUTABLE_SV(unicode_alternate));
8480 rv = newRV_noinc(MUTABLE_SV(av));
8481 n = add_data(pRExC_state, 1, "s");
8482 RExC_rxi->data->data[n] = (void*)rv;
8490 /* reg_skipcomment()
8492 Absorbs an /x style # comments from the input stream.
8493 Returns true if there is more text remaining in the stream.
8494 Will set the REG_SEEN_RUN_ON_COMMENT flag if the comment
8495 terminates the pattern without including a newline.
8497 Note its the callers responsibility to ensure that we are
8503 S_reg_skipcomment(pTHX_ RExC_state_t *pRExC_state)
8507 PERL_ARGS_ASSERT_REG_SKIPCOMMENT;
8509 while (RExC_parse < RExC_end)
8510 if (*RExC_parse++ == '\n') {
8515 /* we ran off the end of the pattern without ending
8516 the comment, so we have to add an \n when wrapping */
8517 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
8525 Advance that parse position, and optionall absorbs
8526 "whitespace" from the inputstream.
8528 Without /x "whitespace" means (?#...) style comments only,
8529 with /x this means (?#...) and # comments and whitespace proper.
8531 Returns the RExC_parse point from BEFORE the scan occurs.
8533 This is the /x friendly way of saying RExC_parse++.
8537 S_nextchar(pTHX_ RExC_state_t *pRExC_state)
8539 char* const retval = RExC_parse++;
8541 PERL_ARGS_ASSERT_NEXTCHAR;
8544 if (*RExC_parse == '(' && RExC_parse[1] == '?' &&
8545 RExC_parse[2] == '#') {
8546 while (*RExC_parse != ')') {
8547 if (RExC_parse == RExC_end)
8548 FAIL("Sequence (?#... not terminated");
8554 if (RExC_flags & RXf_PMf_EXTENDED) {
8555 if (isSPACE(*RExC_parse)) {
8559 else if (*RExC_parse == '#') {
8560 if ( reg_skipcomment( pRExC_state ) )
8569 - reg_node - emit a node
8571 STATIC regnode * /* Location. */
8572 S_reg_node(pTHX_ RExC_state_t *pRExC_state, U8 op)
8575 register regnode *ptr;
8576 regnode * const ret = RExC_emit;
8577 GET_RE_DEBUG_FLAGS_DECL;
8579 PERL_ARGS_ASSERT_REG_NODE;
8582 SIZE_ALIGN(RExC_size);
8586 if (RExC_emit >= RExC_emit_bound)
8587 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8589 NODE_ALIGN_FILL(ret);
8591 FILL_ADVANCE_NODE(ptr, op);
8592 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (ptr) - 1);
8593 #ifdef RE_TRACK_PATTERN_OFFSETS
8594 if (RExC_offsets) { /* MJD */
8595 MJD_OFFSET_DEBUG(("%s:%d: (op %s) %s %"UVuf" (len %"UVuf") (max %"UVuf").\n",
8596 "reg_node", __LINE__,
8598 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0]
8599 ? "Overwriting end of array!\n" : "OK",
8600 (UV)(RExC_emit - RExC_emit_start),
8601 (UV)(RExC_parse - RExC_start),
8602 (UV)RExC_offsets[0]));
8603 Set_Node_Offset(RExC_emit, RExC_parse + (op == END));
8611 - reganode - emit a node with an argument
8613 STATIC regnode * /* Location. */
8614 S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
8617 register regnode *ptr;
8618 regnode * const ret = RExC_emit;
8619 GET_RE_DEBUG_FLAGS_DECL;
8621 PERL_ARGS_ASSERT_REGANODE;
8624 SIZE_ALIGN(RExC_size);
8629 assert(2==regarglen[op]+1);
8631 Anything larger than this has to allocate the extra amount.
8632 If we changed this to be:
8634 RExC_size += (1 + regarglen[op]);
8636 then it wouldn't matter. Its not clear what side effect
8637 might come from that so its not done so far.
8642 if (RExC_emit >= RExC_emit_bound)
8643 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d", op);
8645 NODE_ALIGN_FILL(ret);
8647 FILL_ADVANCE_NODE_ARG(ptr, op, arg);
8648 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (ptr) - 2);
8649 #ifdef RE_TRACK_PATTERN_OFFSETS
8650 if (RExC_offsets) { /* MJD */
8651 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8655 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0] ?
8656 "Overwriting end of array!\n" : "OK",
8657 (UV)(RExC_emit - RExC_emit_start),
8658 (UV)(RExC_parse - RExC_start),
8659 (UV)RExC_offsets[0]));
8660 Set_Cur_Node_Offset;
8668 - reguni - emit (if appropriate) a Unicode character
8671 S_reguni(pTHX_ const RExC_state_t *pRExC_state, UV uv, char* s)
8675 PERL_ARGS_ASSERT_REGUNI;
8677 return SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
8681 - reginsert - insert an operator in front of already-emitted operand
8683 * Means relocating the operand.
8686 S_reginsert(pTHX_ RExC_state_t *pRExC_state, U8 op, regnode *opnd, U32 depth)
8689 register regnode *src;
8690 register regnode *dst;
8691 register regnode *place;
8692 const int offset = regarglen[(U8)op];
8693 const int size = NODE_STEP_REGNODE + offset;
8694 GET_RE_DEBUG_FLAGS_DECL;
8696 PERL_ARGS_ASSERT_REGINSERT;
8697 PERL_UNUSED_ARG(depth);
8698 /* (PL_regkind[(U8)op] == CURLY ? EXTRA_STEP_2ARGS : 0); */
8699 DEBUG_PARSE_FMT("inst"," - %s",PL_reg_name[op]);
8708 if (RExC_open_parens) {
8710 /*DEBUG_PARSE_FMT("inst"," - %"IVdf, (IV)RExC_npar);*/
8711 for ( paren=0 ; paren < RExC_npar ; paren++ ) {
8712 if ( RExC_open_parens[paren] >= opnd ) {
8713 /*DEBUG_PARSE_FMT("open"," - %d",size);*/
8714 RExC_open_parens[paren] += size;
8716 /*DEBUG_PARSE_FMT("open"," - %s","ok");*/
8718 if ( RExC_close_parens[paren] >= opnd ) {
8719 /*DEBUG_PARSE_FMT("close"," - %d",size);*/
8720 RExC_close_parens[paren] += size;
8722 /*DEBUG_PARSE_FMT("close"," - %s","ok");*/
8727 while (src > opnd) {
8728 StructCopy(--src, --dst, regnode);
8729 #ifdef RE_TRACK_PATTERN_OFFSETS
8730 if (RExC_offsets) { /* MJD 20010112 */
8731 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s copy %"UVuf" -> %"UVuf" (max %"UVuf").\n",
8735 (UV)(dst - RExC_emit_start) > RExC_offsets[0]
8736 ? "Overwriting end of array!\n" : "OK",
8737 (UV)(src - RExC_emit_start),
8738 (UV)(dst - RExC_emit_start),
8739 (UV)RExC_offsets[0]));
8740 Set_Node_Offset_To_R(dst-RExC_emit_start, Node_Offset(src));
8741 Set_Node_Length_To_R(dst-RExC_emit_start, Node_Length(src));
8747 place = opnd; /* Op node, where operand used to be. */
8748 #ifdef RE_TRACK_PATTERN_OFFSETS
8749 if (RExC_offsets) { /* MJD */
8750 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
8754 (UV)(place - RExC_emit_start) > RExC_offsets[0]
8755 ? "Overwriting end of array!\n" : "OK",
8756 (UV)(place - RExC_emit_start),
8757 (UV)(RExC_parse - RExC_start),
8758 (UV)RExC_offsets[0]));
8759 Set_Node_Offset(place, RExC_parse);
8760 Set_Node_Length(place, 1);
8763 src = NEXTOPER(place);
8764 FILL_ADVANCE_NODE(place, op);
8765 REH_CALL_COMP_NODE_HOOK(pRExC_state->rx, (place) - 1);
8766 Zero(src, offset, regnode);
8770 - regtail - set the next-pointer at the end of a node chain of p to val.
8771 - SEE ALSO: regtail_study
8773 /* TODO: All three parms should be const */
8775 S_regtail(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8778 register regnode *scan;
8779 GET_RE_DEBUG_FLAGS_DECL;
8781 PERL_ARGS_ASSERT_REGTAIL;
8783 PERL_UNUSED_ARG(depth);
8789 /* Find last node. */
8792 regnode * const temp = regnext(scan);
8794 SV * const mysv=sv_newmortal();
8795 DEBUG_PARSE_MSG((scan==p ? "tail" : ""));
8796 regprop(RExC_rx, mysv, scan);
8797 PerlIO_printf(Perl_debug_log, "~ %s (%d) %s %s\n",
8798 SvPV_nolen_const(mysv), REG_NODE_NUM(scan),
8799 (temp == NULL ? "->" : ""),
8800 (temp == NULL ? PL_reg_name[OP(val)] : "")
8808 if (reg_off_by_arg[OP(scan)]) {
8809 ARG_SET(scan, val - scan);
8812 NEXT_OFF(scan) = val - scan;
8818 - regtail_study - set the next-pointer at the end of a node chain of p to val.
8819 - Look for optimizable sequences at the same time.
8820 - currently only looks for EXACT chains.
8822 This is expermental code. The idea is to use this routine to perform
8823 in place optimizations on branches and groups as they are constructed,
8824 with the long term intention of removing optimization from study_chunk so
8825 that it is purely analytical.
8827 Currently only used when in DEBUG mode. The macro REGTAIL_STUDY() is used
8828 to control which is which.
8831 /* TODO: All four parms should be const */
8834 S_regtail_study(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
8837 register regnode *scan;
8839 #ifdef EXPERIMENTAL_INPLACESCAN
8842 GET_RE_DEBUG_FLAGS_DECL;
8844 PERL_ARGS_ASSERT_REGTAIL_STUDY;
8850 /* Find last node. */
8854 regnode * const temp = regnext(scan);
8855 #ifdef EXPERIMENTAL_INPLACESCAN
8856 if (PL_regkind[OP(scan)] == EXACT)
8857 if (join_exact(pRExC_state,scan,&min,1,val,depth+1))
8865 if( exact == PSEUDO )
8867 else if ( exact != OP(scan) )
8876 SV * const mysv=sv_newmortal();
8877 DEBUG_PARSE_MSG((scan==p ? "tsdy" : ""));
8878 regprop(RExC_rx, mysv, scan);
8879 PerlIO_printf(Perl_debug_log, "~ %s (%d) -> %s\n",
8880 SvPV_nolen_const(mysv),
8882 PL_reg_name[exact]);
8889 SV * const mysv_val=sv_newmortal();
8890 DEBUG_PARSE_MSG("");
8891 regprop(RExC_rx, mysv_val, val);
8892 PerlIO_printf(Perl_debug_log, "~ attach to %s (%"IVdf") offset to %"IVdf"\n",
8893 SvPV_nolen_const(mysv_val),
8894 (IV)REG_NODE_NUM(val),
8898 if (reg_off_by_arg[OP(scan)]) {
8899 ARG_SET(scan, val - scan);
8902 NEXT_OFF(scan) = val - scan;
8910 - regcurly - a little FSA that accepts {\d+,?\d*}
8913 S_regcurly(register const char *s)
8915 PERL_ARGS_ASSERT_REGCURLY;
8934 - regdump - dump a regexp onto Perl_debug_log in vaguely comprehensible form
8938 S_regdump_extflags(pTHX_ const char *lead, const U32 flags)
8943 for (bit=0; bit<32; bit++) {
8944 if (flags & (1<<bit)) {
8946 PerlIO_printf(Perl_debug_log, "%s",lead);
8947 PerlIO_printf(Perl_debug_log, "%s ",PL_reg_extflags_name[bit]);
8952 PerlIO_printf(Perl_debug_log, "\n");
8954 PerlIO_printf(Perl_debug_log, "%s[none-set]\n",lead);
8960 Perl_regdump(pTHX_ const regexp *r)
8964 SV * const sv = sv_newmortal();
8965 SV *dsv= sv_newmortal();
8967 GET_RE_DEBUG_FLAGS_DECL;
8969 PERL_ARGS_ASSERT_REGDUMP;
8971 (void)dumpuntil(r, ri->program, ri->program + 1, NULL, NULL, sv, 0, 0);
8973 /* Header fields of interest. */
8974 if (r->anchored_substr) {
8975 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->anchored_substr),
8976 RE_SV_DUMPLEN(r->anchored_substr), 30);
8977 PerlIO_printf(Perl_debug_log,
8978 "anchored %s%s at %"IVdf" ",
8979 s, RE_SV_TAIL(r->anchored_substr),
8980 (IV)r->anchored_offset);
8981 } else if (r->anchored_utf8) {
8982 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->anchored_utf8),
8983 RE_SV_DUMPLEN(r->anchored_utf8), 30);
8984 PerlIO_printf(Perl_debug_log,
8985 "anchored utf8 %s%s at %"IVdf" ",
8986 s, RE_SV_TAIL(r->anchored_utf8),
8987 (IV)r->anchored_offset);
8989 if (r->float_substr) {
8990 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->float_substr),
8991 RE_SV_DUMPLEN(r->float_substr), 30);
8992 PerlIO_printf(Perl_debug_log,
8993 "floating %s%s at %"IVdf"..%"UVuf" ",
8994 s, RE_SV_TAIL(r->float_substr),
8995 (IV)r->float_min_offset, (UV)r->float_max_offset);
8996 } else if (r->float_utf8) {
8997 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->float_utf8),
8998 RE_SV_DUMPLEN(r->float_utf8), 30);
8999 PerlIO_printf(Perl_debug_log,
9000 "floating utf8 %s%s at %"IVdf"..%"UVuf" ",
9001 s, RE_SV_TAIL(r->float_utf8),
9002 (IV)r->float_min_offset, (UV)r->float_max_offset);
9004 if (r->check_substr || r->check_utf8)
9005 PerlIO_printf(Perl_debug_log,
9007 (r->check_substr == r->float_substr
9008 && r->check_utf8 == r->float_utf8
9009 ? "(checking floating" : "(checking anchored"));
9010 if (r->extflags & RXf_NOSCAN)
9011 PerlIO_printf(Perl_debug_log, " noscan");
9012 if (r->extflags & RXf_CHECK_ALL)
9013 PerlIO_printf(Perl_debug_log, " isall");
9014 if (r->check_substr || r->check_utf8)
9015 PerlIO_printf(Perl_debug_log, ") ");
9017 if (ri->regstclass) {
9018 regprop(r, sv, ri->regstclass);
9019 PerlIO_printf(Perl_debug_log, "stclass %s ", SvPVX_const(sv));
9021 if (r->extflags & RXf_ANCH) {
9022 PerlIO_printf(Perl_debug_log, "anchored");
9023 if (r->extflags & RXf_ANCH_BOL)
9024 PerlIO_printf(Perl_debug_log, "(BOL)");
9025 if (r->extflags & RXf_ANCH_MBOL)
9026 PerlIO_printf(Perl_debug_log, "(MBOL)");
9027 if (r->extflags & RXf_ANCH_SBOL)
9028 PerlIO_printf(Perl_debug_log, "(SBOL)");
9029 if (r->extflags & RXf_ANCH_GPOS)
9030 PerlIO_printf(Perl_debug_log, "(GPOS)");
9031 PerlIO_putc(Perl_debug_log, ' ');
9033 if (r->extflags & RXf_GPOS_SEEN)
9034 PerlIO_printf(Perl_debug_log, "GPOS:%"UVuf" ", (UV)r->gofs);
9035 if (r->intflags & PREGf_SKIP)
9036 PerlIO_printf(Perl_debug_log, "plus ");
9037 if (r->intflags & PREGf_IMPLICIT)
9038 PerlIO_printf(Perl_debug_log, "implicit ");
9039 PerlIO_printf(Perl_debug_log, "minlen %"IVdf" ", (IV)r->minlen);
9040 if (r->extflags & RXf_EVAL_SEEN)
9041 PerlIO_printf(Perl_debug_log, "with eval ");
9042 PerlIO_printf(Perl_debug_log, "\n");
9043 DEBUG_FLAGS_r(regdump_extflags("r->extflags: ",r->extflags));
9045 PERL_ARGS_ASSERT_REGDUMP;
9046 PERL_UNUSED_CONTEXT;
9048 #endif /* DEBUGGING */
9052 - regprop - printable representation of opcode
9054 #define EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags) \
9057 Perl_sv_catpvf(aTHX_ sv,"%s][%s",PL_colors[1],PL_colors[0]); \
9058 if (flags & ANYOF_INVERT) \
9059 /*make sure the invert info is in each */ \
9060 sv_catpvs(sv, "^"); \
9066 Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
9071 RXi_GET_DECL(prog,progi);
9072 GET_RE_DEBUG_FLAGS_DECL;
9074 PERL_ARGS_ASSERT_REGPROP;
9078 if (OP(o) > REGNODE_MAX) /* regnode.type is unsigned */
9079 /* It would be nice to FAIL() here, but this may be called from
9080 regexec.c, and it would be hard to supply pRExC_state. */
9081 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(o), (int)REGNODE_MAX);
9082 sv_catpv(sv, PL_reg_name[OP(o)]); /* Take off const! */
9084 k = PL_regkind[OP(o)];
9088 /* Using is_utf8_string() (via PERL_PV_UNI_DETECT)
9089 * is a crude hack but it may be the best for now since
9090 * we have no flag "this EXACTish node was UTF-8"
9092 pv_pretty(sv, STRING(o), STR_LEN(o), 60, PL_colors[0], PL_colors[1],
9093 PERL_PV_ESCAPE_UNI_DETECT |
9094 PERL_PV_PRETTY_ELLIPSES |
9095 PERL_PV_PRETTY_LTGT |
9096 PERL_PV_PRETTY_NOCLEAR
9098 } else if (k == TRIE) {
9099 /* print the details of the trie in dumpuntil instead, as
9100 * progi->data isn't available here */
9101 const char op = OP(o);
9102 const U32 n = ARG(o);
9103 const reg_ac_data * const ac = IS_TRIE_AC(op) ?
9104 (reg_ac_data *)progi->data->data[n] :
9106 const reg_trie_data * const trie
9107 = (reg_trie_data*)progi->data->data[!IS_TRIE_AC(op) ? n : ac->trie];
9109 Perl_sv_catpvf(aTHX_ sv, "-%s",PL_reg_name[o->flags]);
9110 DEBUG_TRIE_COMPILE_r(
9111 Perl_sv_catpvf(aTHX_ sv,
9112 "<S:%"UVuf"/%"IVdf" W:%"UVuf" L:%"UVuf"/%"UVuf" C:%"UVuf"/%"UVuf">",
9113 (UV)trie->startstate,
9114 (IV)trie->statecount-1, /* -1 because of the unused 0 element */
9115 (UV)trie->wordcount,
9118 (UV)TRIE_CHARCOUNT(trie),
9119 (UV)trie->uniquecharcount
9122 if ( IS_ANYOF_TRIE(op) || trie->bitmap ) {
9124 int rangestart = -1;
9125 U8* bitmap = IS_ANYOF_TRIE(op) ? (U8*)ANYOF_BITMAP(o) : (U8*)TRIE_BITMAP(trie);
9127 for (i = 0; i <= 256; i++) {
9128 if (i < 256 && BITMAP_TEST(bitmap,i)) {
9129 if (rangestart == -1)
9131 } else if (rangestart != -1) {
9132 if (i <= rangestart + 3)
9133 for (; rangestart < i; rangestart++)
9134 put_byte(sv, rangestart);
9136 put_byte(sv, rangestart);
9138 put_byte(sv, i - 1);
9146 } else if (k == CURLY) {
9147 if (OP(o) == CURLYM || OP(o) == CURLYN || OP(o) == CURLYX)
9148 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* Parenth number */
9149 Perl_sv_catpvf(aTHX_ sv, " {%d,%d}", ARG1(o), ARG2(o));
9151 else if (k == WHILEM && o->flags) /* Ordinal/of */
9152 Perl_sv_catpvf(aTHX_ sv, "[%d/%d]", o->flags & 0xf, o->flags>>4);
9153 else if (k == REF || k == OPEN || k == CLOSE || k == GROUPP || OP(o)==ACCEPT) {
9154 Perl_sv_catpvf(aTHX_ sv, "%d", (int)ARG(o)); /* Parenth number */
9155 if ( RXp_PAREN_NAMES(prog) ) {
9156 if ( k != REF || OP(o) < NREF) {
9157 AV *list= MUTABLE_AV(progi->data->data[progi->name_list_idx]);
9158 SV **name= av_fetch(list, ARG(o), 0 );
9160 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9163 AV *list= MUTABLE_AV(progi->data->data[ progi->name_list_idx ]);
9164 SV *sv_dat= MUTABLE_SV(progi->data->data[ ARG( o ) ]);
9165 I32 *nums=(I32*)SvPVX(sv_dat);
9166 SV **name= av_fetch(list, nums[0], 0 );
9169 for ( n=0; n<SvIVX(sv_dat); n++ ) {
9170 Perl_sv_catpvf(aTHX_ sv, "%s%"IVdf,
9171 (n ? "," : ""), (IV)nums[n]);
9173 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
9177 } else if (k == GOSUB)
9178 Perl_sv_catpvf(aTHX_ sv, "%d[%+d]", (int)ARG(o),(int)ARG2L(o)); /* Paren and offset */
9179 else if (k == VERB) {
9181 Perl_sv_catpvf(aTHX_ sv, ":%"SVf,
9182 SVfARG((MUTABLE_SV(progi->data->data[ ARG( o ) ]))));
9183 } else if (k == LOGICAL)
9184 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* 2: embedded, otherwise 1 */
9185 else if (k == FOLDCHAR)
9186 Perl_sv_catpvf(aTHX_ sv, "[0x%"UVXf"]", PTR2UV(ARG(o)) );
9187 else if (k == ANYOF) {
9188 int i, rangestart = -1;
9189 const U8 flags = ANYOF_FLAGS(o);
9192 /* Should be synchronized with * ANYOF_ #xdefines in regcomp.h */
9193 static const char * const anyofs[] = {
9226 if (flags & ANYOF_LOCALE)
9227 sv_catpvs(sv, "{loc}");
9228 if (flags & ANYOF_FOLD)
9229 sv_catpvs(sv, "{i}");
9230 Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
9231 if (flags & ANYOF_INVERT)
9234 /* output what the standard cp 0-255 bitmap matches */
9235 for (i = 0; i <= 256; i++) {
9236 if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
9237 if (rangestart == -1)
9239 } else if (rangestart != -1) {
9240 if (i <= rangestart + 3)
9241 for (; rangestart < i; rangestart++)
9242 put_byte(sv, rangestart);
9244 put_byte(sv, rangestart);
9246 put_byte(sv, i - 1);
9253 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9254 /* output any special charclass tests (used mostly under use locale) */
9255 if (o->flags & ANYOF_CLASS)
9256 for (i = 0; i < (int)(sizeof(anyofs)/sizeof(char*)); i++)
9257 if (ANYOF_CLASS_TEST(o,i)) {
9258 sv_catpv(sv, anyofs[i]);
9262 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
9264 /* output information about the unicode matching */
9265 if (flags & ANYOF_UNICODE)
9266 sv_catpvs(sv, "{unicode}");
9267 else if (flags & ANYOF_UNICODE_ALL)
9268 sv_catpvs(sv, "{unicode_all}");
9272 SV * const sw = regclass_swash(prog, o, FALSE, &lv, 0);
9276 U8 s[UTF8_MAXBYTES_CASE+1];
9278 for (i = 0; i <= 256; i++) { /* just the first 256 */
9279 uvchr_to_utf8(s, i);
9281 if (i < 256 && swash_fetch(sw, s, TRUE)) {
9282 if (rangestart == -1)
9284 } else if (rangestart != -1) {
9285 if (i <= rangestart + 3)
9286 for (; rangestart < i; rangestart++) {
9287 const U8 * const e = uvchr_to_utf8(s,rangestart);
9289 for(p = s; p < e; p++)
9293 const U8 *e = uvchr_to_utf8(s,rangestart);
9295 for (p = s; p < e; p++)
9298 e = uvchr_to_utf8(s, i-1);
9299 for (p = s; p < e; p++)
9306 sv_catpvs(sv, "..."); /* et cetera */
9310 char *s = savesvpv(lv);
9311 char * const origs = s;
9313 while (*s && *s != '\n')
9317 const char * const t = ++s;
9335 Perl_sv_catpvf(aTHX_ sv, "%s]", PL_colors[1]);
9337 else if (k == BRANCHJ && (OP(o) == UNLESSM || OP(o) == IFMATCH))
9338 Perl_sv_catpvf(aTHX_ sv, "[%d]", -(o->flags));
9340 PERL_UNUSED_CONTEXT;
9341 PERL_UNUSED_ARG(sv);
9343 PERL_UNUSED_ARG(prog);
9344 #endif /* DEBUGGING */
9348 Perl_re_intuit_string(pTHX_ REGEXP * const r)
9349 { /* Assume that RE_INTUIT is set */
9351 struct regexp *const prog = (struct regexp *)SvANY(r);
9352 GET_RE_DEBUG_FLAGS_DECL;
9354 PERL_ARGS_ASSERT_RE_INTUIT_STRING;
9355 PERL_UNUSED_CONTEXT;
9359 const char * const s = SvPV_nolen_const(prog->check_substr
9360 ? prog->check_substr : prog->check_utf8);
9362 if (!PL_colorset) reginitcolors();
9363 PerlIO_printf(Perl_debug_log,
9364 "%sUsing REx %ssubstr:%s \"%s%.60s%s%s\"\n",
9366 prog->check_substr ? "" : "utf8 ",
9367 PL_colors[5],PL_colors[0],
9370 (strlen(s) > 60 ? "..." : ""));
9373 return prog->check_substr ? prog->check_substr : prog->check_utf8;
9379 handles refcounting and freeing the perl core regexp structure. When
9380 it is necessary to actually free the structure the first thing it
9381 does is call the 'free' method of the regexp_engine associated to to
9382 the regexp, allowing the handling of the void *pprivate; member
9383 first. (This routine is not overridable by extensions, which is why
9384 the extensions free is called first.)
9386 See regdupe and regdupe_internal if you change anything here.
9388 #ifndef PERL_IN_XSUB_RE
9390 Perl_pregfree(pTHX_ REGEXP *r)
9396 Perl_pregfree2(pTHX_ REGEXP *rx)
9399 struct regexp *const r = (struct regexp *)SvANY(rx);
9400 GET_RE_DEBUG_FLAGS_DECL;
9402 PERL_ARGS_ASSERT_PREGFREE2;
9405 ReREFCNT_dec(r->mother_re);
9407 CALLREGFREE_PVT(rx); /* free the private data */
9408 SvREFCNT_dec(RXp_PAREN_NAMES(r));
9411 SvREFCNT_dec(r->anchored_substr);
9412 SvREFCNT_dec(r->anchored_utf8);
9413 SvREFCNT_dec(r->float_substr);
9414 SvREFCNT_dec(r->float_utf8);
9415 Safefree(r->substrs);
9417 RX_MATCH_COPY_FREE(rx);
9418 #ifdef PERL_OLD_COPY_ON_WRITE
9419 SvREFCNT_dec(r->saved_copy);
9426 This is a hacky workaround to the structural issue of match results
9427 being stored in the regexp structure which is in turn stored in
9428 PL_curpm/PL_reg_curpm. The problem is that due to qr// the pattern
9429 could be PL_curpm in multiple contexts, and could require multiple
9430 result sets being associated with the pattern simultaneously, such
9431 as when doing a recursive match with (??{$qr})
9433 The solution is to make a lightweight copy of the regexp structure
9434 when a qr// is returned from the code executed by (??{$qr}) this
9435 lightweight copy doesnt actually own any of its data except for
9436 the starp/end and the actual regexp structure itself.
9442 Perl_reg_temp_copy (pTHX_ REGEXP *ret_x, REGEXP *rx)
9445 struct regexp *const r = (struct regexp *)SvANY(rx);
9446 register const I32 npar = r->nparens+1;
9448 PERL_ARGS_ASSERT_REG_TEMP_COPY;
9451 ret_x = (REGEXP*) newSV_type(SVt_REGEXP);
9452 ret = (struct regexp *)SvANY(ret_x);
9454 (void)ReREFCNT_inc(rx);
9455 /* We can take advantage of the existing "copied buffer" mechanism in SVs
9456 by pointing directly at the buffer, but flagging that the allocated
9457 space in the copy is zero. As we've just done a struct copy, it's now
9458 a case of zero-ing that, rather than copying the current length. */
9459 SvPV_set(ret_x, RX_WRAPPED(rx));
9460 SvFLAGS(ret_x) |= SvFLAGS(rx) & (SVf_POK|SVp_POK|SVf_UTF8);
9461 memcpy(&(ret->xpv_cur), &(r->xpv_cur),
9462 sizeof(regexp) - STRUCT_OFFSET(regexp, xpv_cur));
9463 SvLEN_set(ret_x, 0);
9464 Newx(ret->offs, npar, regexp_paren_pair);
9465 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9467 Newx(ret->substrs, 1, struct reg_substr_data);
9468 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9470 SvREFCNT_inc_void(ret->anchored_substr);
9471 SvREFCNT_inc_void(ret->anchored_utf8);
9472 SvREFCNT_inc_void(ret->float_substr);
9473 SvREFCNT_inc_void(ret->float_utf8);
9475 /* check_substr and check_utf8, if non-NULL, point to either their
9476 anchored or float namesakes, and don't hold a second reference. */
9478 RX_MATCH_COPIED_off(ret_x);
9479 #ifdef PERL_OLD_COPY_ON_WRITE
9480 ret->saved_copy = NULL;
9482 ret->mother_re = rx;
9488 /* regfree_internal()
9490 Free the private data in a regexp. This is overloadable by
9491 extensions. Perl takes care of the regexp structure in pregfree(),
9492 this covers the *pprivate pointer which technically perldoesnt
9493 know about, however of course we have to handle the
9494 regexp_internal structure when no extension is in use.
9496 Note this is called before freeing anything in the regexp
9501 Perl_regfree_internal(pTHX_ REGEXP * const rx)
9504 struct regexp *const r = (struct regexp *)SvANY(rx);
9506 GET_RE_DEBUG_FLAGS_DECL;
9508 PERL_ARGS_ASSERT_REGFREE_INTERNAL;
9514 SV *dsv= sv_newmortal();
9515 RE_PV_QUOTED_DECL(s, RX_UTF8(rx),
9516 dsv, RX_PRECOMP(rx), RX_PRELEN(rx), 60);
9517 PerlIO_printf(Perl_debug_log,"%sFreeing REx:%s %s\n",
9518 PL_colors[4],PL_colors[5],s);
9521 #ifdef RE_TRACK_PATTERN_OFFSETS
9523 Safefree(ri->u.offsets); /* 20010421 MJD */
9526 int n = ri->data->count;
9527 PAD* new_comppad = NULL;
9532 /* If you add a ->what type here, update the comment in regcomp.h */
9533 switch (ri->data->what[n]) {
9537 SvREFCNT_dec(MUTABLE_SV(ri->data->data[n]));
9540 Safefree(ri->data->data[n]);
9543 new_comppad = MUTABLE_AV(ri->data->data[n]);
9546 if (new_comppad == NULL)
9547 Perl_croak(aTHX_ "panic: pregfree comppad");
9548 PAD_SAVE_LOCAL(old_comppad,
9549 /* Watch out for global destruction's random ordering. */
9550 (SvTYPE(new_comppad) == SVt_PVAV) ? new_comppad : NULL
9553 refcnt = OpREFCNT_dec((OP_4tree*)ri->data->data[n]);
9556 op_free((OP_4tree*)ri->data->data[n]);
9558 PAD_RESTORE_LOCAL(old_comppad);
9559 SvREFCNT_dec(MUTABLE_SV(new_comppad));
9565 { /* Aho Corasick add-on structure for a trie node.
9566 Used in stclass optimization only */
9568 reg_ac_data *aho=(reg_ac_data*)ri->data->data[n];
9570 refcount = --aho->refcount;
9573 PerlMemShared_free(aho->states);
9574 PerlMemShared_free(aho->fail);
9575 /* do this last!!!! */
9576 PerlMemShared_free(ri->data->data[n]);
9577 PerlMemShared_free(ri->regstclass);
9583 /* trie structure. */
9585 reg_trie_data *trie=(reg_trie_data*)ri->data->data[n];
9587 refcount = --trie->refcount;
9590 PerlMemShared_free(trie->charmap);
9591 PerlMemShared_free(trie->states);
9592 PerlMemShared_free(trie->trans);
9594 PerlMemShared_free(trie->bitmap);
9596 PerlMemShared_free(trie->wordlen);
9598 PerlMemShared_free(trie->jump);
9600 PerlMemShared_free(trie->nextword);
9601 /* do this last!!!! */
9602 PerlMemShared_free(ri->data->data[n]);
9607 Perl_croak(aTHX_ "panic: regfree data code '%c'", ri->data->what[n]);
9610 Safefree(ri->data->what);
9617 #define sv_dup_inc(s,t) SvREFCNT_inc(sv_dup(s,t))
9618 #define av_dup_inc(s,t) MUTABLE_AV(SvREFCNT_inc(sv_dup((const SV *)s,t)))
9619 #define hv_dup_inc(s,t) MUTABLE_HV(SvREFCNT_inc(sv_dup((const SV *)s,t)))
9620 #define SAVEPVN(p,n) ((p) ? savepvn(p,n) : NULL)
9623 re_dup - duplicate a regexp.
9625 This routine is expected to clone a given regexp structure. It is only
9626 compiled under USE_ITHREADS.
9628 After all of the core data stored in struct regexp is duplicated
9629 the regexp_engine.dupe method is used to copy any private data
9630 stored in the *pprivate pointer. This allows extensions to handle
9631 any duplication it needs to do.
9633 See pregfree() and regfree_internal() if you change anything here.
9635 #if defined(USE_ITHREADS)
9636 #ifndef PERL_IN_XSUB_RE
9638 Perl_re_dup_guts(pTHX_ const REGEXP *sstr, REGEXP *dstr, CLONE_PARAMS *param)
9642 const struct regexp *r = (const struct regexp *)SvANY(sstr);
9643 struct regexp *ret = (struct regexp *)SvANY(dstr);
9645 PERL_ARGS_ASSERT_RE_DUP_GUTS;
9647 npar = r->nparens+1;
9648 Newx(ret->offs, npar, regexp_paren_pair);
9649 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
9651 /* no need to copy these */
9652 Newx(ret->swap, npar, regexp_paren_pair);
9656 /* Do it this way to avoid reading from *r after the StructCopy().
9657 That way, if any of the sv_dup_inc()s dislodge *r from the L1
9658 cache, it doesn't matter. */
9659 const bool anchored = r->check_substr
9660 ? r->check_substr == r->anchored_substr
9661 : r->check_utf8 == r->anchored_utf8;
9662 Newx(ret->substrs, 1, struct reg_substr_data);
9663 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
9665 ret->anchored_substr = sv_dup_inc(ret->anchored_substr, param);
9666 ret->anchored_utf8 = sv_dup_inc(ret->anchored_utf8, param);
9667 ret->float_substr = sv_dup_inc(ret->float_substr, param);
9668 ret->float_utf8 = sv_dup_inc(ret->float_utf8, param);
9670 /* check_substr and check_utf8, if non-NULL, point to either their
9671 anchored or float namesakes, and don't hold a second reference. */
9673 if (ret->check_substr) {
9675 assert(r->check_utf8 == r->anchored_utf8);
9676 ret->check_substr = ret->anchored_substr;
9677 ret->check_utf8 = ret->anchored_utf8;
9679 assert(r->check_substr == r->float_substr);
9680 assert(r->check_utf8 == r->float_utf8);
9681 ret->check_substr = ret->float_substr;
9682 ret->check_utf8 = ret->float_utf8;
9684 } else if (ret->check_utf8) {
9686 ret->check_utf8 = ret->anchored_utf8;
9688 ret->check_utf8 = ret->float_utf8;
9693 RXp_PAREN_NAMES(ret) = hv_dup_inc(RXp_PAREN_NAMES(ret), param);
9696 RXi_SET(ret,CALLREGDUPE_PVT(dstr,param));
9698 if (RX_MATCH_COPIED(dstr))
9699 ret->subbeg = SAVEPVN(ret->subbeg, ret->sublen);
9702 #ifdef PERL_OLD_COPY_ON_WRITE
9703 ret->saved_copy = NULL;
9706 ret->mother_re = NULL;
9709 #endif /* PERL_IN_XSUB_RE */
9714 This is the internal complement to regdupe() which is used to copy
9715 the structure pointed to by the *pprivate pointer in the regexp.
9716 This is the core version of the extension overridable cloning hook.
9717 The regexp structure being duplicated will be copied by perl prior
9718 to this and will be provided as the regexp *r argument, however
9719 with the /old/ structures pprivate pointer value. Thus this routine
9720 may override any copying normally done by perl.
9722 It returns a pointer to the new regexp_internal structure.
9726 Perl_regdupe_internal(pTHX_ REGEXP * const rx, CLONE_PARAMS *param)
9729 struct regexp *const r = (struct regexp *)SvANY(rx);
9730 regexp_internal *reti;
9734 PERL_ARGS_ASSERT_REGDUPE_INTERNAL;
9736 npar = r->nparens+1;
9739 Newxc(reti, sizeof(regexp_internal) + len*sizeof(regnode), char, regexp_internal);
9740 Copy(ri->program, reti->program, len+1, regnode);
9743 reti->regstclass = NULL;
9747 const int count = ri->data->count;
9750 Newxc(d, sizeof(struct reg_data) + count*sizeof(void *),
9751 char, struct reg_data);
9752 Newx(d->what, count, U8);
9755 for (i = 0; i < count; i++) {
9756 d->what[i] = ri->data->what[i];
9757 switch (d->what[i]) {
9758 /* legal options are one of: sSfpontTu
9759 see also regcomp.h and pregfree() */
9762 case 'p': /* actually an AV, but the dup function is identical. */
9763 case 'u': /* actually an HV, but the dup function is identical. */
9764 d->data[i] = sv_dup_inc((const SV *)ri->data->data[i], param);
9767 /* This is cheating. */
9768 Newx(d->data[i], 1, struct regnode_charclass_class);
9769 StructCopy(ri->data->data[i], d->data[i],
9770 struct regnode_charclass_class);
9771 reti->regstclass = (regnode*)d->data[i];
9774 /* Compiled op trees are readonly and in shared memory,
9775 and can thus be shared without duplication. */
9777 d->data[i] = (void*)OpREFCNT_inc((OP*)ri->data->data[i]);
9781 /* Trie stclasses are readonly and can thus be shared
9782 * without duplication. We free the stclass in pregfree
9783 * when the corresponding reg_ac_data struct is freed.
9785 reti->regstclass= ri->regstclass;
9789 ((reg_trie_data*)ri->data->data[i])->refcount++;
9793 d->data[i] = ri->data->data[i];
9796 Perl_croak(aTHX_ "panic: re_dup unknown data code '%c'", ri->data->what[i]);
9805 reti->name_list_idx = ri->name_list_idx;
9807 #ifdef RE_TRACK_PATTERN_OFFSETS
9808 if (ri->u.offsets) {
9809 Newx(reti->u.offsets, 2*len+1, U32);
9810 Copy(ri->u.offsets, reti->u.offsets, 2*len+1, U32);
9813 SetProgLen(reti,len);
9819 #endif /* USE_ITHREADS */
9821 #ifndef PERL_IN_XSUB_RE
9824 - regnext - dig the "next" pointer out of a node
9827 Perl_regnext(pTHX_ register regnode *p)
9830 register I32 offset;
9835 offset = (reg_off_by_arg[OP(p)] ? ARG(p) : NEXT_OFF(p));
9844 S_re_croak2(pTHX_ const char* pat1,const char* pat2,...)
9847 STRLEN l1 = strlen(pat1);
9848 STRLEN l2 = strlen(pat2);
9851 const char *message;
9853 PERL_ARGS_ASSERT_RE_CROAK2;
9859 Copy(pat1, buf, l1 , char);
9860 Copy(pat2, buf + l1, l2 , char);
9861 buf[l1 + l2] = '\n';
9862 buf[l1 + l2 + 1] = '\0';
9864 /* ANSI variant takes additional second argument */
9865 va_start(args, pat2);
9869 msv = vmess(buf, &args);
9871 message = SvPV_const(msv,l1);
9874 Copy(message, buf, l1 , char);
9875 buf[l1-1] = '\0'; /* Overwrite \n */
9876 Perl_croak(aTHX_ "%s", buf);
9879 /* XXX Here's a total kludge. But we need to re-enter for swash routines. */
9881 #ifndef PERL_IN_XSUB_RE
9883 Perl_save_re_context(pTHX)
9887 struct re_save_state *state;
9889 SAVEVPTR(PL_curcop);
9890 SSGROW(SAVESTACK_ALLOC_FOR_RE_SAVE_STATE + 1);
9892 state = (struct re_save_state *)(PL_savestack + PL_savestack_ix);
9893 PL_savestack_ix += SAVESTACK_ALLOC_FOR_RE_SAVE_STATE;
9894 SSPUSHINT(SAVEt_RE_STATE);
9896 Copy(&PL_reg_state, state, 1, struct re_save_state);
9898 PL_reg_start_tmp = 0;
9899 PL_reg_start_tmpl = 0;
9900 PL_reg_oldsaved = NULL;
9901 PL_reg_oldsavedlen = 0;
9903 PL_reg_leftiter = 0;
9904 PL_reg_poscache = NULL;
9905 PL_reg_poscache_size = 0;
9906 #ifdef PERL_OLD_COPY_ON_WRITE
9910 /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
9912 const REGEXP * const rx = PM_GETRE(PL_curpm);
9915 for (i = 1; i <= RX_NPARENS(rx); i++) {
9916 char digits[TYPE_CHARS(long)];
9917 const STRLEN len = my_snprintf(digits, sizeof(digits), "%lu", (long)i);
9918 GV *const *const gvp
9919 = (GV**)hv_fetch(PL_defstash, digits, len, 0);
9922 GV * const gv = *gvp;
9923 if (SvTYPE(gv) == SVt_PVGV && GvSV(gv))
9933 clear_re(pTHX_ void *r)
9936 ReREFCNT_dec((REGEXP *)r);
9942 S_put_byte(pTHX_ SV *sv, int c)
9944 PERL_ARGS_ASSERT_PUT_BYTE;
9946 /* Our definition of isPRINT() ignores locales, so only bytes that are
9947 not part of UTF-8 are considered printable. I assume that the same
9948 holds for UTF-EBCDIC.
9949 Also, code point 255 is not printable in either (it's E0 in EBCDIC,
9950 which Wikipedia says:
9952 EO, or Eight Ones, is an 8-bit EBCDIC character code represented as all
9953 ones (binary 1111 1111, hexadecimal FF). It is similar, but not
9954 identical, to the ASCII delete (DEL) or rubout control character.
9955 ) So the old condition can be simplified to !isPRINT(c) */
9957 Perl_sv_catpvf(aTHX_ sv, "\\%o", c);
9959 const char string = c;
9960 if (c == '-' || c == ']' || c == '\\' || c == '^')
9961 sv_catpvs(sv, "\\");
9962 sv_catpvn(sv, &string, 1);
9967 #define CLEAR_OPTSTART \
9968 if (optstart) STMT_START { \
9969 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log, " (%"IVdf" nodes)\n", (IV)(node - optstart))); \
9973 #define DUMPUNTIL(b,e) CLEAR_OPTSTART; node=dumpuntil(r,start,(b),(e),last,sv,indent+1,depth+1);
9975 STATIC const regnode *
9976 S_dumpuntil(pTHX_ const regexp *r, const regnode *start, const regnode *node,
9977 const regnode *last, const regnode *plast,
9978 SV* sv, I32 indent, U32 depth)
9981 register U8 op = PSEUDO; /* Arbitrary non-END op. */
9982 register const regnode *next;
9983 const regnode *optstart= NULL;
9986 GET_RE_DEBUG_FLAGS_DECL;
9988 PERL_ARGS_ASSERT_DUMPUNTIL;
9990 #ifdef DEBUG_DUMPUNTIL
9991 PerlIO_printf(Perl_debug_log, "--- %d : %d - %d - %d\n",indent,node-start,
9992 last ? last-start : 0,plast ? plast-start : 0);
9995 if (plast && plast < last)
9998 while (PL_regkind[op] != END && (!last || node < last)) {
9999 /* While that wasn't END last time... */
10002 if (op == CLOSE || op == WHILEM)
10004 next = regnext((regnode *)node);
10007 if (OP(node) == OPTIMIZED) {
10008 if (!optstart && RE_DEBUG_FLAG(RE_DEBUG_COMPILE_OPTIMISE))
10015 regprop(r, sv, node);
10016 PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
10017 (int)(2*indent + 1), "", SvPVX_const(sv));
10019 if (OP(node) != OPTIMIZED) {
10020 if (next == NULL) /* Next ptr. */
10021 PerlIO_printf(Perl_debug_log, " (0)");
10022 else if (PL_regkind[(U8)op] == BRANCH && PL_regkind[OP(next)] != BRANCH )
10023 PerlIO_printf(Perl_debug_log, " (FAIL)");
10025 PerlIO_printf(Perl_debug_log, " (%"IVdf")", (IV)(next - start));
10026 (void)PerlIO_putc(Perl_debug_log, '\n');
10030 if (PL_regkind[(U8)op] == BRANCHJ) {
10033 register const regnode *nnode = (OP(next) == LONGJMP
10034 ? regnext((regnode *)next)
10036 if (last && nnode > last)
10038 DUMPUNTIL(NEXTOPER(NEXTOPER(node)), nnode);
10041 else if (PL_regkind[(U8)op] == BRANCH) {
10043 DUMPUNTIL(NEXTOPER(node), next);
10045 else if ( PL_regkind[(U8)op] == TRIE ) {
10046 const regnode *this_trie = node;
10047 const char op = OP(node);
10048 const U32 n = ARG(node);
10049 const reg_ac_data * const ac = op>=AHOCORASICK ?
10050 (reg_ac_data *)ri->data->data[n] :
10052 const reg_trie_data * const trie =
10053 (reg_trie_data*)ri->data->data[op<AHOCORASICK ? n : ac->trie];
10055 AV *const trie_words = MUTABLE_AV(ri->data->data[n + TRIE_WORDS_OFFSET]);
10057 const regnode *nextbranch= NULL;
10060 for (word_idx= 0; word_idx < (I32)trie->wordcount; word_idx++) {
10061 SV ** const elem_ptr = av_fetch(trie_words,word_idx,0);
10063 PerlIO_printf(Perl_debug_log, "%*s%s ",
10064 (int)(2*(indent+3)), "",
10065 elem_ptr ? pv_pretty(sv, SvPV_nolen_const(*elem_ptr), SvCUR(*elem_ptr), 60,
10066 PL_colors[0], PL_colors[1],
10067 (SvUTF8(*elem_ptr) ? PERL_PV_ESCAPE_UNI : 0) |
10068 PERL_PV_PRETTY_ELLIPSES |
10069 PERL_PV_PRETTY_LTGT
10074 U16 dist= trie->jump[word_idx+1];
10075 PerlIO_printf(Perl_debug_log, "(%"UVuf")\n",
10076 (UV)((dist ? this_trie + dist : next) - start));
10079 nextbranch= this_trie + trie->jump[0];
10080 DUMPUNTIL(this_trie + dist, nextbranch);
10082 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
10083 nextbranch= regnext((regnode *)nextbranch);
10085 PerlIO_printf(Perl_debug_log, "\n");
10088 if (last && next > last)
10093 else if ( op == CURLY ) { /* "next" might be very big: optimizer */
10094 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS,
10095 NEXTOPER(node) + EXTRA_STEP_2ARGS + 1);
10097 else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
10099 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS, next);
10101 else if ( op == PLUS || op == STAR) {
10102 DUMPUNTIL(NEXTOPER(node), NEXTOPER(node) + 1);
10104 else if (op == ANYOF) {
10105 /* arglen 1 + class block */
10106 node += 1 + ((ANYOF_FLAGS(node) & ANYOF_LARGE)
10107 ? ANYOF_CLASS_SKIP : ANYOF_SKIP);
10108 node = NEXTOPER(node);
10110 else if (PL_regkind[(U8)op] == EXACT) {
10111 /* Literal string, where present. */
10112 node += NODE_SZ_STR(node) - 1;
10113 node = NEXTOPER(node);
10116 node = NEXTOPER(node);
10117 node += regarglen[(U8)op];
10119 if (op == CURLYX || op == OPEN)
10123 #ifdef DEBUG_DUMPUNTIL
10124 PerlIO_printf(Perl_debug_log, "--- %d\n", (int)indent);
10129 #endif /* DEBUGGING */
10133 * c-indentation-style: bsd
10134 * c-basic-offset: 4
10135 * indent-tabs-mode: t
10138 * ex: set ts=8 sts=4 sw=4 noet:
projects / perl / modules / re-engine-Hooks.git / blob