5 * 'A fair jaw-cracker dwarf-language must be.' --Samwise Gamgee
7 * [p.285 of _The Lord of the Rings_, II/iii: "The Ring Goes South"]
10 /* This file contains functions for compiling a regular expression. See
11 * also regexec.c which funnily enough, contains functions for executing
12 * a regular expression.
14 * This file is also copied at build time to ext/re/re_comp.c, where
15 * it's built with -DPERL_EXT_RE_BUILD -DPERL_EXT_RE_DEBUG -DPERL_EXT.
16 * This causes the main functions to be compiled under new names and with
17 * debugging support added, which makes "use re 'debug'" work.
20 /* NOTE: this is derived from Henry Spencer's regexp code, and should not
21 * confused with the original package (see point 3 below). Thanks, Henry!
24 /* Additional note: this code is very heavily munged from Henry's version
25 * in places. In some spots I've traded clarity for efficiency, so don't
26 * blame Henry for some of the lack of readability.
29 /* The names of the functions have been changed from regcomp and
30 * regexec to pregcomp and pregexec in order to avoid conflicts
31 * with the POSIX routines of the same names.
34 #ifdef PERL_EXT_RE_BUILD
39 * pregcomp and pregexec -- regsub and regerror are not used in perl
41 * Copyright (c) 1986 by University of Toronto.
42 * Written by Henry Spencer. Not derived from licensed software.
44 * Permission is granted to anyone to use this software for any
45 * purpose on any computer system, and to redistribute it freely,
46 * subject to the following restrictions:
48 * 1. The author is not responsible for the consequences of use of
49 * this software, no matter how awful, even if they arise
52 * 2. The origin of this software must not be misrepresented, either
53 * by explicit claim or by omission.
55 * 3. Altered versions must be plainly marked as such, and must not
56 * be misrepresented as being the original software.
59 **** Alterations to Henry's code are...
61 **** Copyright (C) 1991, 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999,
62 **** 2000, 2001, 2002, 2003, 2004, 2005, 2006, 2007, 2008
63 **** by Larry Wall and others
65 **** You may distribute under the terms of either the GNU General Public
66 **** License or the Artistic License, as specified in the README file.
69 * Beware that some of this code is subtly aware of the way operator
70 * precedence is structured in regular expressions. Serious changes in
71 * regular-expression syntax might require a total rethink.
74 #define PERL_IN_REGCOMP_C
77 #ifndef PERL_IN_XSUB_RE
82 #ifdef PERL_IN_XSUB_RE
88 #include "dquote_static.c"
95 # if defined(BUGGY_MSC6)
96 /* MSC 6.00A breaks on op/regexp.t test 85 unless we turn this off */
97 # pragma optimize("a",off)
98 /* But MSC 6.00A is happy with 'w', for aliases only across function calls*/
99 # pragma optimize("w",on )
100 # endif /* BUGGY_MSC6 */
104 #define STATIC static
107 typedef struct RExC_state_t {
108 U32 flags; /* are we folding, multilining? */
109 char *precomp; /* uncompiled string. */
110 REGEXP *rx_sv; /* The SV that is the regexp. */
111 regexp *rx; /* perl core regexp structure */
112 regexp_internal *rxi; /* internal data for regexp object pprivate field */
113 char *start; /* Start of input for compile */
114 char *end; /* End of input for compile */
115 char *parse; /* Input-scan pointer. */
116 I32 whilem_seen; /* number of WHILEM in this expr */
117 regnode *emit_start; /* Start of emitted-code area */
118 regnode *emit_bound; /* First regnode outside of the allocated space */
119 regnode *emit; /* Code-emit pointer; ®dummy = don't = compiling */
120 I32 naughty; /* How bad is this pattern? */
121 I32 sawback; /* Did we see \1, ...? */
123 I32 size; /* Code size. */
124 I32 npar; /* Capture buffer count, (OPEN). */
125 I32 cpar; /* Capture buffer count, (CLOSE). */
126 I32 nestroot; /* root parens we are in - used by accept */
130 regnode **open_parens; /* pointers to open parens */
131 regnode **close_parens; /* pointers to close parens */
132 regnode *opend; /* END node in program */
133 I32 utf8; /* whether the pattern is utf8 or not */
134 I32 orig_utf8; /* whether the pattern was originally in utf8 */
135 /* XXX use this for future optimisation of case
136 * where pattern must be upgraded to utf8. */
137 I32 uni_semantics; /* If a d charset modifier should use unicode
138 rules, even if the pattern is not in
140 HV *paren_names; /* Paren names */
142 regnode **recurse; /* Recurse regops */
143 I32 recurse_count; /* Number of recurse regops */
146 I32 override_recoding;
148 char *starttry; /* -Dr: where regtry was called. */
149 #define RExC_starttry (pRExC_state->starttry)
152 const char *lastparse;
154 AV *paren_name_list; /* idx -> name */
155 #define RExC_lastparse (pRExC_state->lastparse)
156 #define RExC_lastnum (pRExC_state->lastnum)
157 #define RExC_paren_name_list (pRExC_state->paren_name_list)
161 #define RExC_flags (pRExC_state->flags)
162 #define RExC_precomp (pRExC_state->precomp)
163 #define RExC_rx_sv (pRExC_state->rx_sv)
164 #define RExC_rx (pRExC_state->rx)
165 #define RExC_rxi (pRExC_state->rxi)
166 #define RExC_start (pRExC_state->start)
167 #define RExC_end (pRExC_state->end)
168 #define RExC_parse (pRExC_state->parse)
169 #define RExC_whilem_seen (pRExC_state->whilem_seen)
170 #ifdef RE_TRACK_PATTERN_OFFSETS
171 #define RExC_offsets (pRExC_state->rxi->u.offsets) /* I am not like the others */
173 #define RExC_emit (pRExC_state->emit)
174 #define RExC_emit_start (pRExC_state->emit_start)
175 #define RExC_emit_bound (pRExC_state->emit_bound)
176 #define RExC_naughty (pRExC_state->naughty)
177 #define RExC_sawback (pRExC_state->sawback)
178 #define RExC_seen (pRExC_state->seen)
179 #define RExC_size (pRExC_state->size)
180 #define RExC_npar (pRExC_state->npar)
181 #define RExC_nestroot (pRExC_state->nestroot)
182 #define RExC_extralen (pRExC_state->extralen)
183 #define RExC_seen_zerolen (pRExC_state->seen_zerolen)
184 #define RExC_seen_evals (pRExC_state->seen_evals)
185 #define RExC_utf8 (pRExC_state->utf8)
186 #define RExC_uni_semantics (pRExC_state->uni_semantics)
187 #define RExC_orig_utf8 (pRExC_state->orig_utf8)
188 #define RExC_open_parens (pRExC_state->open_parens)
189 #define RExC_close_parens (pRExC_state->close_parens)
190 #define RExC_opend (pRExC_state->opend)
191 #define RExC_paren_names (pRExC_state->paren_names)
192 #define RExC_recurse (pRExC_state->recurse)
193 #define RExC_recurse_count (pRExC_state->recurse_count)
194 #define RExC_in_lookbehind (pRExC_state->in_lookbehind)
195 #define RExC_contains_locale (pRExC_state->contains_locale)
196 #define RExC_override_recoding (pRExC_state->override_recoding)
199 #define ISMULT1(c) ((c) == '*' || (c) == '+' || (c) == '?')
200 #define ISMULT2(s) ((*s) == '*' || (*s) == '+' || (*s) == '?' || \
201 ((*s) == '{' && regcurly(s)))
204 #undef SPSTART /* dratted cpp namespace... */
207 * Flags to be passed up and down.
209 #define WORST 0 /* Worst case. */
210 #define HASWIDTH 0x01 /* Known to match non-null strings. */
212 /* Simple enough to be STAR/PLUS operand, in an EXACT node must be a single
213 * character, and if utf8, must be invariant. Note that this is not the same thing as REGNODE_SIMPLE */
215 #define SPSTART 0x04 /* Starts with * or +. */
216 #define TRYAGAIN 0x08 /* Weeded out a declaration. */
217 #define POSTPONED 0x10 /* (?1),(?&name), (??{...}) or similar */
219 #define REG_NODE_NUM(x) ((x) ? (int)((x)-RExC_emit_start) : -1)
221 /* whether trie related optimizations are enabled */
222 #if PERL_ENABLE_EXTENDED_TRIE_OPTIMISATION
223 #define TRIE_STUDY_OPT
224 #define FULL_TRIE_STUDY
230 #define PBYTE(u8str,paren) ((U8*)(u8str))[(paren) >> 3]
231 #define PBITVAL(paren) (1 << ((paren) & 7))
232 #define PAREN_TEST(u8str,paren) ( PBYTE(u8str,paren) & PBITVAL(paren))
233 #define PAREN_SET(u8str,paren) PBYTE(u8str,paren) |= PBITVAL(paren)
234 #define PAREN_UNSET(u8str,paren) PBYTE(u8str,paren) &= (~PBITVAL(paren))
236 /* If not already in utf8, do a longjmp back to the beginning */
237 #define UTF8_LONGJMP 42 /* Choose a value not likely to ever conflict */
238 #define REQUIRE_UTF8 STMT_START { \
239 if (! UTF) JMPENV_JUMP(UTF8_LONGJMP); \
242 /* About scan_data_t.
244 During optimisation we recurse through the regexp program performing
245 various inplace (keyhole style) optimisations. In addition study_chunk
246 and scan_commit populate this data structure with information about
247 what strings MUST appear in the pattern. We look for the longest
248 string that must appear at a fixed location, and we look for the
249 longest string that may appear at a floating location. So for instance
254 Both 'FOO' and 'A' are fixed strings. Both 'B' and 'BAR' are floating
255 strings (because they follow a .* construct). study_chunk will identify
256 both FOO and BAR as being the longest fixed and floating strings respectively.
258 The strings can be composites, for instance
262 will result in a composite fixed substring 'foo'.
264 For each string some basic information is maintained:
266 - offset or min_offset
267 This is the position the string must appear at, or not before.
268 It also implicitly (when combined with minlenp) tells us how many
269 characters must match before the string we are searching for.
270 Likewise when combined with minlenp and the length of the string it
271 tells us how many characters must appear after the string we have
275 Only used for floating strings. This is the rightmost point that
276 the string can appear at. If set to I32 max it indicates that the
277 string can occur infinitely far to the right.
280 A pointer to the minimum length of the pattern that the string
281 was found inside. This is important as in the case of positive
282 lookahead or positive lookbehind we can have multiple patterns
287 The minimum length of the pattern overall is 3, the minimum length
288 of the lookahead part is 3, but the minimum length of the part that
289 will actually match is 1. So 'FOO's minimum length is 3, but the
290 minimum length for the F is 1. This is important as the minimum length
291 is used to determine offsets in front of and behind the string being
292 looked for. Since strings can be composites this is the length of the
293 pattern at the time it was committed with a scan_commit. Note that
294 the length is calculated by study_chunk, so that the minimum lengths
295 are not known until the full pattern has been compiled, thus the
296 pointer to the value.
300 In the case of lookbehind the string being searched for can be
301 offset past the start point of the final matching string.
302 If this value was just blithely removed from the min_offset it would
303 invalidate some of the calculations for how many chars must match
304 before or after (as they are derived from min_offset and minlen and
305 the length of the string being searched for).
306 When the final pattern is compiled and the data is moved from the
307 scan_data_t structure into the regexp structure the information
308 about lookbehind is factored in, with the information that would
309 have been lost precalculated in the end_shift field for the
312 The fields pos_min and pos_delta are used to store the minimum offset
313 and the delta to the maximum offset at the current point in the pattern.
317 typedef struct scan_data_t {
318 /*I32 len_min; unused */
319 /*I32 len_delta; unused */
323 I32 last_end; /* min value, <0 unless valid. */
326 SV **longest; /* Either &l_fixed, or &l_float. */
327 SV *longest_fixed; /* longest fixed string found in pattern */
328 I32 offset_fixed; /* offset where it starts */
329 I32 *minlen_fixed; /* pointer to the minlen relevant to the string */
330 I32 lookbehind_fixed; /* is the position of the string modfied by LB */
331 SV *longest_float; /* longest floating string found in pattern */
332 I32 offset_float_min; /* earliest point in string it can appear */
333 I32 offset_float_max; /* latest point in string it can appear */
334 I32 *minlen_float; /* pointer to the minlen relevant to the string */
335 I32 lookbehind_float; /* is the position of the string modified by LB */
339 struct regnode_charclass_class *start_class;
343 * Forward declarations for pregcomp()'s friends.
346 static const scan_data_t zero_scan_data =
347 { 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ,0};
349 #define SF_BEFORE_EOL (SF_BEFORE_SEOL|SF_BEFORE_MEOL)
350 #define SF_BEFORE_SEOL 0x0001
351 #define SF_BEFORE_MEOL 0x0002
352 #define SF_FIX_BEFORE_EOL (SF_FIX_BEFORE_SEOL|SF_FIX_BEFORE_MEOL)
353 #define SF_FL_BEFORE_EOL (SF_FL_BEFORE_SEOL|SF_FL_BEFORE_MEOL)
356 # define SF_FIX_SHIFT_EOL (0+2)
357 # define SF_FL_SHIFT_EOL (0+4)
359 # define SF_FIX_SHIFT_EOL (+2)
360 # define SF_FL_SHIFT_EOL (+4)
363 #define SF_FIX_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FIX_SHIFT_EOL)
364 #define SF_FIX_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FIX_SHIFT_EOL)
366 #define SF_FL_BEFORE_SEOL (SF_BEFORE_SEOL << SF_FL_SHIFT_EOL)
367 #define SF_FL_BEFORE_MEOL (SF_BEFORE_MEOL << SF_FL_SHIFT_EOL) /* 0x20 */
368 #define SF_IS_INF 0x0040
369 #define SF_HAS_PAR 0x0080
370 #define SF_IN_PAR 0x0100
371 #define SF_HAS_EVAL 0x0200
372 #define SCF_DO_SUBSTR 0x0400
373 #define SCF_DO_STCLASS_AND 0x0800
374 #define SCF_DO_STCLASS_OR 0x1000
375 #define SCF_DO_STCLASS (SCF_DO_STCLASS_AND|SCF_DO_STCLASS_OR)
376 #define SCF_WHILEM_VISITED_POS 0x2000
378 #define SCF_TRIE_RESTUDY 0x4000 /* Do restudy? */
379 #define SCF_SEEN_ACCEPT 0x8000
381 #define UTF cBOOL(RExC_utf8)
382 #define LOC (get_regex_charset(RExC_flags) == REGEX_LOCALE_CHARSET)
383 #define UNI_SEMANTICS (get_regex_charset(RExC_flags) == REGEX_UNICODE_CHARSET)
384 #define DEPENDS_SEMANTICS (get_regex_charset(RExC_flags) == REGEX_DEPENDS_CHARSET)
385 #define AT_LEAST_UNI_SEMANTICS (get_regex_charset(RExC_flags) >= REGEX_UNICODE_CHARSET)
386 #define ASCII_RESTRICTED (get_regex_charset(RExC_flags) == REGEX_ASCII_RESTRICTED_CHARSET)
387 #define MORE_ASCII_RESTRICTED (get_regex_charset(RExC_flags) == REGEX_ASCII_MORE_RESTRICTED_CHARSET)
388 #define AT_LEAST_ASCII_RESTRICTED (get_regex_charset(RExC_flags) >= REGEX_ASCII_RESTRICTED_CHARSET)
390 #define FOLD cBOOL(RExC_flags & RXf_PMf_FOLD)
392 #define OOB_UNICODE 12345678
393 #define OOB_NAMEDCLASS -1
395 #define CHR_SVLEN(sv) (UTF ? sv_len_utf8(sv) : SvCUR(sv))
396 #define CHR_DIST(a,b) (UTF ? utf8_distance(a,b) : a - b)
399 /* length of regex to show in messages that don't mark a position within */
400 #define RegexLengthToShowInErrorMessages 127
403 * If MARKER[12] are adjusted, be sure to adjust the constants at the top
404 * of t/op/regmesg.t, the tests in t/op/re_tests, and those in
405 * op/pragma/warn/regcomp.
407 #define MARKER1 "<-- HERE" /* marker as it appears in the description */
408 #define MARKER2 " <-- HERE " /* marker as it appears within the regex */
410 #define REPORT_LOCATION " in regex; marked by " MARKER1 " in m/%.*s" MARKER2 "%s/"
413 * Calls SAVEDESTRUCTOR_X if needed, then calls Perl_croak with the given
414 * arg. Show regex, up to a maximum length. If it's too long, chop and add
417 #define _FAIL(code) STMT_START { \
418 const char *ellipses = ""; \
419 IV len = RExC_end - RExC_precomp; \
422 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
423 if (len > RegexLengthToShowInErrorMessages) { \
424 /* chop 10 shorter than the max, to ensure meaning of "..." */ \
425 len = RegexLengthToShowInErrorMessages - 10; \
431 #define FAIL(msg) _FAIL( \
432 Perl_croak(aTHX_ "%s in regex m/%.*s%s/", \
433 msg, (int)len, RExC_precomp, ellipses))
435 #define FAIL2(msg,arg) _FAIL( \
436 Perl_croak(aTHX_ msg " in regex m/%.*s%s/", \
437 arg, (int)len, RExC_precomp, ellipses))
440 * Simple_vFAIL -- like FAIL, but marks the current location in the scan
442 #define Simple_vFAIL(m) STMT_START { \
443 const IV offset = RExC_parse - RExC_precomp; \
444 Perl_croak(aTHX_ "%s" REPORT_LOCATION, \
445 m, (int)offset, RExC_precomp, RExC_precomp + offset); \
449 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL()
451 #define vFAIL(m) STMT_START { \
453 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
458 * Like Simple_vFAIL(), but accepts two arguments.
460 #define Simple_vFAIL2(m,a1) STMT_START { \
461 const IV offset = RExC_parse - RExC_precomp; \
462 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, \
463 (int)offset, RExC_precomp, RExC_precomp + offset); \
467 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL2().
469 #define vFAIL2(m,a1) STMT_START { \
471 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
472 Simple_vFAIL2(m, a1); \
477 * Like Simple_vFAIL(), but accepts three arguments.
479 #define Simple_vFAIL3(m, a1, a2) STMT_START { \
480 const IV offset = RExC_parse - RExC_precomp; \
481 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, \
482 (int)offset, RExC_precomp, RExC_precomp + offset); \
486 * Calls SAVEDESTRUCTOR_X if needed, then Simple_vFAIL3().
488 #define vFAIL3(m,a1,a2) STMT_START { \
490 SAVEDESTRUCTOR_X(clear_re,(void*)RExC_rx_sv); \
491 Simple_vFAIL3(m, a1, a2); \
495 * Like Simple_vFAIL(), but accepts four arguments.
497 #define Simple_vFAIL4(m, a1, a2, a3) STMT_START { \
498 const IV offset = RExC_parse - RExC_precomp; \
499 S_re_croak2(aTHX_ m, REPORT_LOCATION, a1, a2, a3, \
500 (int)offset, RExC_precomp, RExC_precomp + offset); \
503 #define ckWARNreg(loc,m) STMT_START { \
504 const IV offset = loc - RExC_precomp; \
505 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
506 (int)offset, RExC_precomp, RExC_precomp + offset); \
509 #define ckWARNregdep(loc,m) STMT_START { \
510 const IV offset = loc - RExC_precomp; \
511 Perl_ck_warner_d(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
513 (int)offset, RExC_precomp, RExC_precomp + offset); \
516 #define ckWARN2regdep(loc,m, a1) STMT_START { \
517 const IV offset = loc - RExC_precomp; \
518 Perl_ck_warner_d(aTHX_ packWARN2(WARN_DEPRECATED, WARN_REGEXP), \
520 a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
523 #define ckWARN2reg(loc, m, a1) STMT_START { \
524 const IV offset = loc - RExC_precomp; \
525 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
526 a1, (int)offset, RExC_precomp, RExC_precomp + offset); \
529 #define vWARN3(loc, m, a1, a2) STMT_START { \
530 const IV offset = loc - RExC_precomp; \
531 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
532 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
535 #define ckWARN3reg(loc, m, a1, a2) STMT_START { \
536 const IV offset = loc - RExC_precomp; \
537 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
538 a1, a2, (int)offset, RExC_precomp, RExC_precomp + offset); \
541 #define vWARN4(loc, m, a1, a2, a3) STMT_START { \
542 const IV offset = loc - RExC_precomp; \
543 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
544 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
547 #define ckWARN4reg(loc, m, a1, a2, a3) STMT_START { \
548 const IV offset = loc - RExC_precomp; \
549 Perl_ck_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
550 a1, a2, a3, (int)offset, RExC_precomp, RExC_precomp + offset); \
553 #define vWARN5(loc, m, a1, a2, a3, a4) STMT_START { \
554 const IV offset = loc - RExC_precomp; \
555 Perl_warner(aTHX_ packWARN(WARN_REGEXP), m REPORT_LOCATION, \
556 a1, a2, a3, a4, (int)offset, RExC_precomp, RExC_precomp + offset); \
560 /* Allow for side effects in s */
561 #define REGC(c,s) STMT_START { \
562 if (!SIZE_ONLY) *(s) = (c); else (void)(s); \
565 /* Macros for recording node offsets. 20001227 mjd@plover.com
566 * Nodes are numbered 1, 2, 3, 4. Node #n's position is recorded in
567 * element 2*n-1 of the array. Element #2n holds the byte length node #n.
568 * Element 0 holds the number n.
569 * Position is 1 indexed.
571 #ifndef RE_TRACK_PATTERN_OFFSETS
572 #define Set_Node_Offset_To_R(node,byte)
573 #define Set_Node_Offset(node,byte)
574 #define Set_Cur_Node_Offset
575 #define Set_Node_Length_To_R(node,len)
576 #define Set_Node_Length(node,len)
577 #define Set_Node_Cur_Length(node)
578 #define Node_Offset(n)
579 #define Node_Length(n)
580 #define Set_Node_Offset_Length(node,offset,len)
581 #define ProgLen(ri) ri->u.proglen
582 #define SetProgLen(ri,x) ri->u.proglen = x
584 #define ProgLen(ri) ri->u.offsets[0]
585 #define SetProgLen(ri,x) ri->u.offsets[0] = x
586 #define Set_Node_Offset_To_R(node,byte) STMT_START { \
588 MJD_OFFSET_DEBUG(("** (%d) offset of node %d is %d.\n", \
589 __LINE__, (int)(node), (int)(byte))); \
591 Perl_croak(aTHX_ "value of node is %d in Offset macro", (int)(node)); \
593 RExC_offsets[2*(node)-1] = (byte); \
598 #define Set_Node_Offset(node,byte) \
599 Set_Node_Offset_To_R((node)-RExC_emit_start, (byte)-RExC_start)
600 #define Set_Cur_Node_Offset Set_Node_Offset(RExC_emit, RExC_parse)
602 #define Set_Node_Length_To_R(node,len) STMT_START { \
604 MJD_OFFSET_DEBUG(("** (%d) size of node %d is %d.\n", \
605 __LINE__, (int)(node), (int)(len))); \
607 Perl_croak(aTHX_ "value of node is %d in Length macro", (int)(node)); \
609 RExC_offsets[2*(node)] = (len); \
614 #define Set_Node_Length(node,len) \
615 Set_Node_Length_To_R((node)-RExC_emit_start, len)
616 #define Set_Cur_Node_Length(len) Set_Node_Length(RExC_emit, len)
617 #define Set_Node_Cur_Length(node) \
618 Set_Node_Length(node, RExC_parse - parse_start)
620 /* Get offsets and lengths */
621 #define Node_Offset(n) (RExC_offsets[2*((n)-RExC_emit_start)-1])
622 #define Node_Length(n) (RExC_offsets[2*((n)-RExC_emit_start)])
624 #define Set_Node_Offset_Length(node,offset,len) STMT_START { \
625 Set_Node_Offset_To_R((node)-RExC_emit_start, (offset)); \
626 Set_Node_Length_To_R((node)-RExC_emit_start, (len)); \
630 #if PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS
631 #define EXPERIMENTAL_INPLACESCAN
632 #endif /*PERL_ENABLE_EXPERIMENTAL_REGEX_OPTIMISATIONS*/
634 #define DEBUG_STUDYDATA(str,data,depth) \
635 DEBUG_OPTIMISE_MORE_r(if(data){ \
636 PerlIO_printf(Perl_debug_log, \
637 "%*s" str "Pos:%"IVdf"/%"IVdf \
638 " Flags: 0x%"UVXf" Whilem_c: %"IVdf" Lcp: %"IVdf" %s", \
639 (int)(depth)*2, "", \
640 (IV)((data)->pos_min), \
641 (IV)((data)->pos_delta), \
642 (UV)((data)->flags), \
643 (IV)((data)->whilem_c), \
644 (IV)((data)->last_closep ? *((data)->last_closep) : -1), \
645 is_inf ? "INF " : "" \
647 if ((data)->last_found) \
648 PerlIO_printf(Perl_debug_log, \
649 "Last:'%s' %"IVdf":%"IVdf"/%"IVdf" %sFixed:'%s' @ %"IVdf \
650 " %sFloat: '%s' @ %"IVdf"/%"IVdf"", \
651 SvPVX_const((data)->last_found), \
652 (IV)((data)->last_end), \
653 (IV)((data)->last_start_min), \
654 (IV)((data)->last_start_max), \
655 ((data)->longest && \
656 (data)->longest==&((data)->longest_fixed)) ? "*" : "", \
657 SvPVX_const((data)->longest_fixed), \
658 (IV)((data)->offset_fixed), \
659 ((data)->longest && \
660 (data)->longest==&((data)->longest_float)) ? "*" : "", \
661 SvPVX_const((data)->longest_float), \
662 (IV)((data)->offset_float_min), \
663 (IV)((data)->offset_float_max) \
665 PerlIO_printf(Perl_debug_log,"\n"); \
668 static void clear_re(pTHX_ void *r);
670 /* Mark that we cannot extend a found fixed substring at this point.
671 Update the longest found anchored substring and the longest found
672 floating substrings if needed. */
675 S_scan_commit(pTHX_ const RExC_state_t *pRExC_state, scan_data_t *data, I32 *minlenp, int is_inf)
677 const STRLEN l = CHR_SVLEN(data->last_found);
678 const STRLEN old_l = CHR_SVLEN(*data->longest);
679 GET_RE_DEBUG_FLAGS_DECL;
681 PERL_ARGS_ASSERT_SCAN_COMMIT;
683 if ((l >= old_l) && ((l > old_l) || (data->flags & SF_BEFORE_EOL))) {
684 SvSetMagicSV(*data->longest, data->last_found);
685 if (*data->longest == data->longest_fixed) {
686 data->offset_fixed = l ? data->last_start_min : data->pos_min;
687 if (data->flags & SF_BEFORE_EOL)
689 |= ((data->flags & SF_BEFORE_EOL) << SF_FIX_SHIFT_EOL);
691 data->flags &= ~SF_FIX_BEFORE_EOL;
692 data->minlen_fixed=minlenp;
693 data->lookbehind_fixed=0;
695 else { /* *data->longest == data->longest_float */
696 data->offset_float_min = l ? data->last_start_min : data->pos_min;
697 data->offset_float_max = (l
698 ? data->last_start_max
699 : data->pos_min + data->pos_delta);
700 if (is_inf || (U32)data->offset_float_max > (U32)I32_MAX)
701 data->offset_float_max = I32_MAX;
702 if (data->flags & SF_BEFORE_EOL)
704 |= ((data->flags & SF_BEFORE_EOL) << SF_FL_SHIFT_EOL);
706 data->flags &= ~SF_FL_BEFORE_EOL;
707 data->minlen_float=minlenp;
708 data->lookbehind_float=0;
711 SvCUR_set(data->last_found, 0);
713 SV * const sv = data->last_found;
714 if (SvUTF8(sv) && SvMAGICAL(sv)) {
715 MAGIC * const mg = mg_find(sv, PERL_MAGIC_utf8);
721 data->flags &= ~SF_BEFORE_EOL;
722 DEBUG_STUDYDATA("commit: ",data,0);
725 /* Can match anything (initialization) */
727 S_cl_anything(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
729 PERL_ARGS_ASSERT_CL_ANYTHING;
731 ANYOF_BITMAP_SETALL(cl);
732 cl->flags = ANYOF_CLASS|ANYOF_EOS|ANYOF_UNICODE_ALL
733 |ANYOF_LOC_NONBITMAP_FOLD|ANYOF_NON_UTF8_LATIN1_ALL;
735 /* If any portion of the regex is to operate under locale rules,
736 * initialization includes it. The reason this isn't done for all regexes
737 * is that the optimizer was written under the assumption that locale was
738 * all-or-nothing. Given the complexity and lack of documentation in the
739 * optimizer, and that there are inadequate test cases for locale, so many
740 * parts of it may not work properly, it is safest to avoid locale unless
742 if (RExC_contains_locale) {
743 ANYOF_CLASS_SETALL(cl); /* /l uses class */
744 cl->flags |= ANYOF_LOCALE;
747 ANYOF_CLASS_ZERO(cl); /* Only /l uses class now */
751 /* Can match anything (initialization) */
753 S_cl_is_anything(const struct regnode_charclass_class *cl)
757 PERL_ARGS_ASSERT_CL_IS_ANYTHING;
759 for (value = 0; value <= ANYOF_MAX; value += 2)
760 if (ANYOF_CLASS_TEST(cl, value) && ANYOF_CLASS_TEST(cl, value + 1))
762 if (!(cl->flags & ANYOF_UNICODE_ALL))
764 if (!ANYOF_BITMAP_TESTALLSET((const void*)cl))
769 /* Can match anything (initialization) */
771 S_cl_init(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl)
773 PERL_ARGS_ASSERT_CL_INIT;
775 Zero(cl, 1, struct regnode_charclass_class);
777 cl_anything(pRExC_state, cl);
778 ARG_SET(cl, ANYOF_NONBITMAP_EMPTY);
781 /* These two functions currently do the exact same thing */
782 #define cl_init_zero S_cl_init
784 /* 'AND' a given class with another one. Can create false positives. 'cl'
785 * should not be inverted. 'and_with->flags & ANYOF_CLASS' should be 0 if
786 * 'and_with' is a regnode_charclass instead of a regnode_charclass_class. */
788 S_cl_and(struct regnode_charclass_class *cl,
789 const struct regnode_charclass_class *and_with)
791 PERL_ARGS_ASSERT_CL_AND;
793 assert(and_with->type == ANYOF);
795 /* I (khw) am not sure all these restrictions are necessary XXX */
796 if (!(ANYOF_CLASS_TEST_ANY_SET(and_with))
797 && !(ANYOF_CLASS_TEST_ANY_SET(cl))
798 && (and_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
799 && !(and_with->flags & ANYOF_LOC_NONBITMAP_FOLD)
800 && !(cl->flags & ANYOF_LOC_NONBITMAP_FOLD)) {
803 if (and_with->flags & ANYOF_INVERT)
804 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
805 cl->bitmap[i] &= ~and_with->bitmap[i];
807 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
808 cl->bitmap[i] &= and_with->bitmap[i];
809 } /* XXXX: logic is complicated otherwise, leave it along for a moment. */
811 if (and_with->flags & ANYOF_INVERT) {
813 /* Here, the and'ed node is inverted. Get the AND of the flags that
814 * aren't affected by the inversion. Those that are affected are
815 * handled individually below */
816 U8 affected_flags = cl->flags & ~INVERSION_UNAFFECTED_FLAGS;
817 cl->flags &= (and_with->flags & INVERSION_UNAFFECTED_FLAGS);
818 cl->flags |= affected_flags;
820 /* We currently don't know how to deal with things that aren't in the
821 * bitmap, but we know that the intersection is no greater than what
822 * is already in cl, so let there be false positives that get sorted
823 * out after the synthetic start class succeeds, and the node is
824 * matched for real. */
826 /* The inversion of these two flags indicate that the resulting
827 * intersection doesn't have them */
828 if (and_with->flags & ANYOF_UNICODE_ALL) {
829 cl->flags &= ~ANYOF_UNICODE_ALL;
831 if (and_with->flags & ANYOF_NON_UTF8_LATIN1_ALL) {
832 cl->flags &= ~ANYOF_NON_UTF8_LATIN1_ALL;
835 else { /* and'd node is not inverted */
836 U8 outside_bitmap_but_not_utf8; /* Temp variable */
838 if (! ANYOF_NONBITMAP(and_with)) {
840 /* Here 'and_with' doesn't match anything outside the bitmap
841 * (except possibly ANYOF_UNICODE_ALL), which means the
842 * intersection can't either, except for ANYOF_UNICODE_ALL, in
843 * which case we don't know what the intersection is, but it's no
844 * greater than what cl already has, so can just leave it alone,
845 * with possible false positives */
846 if (! (and_with->flags & ANYOF_UNICODE_ALL)) {
847 ARG_SET(cl, ANYOF_NONBITMAP_EMPTY);
848 cl->flags &= ~ANYOF_NONBITMAP_NON_UTF8;
851 else if (! ANYOF_NONBITMAP(cl)) {
853 /* Here, 'and_with' does match something outside the bitmap, and cl
854 * doesn't have a list of things to match outside the bitmap. If
855 * cl can match all code points above 255, the intersection will
856 * be those above-255 code points that 'and_with' matches. If cl
857 * can't match all Unicode code points, it means that it can't
858 * match anything outside the bitmap (since the 'if' that got us
859 * into this block tested for that), so we leave the bitmap empty.
861 if (cl->flags & ANYOF_UNICODE_ALL) {
862 ARG_SET(cl, ARG(and_with));
864 /* and_with's ARG may match things that don't require UTF8.
865 * And now cl's will too, in spite of this being an 'and'. See
866 * the comments below about the kludge */
867 cl->flags |= and_with->flags & ANYOF_NONBITMAP_NON_UTF8;
871 /* Here, both 'and_with' and cl match something outside the
872 * bitmap. Currently we do not do the intersection, so just match
873 * whatever cl had at the beginning. */
877 /* Take the intersection of the two sets of flags. However, the
878 * ANYOF_NONBITMAP_NON_UTF8 flag is treated as an 'or'. This is a
879 * kludge around the fact that this flag is not treated like the others
880 * which are initialized in cl_anything(). The way the optimizer works
881 * is that the synthetic start class (SSC) is initialized to match
882 * anything, and then the first time a real node is encountered, its
883 * values are AND'd with the SSC's with the result being the values of
884 * the real node. However, there are paths through the optimizer where
885 * the AND never gets called, so those initialized bits are set
886 * inappropriately, which is not usually a big deal, as they just cause
887 * false positives in the SSC, which will just mean a probably
888 * imperceptible slow down in execution. However this bit has a
889 * higher false positive consequence in that it can cause utf8.pm,
890 * utf8_heavy.pl ... to be loaded when not necessary, which is a much
891 * bigger slowdown and also causes significant extra memory to be used.
892 * In order to prevent this, the code now takes a different tack. The
893 * bit isn't set unless some part of the regular expression needs it,
894 * but once set it won't get cleared. This means that these extra
895 * modules won't get loaded unless there was some path through the
896 * pattern that would have required them anyway, and so any false
897 * positives that occur by not ANDing them out when they could be
898 * aren't as severe as they would be if we treated this bit like all
900 outside_bitmap_but_not_utf8 = (cl->flags | and_with->flags)
901 & ANYOF_NONBITMAP_NON_UTF8;
902 cl->flags &= and_with->flags;
903 cl->flags |= outside_bitmap_but_not_utf8;
907 /* 'OR' a given class with another one. Can create false positives. 'cl'
908 * should not be inverted. 'or_with->flags & ANYOF_CLASS' should be 0 if
909 * 'or_with' is a regnode_charclass instead of a regnode_charclass_class. */
911 S_cl_or(const RExC_state_t *pRExC_state, struct regnode_charclass_class *cl, const struct regnode_charclass_class *or_with)
913 PERL_ARGS_ASSERT_CL_OR;
915 if (or_with->flags & ANYOF_INVERT) {
917 /* Here, the or'd node is to be inverted. This means we take the
918 * complement of everything not in the bitmap, but currently we don't
919 * know what that is, so give up and match anything */
920 if (ANYOF_NONBITMAP(or_with)) {
921 cl_anything(pRExC_state, cl);
924 * (B1 | CL1) | (!B2 & !CL2) = (B1 | !B2 & !CL2) | (CL1 | (!B2 & !CL2))
925 * <= (B1 | !B2) | (CL1 | !CL2)
926 * which is wasteful if CL2 is small, but we ignore CL2:
927 * (B1 | CL1) | (!B2 & !CL2) <= (B1 | CL1) | !B2 = (B1 | !B2) | CL1
928 * XXXX Can we handle case-fold? Unclear:
929 * (OK1(i) | OK1(i')) | !(OK1(i) | OK1(i')) =
930 * (OK1(i) | OK1(i')) | (!OK1(i) & !OK1(i'))
932 else if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
933 && !(or_with->flags & ANYOF_LOC_NONBITMAP_FOLD)
934 && !(cl->flags & ANYOF_LOC_NONBITMAP_FOLD) ) {
937 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
938 cl->bitmap[i] |= ~or_with->bitmap[i];
939 } /* XXXX: logic is complicated otherwise */
941 cl_anything(pRExC_state, cl);
944 /* And, we can just take the union of the flags that aren't affected
945 * by the inversion */
946 cl->flags |= or_with->flags & INVERSION_UNAFFECTED_FLAGS;
948 /* For the remaining flags:
949 ANYOF_UNICODE_ALL and inverted means to not match anything above
950 255, which means that the union with cl should just be
951 what cl has in it, so can ignore this flag
952 ANYOF_NON_UTF8_LATIN1_ALL and inverted means if not utf8 and ord
953 is 127-255 to match them, but then invert that, so the
954 union with cl should just be what cl has in it, so can
957 } else { /* 'or_with' is not inverted */
958 /* (B1 | CL1) | (B2 | CL2) = (B1 | B2) | (CL1 | CL2)) */
959 if ( (or_with->flags & ANYOF_LOCALE) == (cl->flags & ANYOF_LOCALE)
960 && (!(or_with->flags & ANYOF_LOC_NONBITMAP_FOLD)
961 || (cl->flags & ANYOF_LOC_NONBITMAP_FOLD)) ) {
964 /* OR char bitmap and class bitmap separately */
965 for (i = 0; i < ANYOF_BITMAP_SIZE; i++)
966 cl->bitmap[i] |= or_with->bitmap[i];
967 if (ANYOF_CLASS_TEST_ANY_SET(or_with)) {
968 for (i = 0; i < ANYOF_CLASSBITMAP_SIZE; i++)
969 cl->classflags[i] |= or_with->classflags[i];
970 cl->flags |= ANYOF_CLASS;
973 else { /* XXXX: logic is complicated, leave it along for a moment. */
974 cl_anything(pRExC_state, cl);
977 if (ANYOF_NONBITMAP(or_with)) {
979 /* Use the added node's outside-the-bit-map match if there isn't a
980 * conflict. If there is a conflict (both nodes match something
981 * outside the bitmap, but what they match outside is not the same
982 * pointer, and hence not easily compared until XXX we extend
983 * inversion lists this far), give up and allow the start class to
984 * match everything outside the bitmap. If that stuff is all above
985 * 255, can just set UNICODE_ALL, otherwise caould be anything. */
986 if (! ANYOF_NONBITMAP(cl)) {
987 ARG_SET(cl, ARG(or_with));
989 else if (ARG(cl) != ARG(or_with)) {
991 if ((or_with->flags & ANYOF_NONBITMAP_NON_UTF8)) {
992 cl_anything(pRExC_state, cl);
995 cl->flags |= ANYOF_UNICODE_ALL;
1000 /* Take the union */
1001 cl->flags |= or_with->flags;
1005 #define TRIE_LIST_ITEM(state,idx) (trie->states[state].trans.list)[ idx ]
1006 #define TRIE_LIST_CUR(state) ( TRIE_LIST_ITEM( state, 0 ).forid )
1007 #define TRIE_LIST_LEN(state) ( TRIE_LIST_ITEM( state, 0 ).newstate )
1008 #define TRIE_LIST_USED(idx) ( trie->states[state].trans.list ? (TRIE_LIST_CUR( idx ) - 1) : 0 )
1013 dump_trie(trie,widecharmap,revcharmap)
1014 dump_trie_interim_list(trie,widecharmap,revcharmap,next_alloc)
1015 dump_trie_interim_table(trie,widecharmap,revcharmap,next_alloc)
1017 These routines dump out a trie in a somewhat readable format.
1018 The _interim_ variants are used for debugging the interim
1019 tables that are used to generate the final compressed
1020 representation which is what dump_trie expects.
1022 Part of the reason for their existence is to provide a form
1023 of documentation as to how the different representations function.
1028 Dumps the final compressed table form of the trie to Perl_debug_log.
1029 Used for debugging make_trie().
1033 S_dump_trie(pTHX_ const struct _reg_trie_data *trie, HV *widecharmap,
1034 AV *revcharmap, U32 depth)
1037 SV *sv=sv_newmortal();
1038 int colwidth= widecharmap ? 6 : 4;
1040 GET_RE_DEBUG_FLAGS_DECL;
1042 PERL_ARGS_ASSERT_DUMP_TRIE;
1044 PerlIO_printf( Perl_debug_log, "%*sChar : %-6s%-6s%-4s ",
1045 (int)depth * 2 + 2,"",
1046 "Match","Base","Ofs" );
1048 for( state = 0 ; state < trie->uniquecharcount ; state++ ) {
1049 SV ** const tmp = av_fetch( revcharmap, state, 0);
1051 PerlIO_printf( Perl_debug_log, "%*s",
1053 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1054 PL_colors[0], PL_colors[1],
1055 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1056 PERL_PV_ESCAPE_FIRSTCHAR
1061 PerlIO_printf( Perl_debug_log, "\n%*sState|-----------------------",
1062 (int)depth * 2 + 2,"");
1064 for( state = 0 ; state < trie->uniquecharcount ; state++ )
1065 PerlIO_printf( Perl_debug_log, "%.*s", colwidth, "--------");
1066 PerlIO_printf( Perl_debug_log, "\n");
1068 for( state = 1 ; state < trie->statecount ; state++ ) {
1069 const U32 base = trie->states[ state ].trans.base;
1071 PerlIO_printf( Perl_debug_log, "%*s#%4"UVXf"|", (int)depth * 2 + 2,"", (UV)state);
1073 if ( trie->states[ state ].wordnum ) {
1074 PerlIO_printf( Perl_debug_log, " W%4X", trie->states[ state ].wordnum );
1076 PerlIO_printf( Perl_debug_log, "%6s", "" );
1079 PerlIO_printf( Perl_debug_log, " @%4"UVXf" ", (UV)base );
1084 while( ( base + ofs < trie->uniquecharcount ) ||
1085 ( base + ofs - trie->uniquecharcount < trie->lasttrans
1086 && trie->trans[ base + ofs - trie->uniquecharcount ].check != state))
1089 PerlIO_printf( Perl_debug_log, "+%2"UVXf"[ ", (UV)ofs);
1091 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
1092 if ( ( base + ofs >= trie->uniquecharcount ) &&
1093 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
1094 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
1096 PerlIO_printf( Perl_debug_log, "%*"UVXf,
1098 (UV)trie->trans[ base + ofs - trie->uniquecharcount ].next );
1100 PerlIO_printf( Perl_debug_log, "%*s",colwidth," ." );
1104 PerlIO_printf( Perl_debug_log, "]");
1107 PerlIO_printf( Perl_debug_log, "\n" );
1109 PerlIO_printf(Perl_debug_log, "%*sword_info N:(prev,len)=", (int)depth*2, "");
1110 for (word=1; word <= trie->wordcount; word++) {
1111 PerlIO_printf(Perl_debug_log, " %d:(%d,%d)",
1112 (int)word, (int)(trie->wordinfo[word].prev),
1113 (int)(trie->wordinfo[word].len));
1115 PerlIO_printf(Perl_debug_log, "\n" );
1118 Dumps a fully constructed but uncompressed trie in list form.
1119 List tries normally only are used for construction when the number of
1120 possible chars (trie->uniquecharcount) is very high.
1121 Used for debugging make_trie().
1124 S_dump_trie_interim_list(pTHX_ const struct _reg_trie_data *trie,
1125 HV *widecharmap, AV *revcharmap, U32 next_alloc,
1129 SV *sv=sv_newmortal();
1130 int colwidth= widecharmap ? 6 : 4;
1131 GET_RE_DEBUG_FLAGS_DECL;
1133 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_LIST;
1135 /* print out the table precompression. */
1136 PerlIO_printf( Perl_debug_log, "%*sState :Word | Transition Data\n%*s%s",
1137 (int)depth * 2 + 2,"", (int)depth * 2 + 2,"",
1138 "------:-----+-----------------\n" );
1140 for( state=1 ; state < next_alloc ; state ++ ) {
1143 PerlIO_printf( Perl_debug_log, "%*s %4"UVXf" :",
1144 (int)depth * 2 + 2,"", (UV)state );
1145 if ( ! trie->states[ state ].wordnum ) {
1146 PerlIO_printf( Perl_debug_log, "%5s| ","");
1148 PerlIO_printf( Perl_debug_log, "W%4x| ",
1149 trie->states[ state ].wordnum
1152 for( charid = 1 ; charid <= TRIE_LIST_USED( state ) ; charid++ ) {
1153 SV ** const tmp = av_fetch( revcharmap, TRIE_LIST_ITEM(state,charid).forid, 0);
1155 PerlIO_printf( Perl_debug_log, "%*s:%3X=%4"UVXf" | ",
1157 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1158 PL_colors[0], PL_colors[1],
1159 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1160 PERL_PV_ESCAPE_FIRSTCHAR
1162 TRIE_LIST_ITEM(state,charid).forid,
1163 (UV)TRIE_LIST_ITEM(state,charid).newstate
1166 PerlIO_printf(Perl_debug_log, "\n%*s| ",
1167 (int)((depth * 2) + 14), "");
1170 PerlIO_printf( Perl_debug_log, "\n");
1175 Dumps a fully constructed but uncompressed trie in table form.
1176 This is the normal DFA style state transition table, with a few
1177 twists to facilitate compression later.
1178 Used for debugging make_trie().
1181 S_dump_trie_interim_table(pTHX_ const struct _reg_trie_data *trie,
1182 HV *widecharmap, AV *revcharmap, U32 next_alloc,
1187 SV *sv=sv_newmortal();
1188 int colwidth= widecharmap ? 6 : 4;
1189 GET_RE_DEBUG_FLAGS_DECL;
1191 PERL_ARGS_ASSERT_DUMP_TRIE_INTERIM_TABLE;
1194 print out the table precompression so that we can do a visual check
1195 that they are identical.
1198 PerlIO_printf( Perl_debug_log, "%*sChar : ",(int)depth * 2 + 2,"" );
1200 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1201 SV ** const tmp = av_fetch( revcharmap, charid, 0);
1203 PerlIO_printf( Perl_debug_log, "%*s",
1205 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), colwidth,
1206 PL_colors[0], PL_colors[1],
1207 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
1208 PERL_PV_ESCAPE_FIRSTCHAR
1214 PerlIO_printf( Perl_debug_log, "\n%*sState+-",(int)depth * 2 + 2,"" );
1216 for( charid=0 ; charid < trie->uniquecharcount ; charid++ ) {
1217 PerlIO_printf( Perl_debug_log, "%.*s", colwidth,"--------");
1220 PerlIO_printf( Perl_debug_log, "\n" );
1222 for( state=1 ; state < next_alloc ; state += trie->uniquecharcount ) {
1224 PerlIO_printf( Perl_debug_log, "%*s%4"UVXf" : ",
1225 (int)depth * 2 + 2,"",
1226 (UV)TRIE_NODENUM( state ) );
1228 for( charid = 0 ; charid < trie->uniquecharcount ; charid++ ) {
1229 UV v=(UV)SAFE_TRIE_NODENUM( trie->trans[ state + charid ].next );
1231 PerlIO_printf( Perl_debug_log, "%*"UVXf, colwidth, v );
1233 PerlIO_printf( Perl_debug_log, "%*s", colwidth, "." );
1235 if ( ! trie->states[ TRIE_NODENUM( state ) ].wordnum ) {
1236 PerlIO_printf( Perl_debug_log, " (%4"UVXf")\n", (UV)trie->trans[ state ].check );
1238 PerlIO_printf( Perl_debug_log, " (%4"UVXf") W%4X\n", (UV)trie->trans[ state ].check,
1239 trie->states[ TRIE_NODENUM( state ) ].wordnum );
1247 /* make_trie(startbranch,first,last,tail,word_count,flags,depth)
1248 startbranch: the first branch in the whole branch sequence
1249 first : start branch of sequence of branch-exact nodes.
1250 May be the same as startbranch
1251 last : Thing following the last branch.
1252 May be the same as tail.
1253 tail : item following the branch sequence
1254 count : words in the sequence
1255 flags : currently the OP() type we will be building one of /EXACT(|F|Fl)/
1256 depth : indent depth
1258 Inplace optimizes a sequence of 2 or more Branch-Exact nodes into a TRIE node.
1260 A trie is an N'ary tree where the branches are determined by digital
1261 decomposition of the key. IE, at the root node you look up the 1st character and
1262 follow that branch repeat until you find the end of the branches. Nodes can be
1263 marked as "accepting" meaning they represent a complete word. Eg:
1267 would convert into the following structure. Numbers represent states, letters
1268 following numbers represent valid transitions on the letter from that state, if
1269 the number is in square brackets it represents an accepting state, otherwise it
1270 will be in parenthesis.
1272 +-h->+-e->[3]-+-r->(8)-+-s->[9]
1276 (1) +-i->(6)-+-s->[7]
1278 +-s->(3)-+-h->(4)-+-e->[5]
1280 Accept Word Mapping: 3=>1 (he),5=>2 (she), 7=>3 (his), 9=>4 (hers)
1282 This shows that when matching against the string 'hers' we will begin at state 1
1283 read 'h' and move to state 2, read 'e' and move to state 3 which is accepting,
1284 then read 'r' and go to state 8 followed by 's' which takes us to state 9 which
1285 is also accepting. Thus we know that we can match both 'he' and 'hers' with a
1286 single traverse. We store a mapping from accepting to state to which word was
1287 matched, and then when we have multiple possibilities we try to complete the
1288 rest of the regex in the order in which they occured in the alternation.
1290 The only prior NFA like behaviour that would be changed by the TRIE support is
1291 the silent ignoring of duplicate alternations which are of the form:
1293 / (DUPE|DUPE) X? (?{ ... }) Y /x
1295 Thus EVAL blocks following a trie may be called a different number of times with
1296 and without the optimisation. With the optimisations dupes will be silently
1297 ignored. This inconsistent behaviour of EVAL type nodes is well established as
1298 the following demonstrates:
1300 'words'=~/(word|word|word)(?{ print $1 })[xyz]/
1302 which prints out 'word' three times, but
1304 'words'=~/(word|word|word)(?{ print $1 })S/
1306 which doesnt print it out at all. This is due to other optimisations kicking in.
1308 Example of what happens on a structural level:
1310 The regexp /(ac|ad|ab)+/ will produce the following debug output:
1312 1: CURLYM[1] {1,32767}(18)
1323 This would be optimizable with startbranch=5, first=5, last=16, tail=16
1324 and should turn into:
1326 1: CURLYM[1] {1,32767}(18)
1328 [Words:3 Chars Stored:6 Unique Chars:4 States:5 NCP:1]
1336 Cases where tail != last would be like /(?foo|bar)baz/:
1346 which would be optimizable with startbranch=1, first=1, last=7, tail=8
1347 and would end up looking like:
1350 [Words:2 Chars Stored:6 Unique Chars:5 States:7 NCP:1]
1357 d = uvuni_to_utf8_flags(d, uv, 0);
1359 is the recommended Unicode-aware way of saying
1364 #define TRIE_STORE_REVCHAR \
1367 SV *zlopp = newSV(2); \
1368 unsigned char *flrbbbbb = (unsigned char *) SvPVX(zlopp); \
1369 unsigned const char *const kapow = uvuni_to_utf8(flrbbbbb, uvc & 0xFF); \
1370 SvCUR_set(zlopp, kapow - flrbbbbb); \
1373 av_push(revcharmap, zlopp); \
1375 char ooooff = (char)uvc; \
1376 av_push(revcharmap, newSVpvn(&ooooff, 1)); \
1380 #define TRIE_READ_CHAR STMT_START { \
1384 if ( foldlen > 0 ) { \
1385 uvc = utf8n_to_uvuni( scan, UTF8_MAXLEN, &len, uniflags ); \
1390 len = UTF8SKIP(uc);\
1391 uvc = to_utf8_fold( uc, foldbuf, &foldlen); \
1392 foldlen -= UNISKIP( uvc ); \
1393 scan = foldbuf + UNISKIP( uvc ); \
1396 uvc = utf8n_to_uvuni( (const U8*)uc, UTF8_MAXLEN, &len, uniflags);\
1406 #define TRIE_LIST_PUSH(state,fid,ns) STMT_START { \
1407 if ( TRIE_LIST_CUR( state ) >=TRIE_LIST_LEN( state ) ) { \
1408 U32 ging = TRIE_LIST_LEN( state ) *= 2; \
1409 Renew( trie->states[ state ].trans.list, ging, reg_trie_trans_le ); \
1411 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).forid = fid; \
1412 TRIE_LIST_ITEM( state, TRIE_LIST_CUR( state ) ).newstate = ns; \
1413 TRIE_LIST_CUR( state )++; \
1416 #define TRIE_LIST_NEW(state) STMT_START { \
1417 Newxz( trie->states[ state ].trans.list, \
1418 4, reg_trie_trans_le ); \
1419 TRIE_LIST_CUR( state ) = 1; \
1420 TRIE_LIST_LEN( state ) = 4; \
1423 #define TRIE_HANDLE_WORD(state) STMT_START { \
1424 U16 dupe= trie->states[ state ].wordnum; \
1425 regnode * const noper_next = regnext( noper ); \
1428 /* store the word for dumping */ \
1430 if (OP(noper) != NOTHING) \
1431 tmp = newSVpvn_utf8(STRING(noper), STR_LEN(noper), UTF); \
1433 tmp = newSVpvn_utf8( "", 0, UTF ); \
1434 av_push( trie_words, tmp ); \
1438 trie->wordinfo[curword].prev = 0; \
1439 trie->wordinfo[curword].len = wordlen; \
1440 trie->wordinfo[curword].accept = state; \
1442 if ( noper_next < tail ) { \
1444 trie->jump = (U16 *) PerlMemShared_calloc( word_count + 1, sizeof(U16) ); \
1445 trie->jump[curword] = (U16)(noper_next - convert); \
1447 jumper = noper_next; \
1449 nextbranch= regnext(cur); \
1453 /* It's a dupe. Pre-insert into the wordinfo[].prev */\
1454 /* chain, so that when the bits of chain are later */\
1455 /* linked together, the dups appear in the chain */\
1456 trie->wordinfo[curword].prev = trie->wordinfo[dupe].prev; \
1457 trie->wordinfo[dupe].prev = curword; \
1459 /* we haven't inserted this word yet. */ \
1460 trie->states[ state ].wordnum = curword; \
1465 #define TRIE_TRANS_STATE(state,base,ucharcount,charid,special) \
1466 ( ( base + charid >= ucharcount \
1467 && base + charid < ubound \
1468 && state == trie->trans[ base - ucharcount + charid ].check \
1469 && trie->trans[ base - ucharcount + charid ].next ) \
1470 ? trie->trans[ base - ucharcount + charid ].next \
1471 : ( state==1 ? special : 0 ) \
1475 #define MADE_JUMP_TRIE 2
1476 #define MADE_EXACT_TRIE 4
1479 S_make_trie(pTHX_ RExC_state_t *pRExC_state, regnode *startbranch, regnode *first, regnode *last, regnode *tail, U32 word_count, U32 flags, U32 depth)
1482 /* first pass, loop through and scan words */
1483 reg_trie_data *trie;
1484 HV *widecharmap = NULL;
1485 AV *revcharmap = newAV();
1487 const U32 uniflags = UTF8_ALLOW_DEFAULT;
1492 regnode *jumper = NULL;
1493 regnode *nextbranch = NULL;
1494 regnode *convert = NULL;
1495 U32 *prev_states; /* temp array mapping each state to previous one */
1496 /* we just use folder as a flag in utf8 */
1497 const U8 * folder = NULL;
1500 const U32 data_slot = add_data( pRExC_state, 4, "tuuu" );
1501 AV *trie_words = NULL;
1502 /* along with revcharmap, this only used during construction but both are
1503 * useful during debugging so we store them in the struct when debugging.
1506 const U32 data_slot = add_data( pRExC_state, 2, "tu" );
1507 STRLEN trie_charcount=0;
1509 SV *re_trie_maxbuff;
1510 GET_RE_DEBUG_FLAGS_DECL;
1512 PERL_ARGS_ASSERT_MAKE_TRIE;
1514 PERL_UNUSED_ARG(depth);
1520 case EXACTFU: folder = PL_fold_latin1; break;
1521 case EXACTF: folder = PL_fold; break;
1522 case EXACTFL: folder = PL_fold_locale; break;
1523 default: Perl_croak( aTHX_ "panic! In trie construction, unknown node type %u", (unsigned) flags );
1526 trie = (reg_trie_data *) PerlMemShared_calloc( 1, sizeof(reg_trie_data) );
1528 trie->startstate = 1;
1529 trie->wordcount = word_count;
1530 RExC_rxi->data->data[ data_slot ] = (void*)trie;
1531 trie->charmap = (U16 *) PerlMemShared_calloc( 256, sizeof(U16) );
1532 if (!(UTF && folder))
1533 trie->bitmap = (char *) PerlMemShared_calloc( ANYOF_BITMAP_SIZE, 1 );
1534 trie->wordinfo = (reg_trie_wordinfo *) PerlMemShared_calloc(
1535 trie->wordcount+1, sizeof(reg_trie_wordinfo));
1538 trie_words = newAV();
1541 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
1542 if (!SvIOK(re_trie_maxbuff)) {
1543 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
1546 PerlIO_printf( Perl_debug_log,
1547 "%*smake_trie start==%d, first==%d, last==%d, tail==%d depth=%d\n",
1548 (int)depth * 2 + 2, "",
1549 REG_NODE_NUM(startbranch),REG_NODE_NUM(first),
1550 REG_NODE_NUM(last), REG_NODE_NUM(tail),
1554 /* Find the node we are going to overwrite */
1555 if ( first == startbranch && OP( last ) != BRANCH ) {
1556 /* whole branch chain */
1559 /* branch sub-chain */
1560 convert = NEXTOPER( first );
1563 /* -- First loop and Setup --
1565 We first traverse the branches and scan each word to determine if it
1566 contains widechars, and how many unique chars there are, this is
1567 important as we have to build a table with at least as many columns as we
1570 We use an array of integers to represent the character codes 0..255
1571 (trie->charmap) and we use a an HV* to store Unicode characters. We use the
1572 native representation of the character value as the key and IV's for the
1575 *TODO* If we keep track of how many times each character is used we can
1576 remap the columns so that the table compression later on is more
1577 efficient in terms of memory by ensuring the most common value is in the
1578 middle and the least common are on the outside. IMO this would be better
1579 than a most to least common mapping as theres a decent chance the most
1580 common letter will share a node with the least common, meaning the node
1581 will not be compressible. With a middle is most common approach the worst
1582 case is when we have the least common nodes twice.
1586 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1587 regnode * const noper = NEXTOPER( cur );
1588 const U8 *uc = (U8*)STRING( noper );
1589 const U8 * const e = uc + STR_LEN( noper );
1591 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1592 const U8 *scan = (U8*)NULL;
1593 U32 wordlen = 0; /* required init */
1595 bool set_bit = trie->bitmap ? 1 : 0; /*store the first char in the bitmap?*/
1597 if (OP(noper) == NOTHING) {
1601 if ( set_bit ) /* bitmap only alloced when !(UTF&&Folding) */
1602 TRIE_BITMAP_SET(trie,*uc); /* store the raw first byte
1603 regardless of encoding */
1605 for ( ; uc < e ; uc += len ) {
1606 TRIE_CHARCOUNT(trie)++;
1610 if ( !trie->charmap[ uvc ] ) {
1611 trie->charmap[ uvc ]=( ++trie->uniquecharcount );
1613 trie->charmap[ folder[ uvc ] ] = trie->charmap[ uvc ];
1617 /* store the codepoint in the bitmap, and its folded
1619 TRIE_BITMAP_SET(trie,uvc);
1621 /* store the folded codepoint */
1622 if ( folder ) TRIE_BITMAP_SET(trie,folder[ uvc ]);
1625 /* store first byte of utf8 representation of
1626 variant codepoints */
1627 if (! UNI_IS_INVARIANT(uvc)) {
1628 TRIE_BITMAP_SET(trie, UTF8_TWO_BYTE_HI(uvc));
1631 set_bit = 0; /* We've done our bit :-) */
1636 widecharmap = newHV();
1638 svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 1 );
1641 Perl_croak( aTHX_ "error creating/fetching widecharmap entry for 0x%"UVXf, uvc );
1643 if ( !SvTRUE( *svpp ) ) {
1644 sv_setiv( *svpp, ++trie->uniquecharcount );
1649 if( cur == first ) {
1652 } else if (chars < trie->minlen) {
1654 } else if (chars > trie->maxlen) {
1658 } /* end first pass */
1659 DEBUG_TRIE_COMPILE_r(
1660 PerlIO_printf( Perl_debug_log, "%*sTRIE(%s): W:%d C:%d Uq:%d Min:%d Max:%d\n",
1661 (int)depth * 2 + 2,"",
1662 ( widecharmap ? "UTF8" : "NATIVE" ), (int)word_count,
1663 (int)TRIE_CHARCOUNT(trie), trie->uniquecharcount,
1664 (int)trie->minlen, (int)trie->maxlen )
1668 We now know what we are dealing with in terms of unique chars and
1669 string sizes so we can calculate how much memory a naive
1670 representation using a flat table will take. If it's over a reasonable
1671 limit (as specified by ${^RE_TRIE_MAXBUF}) we use a more memory
1672 conservative but potentially much slower representation using an array
1675 At the end we convert both representations into the same compressed
1676 form that will be used in regexec.c for matching with. The latter
1677 is a form that cannot be used to construct with but has memory
1678 properties similar to the list form and access properties similar
1679 to the table form making it both suitable for fast searches and
1680 small enough that its feasable to store for the duration of a program.
1682 See the comment in the code where the compressed table is produced
1683 inplace from the flat tabe representation for an explanation of how
1684 the compression works.
1689 Newx(prev_states, TRIE_CHARCOUNT(trie) + 2, U32);
1692 if ( (IV)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1) > SvIV(re_trie_maxbuff) ) {
1694 Second Pass -- Array Of Lists Representation
1696 Each state will be represented by a list of charid:state records
1697 (reg_trie_trans_le) the first such element holds the CUR and LEN
1698 points of the allocated array. (See defines above).
1700 We build the initial structure using the lists, and then convert
1701 it into the compressed table form which allows faster lookups
1702 (but cant be modified once converted).
1705 STRLEN transcount = 1;
1707 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1708 "%*sCompiling trie using list compiler\n",
1709 (int)depth * 2 + 2, ""));
1711 trie->states = (reg_trie_state *)
1712 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1713 sizeof(reg_trie_state) );
1717 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1719 regnode * const noper = NEXTOPER( cur );
1720 U8 *uc = (U8*)STRING( noper );
1721 const U8 * const e = uc + STR_LEN( noper );
1722 U32 state = 1; /* required init */
1723 U16 charid = 0; /* sanity init */
1724 U8 *scan = (U8*)NULL; /* sanity init */
1725 STRLEN foldlen = 0; /* required init */
1726 U32 wordlen = 0; /* required init */
1727 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1729 if (OP(noper) != NOTHING) {
1730 for ( ; uc < e ; uc += len ) {
1735 charid = trie->charmap[ uvc ];
1737 SV** const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1741 charid=(U16)SvIV( *svpp );
1744 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1751 if ( !trie->states[ state ].trans.list ) {
1752 TRIE_LIST_NEW( state );
1754 for ( check = 1; check <= TRIE_LIST_USED( state ); check++ ) {
1755 if ( TRIE_LIST_ITEM( state, check ).forid == charid ) {
1756 newstate = TRIE_LIST_ITEM( state, check ).newstate;
1761 newstate = next_alloc++;
1762 prev_states[newstate] = state;
1763 TRIE_LIST_PUSH( state, charid, newstate );
1768 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1772 TRIE_HANDLE_WORD(state);
1774 } /* end second pass */
1776 /* next alloc is the NEXT state to be allocated */
1777 trie->statecount = next_alloc;
1778 trie->states = (reg_trie_state *)
1779 PerlMemShared_realloc( trie->states,
1781 * sizeof(reg_trie_state) );
1783 /* and now dump it out before we compress it */
1784 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_list(trie, widecharmap,
1785 revcharmap, next_alloc,
1789 trie->trans = (reg_trie_trans *)
1790 PerlMemShared_calloc( transcount, sizeof(reg_trie_trans) );
1797 for( state=1 ; state < next_alloc ; state ++ ) {
1801 DEBUG_TRIE_COMPILE_MORE_r(
1802 PerlIO_printf( Perl_debug_log, "tp: %d zp: %d ",tp,zp)
1806 if (trie->states[state].trans.list) {
1807 U16 minid=TRIE_LIST_ITEM( state, 1).forid;
1811 for( idx = 2 ; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1812 const U16 forid = TRIE_LIST_ITEM( state, idx).forid;
1813 if ( forid < minid ) {
1815 } else if ( forid > maxid ) {
1819 if ( transcount < tp + maxid - minid + 1) {
1821 trie->trans = (reg_trie_trans *)
1822 PerlMemShared_realloc( trie->trans,
1824 * sizeof(reg_trie_trans) );
1825 Zero( trie->trans + (transcount / 2), transcount / 2 , reg_trie_trans );
1827 base = trie->uniquecharcount + tp - minid;
1828 if ( maxid == minid ) {
1830 for ( ; zp < tp ; zp++ ) {
1831 if ( ! trie->trans[ zp ].next ) {
1832 base = trie->uniquecharcount + zp - minid;
1833 trie->trans[ zp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1834 trie->trans[ zp ].check = state;
1840 trie->trans[ tp ].next = TRIE_LIST_ITEM( state, 1).newstate;
1841 trie->trans[ tp ].check = state;
1846 for ( idx=1; idx <= TRIE_LIST_USED( state ) ; idx++ ) {
1847 const U32 tid = base - trie->uniquecharcount + TRIE_LIST_ITEM( state, idx ).forid;
1848 trie->trans[ tid ].next = TRIE_LIST_ITEM( state, idx ).newstate;
1849 trie->trans[ tid ].check = state;
1851 tp += ( maxid - minid + 1 );
1853 Safefree(trie->states[ state ].trans.list);
1856 DEBUG_TRIE_COMPILE_MORE_r(
1857 PerlIO_printf( Perl_debug_log, " base: %d\n",base);
1860 trie->states[ state ].trans.base=base;
1862 trie->lasttrans = tp + 1;
1866 Second Pass -- Flat Table Representation.
1868 we dont use the 0 slot of either trans[] or states[] so we add 1 to each.
1869 We know that we will need Charcount+1 trans at most to store the data
1870 (one row per char at worst case) So we preallocate both structures
1871 assuming worst case.
1873 We then construct the trie using only the .next slots of the entry
1876 We use the .check field of the first entry of the node temporarily to
1877 make compression both faster and easier by keeping track of how many non
1878 zero fields are in the node.
1880 Since trans are numbered from 1 any 0 pointer in the table is a FAIL
1883 There are two terms at use here: state as a TRIE_NODEIDX() which is a
1884 number representing the first entry of the node, and state as a
1885 TRIE_NODENUM() which is the trans number. state 1 is TRIE_NODEIDX(1) and
1886 TRIE_NODENUM(1), state 2 is TRIE_NODEIDX(2) and TRIE_NODENUM(3) if there
1887 are 2 entrys per node. eg:
1895 The table is internally in the right hand, idx form. However as we also
1896 have to deal with the states array which is indexed by nodenum we have to
1897 use TRIE_NODENUM() to convert.
1900 DEBUG_TRIE_COMPILE_MORE_r( PerlIO_printf( Perl_debug_log,
1901 "%*sCompiling trie using table compiler\n",
1902 (int)depth * 2 + 2, ""));
1904 trie->trans = (reg_trie_trans *)
1905 PerlMemShared_calloc( ( TRIE_CHARCOUNT(trie) + 1 )
1906 * trie->uniquecharcount + 1,
1907 sizeof(reg_trie_trans) );
1908 trie->states = (reg_trie_state *)
1909 PerlMemShared_calloc( TRIE_CHARCOUNT(trie) + 2,
1910 sizeof(reg_trie_state) );
1911 next_alloc = trie->uniquecharcount + 1;
1914 for ( cur = first ; cur < last ; cur = regnext( cur ) ) {
1916 regnode * const noper = NEXTOPER( cur );
1917 const U8 *uc = (U8*)STRING( noper );
1918 const U8 * const e = uc + STR_LEN( noper );
1920 U32 state = 1; /* required init */
1922 U16 charid = 0; /* sanity init */
1923 U32 accept_state = 0; /* sanity init */
1924 U8 *scan = (U8*)NULL; /* sanity init */
1926 STRLEN foldlen = 0; /* required init */
1927 U32 wordlen = 0; /* required init */
1928 U8 foldbuf[ UTF8_MAXBYTES_CASE + 1 ];
1930 if ( OP(noper) != NOTHING ) {
1931 for ( ; uc < e ; uc += len ) {
1936 charid = trie->charmap[ uvc ];
1938 SV* const * const svpp = hv_fetch( widecharmap, (char*)&uvc, sizeof( UV ), 0);
1939 charid = svpp ? (U16)SvIV(*svpp) : 0;
1943 if ( !trie->trans[ state + charid ].next ) {
1944 trie->trans[ state + charid ].next = next_alloc;
1945 trie->trans[ state ].check++;
1946 prev_states[TRIE_NODENUM(next_alloc)]
1947 = TRIE_NODENUM(state);
1948 next_alloc += trie->uniquecharcount;
1950 state = trie->trans[ state + charid ].next;
1952 Perl_croak( aTHX_ "panic! In trie construction, no char mapping for %"IVdf, uvc );
1954 /* charid is now 0 if we dont know the char read, or nonzero if we do */
1957 accept_state = TRIE_NODENUM( state );
1958 TRIE_HANDLE_WORD(accept_state);
1960 } /* end second pass */
1962 /* and now dump it out before we compress it */
1963 DEBUG_TRIE_COMPILE_MORE_r(dump_trie_interim_table(trie, widecharmap,
1965 next_alloc, depth+1));
1969 * Inplace compress the table.*
1971 For sparse data sets the table constructed by the trie algorithm will
1972 be mostly 0/FAIL transitions or to put it another way mostly empty.
1973 (Note that leaf nodes will not contain any transitions.)
1975 This algorithm compresses the tables by eliminating most such
1976 transitions, at the cost of a modest bit of extra work during lookup:
1978 - Each states[] entry contains a .base field which indicates the
1979 index in the state[] array wheres its transition data is stored.
1981 - If .base is 0 there are no valid transitions from that node.
1983 - If .base is nonzero then charid is added to it to find an entry in
1986 -If trans[states[state].base+charid].check!=state then the
1987 transition is taken to be a 0/Fail transition. Thus if there are fail
1988 transitions at the front of the node then the .base offset will point
1989 somewhere inside the previous nodes data (or maybe even into a node
1990 even earlier), but the .check field determines if the transition is
1994 The following process inplace converts the table to the compressed
1995 table: We first do not compress the root node 1,and mark all its
1996 .check pointers as 1 and set its .base pointer as 1 as well. This
1997 allows us to do a DFA construction from the compressed table later,
1998 and ensures that any .base pointers we calculate later are greater
2001 - We set 'pos' to indicate the first entry of the second node.
2003 - We then iterate over the columns of the node, finding the first and
2004 last used entry at l and m. We then copy l..m into pos..(pos+m-l),
2005 and set the .check pointers accordingly, and advance pos
2006 appropriately and repreat for the next node. Note that when we copy
2007 the next pointers we have to convert them from the original
2008 NODEIDX form to NODENUM form as the former is not valid post
2011 - If a node has no transitions used we mark its base as 0 and do not
2012 advance the pos pointer.
2014 - If a node only has one transition we use a second pointer into the
2015 structure to fill in allocated fail transitions from other states.
2016 This pointer is independent of the main pointer and scans forward
2017 looking for null transitions that are allocated to a state. When it
2018 finds one it writes the single transition into the "hole". If the
2019 pointer doesnt find one the single transition is appended as normal.
2021 - Once compressed we can Renew/realloc the structures to release the
2024 See "Table-Compression Methods" in sec 3.9 of the Red Dragon,
2025 specifically Fig 3.47 and the associated pseudocode.
2029 const U32 laststate = TRIE_NODENUM( next_alloc );
2032 trie->statecount = laststate;
2034 for ( state = 1 ; state < laststate ; state++ ) {
2036 const U32 stateidx = TRIE_NODEIDX( state );
2037 const U32 o_used = trie->trans[ stateidx ].check;
2038 U32 used = trie->trans[ stateidx ].check;
2039 trie->trans[ stateidx ].check = 0;
2041 for ( charid = 0 ; used && charid < trie->uniquecharcount ; charid++ ) {
2042 if ( flag || trie->trans[ stateidx + charid ].next ) {
2043 if ( trie->trans[ stateidx + charid ].next ) {
2045 for ( ; zp < pos ; zp++ ) {
2046 if ( ! trie->trans[ zp ].next ) {
2050 trie->states[ state ].trans.base = zp + trie->uniquecharcount - charid ;
2051 trie->trans[ zp ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
2052 trie->trans[ zp ].check = state;
2053 if ( ++zp > pos ) pos = zp;
2060 trie->states[ state ].trans.base = pos + trie->uniquecharcount - charid ;
2062 trie->trans[ pos ].next = SAFE_TRIE_NODENUM( trie->trans[ stateidx + charid ].next );
2063 trie->trans[ pos ].check = state;
2068 trie->lasttrans = pos + 1;
2069 trie->states = (reg_trie_state *)
2070 PerlMemShared_realloc( trie->states, laststate
2071 * sizeof(reg_trie_state) );
2072 DEBUG_TRIE_COMPILE_MORE_r(
2073 PerlIO_printf( Perl_debug_log,
2074 "%*sAlloc: %d Orig: %"IVdf" elements, Final:%"IVdf". Savings of %%%5.2f\n",
2075 (int)depth * 2 + 2,"",
2076 (int)( ( TRIE_CHARCOUNT(trie) + 1 ) * trie->uniquecharcount + 1 ),
2079 ( ( next_alloc - pos ) * 100 ) / (double)next_alloc );
2082 } /* end table compress */
2084 DEBUG_TRIE_COMPILE_MORE_r(
2085 PerlIO_printf(Perl_debug_log, "%*sStatecount:%"UVxf" Lasttrans:%"UVxf"\n",
2086 (int)depth * 2 + 2, "",
2087 (UV)trie->statecount,
2088 (UV)trie->lasttrans)
2090 /* resize the trans array to remove unused space */
2091 trie->trans = (reg_trie_trans *)
2092 PerlMemShared_realloc( trie->trans, trie->lasttrans
2093 * sizeof(reg_trie_trans) );
2095 { /* Modify the program and insert the new TRIE node */
2096 U8 nodetype =(U8)(flags & 0xFF);
2100 regnode *optimize = NULL;
2101 #ifdef RE_TRACK_PATTERN_OFFSETS
2104 U32 mjd_nodelen = 0;
2105 #endif /* RE_TRACK_PATTERN_OFFSETS */
2106 #endif /* DEBUGGING */
2108 This means we convert either the first branch or the first Exact,
2109 depending on whether the thing following (in 'last') is a branch
2110 or not and whther first is the startbranch (ie is it a sub part of
2111 the alternation or is it the whole thing.)
2112 Assuming its a sub part we convert the EXACT otherwise we convert
2113 the whole branch sequence, including the first.
2115 /* Find the node we are going to overwrite */
2116 if ( first != startbranch || OP( last ) == BRANCH ) {
2117 /* branch sub-chain */
2118 NEXT_OFF( first ) = (U16)(last - first);
2119 #ifdef RE_TRACK_PATTERN_OFFSETS
2121 mjd_offset= Node_Offset((convert));
2122 mjd_nodelen= Node_Length((convert));
2125 /* whole branch chain */
2127 #ifdef RE_TRACK_PATTERN_OFFSETS
2130 const regnode *nop = NEXTOPER( convert );
2131 mjd_offset= Node_Offset((nop));
2132 mjd_nodelen= Node_Length((nop));
2136 PerlIO_printf(Perl_debug_log, "%*sMJD offset:%"UVuf" MJD length:%"UVuf"\n",
2137 (int)depth * 2 + 2, "",
2138 (UV)mjd_offset, (UV)mjd_nodelen)
2141 /* But first we check to see if there is a common prefix we can
2142 split out as an EXACT and put in front of the TRIE node. */
2143 trie->startstate= 1;
2144 if ( trie->bitmap && !widecharmap && !trie->jump ) {
2146 for ( state = 1 ; state < trie->statecount-1 ; state++ ) {
2150 const U32 base = trie->states[ state ].trans.base;
2152 if ( trie->states[state].wordnum )
2155 for ( ofs = 0 ; ofs < trie->uniquecharcount ; ofs++ ) {
2156 if ( ( base + ofs >= trie->uniquecharcount ) &&
2157 ( base + ofs - trie->uniquecharcount < trie->lasttrans ) &&
2158 trie->trans[ base + ofs - trie->uniquecharcount ].check == state )
2160 if ( ++count > 1 ) {
2161 SV **tmp = av_fetch( revcharmap, ofs, 0);
2162 const U8 *ch = (U8*)SvPV_nolen_const( *tmp );
2163 if ( state == 1 ) break;
2165 Zero(trie->bitmap, ANYOF_BITMAP_SIZE, char);
2167 PerlIO_printf(Perl_debug_log,
2168 "%*sNew Start State=%"UVuf" Class: [",
2169 (int)depth * 2 + 2, "",
2172 SV ** const tmp = av_fetch( revcharmap, idx, 0);
2173 const U8 * const ch = (U8*)SvPV_nolen_const( *tmp );
2175 TRIE_BITMAP_SET(trie,*ch);
2177 TRIE_BITMAP_SET(trie, folder[ *ch ]);
2179 PerlIO_printf(Perl_debug_log, "%s", (char*)ch)
2183 TRIE_BITMAP_SET(trie,*ch);
2185 TRIE_BITMAP_SET(trie,folder[ *ch ]);
2186 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"%s", ch));
2192 SV **tmp = av_fetch( revcharmap, idx, 0);
2194 char *ch = SvPV( *tmp, len );
2196 SV *sv=sv_newmortal();
2197 PerlIO_printf( Perl_debug_log,
2198 "%*sPrefix State: %"UVuf" Idx:%"UVuf" Char='%s'\n",
2199 (int)depth * 2 + 2, "",
2201 pv_pretty(sv, SvPV_nolen_const(*tmp), SvCUR(*tmp), 6,
2202 PL_colors[0], PL_colors[1],
2203 (SvUTF8(*tmp) ? PERL_PV_ESCAPE_UNI : 0) |
2204 PERL_PV_ESCAPE_FIRSTCHAR
2209 OP( convert ) = nodetype;
2210 str=STRING(convert);
2213 STR_LEN(convert) += len;
2219 DEBUG_OPTIMISE_r(PerlIO_printf( Perl_debug_log,"]\n"));
2224 trie->prefixlen = (state-1);
2226 regnode *n = convert+NODE_SZ_STR(convert);
2227 NEXT_OFF(convert) = NODE_SZ_STR(convert);
2228 trie->startstate = state;
2229 trie->minlen -= (state - 1);
2230 trie->maxlen -= (state - 1);
2232 /* At least the UNICOS C compiler choked on this
2233 * being argument to DEBUG_r(), so let's just have
2236 #ifdef PERL_EXT_RE_BUILD
2242 regnode *fix = convert;
2243 U32 word = trie->wordcount;
2245 Set_Node_Offset_Length(convert, mjd_offset, state - 1);
2246 while( ++fix < n ) {
2247 Set_Node_Offset_Length(fix, 0, 0);
2250 SV ** const tmp = av_fetch( trie_words, word, 0 );
2252 if ( STR_LEN(convert) <= SvCUR(*tmp) )
2253 sv_chop(*tmp, SvPV_nolen(*tmp) + STR_LEN(convert));
2255 sv_chop(*tmp, SvPV_nolen(*tmp) + SvCUR(*tmp));
2263 NEXT_OFF(convert) = (U16)(tail - convert);
2264 DEBUG_r(optimize= n);
2270 if ( trie->maxlen ) {
2271 NEXT_OFF( convert ) = (U16)(tail - convert);
2272 ARG_SET( convert, data_slot );
2273 /* Store the offset to the first unabsorbed branch in
2274 jump[0], which is otherwise unused by the jump logic.
2275 We use this when dumping a trie and during optimisation. */
2277 trie->jump[0] = (U16)(nextbranch - convert);
2279 /* If the start state is not accepting (meaning there is no empty string/NOTHING)
2280 * and there is a bitmap
2281 * and the first "jump target" node we found leaves enough room
2282 * then convert the TRIE node into a TRIEC node, with the bitmap
2283 * embedded inline in the opcode - this is hypothetically faster.
2285 if ( !trie->states[trie->startstate].wordnum
2287 && ( (char *)jumper - (char *)convert) >= (int)sizeof(struct regnode_charclass) )
2289 OP( convert ) = TRIEC;
2290 Copy(trie->bitmap, ((struct regnode_charclass *)convert)->bitmap, ANYOF_BITMAP_SIZE, char);
2291 PerlMemShared_free(trie->bitmap);
2294 OP( convert ) = TRIE;
2296 /* store the type in the flags */
2297 convert->flags = nodetype;
2301 + regarglen[ OP( convert ) ];
2303 /* XXX We really should free up the resource in trie now,
2304 as we won't use them - (which resources?) dmq */
2306 /* needed for dumping*/
2307 DEBUG_r(if (optimize) {
2308 regnode *opt = convert;
2310 while ( ++opt < optimize) {
2311 Set_Node_Offset_Length(opt,0,0);
2314 Try to clean up some of the debris left after the
2317 while( optimize < jumper ) {
2318 mjd_nodelen += Node_Length((optimize));
2319 OP( optimize ) = OPTIMIZED;
2320 Set_Node_Offset_Length(optimize,0,0);
2323 Set_Node_Offset_Length(convert,mjd_offset,mjd_nodelen);
2325 } /* end node insert */
2327 /* Finish populating the prev field of the wordinfo array. Walk back
2328 * from each accept state until we find another accept state, and if
2329 * so, point the first word's .prev field at the second word. If the
2330 * second already has a .prev field set, stop now. This will be the
2331 * case either if we've already processed that word's accept state,
2332 * or that state had multiple words, and the overspill words were
2333 * already linked up earlier.
2340 for (word=1; word <= trie->wordcount; word++) {
2342 if (trie->wordinfo[word].prev)
2344 state = trie->wordinfo[word].accept;
2346 state = prev_states[state];
2349 prev = trie->states[state].wordnum;
2353 trie->wordinfo[word].prev = prev;
2355 Safefree(prev_states);
2359 /* and now dump out the compressed format */
2360 DEBUG_TRIE_COMPILE_r(dump_trie(trie, widecharmap, revcharmap, depth+1));
2362 RExC_rxi->data->data[ data_slot + 1 ] = (void*)widecharmap;
2364 RExC_rxi->data->data[ data_slot + TRIE_WORDS_OFFSET ] = (void*)trie_words;
2365 RExC_rxi->data->data[ data_slot + 3 ] = (void*)revcharmap;
2367 SvREFCNT_dec(revcharmap);
2371 : trie->startstate>1
2377 S_make_trie_failtable(pTHX_ RExC_state_t *pRExC_state, regnode *source, regnode *stclass, U32 depth)
2379 /* The Trie is constructed and compressed now so we can build a fail array if it's needed
2381 This is basically the Aho-Corasick algorithm. Its from exercise 3.31 and 3.32 in the
2382 "Red Dragon" -- Compilers, principles, techniques, and tools. Aho, Sethi, Ullman 1985/88
2385 We find the fail state for each state in the trie, this state is the longest proper
2386 suffix of the current state's 'word' that is also a proper prefix of another word in our
2387 trie. State 1 represents the word '' and is thus the default fail state. This allows
2388 the DFA not to have to restart after its tried and failed a word at a given point, it
2389 simply continues as though it had been matching the other word in the first place.
2391 'abcdgu'=~/abcdefg|cdgu/
2392 When we get to 'd' we are still matching the first word, we would encounter 'g' which would
2393 fail, which would bring us to the state representing 'd' in the second word where we would
2394 try 'g' and succeed, proceeding to match 'cdgu'.
2396 /* add a fail transition */
2397 const U32 trie_offset = ARG(source);
2398 reg_trie_data *trie=(reg_trie_data *)RExC_rxi->data->data[trie_offset];
2400 const U32 ucharcount = trie->uniquecharcount;
2401 const U32 numstates = trie->statecount;
2402 const U32 ubound = trie->lasttrans + ucharcount;
2406 U32 base = trie->states[ 1 ].trans.base;
2409 const U32 data_slot = add_data( pRExC_state, 1, "T" );
2410 GET_RE_DEBUG_FLAGS_DECL;
2412 PERL_ARGS_ASSERT_MAKE_TRIE_FAILTABLE;
2414 PERL_UNUSED_ARG(depth);
2418 ARG_SET( stclass, data_slot );
2419 aho = (reg_ac_data *) PerlMemShared_calloc( 1, sizeof(reg_ac_data) );
2420 RExC_rxi->data->data[ data_slot ] = (void*)aho;
2421 aho->trie=trie_offset;
2422 aho->states=(reg_trie_state *)PerlMemShared_malloc( numstates * sizeof(reg_trie_state) );
2423 Copy( trie->states, aho->states, numstates, reg_trie_state );
2424 Newxz( q, numstates, U32);
2425 aho->fail = (U32 *) PerlMemShared_calloc( numstates, sizeof(U32) );
2428 /* initialize fail[0..1] to be 1 so that we always have
2429 a valid final fail state */
2430 fail[ 0 ] = fail[ 1 ] = 1;
2432 for ( charid = 0; charid < ucharcount ; charid++ ) {
2433 const U32 newstate = TRIE_TRANS_STATE( 1, base, ucharcount, charid, 0 );
2435 q[ q_write ] = newstate;
2436 /* set to point at the root */
2437 fail[ q[ q_write++ ] ]=1;
2440 while ( q_read < q_write) {
2441 const U32 cur = q[ q_read++ % numstates ];
2442 base = trie->states[ cur ].trans.base;
2444 for ( charid = 0 ; charid < ucharcount ; charid++ ) {
2445 const U32 ch_state = TRIE_TRANS_STATE( cur, base, ucharcount, charid, 1 );
2447 U32 fail_state = cur;
2450 fail_state = fail[ fail_state ];
2451 fail_base = aho->states[ fail_state ].trans.base;
2452 } while ( !TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 ) );
2454 fail_state = TRIE_TRANS_STATE( fail_state, fail_base, ucharcount, charid, 1 );
2455 fail[ ch_state ] = fail_state;
2456 if ( !aho->states[ ch_state ].wordnum && aho->states[ fail_state ].wordnum )
2458 aho->states[ ch_state ].wordnum = aho->states[ fail_state ].wordnum;
2460 q[ q_write++ % numstates] = ch_state;
2464 /* restore fail[0..1] to 0 so that we "fall out" of the AC loop
2465 when we fail in state 1, this allows us to use the
2466 charclass scan to find a valid start char. This is based on the principle
2467 that theres a good chance the string being searched contains lots of stuff
2468 that cant be a start char.
2470 fail[ 0 ] = fail[ 1 ] = 0;
2471 DEBUG_TRIE_COMPILE_r({
2472 PerlIO_printf(Perl_debug_log,
2473 "%*sStclass Failtable (%"UVuf" states): 0",
2474 (int)(depth * 2), "", (UV)numstates
2476 for( q_read=1; q_read<numstates; q_read++ ) {
2477 PerlIO_printf(Perl_debug_log, ", %"UVuf, (UV)fail[q_read]);
2479 PerlIO_printf(Perl_debug_log, "\n");
2482 /*RExC_seen |= REG_SEEN_TRIEDFA;*/
2487 * There are strange code-generation bugs caused on sparc64 by gcc-2.95.2.
2488 * These need to be revisited when a newer toolchain becomes available.
2490 #if defined(__sparc64__) && defined(__GNUC__)
2491 # if __GNUC__ < 2 || (__GNUC__ == 2 && __GNUC_MINOR__ < 96)
2492 # undef SPARC64_GCC_WORKAROUND
2493 # define SPARC64_GCC_WORKAROUND 1
2497 #define DEBUG_PEEP(str,scan,depth) \
2498 DEBUG_OPTIMISE_r({if (scan){ \
2499 SV * const mysv=sv_newmortal(); \
2500 regnode *Next = regnext(scan); \
2501 regprop(RExC_rx, mysv, scan); \
2502 PerlIO_printf(Perl_debug_log, "%*s" str ">%3d: %s (%d)\n", \
2503 (int)depth*2, "", REG_NODE_NUM(scan), SvPV_nolen_const(mysv),\
2504 Next ? (REG_NODE_NUM(Next)) : 0 ); \
2508 /* The below joins as many adjacent EXACTish nodes as possible into a single
2509 * one, and looks for problematic sequences of characters whose folds vs.
2510 * non-folds have sufficiently different lengths, that the optimizer would be
2511 * fooled into rejecting legitimate matches of them, and the trie construction
2512 * code can't cope with them. The joining is only done if:
2513 * 1) there is room in the current conglomerated node to entirely contain the
2515 * 2) they are the exact same node type
2517 * The adjacent nodes actually may be separated by NOTHING kind nodes, and
2518 * these get optimized out
2520 * If there are problematic code sequences, *min_subtract is set to the delta
2521 * that the minimum size of the node can be less than its actual size. And,
2522 * the node type of the result is changed to reflect that it contains these
2525 * And *has_exactf_sharp_s is set to indicate whether or not the node is EXACTF
2526 * and contains LATIN SMALL LETTER SHARP S
2528 * This is as good a place as any to discuss the design of handling these
2529 * problematic sequences. It's been wrong in Perl for a very long time. There
2530 * are three code points in Unicode whose folded lengths differ so much from
2531 * the un-folded lengths that it causes problems for the optimizer and trie
2532 * construction. Why only these are problematic, and not others where lengths
2533 * also differ is something I (khw) do not understand. New versions of Unicode
2534 * might add more such code points. Hopefully the logic in fold_grind.t that
2535 * figures out what to test (in part by verifying that each size-combination
2536 * gets tested) will catch any that do come along, so they can be added to the
2537 * special handling below. The chances of new ones are actually rather small,
2538 * as most, if not all, of the world's scripts that have casefolding have
2539 * already been encoded by Unicode. Also, a number of Unicode's decisions were
2540 * made to allow compatibility with pre-existing standards, and almost all of
2541 * those have already been dealt with. These would otherwise be the most
2542 * likely candidates for generating further tricky sequences. In other words,
2543 * Unicode by itself is unlikely to add new ones unless it is for compatibility
2544 * with pre-existing standards, and there aren't many of those left.
2546 * The previous designs for dealing with these involved assigning a special
2547 * node for them. This approach doesn't work, as evidenced by this example:
2548 * "\xDFs" =~ /s\xDF/ui # Used to fail before these patches
2549 * Both these fold to "sss", but if the pattern is parsed to create a node of
2550 * that would match just the \xDF, it won't be able to handle the case where a
2551 * successful match would have to cross the node's boundary. The new approach
2552 * that hopefully generally solves the problem generates an EXACTFU_SS node
2555 * There are a number of components to the approach (a lot of work for just
2556 * three code points!):
2557 * 1) This routine examines each EXACTFish node that could contain the
2558 * problematic sequences. It returns in *min_subtract how much to
2559 * subtract from the the actual length of the string to get a real minimum
2560 * for one that could match it. This number is usually 0 except for the
2561 * problematic sequences. This delta is used by the caller to adjust the
2562 * min length of the match, and the delta between min and max, so that the
2563 * optimizer doesn't reject these possibilities based on size constraints.
2564 * 2) These sequences are not currently correctly handled by the trie code
2565 * either, so it changes the joined node type to ops that are not handled
2566 * by trie's, those new ops being EXACTFU_SS and EXACTFU_NO_TRIE.
2567 * 3) This is sufficient for the two Greek sequences (described below), but
2568 * the one involving the Sharp s (\xDF) needs more. The node type
2569 * EXACTFU_SS is used for an EXACTFU node that contains at least one "ss"
2570 * sequence in it. For non-UTF-8 patterns and strings, this is the only
2571 * case where there is a possible fold length change. That means that a
2572 * regular EXACTFU node without UTF-8 involvement doesn't have to concern
2573 * itself with length changes, and so can be processed faster. regexec.c
2574 * takes advantage of this. Generally, an EXACTFish node that is in UTF-8
2575 * is pre-folded by regcomp.c. This saves effort in regex matching.
2576 * However, probably mostly for historical reasons, the pre-folding isn't
2577 * done for non-UTF8 patterns (and it can't be for EXACTF and EXACTFL
2578 * nodes, as what they fold to isn't known until runtime.) The fold
2579 * possibilities for the non-UTF8 patterns are quite simple, except for
2580 * the sharp s. All the ones that don't involve a UTF-8 target string
2581 * are members of a fold-pair, and arrays are set up for all of them
2582 * that quickly find the other member of the pair. It might actually
2583 * be faster to pre-fold these, but it isn't currently done, except for
2584 * the sharp s. Code elsewhere in this file makes sure that it gets
2585 * folded to 'ss', even if the pattern isn't UTF-8. This avoids the
2586 * issues described in the next item.
2587 * 4) A problem remains for the sharp s in EXACTF nodes. Whether it matches
2588 * 'ss' or not is not knowable at compile time. It will match iff the
2589 * target string is in UTF-8, unlike the EXACTFU nodes, where it always
2590 * matches; and the EXACTFL and EXACTFA nodes where it never does. Thus
2591 * it can't be folded to "ss" at compile time, unlike EXACTFU does as
2592 * described in item 3). An assumption that the optimizer part of
2593 * regexec.c (probably unwittingly) makes is that a character in the
2594 * pattern corresponds to at most a single character in the target string.
2595 * (And I do mean character, and not byte here, unlike other parts of the
2596 * documentation that have never been updated to account for multibyte
2597 * Unicode.) This assumption is wrong only in this case, as all other
2598 * cases are either 1-1 folds when no UTF-8 is involved; or is true by
2599 * virtue of having this file pre-fold UTF-8 patterns. I'm
2600 * reluctant to try to change this assumption, so instead the code punts.
2601 * This routine examines EXACTF nodes for the sharp s, and returns a
2602 * boolean indicating whether or not the node is an EXACTF node that
2603 * contains a sharp s. When it is true, the caller sets a flag that later
2604 * causes the optimizer in this file to not set values for the floating
2605 * and fixed string lengths, and thus avoids the optimizer code in
2606 * regexec.c that makes the invalid assumption. Thus, there is no
2607 * optimization based on string lengths for EXACTF nodes that contain the
2608 * sharp s. This only happens for /id rules (which means the pattern
2612 #define JOIN_EXACT(scan,min_subtract,has_exactf_sharp_s, flags) \
2613 if (PL_regkind[OP(scan)] == EXACT) \
2614 join_exact(pRExC_state,(scan),(min_subtract),has_exactf_sharp_s, (flags),NULL,depth+1)
2617 S_join_exact(pTHX_ RExC_state_t *pRExC_state, regnode *scan, UV *min_subtract, bool *has_exactf_sharp_s, U32 flags,regnode *val, U32 depth) {
2618 /* Merge several consecutive EXACTish nodes into one. */
2619 regnode *n = regnext(scan);
2621 regnode *next = scan + NODE_SZ_STR(scan);
2625 regnode *stop = scan;
2626 GET_RE_DEBUG_FLAGS_DECL;
2628 PERL_UNUSED_ARG(depth);
2631 PERL_ARGS_ASSERT_JOIN_EXACT;
2632 #ifndef EXPERIMENTAL_INPLACESCAN
2633 PERL_UNUSED_ARG(flags);
2634 PERL_UNUSED_ARG(val);
2636 DEBUG_PEEP("join",scan,depth);
2638 /* Look through the subsequent nodes in the chain. Skip NOTHING, merge
2639 * EXACT ones that are mergeable to the current one. */
2641 && (PL_regkind[OP(n)] == NOTHING
2642 || (stringok && OP(n) == OP(scan)))
2644 && NEXT_OFF(scan) + NEXT_OFF(n) < I16_MAX)
2647 if (OP(n) == TAIL || n > next)
2649 if (PL_regkind[OP(n)] == NOTHING) {
2650 DEBUG_PEEP("skip:",n,depth);
2651 NEXT_OFF(scan) += NEXT_OFF(n);
2652 next = n + NODE_STEP_REGNODE;
2659 else if (stringok) {
2660 const unsigned int oldl = STR_LEN(scan);
2661 regnode * const nnext = regnext(n);
2663 if (oldl + STR_LEN(n) > U8_MAX)
2666 DEBUG_PEEP("merg",n,depth);
2669 NEXT_OFF(scan) += NEXT_OFF(n);
2670 STR_LEN(scan) += STR_LEN(n);
2671 next = n + NODE_SZ_STR(n);
2672 /* Now we can overwrite *n : */
2673 Move(STRING(n), STRING(scan) + oldl, STR_LEN(n), char);
2681 #ifdef EXPERIMENTAL_INPLACESCAN
2682 if (flags && !NEXT_OFF(n)) {
2683 DEBUG_PEEP("atch", val, depth);
2684 if (reg_off_by_arg[OP(n)]) {
2685 ARG_SET(n, val - n);
2688 NEXT_OFF(n) = val - n;
2696 *has_exactf_sharp_s = FALSE;
2698 /* Here, all the adjacent mergeable EXACTish nodes have been merged. We
2699 * can now analyze for sequences of problematic code points. (Prior to
2700 * this final joining, sequences could have been split over boundaries, and
2701 * hence missed). The sequences only happen in folding, hence for any
2702 * non-EXACT EXACTish node */
2703 if (OP(scan) != EXACT) {
2705 U8 * s0 = (U8*) STRING(scan);
2706 U8 * const s_end = s0 + STR_LEN(scan);
2708 /* The below is perhaps overboard, but this allows us to save a test
2709 * each time through the loop at the expense of a mask. This is
2710 * because on both EBCDIC and ASCII machines, 'S' and 's' differ by a
2711 * single bit. On ASCII they are 32 apart; on EBCDIC, they are 64.
2712 * This uses an exclusive 'or' to find that bit and then inverts it to
2713 * form a mask, with just a single 0, in the bit position where 'S' and
2715 const U8 S_or_s_mask = ~ ('S' ^ 's');
2716 const U8 s_masked = 's' & S_or_s_mask;
2718 /* One pass is made over the node's string looking for all the
2719 * possibilities. to avoid some tests in the loop, there are two main
2720 * cases, for UTF-8 patterns (which can't have EXACTF nodes) and
2724 /* There are two problematic Greek code points in Unicode
2727 * U+0390 - GREEK SMALL LETTER IOTA WITH DIALYTIKA AND TONOS
2728 * U+03B0 - GREEK SMALL LETTER UPSILON WITH DIALYTIKA AND TONOS
2734 * U+03B9 U+0308 U+0301 0xCE 0xB9 0xCC 0x88 0xCC 0x81
2735 * U+03C5 U+0308 U+0301 0xCF 0x85 0xCC 0x88 0xCC 0x81
2737 * This means that in case-insensitive matching (or "loose
2738 * matching", as Unicode calls it), an EXACTF of length six (the
2739 * UTF-8 encoded byte length of the above casefolded versions) can
2740 * match a target string of length two (the byte length of UTF-8
2741 * encoded U+0390 or U+03B0). This would rather mess up the
2742 * minimum length computation. (there are other code points that
2743 * also fold to these two sequences, but the delta is smaller)
2745 * If these sequences are found, the minimum length is decreased by
2746 * four (six minus two).
2748 * Similarly, 'ss' may match the single char and byte LATIN SMALL
2749 * LETTER SHARP S. We decrease the min length by 1 for each
2750 * occurrence of 'ss' found */
2752 #ifdef EBCDIC /* RD tunifold greek 0390 and 03B0 */
2753 # define U390_first_byte 0xb4
2754 const U8 U390_tail[] = "\x68\xaf\x49\xaf\x42";
2755 # define U3B0_first_byte 0xb5
2756 const U8 U3B0_tail[] = "\x46\xaf\x49\xaf\x42";
2758 # define U390_first_byte 0xce
2759 const U8 U390_tail[] = "\xb9\xcc\x88\xcc\x81";
2760 # define U3B0_first_byte 0xcf
2761 const U8 U3B0_tail[] = "\x85\xcc\x88\xcc\x81";
2763 const U8 len = sizeof(U390_tail); /* (-1 for NUL; +1 for 1st byte;
2764 yields a net of 0 */
2765 /* Examine the string for one of the problematic sequences */
2767 s < s_end - 1; /* Can stop 1 before the end, as minimum length
2768 * sequence we are looking for is 2 */
2772 /* Look for the first byte in each problematic sequence */
2774 /* We don't have to worry about other things that fold to
2775 * 's' (such as the long s, U+017F), as all above-latin1
2776 * code points have been pre-folded */
2780 /* Current character is an 's' or 'S'. If next one is
2781 * as well, we have the dreaded sequence */
2782 if (((*(s+1) & S_or_s_mask) == s_masked)
2783 /* These two node types don't have special handling
2785 && OP(scan) != EXACTFL && OP(scan) != EXACTFA)
2788 OP(scan) = EXACTFU_SS;
2789 s++; /* No need to look at this character again */
2793 case U390_first_byte:
2794 if (s_end - s >= len
2796 /* The 1's are because are skipping comparing the
2798 && memEQ(s + 1, U390_tail, len - 1))
2800 goto greek_sequence;
2804 case U3B0_first_byte:
2805 if (! (s_end - s >= len
2806 && memEQ(s + 1, U3B0_tail, len - 1)))
2813 /* This can't currently be handled by trie's, so change
2814 * the node type to indicate this. If EXACTFA and
2815 * EXACTFL were ever to be handled by trie's, this
2816 * would have to be changed. If this node has already
2817 * been changed to EXACTFU_SS in this loop, leave it as
2818 * is. (I (khw) think it doesn't matter in regexec.c
2819 * for UTF patterns, but no need to change it */
2820 if (OP(scan) == EXACTFU) {
2821 OP(scan) = EXACTFU_NO_TRIE;
2823 s += 6; /* We already know what this sequence is. Skip
2829 else if (OP(scan) != EXACTFL && OP(scan) != EXACTFA) {
2831 /* Here, the pattern is not UTF-8. We need to look only for the
2832 * 'ss' sequence, and in the EXACTF case, the sharp s, which can be
2833 * in the final position. Otherwise we can stop looking 1 byte
2834 * earlier because have to find both the first and second 's' */
2835 const U8* upper = (OP(scan) == EXACTF) ? s_end : s_end -1;
2837 for (s = s0; s < upper; s++) {
2842 && ((*(s+1) & S_or_s_mask) == s_masked))
2846 /* EXACTF nodes need to know that the minimum
2847 * length changed so that a sharp s in the string
2848 * can match this ss in the pattern, but they
2849 * remain EXACTF nodes, as they are not trie'able,
2850 * so don't have to invent a new node type to
2851 * exclude them from the trie code */
2852 if (OP(scan) != EXACTF) {
2853 OP(scan) = EXACTFU_SS;
2858 case LATIN_SMALL_LETTER_SHARP_S:
2859 if (OP(scan) == EXACTF) {
2860 *has_exactf_sharp_s = TRUE;
2869 /* Allow dumping but overwriting the collection of skipped
2870 * ops and/or strings with fake optimized ops */
2871 n = scan + NODE_SZ_STR(scan);
2879 DEBUG_OPTIMISE_r(if (merged){DEBUG_PEEP("finl",scan,depth)});
2883 /* REx optimizer. Converts nodes into quicker variants "in place".
2884 Finds fixed substrings. */
2886 /* Stops at toplevel WHILEM as well as at "last". At end *scanp is set
2887 to the position after last scanned or to NULL. */
2889 #define INIT_AND_WITHP \
2890 assert(!and_withp); \
2891 Newx(and_withp,1,struct regnode_charclass_class); \
2892 SAVEFREEPV(and_withp)
2894 /* this is a chain of data about sub patterns we are processing that
2895 need to be handled separately/specially in study_chunk. Its so
2896 we can simulate recursion without losing state. */
2898 typedef struct scan_frame {
2899 regnode *last; /* last node to process in this frame */
2900 regnode *next; /* next node to process when last is reached */
2901 struct scan_frame *prev; /*previous frame*/
2902 I32 stop; /* what stopparen do we use */
2906 #define SCAN_COMMIT(s, data, m) scan_commit(s, data, m, is_inf)
2908 #define CASE_SYNST_FNC(nAmE) \
2910 if (flags & SCF_DO_STCLASS_AND) { \
2911 for (value = 0; value < 256; value++) \
2912 if (!is_ ## nAmE ## _cp(value)) \
2913 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2916 for (value = 0; value < 256; value++) \
2917 if (is_ ## nAmE ## _cp(value)) \
2918 ANYOF_BITMAP_SET(data->start_class, value); \
2922 if (flags & SCF_DO_STCLASS_AND) { \
2923 for (value = 0; value < 256; value++) \
2924 if (is_ ## nAmE ## _cp(value)) \
2925 ANYOF_BITMAP_CLEAR(data->start_class, value); \
2928 for (value = 0; value < 256; value++) \
2929 if (!is_ ## nAmE ## _cp(value)) \
2930 ANYOF_BITMAP_SET(data->start_class, value); \
2937 S_study_chunk(pTHX_ RExC_state_t *pRExC_state, regnode **scanp,
2938 I32 *minlenp, I32 *deltap,
2943 struct regnode_charclass_class *and_withp,
2944 U32 flags, U32 depth)
2945 /* scanp: Start here (read-write). */
2946 /* deltap: Write maxlen-minlen here. */
2947 /* last: Stop before this one. */
2948 /* data: string data about the pattern */
2949 /* stopparen: treat close N as END */
2950 /* recursed: which subroutines have we recursed into */
2951 /* and_withp: Valid if flags & SCF_DO_STCLASS_OR */
2954 I32 min = 0, pars = 0, code;
2955 regnode *scan = *scanp, *next;
2957 int is_inf = (flags & SCF_DO_SUBSTR) && (data->flags & SF_IS_INF);
2958 int is_inf_internal = 0; /* The studied chunk is infinite */
2959 I32 is_par = OP(scan) == OPEN ? ARG(scan) : 0;
2960 scan_data_t data_fake;
2961 SV *re_trie_maxbuff = NULL;
2962 regnode *first_non_open = scan;
2963 I32 stopmin = I32_MAX;
2964 scan_frame *frame = NULL;
2965 GET_RE_DEBUG_FLAGS_DECL;
2967 PERL_ARGS_ASSERT_STUDY_CHUNK;
2970 StructCopy(&zero_scan_data, &data_fake, scan_data_t);
2974 while (first_non_open && OP(first_non_open) == OPEN)
2975 first_non_open=regnext(first_non_open);
2980 while ( scan && OP(scan) != END && scan < last ){
2981 UV min_subtract = 0; /* How much to subtract from the minimum node
2982 length to get a real minimum (because the
2983 folded version may be shorter) */
2984 bool has_exactf_sharp_s = FALSE;
2985 /* Peephole optimizer: */
2986 DEBUG_STUDYDATA("Peep:", data,depth);
2987 DEBUG_PEEP("Peep",scan,depth);
2989 /* Its not clear to khw or hv why this is done here, and not in the
2990 * clauses that deal with EXACT nodes. khw's guess is that it's
2991 * because of a previous design */
2992 JOIN_EXACT(scan,&min_subtract, &has_exactf_sharp_s, 0);
2994 /* Follow the next-chain of the current node and optimize
2995 away all the NOTHINGs from it. */
2996 if (OP(scan) != CURLYX) {
2997 const int max = (reg_off_by_arg[OP(scan)]
2999 /* I32 may be smaller than U16 on CRAYs! */
3000 : (I32_MAX < U16_MAX ? I32_MAX : U16_MAX));
3001 int off = (reg_off_by_arg[OP(scan)] ? ARG(scan) : NEXT_OFF(scan));
3005 /* Skip NOTHING and LONGJMP. */
3006 while ((n = regnext(n))
3007 && ((PL_regkind[OP(n)] == NOTHING && (noff = NEXT_OFF(n)))
3008 || ((OP(n) == LONGJMP) && (noff = ARG(n))))
3009 && off + noff < max)
3011 if (reg_off_by_arg[OP(scan)])
3014 NEXT_OFF(scan) = off;
3019 /* The principal pseudo-switch. Cannot be a switch, since we
3020 look into several different things. */
3021 if (OP(scan) == BRANCH || OP(scan) == BRANCHJ
3022 || OP(scan) == IFTHEN) {
3023 next = regnext(scan);
3025 /* demq: the op(next)==code check is to see if we have "branch-branch" AFAICT */
3027 if (OP(next) == code || code == IFTHEN) {
3028 /* NOTE - There is similar code to this block below for handling
3029 TRIE nodes on a re-study. If you change stuff here check there
3031 I32 max1 = 0, min1 = I32_MAX, num = 0;
3032 struct regnode_charclass_class accum;
3033 regnode * const startbranch=scan;
3035 if (flags & SCF_DO_SUBSTR)
3036 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot merge strings after this. */
3037 if (flags & SCF_DO_STCLASS)
3038 cl_init_zero(pRExC_state, &accum);
3040 while (OP(scan) == code) {
3041 I32 deltanext, minnext, f = 0, fake;
3042 struct regnode_charclass_class this_class;
3045 data_fake.flags = 0;
3047 data_fake.whilem_c = data->whilem_c;
3048 data_fake.last_closep = data->last_closep;
3051 data_fake.last_closep = &fake;
3053 data_fake.pos_delta = delta;
3054 next = regnext(scan);
3055 scan = NEXTOPER(scan);
3057 scan = NEXTOPER(scan);
3058 if (flags & SCF_DO_STCLASS) {
3059 cl_init(pRExC_state, &this_class);
3060 data_fake.start_class = &this_class;
3061 f = SCF_DO_STCLASS_AND;
3063 if (flags & SCF_WHILEM_VISITED_POS)
3064 f |= SCF_WHILEM_VISITED_POS;
3066 /* we suppose the run is continuous, last=next...*/
3067 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
3069 stopparen, recursed, NULL, f,depth+1);
3072 if (max1 < minnext + deltanext)
3073 max1 = minnext + deltanext;
3074 if (deltanext == I32_MAX)
3075 is_inf = is_inf_internal = 1;
3077 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
3079 if (data_fake.flags & SCF_SEEN_ACCEPT) {
3080 if ( stopmin > minnext)
3081 stopmin = min + min1;
3082 flags &= ~SCF_DO_SUBSTR;
3084 data->flags |= SCF_SEEN_ACCEPT;
3087 if (data_fake.flags & SF_HAS_EVAL)
3088 data->flags |= SF_HAS_EVAL;
3089 data->whilem_c = data_fake.whilem_c;
3091 if (flags & SCF_DO_STCLASS)
3092 cl_or(pRExC_state, &accum, &this_class);
3094 if (code == IFTHEN && num < 2) /* Empty ELSE branch */
3096 if (flags & SCF_DO_SUBSTR) {
3097 data->pos_min += min1;
3098 data->pos_delta += max1 - min1;
3099 if (max1 != min1 || is_inf)
3100 data->longest = &(data->longest_float);
3103 delta += max1 - min1;
3104 if (flags & SCF_DO_STCLASS_OR) {
3105 cl_or(pRExC_state, data->start_class, &accum);
3107 cl_and(data->start_class, and_withp);
3108 flags &= ~SCF_DO_STCLASS;
3111 else if (flags & SCF_DO_STCLASS_AND) {
3113 cl_and(data->start_class, &accum);
3114 flags &= ~SCF_DO_STCLASS;
3117 /* Switch to OR mode: cache the old value of
3118 * data->start_class */
3120 StructCopy(data->start_class, and_withp,
3121 struct regnode_charclass_class);
3122 flags &= ~SCF_DO_STCLASS_AND;
3123 StructCopy(&accum, data->start_class,
3124 struct regnode_charclass_class);
3125 flags |= SCF_DO_STCLASS_OR;
3126 data->start_class->flags |= ANYOF_EOS;
3130 if (PERL_ENABLE_TRIE_OPTIMISATION && OP( startbranch ) == BRANCH ) {
3133 Assuming this was/is a branch we are dealing with: 'scan' now
3134 points at the item that follows the branch sequence, whatever
3135 it is. We now start at the beginning of the sequence and look
3142 which would be constructed from a pattern like /A|LIST|OF|WORDS/
3144 If we can find such a subsequence we need to turn the first
3145 element into a trie and then add the subsequent branch exact
3146 strings to the trie.
3150 1. patterns where the whole set of branches can be converted.
3152 2. patterns where only a subset can be converted.
3154 In case 1 we can replace the whole set with a single regop
3155 for the trie. In case 2 we need to keep the start and end
3158 'BRANCH EXACT; BRANCH EXACT; BRANCH X'
3159 becomes BRANCH TRIE; BRANCH X;
3161 There is an additional case, that being where there is a
3162 common prefix, which gets split out into an EXACT like node
3163 preceding the TRIE node.
3165 If x(1..n)==tail then we can do a simple trie, if not we make
3166 a "jump" trie, such that when we match the appropriate word
3167 we "jump" to the appropriate tail node. Essentially we turn
3168 a nested if into a case structure of sorts.
3173 if (!re_trie_maxbuff) {
3174 re_trie_maxbuff = get_sv(RE_TRIE_MAXBUF_NAME, 1);
3175 if (!SvIOK(re_trie_maxbuff))
3176 sv_setiv(re_trie_maxbuff, RE_TRIE_MAXBUF_INIT);
3178 if ( SvIV(re_trie_maxbuff)>=0 ) {
3180 regnode *first = (regnode *)NULL;
3181 regnode *last = (regnode *)NULL;
3182 regnode *tail = scan;
3187 SV * const mysv = sv_newmortal(); /* for dumping */
3189 /* var tail is used because there may be a TAIL
3190 regop in the way. Ie, the exacts will point to the
3191 thing following the TAIL, but the last branch will
3192 point at the TAIL. So we advance tail. If we
3193 have nested (?:) we may have to move through several
3197 while ( OP( tail ) == TAIL ) {
3198 /* this is the TAIL generated by (?:) */
3199 tail = regnext( tail );
3204 regprop(RExC_rx, mysv, tail );
3205 PerlIO_printf( Perl_debug_log, "%*s%s%s\n",
3206 (int)depth * 2 + 2, "",
3207 "Looking for TRIE'able sequences. Tail node is: ",
3208 SvPV_nolen_const( mysv )
3214 step through the branches, cur represents each
3215 branch, noper is the first thing to be matched
3216 as part of that branch and noper_next is the
3217 regnext() of that node. if noper is an EXACT
3218 and noper_next is the same as scan (our current
3219 position in the regex) then the EXACT branch is
3220 a possible optimization target. Once we have
3221 two or more consecutive such branches we can
3222 create a trie of the EXACT's contents and stich
3223 it in place. If the sequence represents all of
3224 the branches we eliminate the whole thing and
3225 replace it with a single TRIE. If it is a
3226 subsequence then we need to stitch it in. This
3227 means the first branch has to remain, and needs
3228 to be repointed at the item on the branch chain
3229 following the last branch optimized. This could
3230 be either a BRANCH, in which case the
3231 subsequence is internal, or it could be the
3232 item following the branch sequence in which
3233 case the subsequence is at the end.
3237 /* dont use tail as the end marker for this traverse */
3238 for ( cur = startbranch ; cur != scan ; cur = regnext( cur ) ) {
3239 regnode * const noper = NEXTOPER( cur );
3240 #if defined(DEBUGGING) || defined(NOJUMPTRIE)
3241 regnode * const noper_next = regnext( noper );
3245 regprop(RExC_rx, mysv, cur);
3246 PerlIO_printf( Perl_debug_log, "%*s- %s (%d)",
3247 (int)depth * 2 + 2,"", SvPV_nolen_const( mysv ), REG_NODE_NUM(cur) );
3249 regprop(RExC_rx, mysv, noper);
3250 PerlIO_printf( Perl_debug_log, " -> %s",
3251 SvPV_nolen_const(mysv));
3254 regprop(RExC_rx, mysv, noper_next );
3255 PerlIO_printf( Perl_debug_log,"\t=> %s\t",
3256 SvPV_nolen_const(mysv));
3258 PerlIO_printf( Perl_debug_log, "(First==%d,Last==%d,Cur==%d)\n",
3259 REG_NODE_NUM(first), REG_NODE_NUM(last), REG_NODE_NUM(cur) );
3261 if ( (((first && optype!=NOTHING) ? OP( noper ) == optype
3262 : PL_regkind[ OP( noper ) ] == EXACT )
3263 || OP(noper) == NOTHING )
3265 && noper_next == tail
3270 if ( !first || optype == NOTHING ) {
3271 if (!first) first = cur;
3272 optype = OP( noper );
3278 Currently the trie logic handles case insensitive matching properly only
3279 when the pattern is UTF-8 and the node is EXACTFU (thus forcing unicode
3282 If/when this is fixed the following define can be swapped
3283 in below to fully enable trie logic.
3285 #define TRIE_TYPE_IS_SAFE 1
3287 Note that join_exact() assumes that the other types of EXACTFish nodes are not
3288 used in tries, so that would have to be updated if this changed
3291 #define TRIE_TYPE_IS_SAFE ((UTF && optype == EXACTFU) || optype==EXACT)
3293 if ( last && TRIE_TYPE_IS_SAFE ) {
3294 make_trie( pRExC_state,
3295 startbranch, first, cur, tail, count,
3298 if ( PL_regkind[ OP( noper ) ] == EXACT
3300 && noper_next == tail
3305 optype = OP( noper );
3315 regprop(RExC_rx, mysv, cur);
3316 PerlIO_printf( Perl_debug_log,
3317 "%*s- %s (%d) <SCAN FINISHED>\n", (int)depth * 2 + 2,
3318 "", SvPV_nolen_const( mysv ),REG_NODE_NUM(cur));
3322 if ( last && TRIE_TYPE_IS_SAFE ) {
3323 made= make_trie( pRExC_state, startbranch, first, scan, tail, count, optype, depth+1 );
3324 #ifdef TRIE_STUDY_OPT
3325 if ( ((made == MADE_EXACT_TRIE &&
3326 startbranch == first)
3327 || ( first_non_open == first )) &&
3329 flags |= SCF_TRIE_RESTUDY;
3330 if ( startbranch == first
3333 RExC_seen &=~REG_TOP_LEVEL_BRANCHES;
3343 else if ( code == BRANCHJ ) { /* single branch is optimized. */
3344 scan = NEXTOPER(NEXTOPER(scan));
3345 } else /* single branch is optimized. */
3346 scan = NEXTOPER(scan);
3348 } else if (OP(scan) == SUSPEND || OP(scan) == GOSUB || OP(scan) == GOSTART) {
3349 scan_frame *newframe = NULL;
3354 if (OP(scan) != SUSPEND) {
3355 /* set the pointer */
3356 if (OP(scan) == GOSUB) {
3358 RExC_recurse[ARG2L(scan)] = scan;
3359 start = RExC_open_parens[paren-1];
3360 end = RExC_close_parens[paren-1];
3363 start = RExC_rxi->program + 1;
3367 Newxz(recursed, (((RExC_npar)>>3) +1), U8);
3368 SAVEFREEPV(recursed);
3370 if (!PAREN_TEST(recursed,paren+1)) {
3371 PAREN_SET(recursed,paren+1);
3372 Newx(newframe,1,scan_frame);
3374 if (flags & SCF_DO_SUBSTR) {
3375 SCAN_COMMIT(pRExC_state,data,minlenp);
3376 data->longest = &(data->longest_float);
3378 is_inf = is_inf_internal = 1;
3379 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
3380 cl_anything(pRExC_state, data->start_class);
3381 flags &= ~SCF_DO_STCLASS;
3384 Newx(newframe,1,scan_frame);
3387 end = regnext(scan);
3392 SAVEFREEPV(newframe);
3393 newframe->next = regnext(scan);
3394 newframe->last = last;
3395 newframe->stop = stopparen;
3396 newframe->prev = frame;
3406 else if (OP(scan) == EXACT) {
3407 I32 l = STR_LEN(scan);
3410 const U8 * const s = (U8*)STRING(scan);
3411 l = utf8_length(s, s + l);
3412 uc = utf8_to_uvchr(s, NULL);
3414 uc = *((U8*)STRING(scan));
3417 if (flags & SCF_DO_SUBSTR) { /* Update longest substr. */
3418 /* The code below prefers earlier match for fixed
3419 offset, later match for variable offset. */
3420 if (data->last_end == -1) { /* Update the start info. */
3421 data->last_start_min = data->pos_min;
3422 data->last_start_max = is_inf
3423 ? I32_MAX : data->pos_min + data->pos_delta;
3425 sv_catpvn(data->last_found, STRING(scan), STR_LEN(scan));
3427 SvUTF8_on(data->last_found);
3429 SV * const sv = data->last_found;
3430 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3431 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3432 if (mg && mg->mg_len >= 0)
3433 mg->mg_len += utf8_length((U8*)STRING(scan),
3434 (U8*)STRING(scan)+STR_LEN(scan));
3436 data->last_end = data->pos_min + l;
3437 data->pos_min += l; /* As in the first entry. */
3438 data->flags &= ~SF_BEFORE_EOL;
3440 if (flags & SCF_DO_STCLASS_AND) {
3441 /* Check whether it is compatible with what we know already! */
3445 /* If compatible, we or it in below. It is compatible if is
3446 * in the bitmp and either 1) its bit or its fold is set, or 2)
3447 * it's for a locale. Even if there isn't unicode semantics
3448 * here, at runtime there may be because of matching against a
3449 * utf8 string, so accept a possible false positive for
3450 * latin1-range folds */
3452 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3453 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3454 && (!(data->start_class->flags & ANYOF_LOC_NONBITMAP_FOLD)
3455 || !ANYOF_BITMAP_TEST(data->start_class, PL_fold_latin1[uc])))
3460 ANYOF_CLASS_ZERO(data->start_class);
3461 ANYOF_BITMAP_ZERO(data->start_class);
3463 ANYOF_BITMAP_SET(data->start_class, uc);
3464 else if (uc >= 0x100) {
3467 /* Some Unicode code points fold to the Latin1 range; as
3468 * XXX temporary code, instead of figuring out if this is
3469 * one, just assume it is and set all the start class bits
3470 * that could be some such above 255 code point's fold
3471 * which will generate fals positives. As the code
3472 * elsewhere that does compute the fold settles down, it
3473 * can be extracted out and re-used here */
3474 for (i = 0; i < 256; i++){
3475 if (_HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(i)) {
3476 ANYOF_BITMAP_SET(data->start_class, i);
3480 data->start_class->flags &= ~ANYOF_EOS;
3482 data->start_class->flags &= ~ANYOF_UNICODE_ALL;
3484 else if (flags & SCF_DO_STCLASS_OR) {
3485 /* false positive possible if the class is case-folded */
3487 ANYOF_BITMAP_SET(data->start_class, uc);
3489 data->start_class->flags |= ANYOF_UNICODE_ALL;
3490 data->start_class->flags &= ~ANYOF_EOS;
3491 cl_and(data->start_class, and_withp);
3493 flags &= ~SCF_DO_STCLASS;
3495 else if (PL_regkind[OP(scan)] == EXACT) { /* But OP != EXACT! */
3496 I32 l = STR_LEN(scan);
3497 UV uc = *((U8*)STRING(scan));
3499 /* Search for fixed substrings supports EXACT only. */
3500 if (flags & SCF_DO_SUBSTR) {
3502 SCAN_COMMIT(pRExC_state, data, minlenp);
3505 const U8 * const s = (U8 *)STRING(scan);
3506 l = utf8_length(s, s + l);
3507 uc = utf8_to_uvchr(s, NULL);
3509 else if (has_exactf_sharp_s) {
3510 RExC_seen |= REG_SEEN_EXACTF_SHARP_S;
3512 min += l - min_subtract;
3516 delta += min_subtract;
3517 if (flags & SCF_DO_SUBSTR) {
3518 data->pos_min += l - min_subtract;
3519 if (data->pos_min < 0) {
3522 data->pos_delta += min_subtract;
3524 data->longest = &(data->longest_float);
3527 if (flags & SCF_DO_STCLASS_AND) {
3528 /* Check whether it is compatible with what we know already! */
3531 (!(data->start_class->flags & (ANYOF_CLASS | ANYOF_LOCALE))
3532 && !ANYOF_BITMAP_TEST(data->start_class, uc)
3533 && !ANYOF_BITMAP_TEST(data->start_class, PL_fold_latin1[uc])))
3537 ANYOF_CLASS_ZERO(data->start_class);
3538 ANYOF_BITMAP_ZERO(data->start_class);
3540 ANYOF_BITMAP_SET(data->start_class, uc);
3541 data->start_class->flags &= ~ANYOF_EOS;
3542 data->start_class->flags |= ANYOF_LOC_NONBITMAP_FOLD;
3543 if (OP(scan) == EXACTFL) {
3544 /* XXX This set is probably no longer necessary, and
3545 * probably wrong as LOCALE now is on in the initial
3547 data->start_class->flags |= ANYOF_LOCALE;
3551 /* Also set the other member of the fold pair. In case
3552 * that unicode semantics is called for at runtime, use
3553 * the full latin1 fold. (Can't do this for locale,
3554 * because not known until runtime) */
3555 ANYOF_BITMAP_SET(data->start_class, PL_fold_latin1[uc]);
3557 /* All other (EXACTFL handled above) folds except under
3558 * /iaa that include s, S, and sharp_s also may include
3560 if (OP(scan) != EXACTFA) {
3561 if (uc == 's' || uc == 'S') {
3562 ANYOF_BITMAP_SET(data->start_class,
3563 LATIN_SMALL_LETTER_SHARP_S);
3565 else if (uc == LATIN_SMALL_LETTER_SHARP_S) {
3566 ANYOF_BITMAP_SET(data->start_class, 's');
3567 ANYOF_BITMAP_SET(data->start_class, 'S');
3572 else if (uc >= 0x100) {
3574 for (i = 0; i < 256; i++){
3575 if (_HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(i)) {
3576 ANYOF_BITMAP_SET(data->start_class, i);
3581 else if (flags & SCF_DO_STCLASS_OR) {
3582 if (data->start_class->flags & ANYOF_LOC_NONBITMAP_FOLD) {
3583 /* false positive possible if the class is case-folded.
3584 Assume that the locale settings are the same... */
3586 ANYOF_BITMAP_SET(data->start_class, uc);
3587 if (OP(scan) != EXACTFL) {
3589 /* And set the other member of the fold pair, but
3590 * can't do that in locale because not known until
3592 ANYOF_BITMAP_SET(data->start_class,
3593 PL_fold_latin1[uc]);
3595 /* All folds except under /iaa that include s, S,
3596 * and sharp_s also may include the others */
3597 if (OP(scan) != EXACTFA) {
3598 if (uc == 's' || uc == 'S') {
3599 ANYOF_BITMAP_SET(data->start_class,
3600 LATIN_SMALL_LETTER_SHARP_S);
3602 else if (uc == LATIN_SMALL_LETTER_SHARP_S) {
3603 ANYOF_BITMAP_SET(data->start_class, 's');
3604 ANYOF_BITMAP_SET(data->start_class, 'S');
3609 data->start_class->flags &= ~ANYOF_EOS;
3611 cl_and(data->start_class, and_withp);
3613 flags &= ~SCF_DO_STCLASS;
3615 else if (REGNODE_VARIES(OP(scan))) {
3616 I32 mincount, maxcount, minnext, deltanext, fl = 0;
3617 I32 f = flags, pos_before = 0;
3618 regnode * const oscan = scan;
3619 struct regnode_charclass_class this_class;
3620 struct regnode_charclass_class *oclass = NULL;
3621 I32 next_is_eval = 0;
3623 switch (PL_regkind[OP(scan)]) {
3624 case WHILEM: /* End of (?:...)* . */
3625 scan = NEXTOPER(scan);
3628 if (flags & (SCF_DO_SUBSTR | SCF_DO_STCLASS)) {
3629 next = NEXTOPER(scan);
3630 if (OP(next) == EXACT || (flags & SCF_DO_STCLASS)) {
3632 maxcount = REG_INFTY;
3633 next = regnext(scan);
3634 scan = NEXTOPER(scan);
3638 if (flags & SCF_DO_SUBSTR)
3643 if (flags & SCF_DO_STCLASS) {
3645 maxcount = REG_INFTY;
3646 next = regnext(scan);
3647 scan = NEXTOPER(scan);
3650 is_inf = is_inf_internal = 1;
3651 scan = regnext(scan);
3652 if (flags & SCF_DO_SUBSTR) {
3653 SCAN_COMMIT(pRExC_state, data, minlenp); /* Cannot extend fixed substrings */
3654 data->longest = &(data->longest_float);
3656 goto optimize_curly_tail;
3658 if (stopparen>0 && (OP(scan)==CURLYN || OP(scan)==CURLYM)
3659 && (scan->flags == stopparen))
3664 mincount = ARG1(scan);
3665 maxcount = ARG2(scan);
3667 next = regnext(scan);
3668 if (OP(scan) == CURLYX) {
3669 I32 lp = (data ? *(data->last_closep) : 0);
3670 scan->flags = ((lp <= (I32)U8_MAX) ? (U8)lp : U8_MAX);
3672 scan = NEXTOPER(scan) + EXTRA_STEP_2ARGS;
3673 next_is_eval = (OP(scan) == EVAL);
3675 if (flags & SCF_DO_SUBSTR) {
3676 if (mincount == 0) SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot extend fixed substrings */
3677 pos_before = data->pos_min;
3681 data->flags &= ~(SF_HAS_PAR|SF_IN_PAR|SF_HAS_EVAL);
3683 data->flags |= SF_IS_INF;
3685 if (flags & SCF_DO_STCLASS) {
3686 cl_init(pRExC_state, &this_class);
3687 oclass = data->start_class;
3688 data->start_class = &this_class;
3689 f |= SCF_DO_STCLASS_AND;
3690 f &= ~SCF_DO_STCLASS_OR;
3692 /* Exclude from super-linear cache processing any {n,m}
3693 regops for which the combination of input pos and regex
3694 pos is not enough information to determine if a match
3697 For example, in the regex /foo(bar\s*){4,8}baz/ with the
3698 regex pos at the \s*, the prospects for a match depend not
3699 only on the input position but also on how many (bar\s*)
3700 repeats into the {4,8} we are. */
3701 if ((mincount > 1) || (maxcount > 1 && maxcount != REG_INFTY))
3702 f &= ~SCF_WHILEM_VISITED_POS;
3704 /* This will finish on WHILEM, setting scan, or on NULL: */
3705 minnext = study_chunk(pRExC_state, &scan, minlenp, &deltanext,
3706 last, data, stopparen, recursed, NULL,
3708 ? (f & ~SCF_DO_SUBSTR) : f),depth+1);
3710 if (flags & SCF_DO_STCLASS)
3711 data->start_class = oclass;
3712 if (mincount == 0 || minnext == 0) {
3713 if (flags & SCF_DO_STCLASS_OR) {
3714 cl_or(pRExC_state, data->start_class, &this_class);
3716 else if (flags & SCF_DO_STCLASS_AND) {
3717 /* Switch to OR mode: cache the old value of
3718 * data->start_class */
3720 StructCopy(data->start_class, and_withp,
3721 struct regnode_charclass_class);
3722 flags &= ~SCF_DO_STCLASS_AND;
3723 StructCopy(&this_class, data->start_class,
3724 struct regnode_charclass_class);
3725 flags |= SCF_DO_STCLASS_OR;
3726 data->start_class->flags |= ANYOF_EOS;
3728 } else { /* Non-zero len */
3729 if (flags & SCF_DO_STCLASS_OR) {
3730 cl_or(pRExC_state, data->start_class, &this_class);
3731 cl_and(data->start_class, and_withp);
3733 else if (flags & SCF_DO_STCLASS_AND)
3734 cl_and(data->start_class, &this_class);
3735 flags &= ~SCF_DO_STCLASS;
3737 if (!scan) /* It was not CURLYX, but CURLY. */
3739 if ( /* ? quantifier ok, except for (?{ ... }) */
3740 (next_is_eval || !(mincount == 0 && maxcount == 1))
3741 && (minnext == 0) && (deltanext == 0)
3742 && data && !(data->flags & (SF_HAS_PAR|SF_IN_PAR))
3743 && maxcount <= REG_INFTY/3) /* Complement check for big count */
3745 ckWARNreg(RExC_parse,
3746 "Quantifier unexpected on zero-length expression");
3749 min += minnext * mincount;
3750 is_inf_internal |= ((maxcount == REG_INFTY
3751 && (minnext + deltanext) > 0)
3752 || deltanext == I32_MAX);
3753 is_inf |= is_inf_internal;
3754 delta += (minnext + deltanext) * maxcount - minnext * mincount;
3756 /* Try powerful optimization CURLYX => CURLYN. */
3757 if ( OP(oscan) == CURLYX && data
3758 && data->flags & SF_IN_PAR
3759 && !(data->flags & SF_HAS_EVAL)
3760 && !deltanext && minnext == 1 ) {
3761 /* Try to optimize to CURLYN. */
3762 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS;
3763 regnode * const nxt1 = nxt;
3770 if (!REGNODE_SIMPLE(OP(nxt))
3771 && !(PL_regkind[OP(nxt)] == EXACT
3772 && STR_LEN(nxt) == 1))
3778 if (OP(nxt) != CLOSE)
3780 if (RExC_open_parens) {
3781 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3782 RExC_close_parens[ARG(nxt1)-1]=nxt+2; /*close->while*/
3784 /* Now we know that nxt2 is the only contents: */
3785 oscan->flags = (U8)ARG(nxt);
3787 OP(nxt1) = NOTHING; /* was OPEN. */
3790 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3791 NEXT_OFF(nxt1+ 1) = 0; /* just for consistency. */
3792 NEXT_OFF(nxt2) = 0; /* just for consistency with CURLY. */
3793 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3794 OP(nxt + 1) = OPTIMIZED; /* was count. */
3795 NEXT_OFF(nxt+ 1) = 0; /* just for consistency. */
3800 /* Try optimization CURLYX => CURLYM. */
3801 if ( OP(oscan) == CURLYX && data
3802 && !(data->flags & SF_HAS_PAR)
3803 && !(data->flags & SF_HAS_EVAL)
3804 && !deltanext /* atom is fixed width */
3805 && minnext != 0 /* CURLYM can't handle zero width */
3807 /* XXXX How to optimize if data == 0? */
3808 /* Optimize to a simpler form. */
3809 regnode *nxt = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN */
3813 while ( (nxt2 = regnext(nxt)) /* skip over embedded stuff*/
3814 && (OP(nxt2) != WHILEM))
3816 OP(nxt2) = SUCCEED; /* Whas WHILEM */
3817 /* Need to optimize away parenths. */
3818 if ((data->flags & SF_IN_PAR) && OP(nxt) == CLOSE) {
3819 /* Set the parenth number. */
3820 regnode *nxt1 = NEXTOPER(oscan) + EXTRA_STEP_2ARGS; /* OPEN*/
3822 oscan->flags = (U8)ARG(nxt);
3823 if (RExC_open_parens) {
3824 RExC_open_parens[ARG(nxt1)-1]=oscan; /*open->CURLYM*/
3825 RExC_close_parens[ARG(nxt1)-1]=nxt2+1; /*close->NOTHING*/
3827 OP(nxt1) = OPTIMIZED; /* was OPEN. */
3828 OP(nxt) = OPTIMIZED; /* was CLOSE. */
3831 OP(nxt1 + 1) = OPTIMIZED; /* was count. */
3832 OP(nxt + 1) = OPTIMIZED; /* was count. */
3833 NEXT_OFF(nxt1 + 1) = 0; /* just for consistency. */
3834 NEXT_OFF(nxt + 1) = 0; /* just for consistency. */
3837 while ( nxt1 && (OP(nxt1) != WHILEM)) {
3838 regnode *nnxt = regnext(nxt1);
3840 if (reg_off_by_arg[OP(nxt1)])
3841 ARG_SET(nxt1, nxt2 - nxt1);
3842 else if (nxt2 - nxt1 < U16_MAX)
3843 NEXT_OFF(nxt1) = nxt2 - nxt1;
3845 OP(nxt) = NOTHING; /* Cannot beautify */
3850 /* Optimize again: */
3851 study_chunk(pRExC_state, &nxt1, minlenp, &deltanext, nxt,
3852 NULL, stopparen, recursed, NULL, 0,depth+1);
3857 else if ((OP(oscan) == CURLYX)
3858 && (flags & SCF_WHILEM_VISITED_POS)
3859 /* See the comment on a similar expression above.
3860 However, this time it's not a subexpression
3861 we care about, but the expression itself. */
3862 && (maxcount == REG_INFTY)
3863 && data && ++data->whilem_c < 16) {
3864 /* This stays as CURLYX, we can put the count/of pair. */
3865 /* Find WHILEM (as in regexec.c) */
3866 regnode *nxt = oscan + NEXT_OFF(oscan);
3868 if (OP(PREVOPER(nxt)) == NOTHING) /* LONGJMP */
3870 PREVOPER(nxt)->flags = (U8)(data->whilem_c
3871 | (RExC_whilem_seen << 4)); /* On WHILEM */
3873 if (data && fl & (SF_HAS_PAR|SF_IN_PAR))
3875 if (flags & SCF_DO_SUBSTR) {
3876 SV *last_str = NULL;
3877 int counted = mincount != 0;
3879 if (data->last_end > 0 && mincount != 0) { /* Ends with a string. */
3880 #if defined(SPARC64_GCC_WORKAROUND)
3883 const char *s = NULL;
3886 if (pos_before >= data->last_start_min)
3889 b = data->last_start_min;
3892 s = SvPV_const(data->last_found, l);
3893 old = b - data->last_start_min;
3896 I32 b = pos_before >= data->last_start_min
3897 ? pos_before : data->last_start_min;
3899 const char * const s = SvPV_const(data->last_found, l);
3900 I32 old = b - data->last_start_min;
3904 old = utf8_hop((U8*)s, old) - (U8*)s;
3906 /* Get the added string: */
3907 last_str = newSVpvn_utf8(s + old, l, UTF);
3908 if (deltanext == 0 && pos_before == b) {
3909 /* What was added is a constant string */
3911 SvGROW(last_str, (mincount * l) + 1);
3912 repeatcpy(SvPVX(last_str) + l,
3913 SvPVX_const(last_str), l, mincount - 1);
3914 SvCUR_set(last_str, SvCUR(last_str) * mincount);
3915 /* Add additional parts. */
3916 SvCUR_set(data->last_found,
3917 SvCUR(data->last_found) - l);
3918 sv_catsv(data->last_found, last_str);
3920 SV * sv = data->last_found;
3922 SvUTF8(sv) && SvMAGICAL(sv) ?
3923 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3924 if (mg && mg->mg_len >= 0)
3925 mg->mg_len += CHR_SVLEN(last_str) - l;
3927 data->last_end += l * (mincount - 1);
3930 /* start offset must point into the last copy */
3931 data->last_start_min += minnext * (mincount - 1);
3932 data->last_start_max += is_inf ? I32_MAX
3933 : (maxcount - 1) * (minnext + data->pos_delta);
3936 /* It is counted once already... */
3937 data->pos_min += minnext * (mincount - counted);
3938 data->pos_delta += - counted * deltanext +
3939 (minnext + deltanext) * maxcount - minnext * mincount;
3940 if (mincount != maxcount) {
3941 /* Cannot extend fixed substrings found inside
3943 SCAN_COMMIT(pRExC_state,data,minlenp);
3944 if (mincount && last_str) {
3945 SV * const sv = data->last_found;
3946 MAGIC * const mg = SvUTF8(sv) && SvMAGICAL(sv) ?
3947 mg_find(sv, PERL_MAGIC_utf8) : NULL;
3951 sv_setsv(sv, last_str);
3952 data->last_end = data->pos_min;
3953 data->last_start_min =
3954 data->pos_min - CHR_SVLEN(last_str);
3955 data->last_start_max = is_inf
3957 : data->pos_min + data->pos_delta
3958 - CHR_SVLEN(last_str);
3960 data->longest = &(data->longest_float);
3962 SvREFCNT_dec(last_str);
3964 if (data && (fl & SF_HAS_EVAL))
3965 data->flags |= SF_HAS_EVAL;
3966 optimize_curly_tail:
3967 if (OP(oscan) != CURLYX) {
3968 while (PL_regkind[OP(next = regnext(oscan))] == NOTHING
3970 NEXT_OFF(oscan) += NEXT_OFF(next);
3973 default: /* REF, ANYOFV, and CLUMP only? */
3974 if (flags & SCF_DO_SUBSTR) {
3975 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
3976 data->longest = &(data->longest_float);
3978 is_inf = is_inf_internal = 1;
3979 if (flags & SCF_DO_STCLASS_OR)
3980 cl_anything(pRExC_state, data->start_class);
3981 flags &= ~SCF_DO_STCLASS;
3985 else if (OP(scan) == LNBREAK) {
3986 if (flags & SCF_DO_STCLASS) {
3988 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
3989 if (flags & SCF_DO_STCLASS_AND) {
3990 for (value = 0; value < 256; value++)
3991 if (!is_VERTWS_cp(value))
3992 ANYOF_BITMAP_CLEAR(data->start_class, value);
3995 for (value = 0; value < 256; value++)
3996 if (is_VERTWS_cp(value))
3997 ANYOF_BITMAP_SET(data->start_class, value);
3999 if (flags & SCF_DO_STCLASS_OR)
4000 cl_and(data->start_class, and_withp);
4001 flags &= ~SCF_DO_STCLASS;
4005 if (flags & SCF_DO_SUBSTR) {
4006 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
4008 data->pos_delta += 1;
4009 data->longest = &(data->longest_float);
4012 else if (REGNODE_SIMPLE(OP(scan))) {
4015 if (flags & SCF_DO_SUBSTR) {
4016 SCAN_COMMIT(pRExC_state,data,minlenp);
4020 if (flags & SCF_DO_STCLASS) {
4021 data->start_class->flags &= ~ANYOF_EOS; /* No match on empty */
4023 /* Some of the logic below assumes that switching
4024 locale on will only add false positives. */
4025 switch (PL_regkind[OP(scan)]) {
4029 /* Perl_croak(aTHX_ "panic: unexpected simple REx opcode %d", OP(scan)); */
4030 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
4031 cl_anything(pRExC_state, data->start_class);
4034 if (OP(scan) == SANY)
4036 if (flags & SCF_DO_STCLASS_OR) { /* Everything but \n */
4037 value = (ANYOF_BITMAP_TEST(data->start_class,'\n')
4038 || ANYOF_CLASS_TEST_ANY_SET(data->start_class));
4039 cl_anything(pRExC_state, data->start_class);
4041 if (flags & SCF_DO_STCLASS_AND || !value)
4042 ANYOF_BITMAP_CLEAR(data->start_class,'\n');
4045 if (flags & SCF_DO_STCLASS_AND)
4046 cl_and(data->start_class,
4047 (struct regnode_charclass_class*)scan);
4049 cl_or(pRExC_state, data->start_class,
4050 (struct regnode_charclass_class*)scan);
4053 if (flags & SCF_DO_STCLASS_AND) {
4054 if (!(data->start_class->flags & ANYOF_LOCALE)) {
4055 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NALNUM);
4056 if (OP(scan) == ALNUMU) {
4057 for (value = 0; value < 256; value++) {
4058 if (!isWORDCHAR_L1(value)) {
4059 ANYOF_BITMAP_CLEAR(data->start_class, value);
4063 for (value = 0; value < 256; value++) {
4064 if (!isALNUM(value)) {
4065 ANYOF_BITMAP_CLEAR(data->start_class, value);
4072 if (data->start_class->flags & ANYOF_LOCALE)
4073 ANYOF_CLASS_SET(data->start_class,ANYOF_ALNUM);
4075 /* Even if under locale, set the bits for non-locale
4076 * in case it isn't a true locale-node. This will
4077 * create false positives if it truly is locale */
4078 if (OP(scan) == ALNUMU) {
4079 for (value = 0; value < 256; value++) {
4080 if (isWORDCHAR_L1(value)) {
4081 ANYOF_BITMAP_SET(data->start_class, value);
4085 for (value = 0; value < 256; value++) {
4086 if (isALNUM(value)) {
4087 ANYOF_BITMAP_SET(data->start_class, value);
4094 if (flags & SCF_DO_STCLASS_AND) {
4095 if (!(data->start_class->flags & ANYOF_LOCALE)) {
4096 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_ALNUM);
4097 if (OP(scan) == NALNUMU) {
4098 for (value = 0; value < 256; value++) {
4099 if (isWORDCHAR_L1(value)) {
4100 ANYOF_BITMAP_CLEAR(data->start_class, value);
4104 for (value = 0; value < 256; value++) {
4105 if (isALNUM(value)) {
4106 ANYOF_BITMAP_CLEAR(data->start_class, value);
4113 if (data->start_class->flags & ANYOF_LOCALE)
4114 ANYOF_CLASS_SET(data->start_class,ANYOF_NALNUM);
4116 /* Even if under locale, set the bits for non-locale in
4117 * case it isn't a true locale-node. This will create
4118 * false positives if it truly is locale */
4119 if (OP(scan) == NALNUMU) {
4120 for (value = 0; value < 256; value++) {
4121 if (! isWORDCHAR_L1(value)) {
4122 ANYOF_BITMAP_SET(data->start_class, value);
4126 for (value = 0; value < 256; value++) {
4127 if (! isALNUM(value)) {
4128 ANYOF_BITMAP_SET(data->start_class, value);
4135 if (flags & SCF_DO_STCLASS_AND) {
4136 if (!(data->start_class->flags & ANYOF_LOCALE)) {
4137 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NSPACE);
4138 if (OP(scan) == SPACEU) {
4139 for (value = 0; value < 256; value++) {
4140 if (!isSPACE_L1(value)) {
4141 ANYOF_BITMAP_CLEAR(data->start_class, value);
4145 for (value = 0; value < 256; value++) {
4146 if (!isSPACE(value)) {
4147 ANYOF_BITMAP_CLEAR(data->start_class, value);
4154 if (data->start_class->flags & ANYOF_LOCALE) {
4155 ANYOF_CLASS_SET(data->start_class,ANYOF_SPACE);
4157 if (OP(scan) == SPACEU) {
4158 for (value = 0; value < 256; value++) {
4159 if (isSPACE_L1(value)) {
4160 ANYOF_BITMAP_SET(data->start_class, value);
4164 for (value = 0; value < 256; value++) {
4165 if (isSPACE(value)) {
4166 ANYOF_BITMAP_SET(data->start_class, value);
4173 if (flags & SCF_DO_STCLASS_AND) {
4174 if (!(data->start_class->flags & ANYOF_LOCALE)) {
4175 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_SPACE);
4176 if (OP(scan) == NSPACEU) {
4177 for (value = 0; value < 256; value++) {
4178 if (isSPACE_L1(value)) {
4179 ANYOF_BITMAP_CLEAR(data->start_class, value);
4183 for (value = 0; value < 256; value++) {
4184 if (isSPACE(value)) {
4185 ANYOF_BITMAP_CLEAR(data->start_class, value);
4192 if (data->start_class->flags & ANYOF_LOCALE)
4193 ANYOF_CLASS_SET(data->start_class,ANYOF_NSPACE);
4194 if (OP(scan) == NSPACEU) {
4195 for (value = 0; value < 256; value++) {
4196 if (!isSPACE_L1(value)) {
4197 ANYOF_BITMAP_SET(data->start_class, value);
4202 for (value = 0; value < 256; value++) {
4203 if (!isSPACE(value)) {
4204 ANYOF_BITMAP_SET(data->start_class, value);
4211 if (flags & SCF_DO_STCLASS_AND) {
4212 if (!(data->start_class->flags & ANYOF_LOCALE)) {
4213 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_NDIGIT);
4214 for (value = 0; value < 256; value++)
4215 if (!isDIGIT(value))
4216 ANYOF_BITMAP_CLEAR(data->start_class, value);
4220 if (data->start_class->flags & ANYOF_LOCALE)
4221 ANYOF_CLASS_SET(data->start_class,ANYOF_DIGIT);
4222 for (value = 0; value < 256; value++)
4224 ANYOF_BITMAP_SET(data->start_class, value);
4228 if (flags & SCF_DO_STCLASS_AND) {
4229 if (!(data->start_class->flags & ANYOF_LOCALE))
4230 ANYOF_CLASS_CLEAR(data->start_class,ANYOF_DIGIT);
4231 for (value = 0; value < 256; value++)
4233 ANYOF_BITMAP_CLEAR(data->start_class, value);
4236 if (data->start_class->flags & ANYOF_LOCALE)
4237 ANYOF_CLASS_SET(data->start_class,ANYOF_NDIGIT);
4238 for (value = 0; value < 256; value++)
4239 if (!isDIGIT(value))
4240 ANYOF_BITMAP_SET(data->start_class, value);
4243 CASE_SYNST_FNC(VERTWS);
4244 CASE_SYNST_FNC(HORIZWS);
4247 if (flags & SCF_DO_STCLASS_OR)
4248 cl_and(data->start_class, and_withp);
4249 flags &= ~SCF_DO_STCLASS;
4252 else if (PL_regkind[OP(scan)] == EOL && flags & SCF_DO_SUBSTR) {
4253 data->flags |= (OP(scan) == MEOL
4257 else if ( PL_regkind[OP(scan)] == BRANCHJ
4258 /* Lookbehind, or need to calculate parens/evals/stclass: */
4259 && (scan->flags || data || (flags & SCF_DO_STCLASS))
4260 && (OP(scan) == IFMATCH || OP(scan) == UNLESSM)) {
4261 if ( !PERL_ENABLE_POSITIVE_ASSERTION_STUDY
4262 || OP(scan) == UNLESSM )
4264 /* Negative Lookahead/lookbehind
4265 In this case we can't do fixed string optimisation.
4268 I32 deltanext, minnext, fake = 0;
4270 struct regnode_charclass_class intrnl;
4273 data_fake.flags = 0;
4275 data_fake.whilem_c = data->whilem_c;
4276 data_fake.last_closep = data->last_closep;
4279 data_fake.last_closep = &fake;
4280 data_fake.pos_delta = delta;
4281 if ( flags & SCF_DO_STCLASS && !scan->flags
4282 && OP(scan) == IFMATCH ) { /* Lookahead */
4283 cl_init(pRExC_state, &intrnl);
4284 data_fake.start_class = &intrnl;
4285 f |= SCF_DO_STCLASS_AND;
4287 if (flags & SCF_WHILEM_VISITED_POS)
4288 f |= SCF_WHILEM_VISITED_POS;
4289 next = regnext(scan);
4290 nscan = NEXTOPER(NEXTOPER(scan));
4291 minnext = study_chunk(pRExC_state, &nscan, minlenp, &deltanext,
4292 last, &data_fake, stopparen, recursed, NULL, f, depth+1);
4295 FAIL("Variable length lookbehind not implemented");
4297 else if (minnext > (I32)U8_MAX) {
4298 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
4300 scan->flags = (U8)minnext;
4303 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
4305 if (data_fake.flags & SF_HAS_EVAL)
4306 data->flags |= SF_HAS_EVAL;
4307 data->whilem_c = data_fake.whilem_c;
4309 if (f & SCF_DO_STCLASS_AND) {
4310 if (flags & SCF_DO_STCLASS_OR) {
4311 /* OR before, AND after: ideally we would recurse with
4312 * data_fake to get the AND applied by study of the
4313 * remainder of the pattern, and then derecurse;
4314 * *** HACK *** for now just treat as "no information".
4315 * See [perl #56690].
4317 cl_init(pRExC_state, data->start_class);
4319 /* AND before and after: combine and continue */
4320 const int was = (data->start_class->flags & ANYOF_EOS);
4322 cl_and(data->start_class, &intrnl);
4324 data->start_class->flags |= ANYOF_EOS;
4328 #if PERL_ENABLE_POSITIVE_ASSERTION_STUDY
4330 /* Positive Lookahead/lookbehind
4331 In this case we can do fixed string optimisation,
4332 but we must be careful about it. Note in the case of
4333 lookbehind the positions will be offset by the minimum
4334 length of the pattern, something we won't know about
4335 until after the recurse.
4337 I32 deltanext, fake = 0;
4339 struct regnode_charclass_class intrnl;
4341 /* We use SAVEFREEPV so that when the full compile
4342 is finished perl will clean up the allocated
4343 minlens when it's all done. This way we don't
4344 have to worry about freeing them when we know
4345 they wont be used, which would be a pain.
4348 Newx( minnextp, 1, I32 );
4349 SAVEFREEPV(minnextp);
4352 StructCopy(data, &data_fake, scan_data_t);
4353 if ((flags & SCF_DO_SUBSTR) && data->last_found) {
4356 SCAN_COMMIT(pRExC_state, &data_fake,minlenp);
4357 data_fake.last_found=newSVsv(data->last_found);
4361 data_fake.last_closep = &fake;
4362 data_fake.flags = 0;
4363 data_fake.pos_delta = delta;
4365 data_fake.flags |= SF_IS_INF;
4366 if ( flags & SCF_DO_STCLASS && !scan->flags
4367 && OP(scan) == IFMATCH ) { /* Lookahead */
4368 cl_init(pRExC_state, &intrnl);
4369 data_fake.start_class = &intrnl;
4370 f |= SCF_DO_STCLASS_AND;
4372 if (flags & SCF_WHILEM_VISITED_POS)
4373 f |= SCF_WHILEM_VISITED_POS;
4374 next = regnext(scan);
4375 nscan = NEXTOPER(NEXTOPER(scan));
4377 *minnextp = study_chunk(pRExC_state, &nscan, minnextp, &deltanext,
4378 last, &data_fake, stopparen, recursed, NULL, f,depth+1);
4381 FAIL("Variable length lookbehind not implemented");
4383 else if (*minnextp > (I32)U8_MAX) {
4384 FAIL2("Lookbehind longer than %"UVuf" not implemented", (UV)U8_MAX);
4386 scan->flags = (U8)*minnextp;
4391 if (f & SCF_DO_STCLASS_AND) {
4392 const int was = (data->start_class->flags & ANYOF_EOS);
4394 cl_and(data->start_class, &intrnl);
4396 data->start_class->flags |= ANYOF_EOS;
4399 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
4401 if (data_fake.flags & SF_HAS_EVAL)
4402 data->flags |= SF_HAS_EVAL;
4403 data->whilem_c = data_fake.whilem_c;
4404 if ((flags & SCF_DO_SUBSTR) && data_fake.last_found) {
4405 if (RExC_rx->minlen<*minnextp)
4406 RExC_rx->minlen=*minnextp;
4407 SCAN_COMMIT(pRExC_state, &data_fake, minnextp);
4408 SvREFCNT_dec(data_fake.last_found);
4410 if ( data_fake.minlen_fixed != minlenp )
4412 data->offset_fixed= data_fake.offset_fixed;
4413 data->minlen_fixed= data_fake.minlen_fixed;
4414 data->lookbehind_fixed+= scan->flags;
4416 if ( data_fake.minlen_float != minlenp )
4418 data->minlen_float= data_fake.minlen_float;
4419 data->offset_float_min=data_fake.offset_float_min;
4420 data->offset_float_max=data_fake.offset_float_max;
4421 data->lookbehind_float+= scan->flags;
4430 else if (OP(scan) == OPEN) {
4431 if (stopparen != (I32)ARG(scan))
4434 else if (OP(scan) == CLOSE) {
4435 if (stopparen == (I32)ARG(scan)) {
4438 if ((I32)ARG(scan) == is_par) {
4439 next = regnext(scan);
4441 if ( next && (OP(next) != WHILEM) && next < last)
4442 is_par = 0; /* Disable optimization */
4445 *(data->last_closep) = ARG(scan);
4447 else if (OP(scan) == EVAL) {
4449 data->flags |= SF_HAS_EVAL;
4451 else if ( PL_regkind[OP(scan)] == ENDLIKE ) {
4452 if (flags & SCF_DO_SUBSTR) {
4453 SCAN_COMMIT(pRExC_state,data,minlenp);
4454 flags &= ~SCF_DO_SUBSTR;
4456 if (data && OP(scan)==ACCEPT) {
4457 data->flags |= SCF_SEEN_ACCEPT;
4462 else if (OP(scan) == LOGICAL && scan->flags == 2) /* Embedded follows */
4464 if (flags & SCF_DO_SUBSTR) {
4465 SCAN_COMMIT(pRExC_state,data,minlenp);
4466 data->longest = &(data->longest_float);
4468 is_inf = is_inf_internal = 1;
4469 if (flags & SCF_DO_STCLASS_OR) /* Allow everything */
4470 cl_anything(pRExC_state, data->start_class);
4471 flags &= ~SCF_DO_STCLASS;
4473 else if (OP(scan) == GPOS) {
4474 if (!(RExC_rx->extflags & RXf_GPOS_FLOAT) &&
4475 !(delta || is_inf || (data && data->pos_delta)))
4477 if (!(RExC_rx->extflags & RXf_ANCH) && (flags & SCF_DO_SUBSTR))
4478 RExC_rx->extflags |= RXf_ANCH_GPOS;
4479 if (RExC_rx->gofs < (U32)min)
4480 RExC_rx->gofs = min;
4482 RExC_rx->extflags |= RXf_GPOS_FLOAT;
4486 #ifdef TRIE_STUDY_OPT
4487 #ifdef FULL_TRIE_STUDY
4488 else if (PL_regkind[OP(scan)] == TRIE) {
4489 /* NOTE - There is similar code to this block above for handling
4490 BRANCH nodes on the initial study. If you change stuff here
4492 regnode *trie_node= scan;
4493 regnode *tail= regnext(scan);
4494 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
4495 I32 max1 = 0, min1 = I32_MAX;
4496 struct regnode_charclass_class accum;
4498 if (flags & SCF_DO_SUBSTR) /* XXXX Add !SUSPEND? */
4499 SCAN_COMMIT(pRExC_state, data,minlenp); /* Cannot merge strings after this. */
4500 if (flags & SCF_DO_STCLASS)
4501 cl_init_zero(pRExC_state, &accum);
4507 const regnode *nextbranch= NULL;
4510 for ( word=1 ; word <= trie->wordcount ; word++)
4512 I32 deltanext=0, minnext=0, f = 0, fake;
4513 struct regnode_charclass_class this_class;
4515 data_fake.flags = 0;
4517 data_fake.whilem_c = data->whilem_c;
4518 data_fake.last_closep = data->last_closep;
4521 data_fake.last_closep = &fake;
4522 data_fake.pos_delta = delta;
4523 if (flags & SCF_DO_STCLASS) {
4524 cl_init(pRExC_state, &this_class);
4525 data_fake.start_class = &this_class;
4526 f = SCF_DO_STCLASS_AND;
4528 if (flags & SCF_WHILEM_VISITED_POS)
4529 f |= SCF_WHILEM_VISITED_POS;
4531 if (trie->jump[word]) {
4533 nextbranch = trie_node + trie->jump[0];
4534 scan= trie_node + trie->jump[word];
4535 /* We go from the jump point to the branch that follows
4536 it. Note this means we need the vestigal unused branches
4537 even though they arent otherwise used.
4539 minnext = study_chunk(pRExC_state, &scan, minlenp,
4540 &deltanext, (regnode *)nextbranch, &data_fake,
4541 stopparen, recursed, NULL, f,depth+1);
4543 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
4544 nextbranch= regnext((regnode*)nextbranch);
4546 if (min1 > (I32)(minnext + trie->minlen))
4547 min1 = minnext + trie->minlen;
4548 if (max1 < (I32)(minnext + deltanext + trie->maxlen))
4549 max1 = minnext + deltanext + trie->maxlen;
4550 if (deltanext == I32_MAX)
4551 is_inf = is_inf_internal = 1;
4553 if (data_fake.flags & (SF_HAS_PAR|SF_IN_PAR))
4555 if (data_fake.flags & SCF_SEEN_ACCEPT) {
4556 if ( stopmin > min + min1)
4557 stopmin = min + min1;
4558 flags &= ~SCF_DO_SUBSTR;
4560 data->flags |= SCF_SEEN_ACCEPT;
4563 if (data_fake.flags & SF_HAS_EVAL)
4564 data->flags |= SF_HAS_EVAL;
4565 data->whilem_c = data_fake.whilem_c;
4567 if (flags & SCF_DO_STCLASS)
4568 cl_or(pRExC_state, &accum, &this_class);
4571 if (flags & SCF_DO_SUBSTR) {
4572 data->pos_min += min1;
4573 data->pos_delta += max1 - min1;
4574 if (max1 != min1 || is_inf)
4575 data->longest = &(data->longest_float);
4578 delta += max1 - min1;
4579 if (flags & SCF_DO_STCLASS_OR) {
4580 cl_or(pRExC_state, data->start_class, &accum);
4582 cl_and(data->start_class, and_withp);
4583 flags &= ~SCF_DO_STCLASS;
4586 else if (flags & SCF_DO_STCLASS_AND) {
4588 cl_and(data->start_class, &accum);
4589 flags &= ~SCF_DO_STCLASS;
4592 /* Switch to OR mode: cache the old value of
4593 * data->start_class */
4595 StructCopy(data->start_class, and_withp,
4596 struct regnode_charclass_class);
4597 flags &= ~SCF_DO_STCLASS_AND;
4598 StructCopy(&accum, data->start_class,
4599 struct regnode_charclass_class);
4600 flags |= SCF_DO_STCLASS_OR;
4601 data->start_class->flags |= ANYOF_EOS;
4608 else if (PL_regkind[OP(scan)] == TRIE) {
4609 reg_trie_data *trie = (reg_trie_data*)RExC_rxi->data->data[ ARG(scan) ];
4612 min += trie->minlen;
4613 delta += (trie->maxlen - trie->minlen);
4614 flags &= ~SCF_DO_STCLASS; /* xxx */
4615 if (flags & SCF_DO_SUBSTR) {
4616 SCAN_COMMIT(pRExC_state,data,minlenp); /* Cannot expect anything... */
4617 data->pos_min += trie->minlen;
4618 data->pos_delta += (trie->maxlen - trie->minlen);
4619 if (trie->maxlen != trie->minlen)
4620 data->longest = &(data->longest_float);
4622 if (trie->jump) /* no more substrings -- for now /grr*/
4623 flags &= ~SCF_DO_SUBSTR;
4625 #endif /* old or new */
4626 #endif /* TRIE_STUDY_OPT */
4628 /* Else: zero-length, ignore. */
4629 scan = regnext(scan);
4634 stopparen = frame->stop;
4635 frame = frame->prev;
4636 goto fake_study_recurse;
4641 DEBUG_STUDYDATA("pre-fin:",data,depth);
4644 *deltap = is_inf_internal ? I32_MAX : delta;
4645 if (flags & SCF_DO_SUBSTR && is_inf)
4646 data->pos_delta = I32_MAX - data->pos_min;
4647 if (is_par > (I32)U8_MAX)
4649 if (is_par && pars==1 && data) {
4650 data->flags |= SF_IN_PAR;
4651 data->flags &= ~SF_HAS_PAR;
4653 else if (pars && data) {
4654 data->flags |= SF_HAS_PAR;
4655 data->flags &= ~SF_IN_PAR;
4657 if (flags & SCF_DO_STCLASS_OR)
4658 cl_and(data->start_class, and_withp);
4659 if (flags & SCF_TRIE_RESTUDY)
4660 data->flags |= SCF_TRIE_RESTUDY;
4662 DEBUG_STUDYDATA("post-fin:",data,depth);
4664 return min < stopmin ? min : stopmin;
4668 S_add_data(RExC_state_t *pRExC_state, U32 n, const char *s)
4670 U32 count = RExC_rxi->data ? RExC_rxi->data->count : 0;
4672 PERL_ARGS_ASSERT_ADD_DATA;
4674 Renewc(RExC_rxi->data,
4675 sizeof(*RExC_rxi->data) + sizeof(void*) * (count + n - 1),
4676 char, struct reg_data);
4678 Renew(RExC_rxi->data->what, count + n, U8);
4680 Newx(RExC_rxi->data->what, n, U8);
4681 RExC_rxi->data->count = count + n;
4682 Copy(s, RExC_rxi->data->what + count, n, U8);
4686 /*XXX: todo make this not included in a non debugging perl */
4687 #ifndef PERL_IN_XSUB_RE
4689 Perl_reginitcolors(pTHX)
4692 const char * const s = PerlEnv_getenv("PERL_RE_COLORS");
4694 char *t = savepv(s);
4698 t = strchr(t, '\t');
4704 PL_colors[i] = t = (char *)"";
4709 PL_colors[i++] = (char *)"";
4716 #ifdef TRIE_STUDY_OPT
4717 #define CHECK_RESTUDY_GOTO \
4719 (data.flags & SCF_TRIE_RESTUDY) \
4723 #define CHECK_RESTUDY_GOTO
4727 - pregcomp - compile a regular expression into internal code
4729 * We can't allocate space until we know how big the compiled form will be,
4730 * but we can't compile it (and thus know how big it is) until we've got a
4731 * place to put the code. So we cheat: we compile it twice, once with code
4732 * generation turned off and size counting turned on, and once "for real".
4733 * This also means that we don't allocate space until we are sure that the
4734 * thing really will compile successfully, and we never have to move the
4735 * code and thus invalidate pointers into it. (Note that it has to be in
4736 * one piece because free() must be able to free it all.) [NB: not true in perl]
4738 * Beware that the optimization-preparation code in here knows about some
4739 * of the structure of the compiled regexp. [I'll say.]
4744 #ifndef PERL_IN_XSUB_RE
4745 #define RE_ENGINE_PTR &PL_core_reg_engine
4747 extern const struct regexp_engine my_reg_engine;
4748 #define RE_ENGINE_PTR &my_reg_engine
4751 #ifndef PERL_IN_XSUB_RE
4753 Perl_pregcomp(pTHX_ SV * const pattern, const U32 flags)
4756 HV * const table = GvHV(PL_hintgv);
4758 PERL_ARGS_ASSERT_PREGCOMP;
4760 /* Dispatch a request to compile a regexp to correct
4763 SV **ptr= hv_fetchs(table, "regcomp", FALSE);
4764 GET_RE_DEBUG_FLAGS_DECL;
4765 if (ptr && SvIOK(*ptr) && SvIV(*ptr)) {
4766 const regexp_engine *eng=INT2PTR(regexp_engine*,SvIV(*ptr));
4768 PerlIO_printf(Perl_debug_log, "Using engine %"UVxf"\n",
4771 return CALLREGCOMP_ENG(eng, pattern, flags);
4774 return Perl_re_compile(aTHX_ pattern, flags);
4779 Perl_re_compile(pTHX_ SV * const pattern, U32 orig_pm_flags)
4784 register regexp_internal *ri;
4793 /* these are all flags - maybe they should be turned
4794 * into a single int with different bit masks */
4795 I32 sawlookahead = 0;
4798 bool used_setjump = FALSE;
4799 regex_charset initial_charset = get_regex_charset(orig_pm_flags);
4804 RExC_state_t RExC_state;
4805 RExC_state_t * const pRExC_state = &RExC_state;
4806 #ifdef TRIE_STUDY_OPT
4808 RExC_state_t copyRExC_state;
4810 GET_RE_DEBUG_FLAGS_DECL;
4812 PERL_ARGS_ASSERT_RE_COMPILE;
4814 DEBUG_r(if (!PL_colorset) reginitcolors());
4816 exp = SvPV(pattern, plen);
4818 if (plen == 0) { /* ignore the utf8ness if the pattern is 0 length */
4819 RExC_utf8 = RExC_orig_utf8 = 0;
4822 RExC_utf8 = RExC_orig_utf8 = SvUTF8(pattern);
4824 RExC_uni_semantics = 0;
4825 RExC_contains_locale = 0;
4827 /****************** LONG JUMP TARGET HERE***********************/
4828 /* Longjmp back to here if have to switch in midstream to utf8 */
4829 if (! RExC_orig_utf8) {
4830 JMPENV_PUSH(jump_ret);
4831 used_setjump = TRUE;
4834 if (jump_ret == 0) { /* First time through */
4838 SV *dsv= sv_newmortal();
4839 RE_PV_QUOTED_DECL(s, RExC_utf8,
4840 dsv, exp, plen, 60);
4841 PerlIO_printf(Perl_debug_log, "%sCompiling REx%s %s\n",
4842 PL_colors[4],PL_colors[5],s);
4845 else { /* longjumped back */
4848 /* If the cause for the longjmp was other than changing to utf8, pop
4849 * our own setjmp, and longjmp to the correct handler */
4850 if (jump_ret != UTF8_LONGJMP) {
4852 JMPENV_JUMP(jump_ret);
4857 /* It's possible to write a regexp in ascii that represents Unicode
4858 codepoints outside of the byte range, such as via \x{100}. If we
4859 detect such a sequence we have to convert the entire pattern to utf8
4860 and then recompile, as our sizing calculation will have been based
4861 on 1 byte == 1 character, but we will need to use utf8 to encode
4862 at least some part of the pattern, and therefore must convert the whole
4865 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log,
4866 "UTF8 mismatch! Converting to utf8 for resizing and compile\n"));
4867 exp = (char*)Perl_bytes_to_utf8(aTHX_
4868 (U8*)SvPV_nomg(pattern, plen),
4871 RExC_orig_utf8 = RExC_utf8 = 1;
4875 #ifdef TRIE_STUDY_OPT
4879 pm_flags = orig_pm_flags;
4881 if (initial_charset == REGEX_LOCALE_CHARSET) {
4882 RExC_contains_locale = 1;
4884 else if (RExC_utf8 && initial_charset == REGEX_DEPENDS_CHARSET) {
4886 /* Set to use unicode semantics if the pattern is in utf8 and has the
4887 * 'depends' charset specified, as it means unicode when utf8 */
4888 set_regex_charset(&pm_flags, REGEX_UNICODE_CHARSET);
4892 RExC_flags = pm_flags;
4896 RExC_in_lookbehind = 0;
4897 RExC_seen_zerolen = *exp == '^' ? -1 : 0;
4898 RExC_seen_evals = 0;
4900 RExC_override_recoding = 0;
4902 /* First pass: determine size, legality. */
4910 RExC_emit = &PL_regdummy;
4911 RExC_whilem_seen = 0;
4912 RExC_open_parens = NULL;
4913 RExC_close_parens = NULL;
4915 RExC_paren_names = NULL;
4917 RExC_paren_name_list = NULL;
4919 RExC_recurse = NULL;
4920 RExC_recurse_count = 0;
4922 #if 0 /* REGC() is (currently) a NOP at the first pass.
4923 * Clever compilers notice this and complain. --jhi */
4924 REGC((U8)REG_MAGIC, (char*)RExC_emit);
4927 PerlIO_printf(Perl_debug_log, "Starting first pass (sizing)\n");
4929 RExC_lastparse=NULL;
4931 if (reg(pRExC_state, 0, &flags,1) == NULL) {
4932 RExC_precomp = NULL;
4936 /* Here, finished first pass. Get rid of any added setjmp */
4942 PerlIO_printf(Perl_debug_log,
4943 "Required size %"IVdf" nodes\n"
4944 "Starting second pass (creation)\n",
4947 RExC_lastparse=NULL;
4950 /* The first pass could have found things that force Unicode semantics */
4951 if ((RExC_utf8 || RExC_uni_semantics)
4952 && get_regex_charset(pm_flags) == REGEX_DEPENDS_CHARSET)
4954 set_regex_charset(&pm_flags, REGEX_UNICODE_CHARSET);
4957 /* Small enough for pointer-storage convention?
4958 If extralen==0, this means that we will not need long jumps. */
4959 if (RExC_size >= 0x10000L && RExC_extralen)
4960 RExC_size += RExC_extralen;
4963 if (RExC_whilem_seen > 15)
4964 RExC_whilem_seen = 15;
4966 /* Allocate space and zero-initialize. Note, the two step process
4967 of zeroing when in debug mode, thus anything assigned has to
4968 happen after that */
4969 rx = (REGEXP*) newSV_type(SVt_REGEXP);
4970 r = (struct regexp*)SvANY(rx);
4971 Newxc(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode),
4972 char, regexp_internal);
4973 if ( r == NULL || ri == NULL )
4974 FAIL("Regexp out of space");
4976 /* avoid reading uninitialized memory in DEBUGGING code in study_chunk() */
4977 Zero(ri, sizeof(regexp_internal) + (unsigned)RExC_size * sizeof(regnode), char);
4979 /* bulk initialize base fields with 0. */
4980 Zero(ri, sizeof(regexp_internal), char);
4983 /* non-zero initialization begins here */
4985 r->engine= RE_ENGINE_PTR;
4986 r->extflags = pm_flags;
4988 bool has_p = ((r->extflags & RXf_PMf_KEEPCOPY) == RXf_PMf_KEEPCOPY);
4989 bool has_charset = (get_regex_charset(r->extflags) != REGEX_DEPENDS_CHARSET);
4991 /* The caret is output if there are any defaults: if not all the STD
4992 * flags are set, or if no character set specifier is needed */
4994 (((r->extflags & RXf_PMf_STD_PMMOD) != RXf_PMf_STD_PMMOD)
4996 bool has_runon = ((RExC_seen & REG_SEEN_RUN_ON_COMMENT)==REG_SEEN_RUN_ON_COMMENT);
4997 U16 reganch = (U16)((r->extflags & RXf_PMf_STD_PMMOD)
4998 >> RXf_PMf_STD_PMMOD_SHIFT);
4999 const char *fptr = STD_PAT_MODS; /*"msix"*/
5001 /* Allocate for the worst case, which is all the std flags are turned
5002 * on. If more precision is desired, we could do a population count of
5003 * the flags set. This could be done with a small lookup table, or by
5004 * shifting, masking and adding, or even, when available, assembly
5005 * language for a machine-language population count.
5006 * We never output a minus, as all those are defaults, so are
5007 * covered by the caret */
5008 const STRLEN wraplen = plen + has_p + has_runon
5009 + has_default /* If needs a caret */
5011 /* If needs a character set specifier */
5012 + ((has_charset) ? MAX_CHARSET_NAME_LENGTH : 0)
5013 + (sizeof(STD_PAT_MODS) - 1)
5014 + (sizeof("(?:)") - 1);
5016 p = sv_grow(MUTABLE_SV(rx), wraplen + 1); /* +1 for the ending NUL */
5018 SvFLAGS(rx) |= SvUTF8(pattern);
5021 /* If a default, cover it using the caret */
5023 *p++= DEFAULT_PAT_MOD;
5027 const char* const name = get_regex_charset_name(r->extflags, &len);
5028 Copy(name, p, len, char);
5032 *p++ = KEEPCOPY_PAT_MOD; /*'p'*/
5035 while((ch = *fptr++)) {
5043 Copy(RExC_precomp, p, plen, char);
5044 assert ((RX_WRAPPED(rx) - p) < 16);
5045 r->pre_prefix = p - RX_WRAPPED(rx);
5051 SvCUR_set(rx, p - SvPVX_const(rx));
5055 r->nparens = RExC_npar - 1; /* set early to validate backrefs */
5057 if (RExC_seen & REG_SEEN_RECURSE) {
5058 Newxz(RExC_open_parens, RExC_npar,regnode *);
5059 SAVEFREEPV(RExC_open_parens);
5060 Newxz(RExC_close_parens,RExC_npar,regnode *);
5061 SAVEFREEPV(RExC_close_parens);
5064 /* Useful during FAIL. */
5065 #ifdef RE_TRACK_PATTERN_OFFSETS
5066 Newxz(ri->u.offsets, 2*RExC_size+1, U32); /* MJD 20001228 */
5067 DEBUG_OFFSETS_r(PerlIO_printf(Perl_debug_log,
5068 "%s %"UVuf" bytes for offset annotations.\n",
5069 ri->u.offsets ? "Got" : "Couldn't get",
5070 (UV)((2*RExC_size+1) * sizeof(U32))));
5072 SetProgLen(ri,RExC_size);
5077 /* Second pass: emit code. */
5078 RExC_flags = pm_flags; /* don't let top level (?i) bleed */
5083 RExC_emit_start = ri->program;
5084 RExC_emit = ri->program;
5085 RExC_emit_bound = ri->program + RExC_size + 1;
5087 /* Store the count of eval-groups for security checks: */
5088 RExC_rx->seen_evals = RExC_seen_evals;
5089 REGC((U8)REG_MAGIC, (char*) RExC_emit++);
5090 if (reg(pRExC_state, 0, &flags,1) == NULL) {
5094 /* XXXX To minimize changes to RE engine we always allocate
5095 3-units-long substrs field. */
5096 Newx(r->substrs, 1, struct reg_substr_data);
5097 if (RExC_recurse_count) {
5098 Newxz(RExC_recurse,RExC_recurse_count,regnode *);
5099 SAVEFREEPV(RExC_recurse);
5103 r->minlen = minlen = sawlookahead = sawplus = sawopen = 0;
5104 Zero(r->substrs, 1, struct reg_substr_data);
5106 #ifdef TRIE_STUDY_OPT
5108 StructCopy(&zero_scan_data, &data, scan_data_t);
5109 copyRExC_state = RExC_state;
5112 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log,"Restudying\n"));
5114 RExC_state = copyRExC_state;
5115 if (seen & REG_TOP_LEVEL_BRANCHES)
5116 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
5118 RExC_seen &= ~REG_TOP_LEVEL_BRANCHES;
5119 if (data.last_found) {
5120 SvREFCNT_dec(data.longest_fixed);
5121 SvREFCNT_dec(data.longest_float);
5122 SvREFCNT_dec(data.last_found);
5124 StructCopy(&zero_scan_data, &data, scan_data_t);
5127 StructCopy(&zero_scan_data, &data, scan_data_t);
5130 /* Dig out information for optimizations. */
5131 r->extflags = RExC_flags; /* was pm_op */
5132 /*dmq: removed as part of de-PMOP: pm->op_pmflags = RExC_flags; */
5135 SvUTF8_on(rx); /* Unicode in it? */
5136 ri->regstclass = NULL;
5137 if (RExC_naughty >= 10) /* Probably an expensive pattern. */
5138 r->intflags |= PREGf_NAUGHTY;
5139 scan = ri->program + 1; /* First BRANCH. */
5141 /* testing for BRANCH here tells us whether there is "must appear"
5142 data in the pattern. If there is then we can use it for optimisations */
5143 if (!(RExC_seen & REG_TOP_LEVEL_BRANCHES)) { /* Only one top-level choice. */
5145 STRLEN longest_float_length, longest_fixed_length;
5146 struct regnode_charclass_class ch_class; /* pointed to by data */
5148 I32 last_close = 0; /* pointed to by data */
5149 regnode *first= scan;
5150 regnode *first_next= regnext(first);
5152 * Skip introductions and multiplicators >= 1
5153 * so that we can extract the 'meat' of the pattern that must
5154 * match in the large if() sequence following.
5155 * NOTE that EXACT is NOT covered here, as it is normally
5156 * picked up by the optimiser separately.
5158 * This is unfortunate as the optimiser isnt handling lookahead
5159 * properly currently.
5162 while ((OP(first) == OPEN && (sawopen = 1)) ||
5163 /* An OR of *one* alternative - should not happen now. */
5164 (OP(first) == BRANCH && OP(first_next) != BRANCH) ||
5165 /* for now we can't handle lookbehind IFMATCH*/
5166 (OP(first) == IFMATCH && !first->flags && (sawlookahead = 1)) ||
5167 (OP(first) == PLUS) ||
5168 (OP(first) == MINMOD) ||
5169 /* An {n,m} with n>0 */
5170 (PL_regkind[OP(first)] == CURLY && ARG1(first) > 0) ||
5171 (OP(first) == NOTHING && PL_regkind[OP(first_next)] != END ))
5174 * the only op that could be a regnode is PLUS, all the rest
5175 * will be regnode_1 or regnode_2.
5178 if (OP(first) == PLUS)
5181 first += regarglen[OP(first)];
5183 first = NEXTOPER(first);
5184 first_next= regnext(first);
5187 /* Starting-point info. */
5189 DEBUG_PEEP("first:",first,0);
5190 /* Ignore EXACT as we deal with it later. */
5191 if (PL_regkind[OP(first)] == EXACT) {
5192 if (OP(first) == EXACT)
5193 NOOP; /* Empty, get anchored substr later. */
5195 ri->regstclass = first;
5198 else if (PL_regkind[OP(first)] == TRIE &&
5199 ((reg_trie_data *)ri->data->data[ ARG(first) ])->minlen>0)
5202 /* this can happen only on restudy */
5203 if ( OP(first) == TRIE ) {
5204 struct regnode_1 *trieop = (struct regnode_1 *)
5205 PerlMemShared_calloc(1, sizeof(struct regnode_1));
5206 StructCopy(first,trieop,struct regnode_1);
5207 trie_op=(regnode *)trieop;
5209 struct regnode_charclass *trieop = (struct regnode_charclass *)
5210 PerlMemShared_calloc(1, sizeof(struct regnode_charclass));
5211 StructCopy(first,trieop,struct regnode_charclass);
5212 trie_op=(regnode *)trieop;
5215 make_trie_failtable(pRExC_state, (regnode *)first, trie_op, 0);
5216 ri->regstclass = trie_op;
5219 else if (REGNODE_SIMPLE(OP(first)))
5220 ri->regstclass = first;
5221 else if (PL_regkind[OP(first)] == BOUND ||
5222 PL_regkind[OP(first)] == NBOUND)
5223 ri->regstclass = first;
5224 else if (PL_regkind[OP(first)] == BOL) {
5225 r->extflags |= (OP(first) == MBOL
5227 : (OP(first) == SBOL
5230 first = NEXTOPER(first);
5233 else if (OP(first) == GPOS) {
5234 r->extflags |= RXf_ANCH_GPOS;
5235 first = NEXTOPER(first);
5238 else if ((!sawopen || !RExC_sawback) &&
5239 (OP(first) == STAR &&
5240 PL_regkind[OP(NEXTOPER(first))] == REG_ANY) &&
5241 !(r->extflags & RXf_ANCH) && !(RExC_seen & REG_SEEN_EVAL))
5243 /* turn .* into ^.* with an implied $*=1 */
5245 (OP(NEXTOPER(first)) == REG_ANY)
5248 r->extflags |= type;
5249 r->intflags |= PREGf_IMPLICIT;
5250 first = NEXTOPER(first);
5253 if (sawplus && !sawlookahead && (!sawopen || !RExC_sawback)
5254 && !(RExC_seen & REG_SEEN_EVAL)) /* May examine pos and $& */
5255 /* x+ must match at the 1st pos of run of x's */
5256 r->intflags |= PREGf_SKIP;
5258 /* Scan is after the zeroth branch, first is atomic matcher. */
5259 #ifdef TRIE_STUDY_OPT
5262 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
5263 (IV)(first - scan + 1))
5267 PerlIO_printf(Perl_debug_log, "first at %"IVdf"\n",
5268 (IV)(first - scan + 1))
5274 * If there's something expensive in the r.e., find the
5275 * longest literal string that must appear and make it the
5276 * regmust. Resolve ties in favor of later strings, since
5277 * the regstart check works with the beginning of the r.e.
5278 * and avoiding duplication strengthens checking. Not a
5279 * strong reason, but sufficient in the absence of others.
5280 * [Now we resolve ties in favor of the earlier string if
5281 * it happens that c_offset_min has been invalidated, since the
5282 * earlier string may buy us something the later one won't.]
5285 data.longest_fixed = newSVpvs("");
5286 data.longest_float = newSVpvs("");
5287 data.last_found = newSVpvs("");
5288 data.longest = &(data.longest_fixed);
5290 if (!ri->regstclass) {
5291 cl_init(pRExC_state, &ch_class);
5292 data.start_class = &ch_class;
5293 stclass_flag = SCF_DO_STCLASS_AND;
5294 } else /* XXXX Check for BOUND? */
5296 data.last_closep = &last_close;
5298 minlen = study_chunk(pRExC_state, &first, &minlen, &fake, scan + RExC_size, /* Up to end */
5299 &data, -1, NULL, NULL,
5300 SCF_DO_SUBSTR | SCF_WHILEM_VISITED_POS | stclass_flag,0);
5306 if ( RExC_npar == 1 && data.longest == &(data.longest_fixed)
5307 && data.last_start_min == 0 && data.last_end > 0
5308 && !RExC_seen_zerolen
5309 && !(RExC_seen & REG_SEEN_VERBARG)
5310 && (!(RExC_seen & REG_SEEN_GPOS) || (r->extflags & RXf_ANCH_GPOS)))
5311 r->extflags |= RXf_CHECK_ALL;
5312 scan_commit(pRExC_state, &data,&minlen,0);
5313 SvREFCNT_dec(data.last_found);
5315 /* Note that code very similar to this but for anchored string
5316 follows immediately below, changes may need to be made to both.
5319 longest_float_length = CHR_SVLEN(data.longest_float);
5320 if (longest_float_length
5321 || (data.flags & SF_FL_BEFORE_EOL
5322 && (!(data.flags & SF_FL_BEFORE_MEOL)
5323 || (RExC_flags & RXf_PMf_MULTILINE))))
5327 /* See comments for join_exact for why REG_SEEN_EXACTF_SHARP_S */
5328 if ((RExC_seen & REG_SEEN_EXACTF_SHARP_S)
5329 || (SvCUR(data.longest_fixed) /* ok to leave SvCUR */
5330 && data.offset_fixed == data.offset_float_min
5331 && SvCUR(data.longest_fixed) == SvCUR(data.longest_float)))
5332 goto remove_float; /* As in (a)+. */
5334 /* copy the information about the longest float from the reg_scan_data
5335 over to the program. */
5336 if (SvUTF8(data.longest_float)) {
5337 r->float_utf8 = data.longest_float;
5338 r->float_substr = NULL;
5340 r->float_substr = data.longest_float;
5341 r->float_utf8 = NULL;
5343 /* float_end_shift is how many chars that must be matched that
5344 follow this item. We calculate it ahead of time as once the
5345 lookbehind offset is added in we lose the ability to correctly
5347 ml = data.minlen_float ? *(data.minlen_float)
5348 : (I32)longest_float_length;
5349 r->float_end_shift = ml - data.offset_float_min
5350 - longest_float_length + (SvTAIL(data.longest_float) != 0)
5351 + data.lookbehind_float;
5352 r->float_min_offset = data.offset_float_min - data.lookbehind_float;
5353 r->float_max_offset = data.offset_float_max;
5354 if (data.offset_float_max < I32_MAX) /* Don't offset infinity */
5355 r->float_max_offset -= data.lookbehind_float;
5357 t = (data.flags & SF_FL_BEFORE_EOL /* Can't have SEOL and MULTI */
5358 && (!(data.flags & SF_FL_BEFORE_MEOL)
5359 || (RExC_flags & RXf_PMf_MULTILINE)));
5360 fbm_compile(data.longest_float, t ? FBMcf_TAIL : 0);
5364 r->float_substr = r->float_utf8 = NULL;
5365 SvREFCNT_dec(data.longest_float);
5366 longest_float_length = 0;
5369 /* Note that code very similar to this but for floating string
5370 is immediately above, changes may need to be made to both.
5373 longest_fixed_length = CHR_SVLEN(data.longest_fixed);
5375 /* See comments for join_exact for why REG_SEEN_EXACTF_SHARP_S */
5376 if (! (RExC_seen & REG_SEEN_EXACTF_SHARP_S)
5377 && (longest_fixed_length
5378 || (data.flags & SF_FIX_BEFORE_EOL /* Cannot have SEOL and MULTI */
5379 && (!(data.flags & SF_FIX_BEFORE_MEOL)
5380 || (RExC_flags & RXf_PMf_MULTILINE)))) )
5384 /* copy the information about the longest fixed
5385 from the reg_scan_data over to the program. */
5386 if (SvUTF8(data.longest_fixed)) {
5387 r->anchored_utf8 = data.longest_fixed;
5388 r->anchored_substr = NULL;
5390 r->anchored_substr = data.longest_fixed;
5391 r->anchored_utf8 = NULL;
5393 /* fixed_end_shift is how many chars that must be matched that
5394 follow this item. We calculate it ahead of time as once the
5395 lookbehind offset is added in we lose the ability to correctly
5397 ml = data.minlen_fixed ? *(data.minlen_fixed)
5398 : (I32)longest_fixed_length;
5399 r->anchored_end_shift = ml - data.offset_fixed
5400 - longest_fixed_length + (SvTAIL(data.longest_fixed) != 0)
5401 + data.lookbehind_fixed;
5402 r->anchored_offset = data.offset_fixed - data.lookbehind_fixed;
5404 t = (data.flags & SF_FIX_BEFORE_EOL /* Can't have SEOL and MULTI */
5405 && (!(data.flags & SF_FIX_BEFORE_MEOL)
5406 || (RExC_flags & RXf_PMf_MULTILINE)));
5407 fbm_compile(data.longest_fixed, t ? FBMcf_TAIL : 0);
5410 r->anchored_substr = r->anchored_utf8 = NULL;
5411 SvREFCNT_dec(data.longest_fixed);
5412 longest_fixed_length = 0;
5415 && (OP(ri->regstclass) == REG_ANY || OP(ri->regstclass) == SANY))
5416 ri->regstclass = NULL;
5418 if ((!(r->anchored_substr || r->anchored_utf8) || r->anchored_offset)
5420 && !(data.start_class->flags & ANYOF_EOS)
5421 && !cl_is_anything(data.start_class))
5423 const U32 n = add_data(pRExC_state, 1, "f");
5424 data.start_class->flags |= ANYOF_IS_SYNTHETIC;
5426 Newx(RExC_rxi->data->data[n], 1,
5427 struct regnode_charclass_class);
5428 StructCopy(data.start_class,
5429 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
5430 struct regnode_charclass_class);
5431 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
5432 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
5433 DEBUG_COMPILE_r({ SV *sv = sv_newmortal();
5434 regprop(r, sv, (regnode*)data.start_class);
5435 PerlIO_printf(Perl_debug_log,
5436 "synthetic stclass \"%s\".\n",
5437 SvPVX_const(sv));});
5440 /* A temporary algorithm prefers floated substr to fixed one to dig more info. */
5441 if (longest_fixed_length > longest_float_length) {
5442 r->check_end_shift = r->anchored_end_shift;
5443 r->check_substr = r->anchored_substr;
5444 r->check_utf8 = r->anchored_utf8;
5445 r->check_offset_min = r->check_offset_max = r->anchored_offset;
5446 if (r->extflags & RXf_ANCH_SINGLE)
5447 r->extflags |= RXf_NOSCAN;
5450 r->check_end_shift = r->float_end_shift;
5451 r->check_substr = r->float_substr;
5452 r->check_utf8 = r->float_utf8;
5453 r->check_offset_min = r->float_min_offset;
5454 r->check_offset_max = r->float_max_offset;
5456 /* XXXX Currently intuiting is not compatible with ANCH_GPOS.
5457 This should be changed ASAP! */
5458 if ((r->check_substr || r->check_utf8) && !(r->extflags & RXf_ANCH_GPOS)) {
5459 r->extflags |= RXf_USE_INTUIT;
5460 if (SvTAIL(r->check_substr ? r->check_substr : r->check_utf8))
5461 r->extflags |= RXf_INTUIT_TAIL;
5463 /* XXX Unneeded? dmq (shouldn't as this is handled elsewhere)
5464 if ( (STRLEN)minlen < longest_float_length )
5465 minlen= longest_float_length;
5466 if ( (STRLEN)minlen < longest_fixed_length )
5467 minlen= longest_fixed_length;
5471 /* Several toplevels. Best we can is to set minlen. */
5473 struct regnode_charclass_class ch_class;
5476 DEBUG_PARSE_r(PerlIO_printf(Perl_debug_log, "\nMulti Top Level\n"));
5478 scan = ri->program + 1;
5479 cl_init(pRExC_state, &ch_class);
5480 data.start_class = &ch_class;
5481 data.last_closep = &last_close;
5484 minlen = study_chunk(pRExC_state, &scan, &minlen, &fake, scan + RExC_size,
5485 &data, -1, NULL, NULL, SCF_DO_STCLASS_AND|SCF_WHILEM_VISITED_POS,0);
5489 r->check_substr = r->check_utf8 = r->anchored_substr = r->anchored_utf8
5490 = r->float_substr = r->float_utf8 = NULL;
5492 if (!(data.start_class->flags & ANYOF_EOS)
5493 && !cl_is_anything(data.start_class))
5495 const U32 n = add_data(pRExC_state, 1, "f");
5496 data.start_class->flags |= ANYOF_IS_SYNTHETIC;
5498 Newx(RExC_rxi->data->data[n], 1,
5499 struct regnode_charclass_class);
5500 StructCopy(data.start_class,
5501 (struct regnode_charclass_class*)RExC_rxi->data->data[n],
5502 struct regnode_charclass_class);
5503 ri->regstclass = (regnode*)RExC_rxi->data->data[n];
5504 r->intflags &= ~PREGf_SKIP; /* Used in find_byclass(). */
5505 DEBUG_COMPILE_r({ SV* sv = sv_newmortal();
5506 regprop(r, sv, (regnode*)data.start_class);
5507 PerlIO_printf(Perl_debug_log,
5508 "synthetic stclass \"%s\".\n",
5509 SvPVX_const(sv));});
5513 /* Guard against an embedded (?=) or (?<=) with a longer minlen than
5514 the "real" pattern. */
5516 PerlIO_printf(Perl_debug_log,"minlen: %"IVdf" r->minlen:%"IVdf"\n",
5517 (IV)minlen, (IV)r->minlen);
5519 r->minlenret = minlen;
5520 if (r->minlen < minlen)
5523 if (RExC_seen & REG_SEEN_GPOS)
5524 r->extflags |= RXf_GPOS_SEEN;
5525 if (RExC_seen & REG_SEEN_LOOKBEHIND)
5526 r->extflags |= RXf_LOOKBEHIND_SEEN;
5527 if (RExC_seen & REG_SEEN_EVAL)
5528 r->extflags |= RXf_EVAL_SEEN;
5529 if (RExC_seen & REG_SEEN_CANY)
5530 r->extflags |= RXf_CANY_SEEN;
5531 if (RExC_seen & REG_SEEN_VERBARG)
5532 r->intflags |= PREGf_VERBARG_SEEN;
5533 if (RExC_seen & REG_SEEN_CUTGROUP)
5534 r->intflags |= PREGf_CUTGROUP_SEEN;
5535 if (RExC_paren_names)
5536 RXp_PAREN_NAMES(r) = MUTABLE_HV(SvREFCNT_inc(RExC_paren_names));
5538 RXp_PAREN_NAMES(r) = NULL;
5540 #ifdef STUPID_PATTERN_CHECKS
5541 if (RX_PRELEN(rx) == 0)
5542 r->extflags |= RXf_NULL;
5543 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
5544 /* XXX: this should happen BEFORE we compile */
5545 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
5546 else if (RX_PRELEN(rx) == 3 && memEQ("\\s+", RX_PRECOMP(rx), 3))
5547 r->extflags |= RXf_WHITE;
5548 else if (RX_PRELEN(rx) == 1 && RXp_PRECOMP(rx)[0] == '^')
5549 r->extflags |= RXf_START_ONLY;
5551 if (r->extflags & RXf_SPLIT && RX_PRELEN(rx) == 1 && RX_PRECOMP(rx)[0] == ' ')
5552 /* XXX: this should happen BEFORE we compile */
5553 r->extflags |= (RXf_SKIPWHITE|RXf_WHITE);
5555 regnode *first = ri->program + 1;
5558 if (PL_regkind[fop] == NOTHING && OP(NEXTOPER(first)) == END)
5559 r->extflags |= RXf_NULL;
5560 else if (PL_regkind[fop] == BOL && OP(NEXTOPER(first)) == END)
5561 r->extflags |= RXf_START_ONLY;
5562 else if (fop == PLUS && OP(NEXTOPER(first)) == SPACE
5563 && OP(regnext(first)) == END)
5564 r->extflags |= RXf_WHITE;
5568 if (RExC_paren_names) {
5569 ri->name_list_idx = add_data( pRExC_state, 1, "a" );
5570 ri->data->data[ri->name_list_idx] = (void*)SvREFCNT_inc(RExC_paren_name_list);
5573 ri->name_list_idx = 0;
5575 if (RExC_recurse_count) {
5576 for ( ; RExC_recurse_count ; RExC_recurse_count-- ) {
5577 const regnode *scan = RExC_recurse[RExC_recurse_count-1];
5578 ARG2L_SET( scan, RExC_open_parens[ARG(scan)-1] - scan );
5581 Newxz(r->offs, RExC_npar, regexp_paren_pair);
5582 /* assume we don't need to swap parens around before we match */
5585 PerlIO_printf(Perl_debug_log,"Final program:\n");
5588 #ifdef RE_TRACK_PATTERN_OFFSETS
5589 DEBUG_OFFSETS_r(if (ri->u.offsets) {
5590 const U32 len = ri->u.offsets[0];
5592 GET_RE_DEBUG_FLAGS_DECL;
5593 PerlIO_printf(Perl_debug_log, "Offsets: [%"UVuf"]\n\t", (UV)ri->u.offsets[0]);
5594 for (i = 1; i <= len; i++) {
5595 if (ri->u.offsets[i*2-1] || ri->u.offsets[i*2])
5596 PerlIO_printf(Perl_debug_log, "%"UVuf":%"UVuf"[%"UVuf"] ",
5597 (UV)i, (UV)ri->u.offsets[i*2-1], (UV)ri->u.offsets[i*2]);
5599 PerlIO_printf(Perl_debug_log, "\n");
5605 #undef RE_ENGINE_PTR
5609 Perl_reg_named_buff(pTHX_ REGEXP * const rx, SV * const key, SV * const value,
5612 PERL_ARGS_ASSERT_REG_NAMED_BUFF;
5614 PERL_UNUSED_ARG(value);
5616 if (flags & RXapif_FETCH) {
5617 return reg_named_buff_fetch(rx, key, flags);
5618 } else if (flags & (RXapif_STORE | RXapif_DELETE | RXapif_CLEAR)) {
5619 Perl_croak_no_modify(aTHX);
5621 } else if (flags & RXapif_EXISTS) {
5622 return reg_named_buff_exists(rx, key, flags)
5625 } else if (flags & RXapif_REGNAMES) {
5626 return reg_named_buff_all(rx, flags);
5627 } else if (flags & (RXapif_SCALAR | RXapif_REGNAMES_COUNT)) {
5628 return reg_named_buff_scalar(rx, flags);
5630 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff", (int)flags);
5636 Perl_reg_named_buff_iter(pTHX_ REGEXP * const rx, const SV * const lastkey,
5639 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ITER;
5640 PERL_UNUSED_ARG(lastkey);
5642 if (flags & RXapif_FIRSTKEY)
5643 return reg_named_buff_firstkey(rx, flags);
5644 else if (flags & RXapif_NEXTKEY)
5645 return reg_named_buff_nextkey(rx, flags);
5647 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_iter", (int)flags);
5653 Perl_reg_named_buff_fetch(pTHX_ REGEXP * const r, SV * const namesv,
5656 AV *retarray = NULL;
5658 struct regexp *const rx = (struct regexp *)SvANY(r);
5660 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FETCH;
5662 if (flags & RXapif_ALL)
5665 if (rx && RXp_PAREN_NAMES(rx)) {
5666 HE *he_str = hv_fetch_ent( RXp_PAREN_NAMES(rx), namesv, 0, 0 );
5669 SV* sv_dat=HeVAL(he_str);
5670 I32 *nums=(I32*)SvPVX(sv_dat);
5671 for ( i=0; i<SvIVX(sv_dat); i++ ) {
5672 if ((I32)(rx->nparens) >= nums[i]
5673 && rx->offs[nums[i]].start != -1
5674 && rx->offs[nums[i]].end != -1)
5677 CALLREG_NUMBUF_FETCH(r,nums[i],ret);
5682 ret = newSVsv(&PL_sv_undef);
5685 av_push(retarray, ret);
5688 return newRV_noinc(MUTABLE_SV(retarray));
5695 Perl_reg_named_buff_exists(pTHX_ REGEXP * const r, SV * const key,
5698 struct regexp *const rx = (struct regexp *)SvANY(r);
5700 PERL_ARGS_ASSERT_REG_NAMED_BUFF_EXISTS;
5702 if (rx && RXp_PAREN_NAMES(rx)) {
5703 if (flags & RXapif_ALL) {
5704 return hv_exists_ent(RXp_PAREN_NAMES(rx), key, 0);
5706 SV *sv = CALLREG_NAMED_BUFF_FETCH(r, key, flags);
5720 Perl_reg_named_buff_firstkey(pTHX_ REGEXP * const r, const U32 flags)
5722 struct regexp *const rx = (struct regexp *)SvANY(r);
5724 PERL_ARGS_ASSERT_REG_NAMED_BUFF_FIRSTKEY;
5726 if ( rx && RXp_PAREN_NAMES(rx) ) {
5727 (void)hv_iterinit(RXp_PAREN_NAMES(rx));
5729 return CALLREG_NAMED_BUFF_NEXTKEY(r, NULL, flags & ~RXapif_FIRSTKEY);
5736 Perl_reg_named_buff_nextkey(pTHX_ REGEXP * const r, const U32 flags)
5738 struct regexp *const rx = (struct regexp *)SvANY(r);
5739 GET_RE_DEBUG_FLAGS_DECL;
5741 PERL_ARGS_ASSERT_REG_NAMED_BUFF_NEXTKEY;
5743 if (rx && RXp_PAREN_NAMES(rx)) {
5744 HV *hv = RXp_PAREN_NAMES(rx);
5746 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5749 SV* sv_dat = HeVAL(temphe);
5750 I32 *nums = (I32*)SvPVX(sv_dat);
5751 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5752 if ((I32)(rx->lastparen) >= nums[i] &&
5753 rx->offs[nums[i]].start != -1 &&
5754 rx->offs[nums[i]].end != -1)
5760 if (parno || flags & RXapif_ALL) {
5761 return newSVhek(HeKEY_hek(temphe));
5769 Perl_reg_named_buff_scalar(pTHX_ REGEXP * const r, const U32 flags)
5774 struct regexp *const rx = (struct regexp *)SvANY(r);
5776 PERL_ARGS_ASSERT_REG_NAMED_BUFF_SCALAR;
5778 if (rx && RXp_PAREN_NAMES(rx)) {
5779 if (flags & (RXapif_ALL | RXapif_REGNAMES_COUNT)) {
5780 return newSViv(HvTOTALKEYS(RXp_PAREN_NAMES(rx)));
5781 } else if (flags & RXapif_ONE) {
5782 ret = CALLREG_NAMED_BUFF_ALL(r, (flags | RXapif_REGNAMES));
5783 av = MUTABLE_AV(SvRV(ret));
5784 length = av_len(av);
5786 return newSViv(length + 1);
5788 Perl_croak(aTHX_ "panic: Unknown flags %d in named_buff_scalar", (int)flags);
5792 return &PL_sv_undef;
5796 Perl_reg_named_buff_all(pTHX_ REGEXP * const r, const U32 flags)
5798 struct regexp *const rx = (struct regexp *)SvANY(r);
5801 PERL_ARGS_ASSERT_REG_NAMED_BUFF_ALL;
5803 if (rx && RXp_PAREN_NAMES(rx)) {
5804 HV *hv= RXp_PAREN_NAMES(rx);
5806 (void)hv_iterinit(hv);
5807 while ( (temphe = hv_iternext_flags(hv,0)) ) {
5810 SV* sv_dat = HeVAL(temphe);
5811 I32 *nums = (I32*)SvPVX(sv_dat);
5812 for ( i = 0; i < SvIVX(sv_dat); i++ ) {
5813 if ((I32)(rx->lastparen) >= nums[i] &&
5814 rx->offs[nums[i]].start != -1 &&
5815 rx->offs[nums[i]].end != -1)
5821 if (parno || flags & RXapif_ALL) {
5822 av_push(av, newSVhek(HeKEY_hek(temphe)));
5827 return newRV_noinc(MUTABLE_SV(av));
5831 Perl_reg_numbered_buff_fetch(pTHX_ REGEXP * const r, const I32 paren,
5834 struct regexp *const rx = (struct regexp *)SvANY(r);
5839 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_FETCH;
5842 sv_setsv(sv,&PL_sv_undef);
5846 if (paren == RX_BUFF_IDX_PREMATCH && rx->offs[0].start != -1) {
5848 i = rx->offs[0].start;
5852 if (paren == RX_BUFF_IDX_POSTMATCH && rx->offs[0].end != -1) {
5854 s = rx->subbeg + rx->offs[0].end;
5855 i = rx->sublen - rx->offs[0].end;
5858 if ( 0 <= paren && paren <= (I32)rx->nparens &&
5859 (s1 = rx->offs[paren].start) != -1 &&
5860 (t1 = rx->offs[paren].end) != -1)
5864 s = rx->subbeg + s1;
5866 sv_setsv(sv,&PL_sv_undef);
5869 assert(rx->sublen >= (s - rx->subbeg) + i );
5871 const int oldtainted = PL_tainted;
5873 sv_setpvn(sv, s, i);
5874 PL_tainted = oldtainted;
5875 if ( (rx->extflags & RXf_CANY_SEEN)
5876 ? (RXp_MATCH_UTF8(rx)
5877 && (!i || is_utf8_string((U8*)s, i)))
5878 : (RXp_MATCH_UTF8(rx)) )
5885 if (RXp_MATCH_TAINTED(rx)) {
5886 if (SvTYPE(sv) >= SVt_PVMG) {
5887 MAGIC* const mg = SvMAGIC(sv);
5890 SvMAGIC_set(sv, mg->mg_moremagic);
5892 if ((mgt = SvMAGIC(sv))) {
5893 mg->mg_moremagic = mgt;
5894 SvMAGIC_set(sv, mg);
5904 sv_setsv(sv,&PL_sv_undef);
5910 Perl_reg_numbered_buff_store(pTHX_ REGEXP * const rx, const I32 paren,
5911 SV const * const value)
5913 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_STORE;
5915 PERL_UNUSED_ARG(rx);
5916 PERL_UNUSED_ARG(paren);
5917 PERL_UNUSED_ARG(value);
5920 Perl_croak_no_modify(aTHX);
5924 Perl_reg_numbered_buff_length(pTHX_ REGEXP * const r, const SV * const sv,
5927 struct regexp *const rx = (struct regexp *)SvANY(r);
5931 PERL_ARGS_ASSERT_REG_NUMBERED_BUFF_LENGTH;
5933 /* Some of this code was originally in C<Perl_magic_len> in F<mg.c> */
5935 /* $` / ${^PREMATCH} */
5936 case RX_BUFF_IDX_PREMATCH:
5937 if (rx->offs[0].start != -1) {
5938 i = rx->offs[0].start;
5946 /* $' / ${^POSTMATCH} */
5947 case RX_BUFF_IDX_POSTMATCH:
5948 if (rx->offs[0].end != -1) {
5949 i = rx->sublen - rx->offs[0].end;
5951 s1 = rx->offs[0].end;
5957 /* $& / ${^MATCH}, $1, $2, ... */
5959 if (paren <= (I32)rx->nparens &&
5960 (s1 = rx->offs[paren].start) != -1 &&
5961 (t1 = rx->offs[paren].end) != -1)
5966 if (ckWARN(WARN_UNINITIALIZED))
5967 report_uninit((const SV *)sv);
5972 if (i > 0 && RXp_MATCH_UTF8(rx)) {
5973 const char * const s = rx->subbeg + s1;
5978 if (is_utf8_string_loclen((U8*)s, i, &ep, &el))
5985 Perl_reg_qr_package(pTHX_ REGEXP * const rx)
5987 PERL_ARGS_ASSERT_REG_QR_PACKAGE;
5988 PERL_UNUSED_ARG(rx);
5992 return newSVpvs("Regexp");
5995 /* Scans the name of a named buffer from the pattern.
5996 * If flags is REG_RSN_RETURN_NULL returns null.
5997 * If flags is REG_RSN_RETURN_NAME returns an SV* containing the name
5998 * If flags is REG_RSN_RETURN_DATA returns the data SV* corresponding
5999 * to the parsed name as looked up in the RExC_paren_names hash.
6000 * If there is an error throws a vFAIL().. type exception.
6003 #define REG_RSN_RETURN_NULL 0
6004 #define REG_RSN_RETURN_NAME 1
6005 #define REG_RSN_RETURN_DATA 2
6008 S_reg_scan_name(pTHX_ RExC_state_t *pRExC_state, U32 flags)
6010 char *name_start = RExC_parse;
6012 PERL_ARGS_ASSERT_REG_SCAN_NAME;
6014 if (isIDFIRST_lazy_if(RExC_parse, UTF)) {
6015 /* skip IDFIRST by using do...while */
6018 RExC_parse += UTF8SKIP(RExC_parse);
6019 } while (isALNUM_utf8((U8*)RExC_parse));
6023 } while (isALNUM(*RExC_parse));
6028 = newSVpvn_flags(name_start, (int)(RExC_parse - name_start),
6029 SVs_TEMP | (UTF ? SVf_UTF8 : 0));
6030 if ( flags == REG_RSN_RETURN_NAME)
6032 else if (flags==REG_RSN_RETURN_DATA) {
6035 if ( ! sv_name ) /* should not happen*/
6036 Perl_croak(aTHX_ "panic: no svname in reg_scan_name");
6037 if (RExC_paren_names)
6038 he_str = hv_fetch_ent( RExC_paren_names, sv_name, 0, 0 );
6040 sv_dat = HeVAL(he_str);
6042 vFAIL("Reference to nonexistent named group");
6046 Perl_croak(aTHX_ "panic: bad flag %lx in reg_scan_name",
6047 (unsigned long) flags);
6054 #define DEBUG_PARSE_MSG(funcname) DEBUG_PARSE_r({ \
6055 int rem=(int)(RExC_end - RExC_parse); \
6064 if (RExC_lastparse!=RExC_parse) \
6065 PerlIO_printf(Perl_debug_log," >%.*s%-*s", \
6068 iscut ? "..." : "<" \
6071 PerlIO_printf(Perl_debug_log,"%16s",""); \
6074 num = RExC_size + 1; \
6076 num=REG_NODE_NUM(RExC_emit); \
6077 if (RExC_lastnum!=num) \
6078 PerlIO_printf(Perl_debug_log,"|%4d",num); \
6080 PerlIO_printf(Perl_debug_log,"|%4s",""); \
6081 PerlIO_printf(Perl_debug_log,"|%*s%-4s", \
6082 (int)((depth*2)), "", \
6086 RExC_lastparse=RExC_parse; \
6091 #define DEBUG_PARSE(funcname) DEBUG_PARSE_r({ \
6092 DEBUG_PARSE_MSG((funcname)); \
6093 PerlIO_printf(Perl_debug_log,"%4s","\n"); \
6095 #define DEBUG_PARSE_FMT(funcname,fmt,args) DEBUG_PARSE_r({ \
6096 DEBUG_PARSE_MSG((funcname)); \
6097 PerlIO_printf(Perl_debug_log,fmt "\n",args); \
6100 /* This section of code defines the inversion list object and its methods. The
6101 * interfaces are highly subject to change, so as much as possible is static to
6102 * this file. An inversion list is here implemented as a malloc'd C UV array
6103 * with some added info that is placed as UVs at the beginning in a header
6104 * portion. An inversion list for Unicode is an array of code points, sorted
6105 * by ordinal number. The zeroth element is the first code point in the list.
6106 * The 1th element is the first element beyond that not in the list. In other
6107 * words, the first range is
6108 * invlist[0]..(invlist[1]-1)
6109 * The other ranges follow. Thus every element whose index is divisible by two
6110 * marks the beginning of a range that is in the list, and every element not
6111 * divisible by two marks the beginning of a range not in the list. A single
6112 * element inversion list that contains the single code point N generally
6113 * consists of two elements
6116 * (The exception is when N is the highest representable value on the
6117 * machine, in which case the list containing just it would be a single
6118 * element, itself. By extension, if the last range in the list extends to
6119 * infinity, then the first element of that range will be in the inversion list
6120 * at a position that is divisible by two, and is the final element in the
6122 * Taking the complement (inverting) an inversion list is quite simple, if the
6123 * first element is 0, remove it; otherwise add a 0 element at the beginning.
6124 * This implementation reserves an element at the beginning of each inversion list
6125 * to contain 0 when the list contains 0, and contains 1 otherwise. The actual
6126 * beginning of the list is either that element if 0, or the next one if 1.
6128 * More about inversion lists can be found in "Unicode Demystified"
6129 * Chapter 13 by Richard Gillam, published by Addison-Wesley.
6130 * More will be coming when functionality is added later.
6132 * The inversion list data structure is currently implemented as an SV pointing
6133 * to an array of UVs that the SV thinks are bytes. This allows us to have an
6134 * array of UV whose memory management is automatically handled by the existing
6135 * facilities for SV's.
6137 * Some of the methods should always be private to the implementation, and some
6138 * should eventually be made public */
6140 #define INVLIST_LEN_OFFSET 0 /* Number of elements in the inversion list */
6141 #define INVLIST_ITER_OFFSET 1 /* Current iteration position */
6143 #define INVLIST_ZERO_OFFSET 2 /* 0 or 1; must be last element in header */
6144 /* The UV at position ZERO contains either 0 or 1. If 0, the inversion list
6145 * contains the code point U+00000, and begins here. If 1, the inversion list
6146 * doesn't contain U+0000, and it begins at the next UV in the array.
6147 * Inverting an inversion list consists of adding or removing the 0 at the
6148 * beginning of it. By reserving a space for that 0, inversion can be made
6151 #define HEADER_LENGTH (INVLIST_ZERO_OFFSET + 1)
6153 /* Internally things are UVs */
6154 #define TO_INTERNAL_SIZE(x) ((x + HEADER_LENGTH) * sizeof(UV))
6155 #define FROM_INTERNAL_SIZE(x) ((x / sizeof(UV)) - HEADER_LENGTH)
6157 #define INVLIST_INITIAL_LEN 10
6159 PERL_STATIC_INLINE UV*
6160 S__invlist_array_init(pTHX_ SV* const invlist, const bool will_have_0)
6162 /* Returns a pointer to the first element in the inversion list's array.
6163 * This is called upon initialization of an inversion list. Where the
6164 * array begins depends on whether the list has the code point U+0000
6165 * in it or not. The other parameter tells it whether the code that
6166 * follows this call is about to put a 0 in the inversion list or not.
6167 * The first element is either the element with 0, if 0, or the next one,
6170 UV* zero = get_invlist_zero_addr(invlist);
6172 PERL_ARGS_ASSERT__INVLIST_ARRAY_INIT;
6175 assert(! *get_invlist_len_addr(invlist));
6177 /* 1^1 = 0; 1^0 = 1 */
6178 *zero = 1 ^ will_have_0;
6179 return zero + *zero;
6182 PERL_STATIC_INLINE UV*
6183 S_invlist_array(pTHX_ SV* const invlist)
6185 /* Returns the pointer to the inversion list's array. Every time the
6186 * length changes, this needs to be called in case malloc or realloc moved
6189 PERL_ARGS_ASSERT_INVLIST_ARRAY;
6191 /* Must not be empty. If these fail, you probably didn't check for <len>
6192 * being non-zero before trying to get the array */
6193 assert(*get_invlist_len_addr(invlist));
6194 assert(*get_invlist_zero_addr(invlist) == 0
6195 || *get_invlist_zero_addr(invlist) == 1);
6197 /* The array begins either at the element reserved for zero if the
6198 * list contains 0 (that element will be set to 0), or otherwise the next
6199 * element (in which case the reserved element will be set to 1). */
6200 return (UV *) (get_invlist_zero_addr(invlist)
6201 + *get_invlist_zero_addr(invlist));
6204 PERL_STATIC_INLINE UV*
6205 S_get_invlist_len_addr(pTHX_ SV* invlist)
6207 /* Return the address of the UV that contains the current number
6208 * of used elements in the inversion list */
6210 PERL_ARGS_ASSERT_GET_INVLIST_LEN_ADDR;
6212 return (UV *) (SvPVX(invlist) + (INVLIST_LEN_OFFSET * sizeof (UV)));
6215 PERL_STATIC_INLINE UV
6216 S_invlist_len(pTHX_ SV* const invlist)
6218 /* Returns the current number of elements stored in the inversion list's
6221 PERL_ARGS_ASSERT_INVLIST_LEN;
6223 return *get_invlist_len_addr(invlist);
6226 PERL_STATIC_INLINE void
6227 S_invlist_set_len(pTHX_ SV* const invlist, const UV len)
6229 /* Sets the current number of elements stored in the inversion list */
6231 PERL_ARGS_ASSERT_INVLIST_SET_LEN;
6233 *get_invlist_len_addr(invlist) = len;
6235 assert(len <= SvLEN(invlist));
6237 SvCUR_set(invlist, TO_INTERNAL_SIZE(len));
6238 /* If the list contains U+0000, that element is part of the header,
6239 * and should not be counted as part of the array. It will contain
6240 * 0 in that case, and 1 otherwise. So we could flop 0=>1, 1=>0 and
6242 * SvCUR_set(invlist,
6243 * TO_INTERNAL_SIZE(len
6244 * - (*get_invlist_zero_addr(inv_list) ^ 1)));
6245 * But, this is only valid if len is not 0. The consequences of not doing
6246 * this is that the memory allocation code may think that 1 more UV is
6247 * being used than actually is, and so might do an unnecessary grow. That
6248 * seems worth not bothering to make this the precise amount.
6250 * Note that when inverting, SvCUR shouldn't change */
6253 PERL_STATIC_INLINE UV
6254 S_invlist_max(pTHX_ SV* const invlist)
6256 /* Returns the maximum number of elements storable in the inversion list's
6257 * array, without having to realloc() */
6259 PERL_ARGS_ASSERT_INVLIST_MAX;
6261 return FROM_INTERNAL_SIZE(SvLEN(invlist));
6264 PERL_STATIC_INLINE UV*
6265 S_get_invlist_zero_addr(pTHX_ SV* invlist)
6267 /* Return the address of the UV that is reserved to hold 0 if the inversion
6268 * list contains 0. This has to be the last element of the heading, as the
6269 * list proper starts with either it if 0, or the next element if not.
6270 * (But we force it to contain either 0 or 1) */
6272 PERL_ARGS_ASSERT_GET_INVLIST_ZERO_ADDR;
6274 return (UV *) (SvPVX(invlist) + (INVLIST_ZERO_OFFSET * sizeof (UV)));
6277 #ifndef PERL_IN_XSUB_RE
6279 Perl__new_invlist(pTHX_ IV initial_size)
6282 /* Return a pointer to a newly constructed inversion list, with enough
6283 * space to store 'initial_size' elements. If that number is negative, a
6284 * system default is used instead */
6288 if (initial_size < 0) {
6289 initial_size = INVLIST_INITIAL_LEN;
6292 /* Allocate the initial space */
6293 new_list = newSV(TO_INTERNAL_SIZE(initial_size));
6294 invlist_set_len(new_list, 0);
6296 /* Force iterinit() to be used to get iteration to work */
6297 *get_invlist_iter_addr(new_list) = UV_MAX;
6299 /* This should force a segfault if a method doesn't initialize this
6301 *get_invlist_zero_addr(new_list) = UV_MAX;
6308 S_invlist_extend(pTHX_ SV* const invlist, const UV new_max)
6310 /* Grow the maximum size of an inversion list */
6312 PERL_ARGS_ASSERT_INVLIST_EXTEND;
6314 SvGROW((SV *)invlist, TO_INTERNAL_SIZE(new_max));
6317 PERL_STATIC_INLINE void
6318 S_invlist_trim(pTHX_ SV* const invlist)
6320 PERL_ARGS_ASSERT_INVLIST_TRIM;
6322 /* Change the length of the inversion list to how many entries it currently
6325 SvPV_shrink_to_cur((SV *) invlist);
6328 /* An element is in an inversion list iff its index is even numbered: 0, 2, 4,
6330 #define ELEMENT_RANGE_MATCHES_INVLIST(i) (! ((i) & 1))
6331 #define PREV_RANGE_MATCHES_INVLIST(i) (! ELEMENT_RANGE_MATCHES_INVLIST(i))
6333 #ifndef PERL_IN_XSUB_RE
6335 Perl__append_range_to_invlist(pTHX_ SV* const invlist, const UV start, const UV end)
6337 /* Subject to change or removal. Append the range from 'start' to 'end' at
6338 * the end of the inversion list. The range must be above any existing
6342 UV max = invlist_max(invlist);
6343 UV len = invlist_len(invlist);
6345 PERL_ARGS_ASSERT__APPEND_RANGE_TO_INVLIST;
6347 if (len == 0) { /* Empty lists must be initialized */
6348 array = _invlist_array_init(invlist, start == 0);
6351 /* Here, the existing list is non-empty. The current max entry in the
6352 * list is generally the first value not in the set, except when the
6353 * set extends to the end of permissible values, in which case it is
6354 * the first entry in that final set, and so this call is an attempt to
6355 * append out-of-order */
6357 UV final_element = len - 1;
6358 array = invlist_array(invlist);
6359 if (array[final_element] > start
6360 || ELEMENT_RANGE_MATCHES_INVLIST(final_element))
6362 Perl_croak(aTHX_ "panic: attempting to append to an inversion list, but wasn't at the end of the list, final=%"UVuf", start=%"UVuf", match=%c",
6363 array[final_element], start,
6364 ELEMENT_RANGE_MATCHES_INVLIST(final_element) ? 't' : 'f');
6367 /* Here, it is a legal append. If the new range begins with the first
6368 * value not in the set, it is extending the set, so the new first
6369 * value not in the set is one greater than the newly extended range.
6371 if (array[final_element] == start) {
6372 if (end != UV_MAX) {
6373 array[final_element] = end + 1;
6376 /* But if the end is the maximum representable on the machine,
6377 * just let the range that this would extend to have no end */
6378 invlist_set_len(invlist, len - 1);
6384 /* Here the new range doesn't extend any existing set. Add it */
6386 len += 2; /* Includes an element each for the start and end of range */
6388 /* If overflows the existing space, extend, which may cause the array to be
6391 invlist_extend(invlist, len);
6392 invlist_set_len(invlist, len); /* Have to set len here to avoid assert
6393 failure in invlist_array() */
6394 array = invlist_array(invlist);
6397 invlist_set_len(invlist, len);
6400 /* The next item on the list starts the range, the one after that is
6401 * one past the new range. */
6402 array[len - 2] = start;
6403 if (end != UV_MAX) {
6404 array[len - 1] = end + 1;
6407 /* But if the end is the maximum representable on the machine, just let
6408 * the range have no end */
6409 invlist_set_len(invlist, len - 1);
6414 S_invlist_search(pTHX_ SV* const invlist, const UV cp)
6416 /* Searches the inversion list for the entry that contains the input code
6417 * point <cp>. If <cp> is not in the list, -1 is returned. Otherwise, the
6418 * return value is the index into the list's array of the range that
6422 IV high = invlist_len(invlist);
6423 const UV * const array = invlist_array(invlist);
6425 PERL_ARGS_ASSERT_INVLIST_SEARCH;
6427 /* If list is empty or the code point is before the first element, return
6429 if (high == 0 || cp < array[0]) {
6433 /* Binary search. What we are looking for is <i> such that
6434 * array[i] <= cp < array[i+1]
6435 * The loop below converges on the i+1. */
6436 while (low < high) {
6437 IV mid = (low + high) / 2;
6438 if (array[mid] <= cp) {
6441 /* We could do this extra test to exit the loop early.
6442 if (cp < array[low]) {
6447 else { /* cp < array[mid] */
6456 Perl__invlist_populate_swatch(pTHX_ SV* const invlist, const UV start, const UV end, U8* swatch)
6458 /* populates a swatch of a swash the same way swatch_get() does in utf8.c,
6459 * but is used when the swash has an inversion list. This makes this much
6460 * faster, as it uses a binary search instead of a linear one. This is
6461 * intimately tied to that function, and perhaps should be in utf8.c,
6462 * except it is intimately tied to inversion lists as well. It assumes
6463 * that <swatch> is all 0's on input */
6466 const IV len = invlist_len(invlist);
6470 PERL_ARGS_ASSERT__INVLIST_POPULATE_SWATCH;
6472 if (len == 0) { /* Empty inversion list */
6476 array = invlist_array(invlist);
6478 /* Find which element it is */
6479 i = invlist_search(invlist, start);
6481 /* We populate from <start> to <end> */
6482 while (current < end) {
6485 /* The inversion list gives the results for every possible code point
6486 * after the first one in the list. Only those ranges whose index is
6487 * even are ones that the inversion list matches. For the odd ones,
6488 * and if the initial code point is not in the list, we have to skip
6489 * forward to the next element */
6490 if (i == -1 || ! ELEMENT_RANGE_MATCHES_INVLIST(i)) {
6492 if (i >= len) { /* Finished if beyond the end of the array */
6496 if (current >= end) { /* Finished if beyond the end of what we
6501 assert(current >= start);
6503 /* The current range ends one below the next one, except don't go past
6506 upper = (i < len && array[i] < end) ? array[i] : end;
6508 /* Here we are in a range that matches. Populate a bit in the 3-bit U8
6509 * for each code point in it */
6510 for (; current < upper; current++) {
6511 const STRLEN offset = (STRLEN)(current - start);
6512 swatch[offset >> 3] |= 1 << (offset & 7);
6515 /* Quit if at the end of the list */
6518 /* But first, have to deal with the highest possible code point on
6519 * the platform. The previous code assumes that <end> is one
6520 * beyond where we want to populate, but that is impossible at the
6521 * platform's infinity, so have to handle it specially */
6522 if (UNLIKELY(end == UV_MAX && ELEMENT_RANGE_MATCHES_INVLIST(len-1)))
6524 const STRLEN offset = (STRLEN)(end - start);
6525 swatch[offset >> 3] |= 1 << (offset & 7);
6530 /* Advance to the next range, which will be for code points not in the
6539 Perl__invlist_union(pTHX_ SV* const a, SV* const b, SV** output)
6541 /* Take the union of two inversion lists and point <output> to it. *output
6542 * should be defined upon input, and if it points to one of the two lists,
6543 * the reference count to that list will be decremented.
6544 * The basis for this comes from "Unicode Demystified" Chapter 13 by
6545 * Richard Gillam, published by Addison-Wesley, and explained at some
6546 * length there. The preface says to incorporate its examples into your
6547 * code at your own risk.
6549 * The algorithm is like a merge sort.
6551 * XXX A potential performance improvement is to keep track as we go along
6552 * if only one of the inputs contributes to the result, meaning the other
6553 * is a subset of that one. In that case, we can skip the final copy and
6554 * return the larger of the input lists, but then outside code might need
6555 * to keep track of whether to free the input list or not */
6557 UV* array_a; /* a's array */
6559 UV len_a; /* length of a's array */
6562 SV* u; /* the resulting union */
6566 UV i_a = 0; /* current index into a's array */
6570 /* running count, as explained in the algorithm source book; items are
6571 * stopped accumulating and are output when the count changes to/from 0.
6572 * The count is incremented when we start a range that's in the set, and
6573 * decremented when we start a range that's not in the set. So its range
6574 * is 0 to 2. Only when the count is zero is something not in the set.
6578 PERL_ARGS_ASSERT__INVLIST_UNION;
6581 /* If either one is empty, the union is the other one */
6582 len_a = invlist_len(a);
6588 *output = invlist_clone(b);
6589 } /* else *output already = b; */
6592 else if ((len_b = invlist_len(b)) == 0) {
6597 *output = invlist_clone(a);
6599 /* else *output already = a; */
6603 /* Here both lists exist and are non-empty */
6604 array_a = invlist_array(a);
6605 array_b = invlist_array(b);
6607 /* Size the union for the worst case: that the sets are completely
6609 u = _new_invlist(len_a + len_b);
6611 /* Will contain U+0000 if either component does */
6612 array_u = _invlist_array_init(u, (len_a > 0 && array_a[0] == 0)
6613 || (len_b > 0 && array_b[0] == 0));
6615 /* Go through each list item by item, stopping when exhausted one of
6617 while (i_a < len_a && i_b < len_b) {
6618 UV cp; /* The element to potentially add to the union's array */
6619 bool cp_in_set; /* is it in the the input list's set or not */
6621 /* We need to take one or the other of the two inputs for the union.
6622 * Since we are merging two sorted lists, we take the smaller of the
6623 * next items. In case of a tie, we take the one that is in its set
6624 * first. If we took one not in the set first, it would decrement the
6625 * count, possibly to 0 which would cause it to be output as ending the
6626 * range, and the next time through we would take the same number, and
6627 * output it again as beginning the next range. By doing it the
6628 * opposite way, there is no possibility that the count will be
6629 * momentarily decremented to 0, and thus the two adjoining ranges will
6630 * be seamlessly merged. (In a tie and both are in the set or both not
6631 * in the set, it doesn't matter which we take first.) */
6632 if (array_a[i_a] < array_b[i_b]
6633 || (array_a[i_a] == array_b[i_b]
6634 && ELEMENT_RANGE_MATCHES_INVLIST(i_a)))
6636 cp_in_set = ELEMENT_RANGE_MATCHES_INVLIST(i_a);
6640 cp_in_set = ELEMENT_RANGE_MATCHES_INVLIST(i_b);
6644 /* Here, have chosen which of the two inputs to look at. Only output
6645 * if the running count changes to/from 0, which marks the
6646 * beginning/end of a range in that's in the set */
6649 array_u[i_u++] = cp;
6656 array_u[i_u++] = cp;
6661 /* Here, we are finished going through at least one of the lists, which
6662 * means there is something remaining in at most one. We check if the list
6663 * that hasn't been exhausted is positioned such that we are in the middle
6664 * of a range in its set or not. (i_a and i_b point to the element beyond
6665 * the one we care about.) If in the set, we decrement 'count'; if 0, there
6666 * is potentially more to output.
6667 * There are four cases:
6668 * 1) Both weren't in their sets, count is 0, and remains 0. What's left
6669 * in the union is entirely from the non-exhausted set.
6670 * 2) Both were in their sets, count is 2. Nothing further should
6671 * be output, as everything that remains will be in the exhausted
6672 * list's set, hence in the union; decrementing to 1 but not 0 insures
6674 * 3) the exhausted was in its set, non-exhausted isn't, count is 1.
6675 * Nothing further should be output because the union includes
6676 * everything from the exhausted set. Not decrementing ensures that.
6677 * 4) the exhausted wasn't in its set, non-exhausted is, count is 1;
6678 * decrementing to 0 insures that we look at the remainder of the
6679 * non-exhausted set */
6680 if ((i_a != len_a && PREV_RANGE_MATCHES_INVLIST(i_a))
6681 || (i_b != len_b && PREV_RANGE_MATCHES_INVLIST(i_b)))
6686 /* The final length is what we've output so far, plus what else is about to
6687 * be output. (If 'count' is non-zero, then the input list we exhausted
6688 * has everything remaining up to the machine's limit in its set, and hence
6689 * in the union, so there will be no further output. */
6692 /* At most one of the subexpressions will be non-zero */
6693 len_u += (len_a - i_a) + (len_b - i_b);
6696 /* Set result to final length, which can change the pointer to array_u, so
6698 if (len_u != invlist_len(u)) {
6699 invlist_set_len(u, len_u);
6701 array_u = invlist_array(u);
6704 /* When 'count' is 0, the list that was exhausted (if one was shorter than
6705 * the other) ended with everything above it not in its set. That means
6706 * that the remaining part of the union is precisely the same as the
6707 * non-exhausted list, so can just copy it unchanged. (If both list were
6708 * exhausted at the same time, then the operations below will be both 0.)
6711 IV copy_count; /* At most one will have a non-zero copy count */
6712 if ((copy_count = len_a - i_a) > 0) {
6713 Copy(array_a + i_a, array_u + i_u, copy_count, UV);
6715 else if ((copy_count = len_b - i_b) > 0) {
6716 Copy(array_b + i_b, array_u + i_u, copy_count, UV);
6720 /* We may be removing a reference to one of the inputs */
6721 if (a == *output || b == *output) {
6722 SvREFCNT_dec(*output);
6730 Perl__invlist_intersection(pTHX_ SV* const a, SV* const b, SV** i)
6732 /* Take the intersection of two inversion lists and point <i> to it. *i
6733 * should be defined upon input, and if it points to one of the two lists,
6734 * the reference count to that list will be decremented.
6735 * The basis for this comes from "Unicode Demystified" Chapter 13 by
6736 * Richard Gillam, published by Addison-Wesley, and explained at some
6737 * length there. The preface says to incorporate its examples into your
6738 * code at your own risk. In fact, it had bugs
6740 * The algorithm is like a merge sort, and is essentially the same as the
6744 UV* array_a; /* a's array */
6746 UV len_a; /* length of a's array */
6749 SV* r; /* the resulting intersection */
6753 UV i_a = 0; /* current index into a's array */
6757 /* running count, as explained in the algorithm source book; items are
6758 * stopped accumulating and are output when the count changes to/from 2.
6759 * The count is incremented when we start a range that's in the set, and
6760 * decremented when we start a range that's not in the set. So its range
6761 * is 0 to 2. Only when the count is 2 is something in the intersection.
6765 PERL_ARGS_ASSERT__INVLIST_INTERSECTION;
6768 /* If either one is empty, the intersection is null */
6769 len_a = invlist_len(a);
6770 if ((len_a == 0) || ((len_b = invlist_len(b)) == 0)) {
6772 /* If the result is the same as one of the inputs, the input is being
6781 *i = _new_invlist(0);
6785 /* Here both lists exist and are non-empty */
6786 array_a = invlist_array(a);
6787 array_b = invlist_array(b);
6789 /* Size the intersection for the worst case: that the intersection ends up
6790 * fragmenting everything to be completely disjoint */
6791 r= _new_invlist(len_a + len_b);
6793 /* Will contain U+0000 iff both components do */
6794 array_r = _invlist_array_init(r, len_a > 0 && array_a[0] == 0
6795 && len_b > 0 && array_b[0] == 0);
6797 /* Go through each list item by item, stopping when exhausted one of
6799 while (i_a < len_a && i_b < len_b) {
6800 UV cp; /* The element to potentially add to the intersection's
6802 bool cp_in_set; /* Is it in the input list's set or not */
6804 /* We need to take one or the other of the two inputs for the
6805 * intersection. Since we are merging two sorted lists, we take the
6806 * smaller of the next items. In case of a tie, we take the one that
6807 * is not in its set first (a difference from the union algorithm). If
6808 * we took one in the set first, it would increment the count, possibly
6809 * to 2 which would cause it to be output as starting a range in the
6810 * intersection, and the next time through we would take that same
6811 * number, and output it again as ending the set. By doing it the
6812 * opposite of this, there is no possibility that the count will be
6813 * momentarily incremented to 2. (In a tie and both are in the set or
6814 * both not in the set, it doesn't matter which we take first.) */
6815 if (array_a[i_a] < array_b[i_b]
6816 || (array_a[i_a] == array_b[i_b]
6817 && ! ELEMENT_RANGE_MATCHES_INVLIST(i_a)))
6819 cp_in_set = ELEMENT_RANGE_MATCHES_INVLIST(i_a);
6823 cp_in_set = ELEMENT_RANGE_MATCHES_INVLIST(i_b);
6827 /* Here, have chosen which of the two inputs to look at. Only output
6828 * if the running count changes to/from 2, which marks the
6829 * beginning/end of a range that's in the intersection */
6833 array_r[i_r++] = cp;
6838 array_r[i_r++] = cp;
6844 /* Here, we are finished going through at least one of the lists, which
6845 * means there is something remaining in at most one. We check if the list
6846 * that has been exhausted is positioned such that we are in the middle
6847 * of a range in its set or not. (i_a and i_b point to elements 1 beyond
6848 * the ones we care about.) There are four cases:
6849 * 1) Both weren't in their sets, count is 0, and remains 0. There's
6850 * nothing left in the intersection.
6851 * 2) Both were in their sets, count is 2 and perhaps is incremented to
6852 * above 2. What should be output is exactly that which is in the
6853 * non-exhausted set, as everything it has is also in the intersection
6854 * set, and everything it doesn't have can't be in the intersection
6855 * 3) The exhausted was in its set, non-exhausted isn't, count is 1, and
6856 * gets incremented to 2. Like the previous case, the intersection is
6857 * everything that remains in the non-exhausted set.
6858 * 4) the exhausted wasn't in its set, non-exhausted is, count is 1, and
6859 * remains 1. And the intersection has nothing more. */
6860 if ((i_a == len_a && PREV_RANGE_MATCHES_INVLIST(i_a))
6861 || (i_b == len_b && PREV_RANGE_MATCHES_INVLIST(i_b)))
6866 /* The final length is what we've output so far plus what else is in the
6867 * intersection. At most one of the subexpressions below will be non-zero */
6870 len_r += (len_a - i_a) + (len_b - i_b);
6873 /* Set result to final length, which can change the pointer to array_r, so
6875 if (len_r != invlist_len(r)) {
6876 invlist_set_len(r, len_r);
6878 array_r = invlist_array(r);
6881 /* Finish outputting any remaining */
6882 if (count >= 2) { /* At most one will have a non-zero copy count */
6884 if ((copy_count = len_a - i_a) > 0) {
6885 Copy(array_a + i_a, array_r + i_r, copy_count, UV);
6887 else if ((copy_count = len_b - i_b) > 0) {
6888 Copy(array_b + i_b, array_r + i_r, copy_count, UV);
6892 /* We may be removing a reference to one of the inputs */
6893 if (a == *i || b == *i) {
6904 S_add_range_to_invlist(pTHX_ SV* invlist, const UV start, const UV end)
6906 /* Add the range from 'start' to 'end' inclusive to the inversion list's
6907 * set. A pointer to the inversion list is returned. This may actually be
6908 * a new list, in which case the passed in one has been destroyed. The
6909 * passed in inversion list can be NULL, in which case a new one is created
6910 * with just the one range in it */
6915 if (invlist == NULL) {
6916 invlist = _new_invlist(2);
6920 len = invlist_len(invlist);
6923 /* If comes after the final entry, can just append it to the end */
6925 || start >= invlist_array(invlist)
6926 [invlist_len(invlist) - 1])
6928 _append_range_to_invlist(invlist, start, end);
6932 /* Here, can't just append things, create and return a new inversion list
6933 * which is the union of this range and the existing inversion list */
6934 range_invlist = _new_invlist(2);
6935 _append_range_to_invlist(range_invlist, start, end);
6937 _invlist_union(invlist, range_invlist, &invlist);
6939 /* The temporary can be freed */
6940 SvREFCNT_dec(range_invlist);
6945 PERL_STATIC_INLINE SV*
6946 S_add_cp_to_invlist(pTHX_ SV* invlist, const UV cp) {
6947 return add_range_to_invlist(invlist, cp, cp);
6950 #ifndef PERL_IN_XSUB_RE
6952 Perl__invlist_invert(pTHX_ SV* const invlist)
6954 /* Complement the input inversion list. This adds a 0 if the list didn't
6955 * have a zero; removes it otherwise. As described above, the data
6956 * structure is set up so that this is very efficient */
6958 UV* len_pos = get_invlist_len_addr(invlist);
6960 PERL_ARGS_ASSERT__INVLIST_INVERT;
6962 /* The inverse of matching nothing is matching everything */
6963 if (*len_pos == 0) {
6964 _append_range_to_invlist(invlist, 0, UV_MAX);
6968 /* The exclusive or complents 0 to 1; and 1 to 0. If the result is 1, the
6969 * zero element was a 0, so it is being removed, so the length decrements
6970 * by 1; and vice-versa. SvCUR is unaffected */
6971 if (*get_invlist_zero_addr(invlist) ^= 1) {
6980 Perl__invlist_invert_prop(pTHX_ SV* const invlist)
6982 /* Complement the input inversion list (which must be a Unicode property,
6983 * all of which don't match above the Unicode maximum code point.) And
6984 * Perl has chosen to not have the inversion match above that either. This
6985 * adds a 0x110000 if the list didn't end with it, and removes it if it did
6991 PERL_ARGS_ASSERT__INVLIST_INVERT_PROP;
6993 _invlist_invert(invlist);
6995 len = invlist_len(invlist);
6997 if (len != 0) { /* If empty do nothing */
6998 array = invlist_array(invlist);
6999 if (array[len - 1] != PERL_UNICODE_MAX + 1) {
7000 /* Add 0x110000. First, grow if necessary */
7002 if (invlist_max(invlist) < len) {
7003 invlist_extend(invlist, len);
7004 array = invlist_array(invlist);
7006 invlist_set_len(invlist, len);
7007 array[len - 1] = PERL_UNICODE_MAX + 1;
7009 else { /* Remove the 0x110000 */
7010 invlist_set_len(invlist, len - 1);
7018 PERL_STATIC_INLINE SV*
7019 S_invlist_clone(pTHX_ SV* const invlist)
7022 /* Return a new inversion list that is a copy of the input one, which is
7025 /* Need to allocate extra space to accommodate Perl's addition of a
7026 * trailing NUL to SvPV's, since it thinks they are always strings */
7027 SV* new_invlist = _new_invlist(invlist_len(invlist) + 1);
7028 STRLEN length = SvCUR(invlist);
7030 PERL_ARGS_ASSERT_INVLIST_CLONE;
7032 SvCUR_set(new_invlist, length); /* This isn't done automatically */
7033 Copy(SvPVX(invlist), SvPVX(new_invlist), length, char);
7038 #ifndef PERL_IN_XSUB_RE
7040 Perl__invlist_subtract(pTHX_ SV* const a, SV* const b, SV** result)
7042 /* Point <result> to an inversion list which consists of all elements in
7043 * <a> that aren't also in <b>. *result should be defined upon input, and
7044 * if it points to C<b> its reference count will be decremented. */
7046 PERL_ARGS_ASSERT__INVLIST_SUBTRACT;
7049 /* Subtracting nothing retains the original */
7050 if (invlist_len(b) == 0) {
7056 /* If the result is not to be the same variable as the original, create
7059 *result = invlist_clone(a);
7062 SV *b_copy = invlist_clone(b);
7063 _invlist_invert(b_copy); /* Everything not in 'b' */
7069 _invlist_intersection(a, b_copy, result); /* Everything in 'a' not in
7071 SvREFCNT_dec(b_copy);
7078 PERL_STATIC_INLINE UV*
7079 S_get_invlist_iter_addr(pTHX_ SV* invlist)
7081 /* Return the address of the UV that contains the current iteration
7084 PERL_ARGS_ASSERT_GET_INVLIST_ITER_ADDR;
7086 return (UV *) (SvPVX(invlist) + (INVLIST_ITER_OFFSET * sizeof (UV)));
7089 PERL_STATIC_INLINE void
7090 S_invlist_iterinit(pTHX_ SV* invlist) /* Initialize iterator for invlist */
7092 PERL_ARGS_ASSERT_INVLIST_ITERINIT;
7094 *get_invlist_iter_addr(invlist) = 0;
7098 S_invlist_iternext(pTHX_ SV* invlist, UV* start, UV* end)
7100 /* An C<invlist_iterinit> call on <invlist> must be used to set this up.
7101 * This call sets in <*start> and <*end>, the next range in <invlist>.
7102 * Returns <TRUE> if successful and the next call will return the next
7103 * range; <FALSE> if was already at the end of the list. If the latter,
7104 * <*start> and <*end> are unchanged, and the next call to this function
7105 * will start over at the beginning of the list */
7107 UV* pos = get_invlist_iter_addr(invlist);
7108 UV len = invlist_len(invlist);
7111 PERL_ARGS_ASSERT_INVLIST_ITERNEXT;
7114 *pos = UV_MAX; /* Force iternit() to be required next time */
7118 array = invlist_array(invlist);
7120 *start = array[(*pos)++];
7126 *end = array[(*pos)++] - 1;
7132 #ifndef PERL_IN_XSUB_RE
7134 Perl__invlist_contents(pTHX_ SV* const invlist)
7136 /* Get the contents of an inversion list into a string SV so that they can
7137 * be printed out. It uses the format traditionally done for debug tracing
7141 SV* output = newSVpvs("\n");
7143 PERL_ARGS_ASSERT__INVLIST_CONTENTS;
7145 invlist_iterinit(invlist);
7146 while (invlist_iternext(invlist, &start, &end)) {
7147 if (end == UV_MAX) {
7148 Perl_sv_catpvf(aTHX_ output, "%04"UVXf"\tINFINITY\n", start);
7150 else if (end != start) {
7151 Perl_sv_catpvf(aTHX_ output, "%04"UVXf"\t%04"UVXf"\n",
7155 Perl_sv_catpvf(aTHX_ output, "%04"UVXf"\n", start);
7165 S_invlist_dump(pTHX_ SV* const invlist, const char * const header)
7167 /* Dumps out the ranges in an inversion list. The string 'header'
7168 * if present is output on a line before the first range */
7172 if (header && strlen(header)) {
7173 PerlIO_printf(Perl_debug_log, "%s\n", header);
7175 invlist_iterinit(invlist);
7176 while (invlist_iternext(invlist, &start, &end)) {
7177 if (end == UV_MAX) {
7178 PerlIO_printf(Perl_debug_log, "0x%04"UVXf" .. INFINITY\n", start);
7181 PerlIO_printf(Perl_debug_log, "0x%04"UVXf" .. 0x%04"UVXf"\n", start, end);
7187 #undef HEADER_LENGTH
7188 #undef INVLIST_INITIAL_LENGTH
7189 #undef TO_INTERNAL_SIZE
7190 #undef FROM_INTERNAL_SIZE
7191 #undef INVLIST_LEN_OFFSET
7192 #undef INVLIST_ZERO_OFFSET
7193 #undef INVLIST_ITER_OFFSET
7195 /* End of inversion list object */
7198 - reg - regular expression, i.e. main body or parenthesized thing
7200 * Caller must absorb opening parenthesis.
7202 * Combining parenthesis handling with the base level of regular expression
7203 * is a trifle forced, but the need to tie the tails of the branches to what
7204 * follows makes it hard to avoid.
7206 #define REGTAIL(x,y,z) regtail((x),(y),(z),depth+1)
7208 #define REGTAIL_STUDY(x,y,z) regtail_study((x),(y),(z),depth+1)
7210 #define REGTAIL_STUDY(x,y,z) regtail((x),(y),(z),depth+1)
7214 S_reg(pTHX_ RExC_state_t *pRExC_state, I32 paren, I32 *flagp,U32 depth)
7215 /* paren: Parenthesized? 0=top, 1=(, inside: changed to letter. */
7218 register regnode *ret; /* Will be the head of the group. */
7219 register regnode *br;
7220 register regnode *lastbr;
7221 register regnode *ender = NULL;
7222 register I32 parno = 0;
7224 U32 oregflags = RExC_flags;
7225 bool have_branch = 0;
7227 I32 freeze_paren = 0;
7228 I32 after_freeze = 0;
7230 /* for (?g), (?gc), and (?o) warnings; warning
7231 about (?c) will warn about (?g) -- japhy */
7233 #define WASTED_O 0x01
7234 #define WASTED_G 0x02
7235 #define WASTED_C 0x04
7236 #define WASTED_GC (0x02|0x04)
7237 I32 wastedflags = 0x00;
7239 char * parse_start = RExC_parse; /* MJD */
7240 char * const oregcomp_parse = RExC_parse;
7242 GET_RE_DEBUG_FLAGS_DECL;
7244 PERL_ARGS_ASSERT_REG;
7245 DEBUG_PARSE("reg ");
7247 *flagp = 0; /* Tentatively. */
7250 /* Make an OPEN node, if parenthesized. */
7252 if ( *RExC_parse == '*') { /* (*VERB:ARG) */
7253 char *start_verb = RExC_parse;
7254 STRLEN verb_len = 0;
7255 char *start_arg = NULL;
7256 unsigned char op = 0;
7258 int internal_argval = 0; /* internal_argval is only useful if !argok */
7259 while ( *RExC_parse && *RExC_parse != ')' ) {
7260 if ( *RExC_parse == ':' ) {
7261 start_arg = RExC_parse + 1;
7267 verb_len = RExC_parse - start_verb;
7270 while ( *RExC_parse && *RExC_parse != ')' )
7272 if ( *RExC_parse != ')' )
7273 vFAIL("Unterminated verb pattern argument");
7274 if ( RExC_parse == start_arg )
7277 if ( *RExC_parse != ')' )
7278 vFAIL("Unterminated verb pattern");
7281 switch ( *start_verb ) {
7282 case 'A': /* (*ACCEPT) */
7283 if ( memEQs(start_verb,verb_len,"ACCEPT") ) {
7285 internal_argval = RExC_nestroot;
7288 case 'C': /* (*COMMIT) */
7289 if ( memEQs(start_verb,verb_len,"COMMIT") )
7292 case 'F': /* (*FAIL) */
7293 if ( verb_len==1 || memEQs(start_verb,verb_len,"FAIL") ) {
7298 case ':': /* (*:NAME) */
7299 case 'M': /* (*MARK:NAME) */
7300 if ( verb_len==0 || memEQs(start_verb,verb_len,"MARK") ) {
7305 case 'P': /* (*PRUNE) */
7306 if ( memEQs(start_verb,verb_len,"PRUNE") )
7309 case 'S': /* (*SKIP) */
7310 if ( memEQs(start_verb,verb_len,"SKIP") )
7313 case 'T': /* (*THEN) */
7314 /* [19:06] <TimToady> :: is then */
7315 if ( memEQs(start_verb,verb_len,"THEN") ) {
7317 RExC_seen |= REG_SEEN_CUTGROUP;
7323 vFAIL3("Unknown verb pattern '%.*s'",
7324 verb_len, start_verb);
7327 if ( start_arg && internal_argval ) {
7328 vFAIL3("Verb pattern '%.*s' may not have an argument",
7329 verb_len, start_verb);
7330 } else if ( argok < 0 && !start_arg ) {
7331 vFAIL3("Verb pattern '%.*s' has a mandatory argument",
7332 verb_len, start_verb);
7334 ret = reganode(pRExC_state, op, internal_argval);
7335 if ( ! internal_argval && ! SIZE_ONLY ) {
7337 SV *sv = newSVpvn( start_arg, RExC_parse - start_arg);
7338 ARG(ret) = add_data( pRExC_state, 1, "S" );
7339 RExC_rxi->data->data[ARG(ret)]=(void*)sv;
7346 if (!internal_argval)
7347 RExC_seen |= REG_SEEN_VERBARG;
7348 } else if ( start_arg ) {
7349 vFAIL3("Verb pattern '%.*s' may not have an argument",
7350 verb_len, start_verb);
7352 ret = reg_node(pRExC_state, op);
7354 nextchar(pRExC_state);
7357 if (*RExC_parse == '?') { /* (?...) */
7358 bool is_logical = 0;
7359 const char * const seqstart = RExC_parse;
7360 bool has_use_defaults = FALSE;
7363 paren = *RExC_parse++;
7364 ret = NULL; /* For look-ahead/behind. */
7367 case 'P': /* (?P...) variants for those used to PCRE/Python */
7368 paren = *RExC_parse++;
7369 if ( paren == '<') /* (?P<...>) named capture */
7371 else if (paren == '>') { /* (?P>name) named recursion */
7372 goto named_recursion;
7374 else if (paren == '=') { /* (?P=...) named backref */
7375 /* this pretty much dupes the code for \k<NAME> in regatom(), if
7376 you change this make sure you change that */
7377 char* name_start = RExC_parse;
7379 SV *sv_dat = reg_scan_name(pRExC_state,
7380 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7381 if (RExC_parse == name_start || *RExC_parse != ')')
7382 vFAIL2("Sequence %.3s... not terminated",parse_start);
7385 num = add_data( pRExC_state, 1, "S" );
7386 RExC_rxi->data->data[num]=(void*)sv_dat;
7387 SvREFCNT_inc_simple_void(sv_dat);
7390 ret = reganode(pRExC_state,
7393 : (MORE_ASCII_RESTRICTED)
7395 : (AT_LEAST_UNI_SEMANTICS)
7403 Set_Node_Offset(ret, parse_start+1);
7404 Set_Node_Cur_Length(ret); /* MJD */
7406 nextchar(pRExC_state);
7410 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
7412 case '<': /* (?<...) */
7413 if (*RExC_parse == '!')
7415 else if (*RExC_parse != '=')
7421 case '\'': /* (?'...') */
7422 name_start= RExC_parse;
7423 svname = reg_scan_name(pRExC_state,
7424 SIZE_ONLY ? /* reverse test from the others */
7425 REG_RSN_RETURN_NAME :
7426 REG_RSN_RETURN_NULL);
7427 if (RExC_parse == name_start) {
7429 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
7432 if (*RExC_parse != paren)
7433 vFAIL2("Sequence (?%c... not terminated",
7434 paren=='>' ? '<' : paren);
7438 if (!svname) /* shouldn't happen */
7440 "panic: reg_scan_name returned NULL");
7441 if (!RExC_paren_names) {
7442 RExC_paren_names= newHV();
7443 sv_2mortal(MUTABLE_SV(RExC_paren_names));
7445 RExC_paren_name_list= newAV();
7446 sv_2mortal(MUTABLE_SV(RExC_paren_name_list));
7449 he_str = hv_fetch_ent( RExC_paren_names, svname, 1, 0 );
7451 sv_dat = HeVAL(he_str);
7453 /* croak baby croak */
7455 "panic: paren_name hash element allocation failed");
7456 } else if ( SvPOK(sv_dat) ) {
7457 /* (?|...) can mean we have dupes so scan to check
7458 its already been stored. Maybe a flag indicating
7459 we are inside such a construct would be useful,
7460 but the arrays are likely to be quite small, so
7461 for now we punt -- dmq */
7462 IV count = SvIV(sv_dat);
7463 I32 *pv = (I32*)SvPVX(sv_dat);
7465 for ( i = 0 ; i < count ; i++ ) {
7466 if ( pv[i] == RExC_npar ) {
7472 pv = (I32*)SvGROW(sv_dat, SvCUR(sv_dat) + sizeof(I32)+1);
7473 SvCUR_set(sv_dat, SvCUR(sv_dat) + sizeof(I32));
7474 pv[count] = RExC_npar;
7475 SvIV_set(sv_dat, SvIVX(sv_dat) + 1);
7478 (void)SvUPGRADE(sv_dat,SVt_PVNV);
7479 sv_setpvn(sv_dat, (char *)&(RExC_npar), sizeof(I32));
7481 SvIV_set(sv_dat, 1);
7484 /* Yes this does cause a memory leak in debugging Perls */
7485 if (!av_store(RExC_paren_name_list, RExC_npar, SvREFCNT_inc(svname)))
7486 SvREFCNT_dec(svname);
7489 /*sv_dump(sv_dat);*/
7491 nextchar(pRExC_state);
7493 goto capturing_parens;
7495 RExC_seen |= REG_SEEN_LOOKBEHIND;
7496 RExC_in_lookbehind++;
7498 case '=': /* (?=...) */
7499 RExC_seen_zerolen++;
7501 case '!': /* (?!...) */
7502 RExC_seen_zerolen++;
7503 if (*RExC_parse == ')') {
7504 ret=reg_node(pRExC_state, OPFAIL);
7505 nextchar(pRExC_state);
7509 case '|': /* (?|...) */
7510 /* branch reset, behave like a (?:...) except that
7511 buffers in alternations share the same numbers */
7513 after_freeze = freeze_paren = RExC_npar;
7515 case ':': /* (?:...) */
7516 case '>': /* (?>...) */
7518 case '$': /* (?$...) */
7519 case '@': /* (?@...) */
7520 vFAIL2("Sequence (?%c...) not implemented", (int)paren);
7522 case '#': /* (?#...) */
7523 while (*RExC_parse && *RExC_parse != ')')
7525 if (*RExC_parse != ')')
7526 FAIL("Sequence (?#... not terminated");
7527 nextchar(pRExC_state);
7530 case '0' : /* (?0) */
7531 case 'R' : /* (?R) */
7532 if (*RExC_parse != ')')
7533 FAIL("Sequence (?R) not terminated");
7534 ret = reg_node(pRExC_state, GOSTART);
7535 *flagp |= POSTPONED;
7536 nextchar(pRExC_state);
7539 { /* named and numeric backreferences */
7541 case '&': /* (?&NAME) */
7542 parse_start = RExC_parse - 1;
7545 SV *sv_dat = reg_scan_name(pRExC_state,
7546 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7547 num = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
7549 goto gen_recurse_regop;
7552 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
7554 vFAIL("Illegal pattern");
7556 goto parse_recursion;
7558 case '-': /* (?-1) */
7559 if (!(RExC_parse[0] >= '1' && RExC_parse[0] <= '9')) {
7560 RExC_parse--; /* rewind to let it be handled later */
7564 case '1': case '2': case '3': case '4': /* (?1) */
7565 case '5': case '6': case '7': case '8': case '9':
7568 num = atoi(RExC_parse);
7569 parse_start = RExC_parse - 1; /* MJD */
7570 if (*RExC_parse == '-')
7572 while (isDIGIT(*RExC_parse))
7574 if (*RExC_parse!=')')
7575 vFAIL("Expecting close bracket");
7578 if ( paren == '-' ) {
7580 Diagram of capture buffer numbering.
7581 Top line is the normal capture buffer numbers
7582 Bottom line is the negative indexing as from
7586 /(a(x)y)(a(b(c(?-2)d)e)f)(g(h))/
7590 num = RExC_npar + num;
7593 vFAIL("Reference to nonexistent group");
7595 } else if ( paren == '+' ) {
7596 num = RExC_npar + num - 1;
7599 ret = reganode(pRExC_state, GOSUB, num);
7601 if (num > (I32)RExC_rx->nparens) {
7603 vFAIL("Reference to nonexistent group");
7605 ARG2L_SET( ret, RExC_recurse_count++);
7607 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
7608 "Recurse #%"UVuf" to %"IVdf"\n", (UV)ARG(ret), (IV)ARG2L(ret)));
7612 RExC_seen |= REG_SEEN_RECURSE;
7613 Set_Node_Length(ret, 1 + regarglen[OP(ret)]); /* MJD */
7614 Set_Node_Offset(ret, parse_start); /* MJD */
7616 *flagp |= POSTPONED;
7617 nextchar(pRExC_state);
7619 } /* named and numeric backreferences */
7622 case '?': /* (??...) */
7624 if (*RExC_parse != '{') {
7626 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
7629 *flagp |= POSTPONED;
7630 paren = *RExC_parse++;
7632 case '{': /* (?{...}) */
7637 char *s = RExC_parse;
7639 RExC_seen_zerolen++;
7640 RExC_seen |= REG_SEEN_EVAL;
7641 while (count && (c = *RExC_parse)) {
7652 if (*RExC_parse != ')') {
7654 vFAIL("Sequence (?{...}) not terminated or not {}-balanced");
7658 OP_4tree *sop, *rop;
7659 SV * const sv = newSVpvn(s, RExC_parse - 1 - s);
7662 Perl_save_re_context(aTHX);
7663 rop = Perl_sv_compile_2op_is_broken(aTHX_ sv, &sop, "re", &pad);
7664 sop->op_private |= OPpREFCOUNTED;
7665 /* re_dup will OpREFCNT_inc */
7666 OpREFCNT_set(sop, 1);
7669 n = add_data(pRExC_state, 3, "nop");
7670 RExC_rxi->data->data[n] = (void*)rop;
7671 RExC_rxi->data->data[n+1] = (void*)sop;
7672 RExC_rxi->data->data[n+2] = (void*)pad;
7675 else { /* First pass */
7676 if (PL_reginterp_cnt < ++RExC_seen_evals
7678 /* No compiled RE interpolated, has runtime
7679 components ===> unsafe. */
7680 FAIL("Eval-group not allowed at runtime, use re 'eval'");
7681 if (PL_tainting && PL_tainted)
7682 FAIL("Eval-group in insecure regular expression");
7683 #if PERL_VERSION > 8
7684 if (IN_PERL_COMPILETIME)
7689 nextchar(pRExC_state);
7691 ret = reg_node(pRExC_state, LOGICAL);
7694 REGTAIL(pRExC_state, ret, reganode(pRExC_state, EVAL, n));
7695 /* deal with the length of this later - MJD */
7698 ret = reganode(pRExC_state, EVAL, n);
7699 Set_Node_Length(ret, RExC_parse - parse_start + 1);
7700 Set_Node_Offset(ret, parse_start);
7703 case '(': /* (?(?{...})...) and (?(?=...)...) */
7706 if (RExC_parse[0] == '?') { /* (?(?...)) */
7707 if (RExC_parse[1] == '=' || RExC_parse[1] == '!'
7708 || RExC_parse[1] == '<'
7709 || RExC_parse[1] == '{') { /* Lookahead or eval. */
7712 ret = reg_node(pRExC_state, LOGICAL);
7715 REGTAIL(pRExC_state, ret, reg(pRExC_state, 1, &flag,depth+1));
7719 else if ( RExC_parse[0] == '<' /* (?(<NAME>)...) */
7720 || RExC_parse[0] == '\'' ) /* (?('NAME')...) */
7722 char ch = RExC_parse[0] == '<' ? '>' : '\'';
7723 char *name_start= RExC_parse++;
7725 SV *sv_dat=reg_scan_name(pRExC_state,
7726 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7727 if (RExC_parse == name_start || *RExC_parse != ch)
7728 vFAIL2("Sequence (?(%c... not terminated",
7729 (ch == '>' ? '<' : ch));
7732 num = add_data( pRExC_state, 1, "S" );
7733 RExC_rxi->data->data[num]=(void*)sv_dat;
7734 SvREFCNT_inc_simple_void(sv_dat);
7736 ret = reganode(pRExC_state,NGROUPP,num);
7737 goto insert_if_check_paren;
7739 else if (RExC_parse[0] == 'D' &&
7740 RExC_parse[1] == 'E' &&
7741 RExC_parse[2] == 'F' &&
7742 RExC_parse[3] == 'I' &&
7743 RExC_parse[4] == 'N' &&
7744 RExC_parse[5] == 'E')
7746 ret = reganode(pRExC_state,DEFINEP,0);
7749 goto insert_if_check_paren;
7751 else if (RExC_parse[0] == 'R') {
7754 if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
7755 parno = atoi(RExC_parse++);
7756 while (isDIGIT(*RExC_parse))
7758 } else if (RExC_parse[0] == '&') {
7761 sv_dat = reg_scan_name(pRExC_state,
7762 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
7763 parno = sv_dat ? *((I32 *)SvPVX(sv_dat)) : 0;
7765 ret = reganode(pRExC_state,INSUBP,parno);
7766 goto insert_if_check_paren;
7768 else if (RExC_parse[0] >= '1' && RExC_parse[0] <= '9' ) {
7771 parno = atoi(RExC_parse++);
7773 while (isDIGIT(*RExC_parse))
7775 ret = reganode(pRExC_state, GROUPP, parno);
7777 insert_if_check_paren:
7778 if ((c = *nextchar(pRExC_state)) != ')')
7779 vFAIL("Switch condition not recognized");
7781 REGTAIL(pRExC_state, ret, reganode(pRExC_state, IFTHEN, 0));
7782 br = regbranch(pRExC_state, &flags, 1,depth+1);
7784 br = reganode(pRExC_state, LONGJMP, 0);
7786 REGTAIL(pRExC_state, br, reganode(pRExC_state, LONGJMP, 0));
7787 c = *nextchar(pRExC_state);
7792 vFAIL("(?(DEFINE)....) does not allow branches");
7793 lastbr = reganode(pRExC_state, IFTHEN, 0); /* Fake one for optimizer. */
7794 regbranch(pRExC_state, &flags, 1,depth+1);
7795 REGTAIL(pRExC_state, ret, lastbr);
7798 c = *nextchar(pRExC_state);
7803 vFAIL("Switch (?(condition)... contains too many branches");
7804 ender = reg_node(pRExC_state, TAIL);
7805 REGTAIL(pRExC_state, br, ender);
7807 REGTAIL(pRExC_state, lastbr, ender);
7808 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender);
7811 REGTAIL(pRExC_state, ret, ender);
7812 RExC_size++; /* XXX WHY do we need this?!!
7813 For large programs it seems to be required
7814 but I can't figure out why. -- dmq*/
7818 vFAIL2("Unknown switch condition (?(%.2s", RExC_parse);
7822 RExC_parse--; /* for vFAIL to print correctly */
7823 vFAIL("Sequence (? incomplete");
7825 case DEFAULT_PAT_MOD: /* Use default flags with the exceptions
7827 has_use_defaults = TRUE;
7828 STD_PMMOD_FLAGS_CLEAR(&RExC_flags);
7829 set_regex_charset(&RExC_flags, (RExC_utf8 || RExC_uni_semantics)
7830 ? REGEX_UNICODE_CHARSET
7831 : REGEX_DEPENDS_CHARSET);
7835 parse_flags: /* (?i) */
7837 U32 posflags = 0, negflags = 0;
7838 U32 *flagsp = &posflags;
7839 char has_charset_modifier = '\0';
7840 regex_charset cs = (RExC_utf8 || RExC_uni_semantics)
7841 ? REGEX_UNICODE_CHARSET
7842 : REGEX_DEPENDS_CHARSET;
7844 while (*RExC_parse) {
7845 /* && strchr("iogcmsx", *RExC_parse) */
7846 /* (?g), (?gc) and (?o) are useless here
7847 and must be globally applied -- japhy */
7848 switch (*RExC_parse) {
7849 CASE_STD_PMMOD_FLAGS_PARSE_SET(flagsp);
7850 case LOCALE_PAT_MOD:
7851 if (has_charset_modifier) {
7852 goto excess_modifier;
7854 else if (flagsp == &negflags) {
7857 cs = REGEX_LOCALE_CHARSET;
7858 has_charset_modifier = LOCALE_PAT_MOD;
7859 RExC_contains_locale = 1;
7861 case UNICODE_PAT_MOD:
7862 if (has_charset_modifier) {
7863 goto excess_modifier;
7865 else if (flagsp == &negflags) {
7868 cs = REGEX_UNICODE_CHARSET;
7869 has_charset_modifier = UNICODE_PAT_MOD;
7871 case ASCII_RESTRICT_PAT_MOD:
7872 if (flagsp == &negflags) {
7875 if (has_charset_modifier) {
7876 if (cs != REGEX_ASCII_RESTRICTED_CHARSET) {
7877 goto excess_modifier;
7879 /* Doubled modifier implies more restricted */
7880 cs = REGEX_ASCII_MORE_RESTRICTED_CHARSET;
7883 cs = REGEX_ASCII_RESTRICTED_CHARSET;
7885 has_charset_modifier = ASCII_RESTRICT_PAT_MOD;
7887 case DEPENDS_PAT_MOD:
7888 if (has_use_defaults) {
7889 goto fail_modifiers;
7891 else if (flagsp == &negflags) {
7894 else if (has_charset_modifier) {
7895 goto excess_modifier;
7898 /* The dual charset means unicode semantics if the
7899 * pattern (or target, not known until runtime) are
7900 * utf8, or something in the pattern indicates unicode
7902 cs = (RExC_utf8 || RExC_uni_semantics)
7903 ? REGEX_UNICODE_CHARSET
7904 : REGEX_DEPENDS_CHARSET;
7905 has_charset_modifier = DEPENDS_PAT_MOD;
7909 if (has_charset_modifier == ASCII_RESTRICT_PAT_MOD) {
7910 vFAIL2("Regexp modifier \"%c\" may appear a maximum of twice", ASCII_RESTRICT_PAT_MOD);
7912 else if (has_charset_modifier == *(RExC_parse - 1)) {
7913 vFAIL2("Regexp modifier \"%c\" may not appear twice", *(RExC_parse - 1));
7916 vFAIL3("Regexp modifiers \"%c\" and \"%c\" are mutually exclusive", has_charset_modifier, *(RExC_parse - 1));
7921 vFAIL2("Regexp modifier \"%c\" may not appear after the \"-\"", *(RExC_parse - 1));
7923 case ONCE_PAT_MOD: /* 'o' */
7924 case GLOBAL_PAT_MOD: /* 'g' */
7925 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
7926 const I32 wflagbit = *RExC_parse == 'o' ? WASTED_O : WASTED_G;
7927 if (! (wastedflags & wflagbit) ) {
7928 wastedflags |= wflagbit;
7931 "Useless (%s%c) - %suse /%c modifier",
7932 flagsp == &negflags ? "?-" : "?",
7934 flagsp == &negflags ? "don't " : "",
7941 case CONTINUE_PAT_MOD: /* 'c' */
7942 if (SIZE_ONLY && ckWARN(WARN_REGEXP)) {
7943 if (! (wastedflags & WASTED_C) ) {
7944 wastedflags |= WASTED_GC;
7947 "Useless (%sc) - %suse /gc modifier",
7948 flagsp == &negflags ? "?-" : "?",
7949 flagsp == &negflags ? "don't " : ""
7954 case KEEPCOPY_PAT_MOD: /* 'p' */
7955 if (flagsp == &negflags) {
7957 ckWARNreg(RExC_parse + 1,"Useless use of (?-p)");
7959 *flagsp |= RXf_PMf_KEEPCOPY;
7963 /* A flag is a default iff it is following a minus, so
7964 * if there is a minus, it means will be trying to
7965 * re-specify a default which is an error */
7966 if (has_use_defaults || flagsp == &negflags) {
7969 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
7973 wastedflags = 0; /* reset so (?g-c) warns twice */
7979 RExC_flags |= posflags;
7980 RExC_flags &= ~negflags;
7981 set_regex_charset(&RExC_flags, cs);
7983 oregflags |= posflags;
7984 oregflags &= ~negflags;
7985 set_regex_charset(&oregflags, cs);
7987 nextchar(pRExC_state);
7998 vFAIL3("Sequence (%.*s...) not recognized", RExC_parse-seqstart, seqstart);
8003 }} /* one for the default block, one for the switch */
8010 ret = reganode(pRExC_state, OPEN, parno);
8013 RExC_nestroot = parno;
8014 if (RExC_seen & REG_SEEN_RECURSE
8015 && !RExC_open_parens[parno-1])
8017 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
8018 "Setting open paren #%"IVdf" to %d\n",
8019 (IV)parno, REG_NODE_NUM(ret)));
8020 RExC_open_parens[parno-1]= ret;
8023 Set_Node_Length(ret, 1); /* MJD */
8024 Set_Node_Offset(ret, RExC_parse); /* MJD */
8032 /* Pick up the branches, linking them together. */
8033 parse_start = RExC_parse; /* MJD */
8034 br = regbranch(pRExC_state, &flags, 1,depth+1);
8036 /* branch_len = (paren != 0); */
8040 if (*RExC_parse == '|') {
8041 if (!SIZE_ONLY && RExC_extralen) {
8042 reginsert(pRExC_state, BRANCHJ, br, depth+1);
8045 reginsert(pRExC_state, BRANCH, br, depth+1);
8046 Set_Node_Length(br, paren != 0);
8047 Set_Node_Offset_To_R(br-RExC_emit_start, parse_start-RExC_start);
8051 RExC_extralen += 1; /* For BRANCHJ-BRANCH. */
8053 else if (paren == ':') {
8054 *flagp |= flags&SIMPLE;
8056 if (is_open) { /* Starts with OPEN. */
8057 REGTAIL(pRExC_state, ret, br); /* OPEN -> first. */
8059 else if (paren != '?') /* Not Conditional */
8061 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
8063 while (*RExC_parse == '|') {
8064 if (!SIZE_ONLY && RExC_extralen) {
8065 ender = reganode(pRExC_state, LONGJMP,0);
8066 REGTAIL(pRExC_state, NEXTOPER(NEXTOPER(lastbr)), ender); /* Append to the previous. */
8069 RExC_extralen += 2; /* Account for LONGJMP. */
8070 nextchar(pRExC_state);
8072 if (RExC_npar > after_freeze)
8073 after_freeze = RExC_npar;
8074 RExC_npar = freeze_paren;
8076 br = regbranch(pRExC_state, &flags, 0, depth+1);
8080 REGTAIL(pRExC_state, lastbr, br); /* BRANCH -> BRANCH. */
8082 *flagp |= flags & (SPSTART | HASWIDTH | POSTPONED);
8085 if (have_branch || paren != ':') {
8086 /* Make a closing node, and hook it on the end. */
8089 ender = reg_node(pRExC_state, TAIL);
8092 ender = reganode(pRExC_state, CLOSE, parno);
8093 if (!SIZE_ONLY && RExC_seen & REG_SEEN_RECURSE) {
8094 DEBUG_OPTIMISE_MORE_r(PerlIO_printf(Perl_debug_log,
8095 "Setting close paren #%"IVdf" to %d\n",
8096 (IV)parno, REG_NODE_NUM(ender)));
8097 RExC_close_parens[parno-1]= ender;
8098 if (RExC_nestroot == parno)
8101 Set_Node_Offset(ender,RExC_parse+1); /* MJD */
8102 Set_Node_Length(ender,1); /* MJD */
8108 *flagp &= ~HASWIDTH;
8111 ender = reg_node(pRExC_state, SUCCEED);
8114 ender = reg_node(pRExC_state, END);
8116 assert(!RExC_opend); /* there can only be one! */
8121 REGTAIL(pRExC_state, lastbr, ender);
8123 if (have_branch && !SIZE_ONLY) {
8125 RExC_seen |= REG_TOP_LEVEL_BRANCHES;
8127 /* Hook the tails of the branches to the closing node. */
8128 for (br = ret; br; br = regnext(br)) {
8129 const U8 op = PL_regkind[OP(br)];
8131 REGTAIL_STUDY(pRExC_state, NEXTOPER(br), ender);
8133 else if (op == BRANCHJ) {
8134 REGTAIL_STUDY(pRExC_state, NEXTOPER(NEXTOPER(br)), ender);
8142 static const char parens[] = "=!<,>";
8144 if (paren && (p = strchr(parens, paren))) {
8145 U8 node = ((p - parens) % 2) ? UNLESSM : IFMATCH;
8146 int flag = (p - parens) > 1;
8149 node = SUSPEND, flag = 0;
8150 reginsert(pRExC_state, node,ret, depth+1);
8151 Set_Node_Cur_Length(ret);
8152 Set_Node_Offset(ret, parse_start + 1);
8154 REGTAIL_STUDY(pRExC_state, ret, reg_node(pRExC_state, TAIL));
8158 /* Check for proper termination. */
8160 RExC_flags = oregflags;
8161 if (RExC_parse >= RExC_end || *nextchar(pRExC_state) != ')') {
8162 RExC_parse = oregcomp_parse;
8163 vFAIL("Unmatched (");
8166 else if (!paren && RExC_parse < RExC_end) {
8167 if (*RExC_parse == ')') {
8169 vFAIL("Unmatched )");
8172 FAIL("Junk on end of regexp"); /* "Can't happen". */
8176 if (RExC_in_lookbehind) {
8177 RExC_in_lookbehind--;
8179 if (after_freeze > RExC_npar)
8180 RExC_npar = after_freeze;
8185 - regbranch - one alternative of an | operator
8187 * Implements the concatenation operator.
8190 S_regbranch(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, I32 first, U32 depth)
8193 register regnode *ret;
8194 register regnode *chain = NULL;
8195 register regnode *latest;
8196 I32 flags = 0, c = 0;
8197 GET_RE_DEBUG_FLAGS_DECL;
8199 PERL_ARGS_ASSERT_REGBRANCH;
8201 DEBUG_PARSE("brnc");
8206 if (!SIZE_ONLY && RExC_extralen)
8207 ret = reganode(pRExC_state, BRANCHJ,0);
8209 ret = reg_node(pRExC_state, BRANCH);
8210 Set_Node_Length(ret, 1);
8214 if (!first && SIZE_ONLY)
8215 RExC_extralen += 1; /* BRANCHJ */
8217 *flagp = WORST; /* Tentatively. */
8220 nextchar(pRExC_state);
8221 while (RExC_parse < RExC_end && *RExC_parse != '|' && *RExC_parse != ')') {
8223 latest = regpiece(pRExC_state, &flags,depth+1);
8224 if (latest == NULL) {
8225 if (flags & TRYAGAIN)
8229 else if (ret == NULL)
8231 *flagp |= flags&(HASWIDTH|POSTPONED);
8232 if (chain == NULL) /* First piece. */
8233 *flagp |= flags&SPSTART;
8236 REGTAIL(pRExC_state, chain, latest);
8241 if (chain == NULL) { /* Loop ran zero times. */
8242 chain = reg_node(pRExC_state, NOTHING);
8247 *flagp |= flags&SIMPLE;
8254 - regpiece - something followed by possible [*+?]
8256 * Note that the branching code sequences used for ? and the general cases
8257 * of * and + are somewhat optimized: they use the same NOTHING node as
8258 * both the endmarker for their branch list and the body of the last branch.
8259 * It might seem that this node could be dispensed with entirely, but the
8260 * endmarker role is not redundant.
8263 S_regpiece(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
8266 register regnode *ret;
8268 register char *next;
8270 const char * const origparse = RExC_parse;
8272 I32 max = REG_INFTY;
8273 #ifdef RE_TRACK_PATTERN_OFFSETS
8276 const char *maxpos = NULL;
8277 GET_RE_DEBUG_FLAGS_DECL;
8279 PERL_ARGS_ASSERT_REGPIECE;
8281 DEBUG_PARSE("piec");
8283 ret = regatom(pRExC_state, &flags,depth+1);
8285 if (flags & TRYAGAIN)
8292 if (op == '{' && regcurly(RExC_parse)) {
8294 #ifdef RE_TRACK_PATTERN_OFFSETS
8295 parse_start = RExC_parse; /* MJD */
8297 next = RExC_parse + 1;
8298 while (isDIGIT(*next) || *next == ',') {
8307 if (*next == '}') { /* got one */
8311 min = atoi(RExC_parse);
8315 maxpos = RExC_parse;
8317 if (!max && *maxpos != '0')
8318 max = REG_INFTY; /* meaning "infinity" */
8319 else if (max >= REG_INFTY)
8320 vFAIL2("Quantifier in {,} bigger than %d", REG_INFTY - 1);
8322 nextchar(pRExC_state);
8325 if ((flags&SIMPLE)) {
8326 RExC_naughty += 2 + RExC_naughty / 2;
8327 reginsert(pRExC_state, CURLY, ret, depth+1);
8328 Set_Node_Offset(ret, parse_start+1); /* MJD */
8329 Set_Node_Cur_Length(ret);
8332 regnode * const w = reg_node(pRExC_state, WHILEM);
8335 REGTAIL(pRExC_state, ret, w);
8336 if (!SIZE_ONLY && RExC_extralen) {
8337 reginsert(pRExC_state, LONGJMP,ret, depth+1);
8338 reginsert(pRExC_state, NOTHING,ret, depth+1);
8339 NEXT_OFF(ret) = 3; /* Go over LONGJMP. */
8341 reginsert(pRExC_state, CURLYX,ret, depth+1);
8343 Set_Node_Offset(ret, parse_start+1);
8344 Set_Node_Length(ret,
8345 op == '{' ? (RExC_parse - parse_start) : 1);
8347 if (!SIZE_ONLY && RExC_extralen)
8348 NEXT_OFF(ret) = 3; /* Go over NOTHING to LONGJMP. */
8349 REGTAIL(pRExC_state, ret, reg_node(pRExC_state, NOTHING));
8351 RExC_whilem_seen++, RExC_extralen += 3;
8352 RExC_naughty += 4 + RExC_naughty; /* compound interest */
8361 vFAIL("Can't do {n,m} with n > m");
8363 ARG1_SET(ret, (U16)min);
8364 ARG2_SET(ret, (U16)max);
8376 #if 0 /* Now runtime fix should be reliable. */
8378 /* if this is reinstated, don't forget to put this back into perldiag:
8380 =item Regexp *+ operand could be empty at {#} in regex m/%s/
8382 (F) The part of the regexp subject to either the * or + quantifier
8383 could match an empty string. The {#} shows in the regular
8384 expression about where the problem was discovered.
8388 if (!(flags&HASWIDTH) && op != '?')
8389 vFAIL("Regexp *+ operand could be empty");
8392 #ifdef RE_TRACK_PATTERN_OFFSETS
8393 parse_start = RExC_parse;
8395 nextchar(pRExC_state);
8397 *flagp = (op != '+') ? (WORST|SPSTART|HASWIDTH) : (WORST|HASWIDTH);
8399 if (op == '*' && (flags&SIMPLE)) {
8400 reginsert(pRExC_state, STAR, ret, depth+1);
8404 else if (op == '*') {
8408 else if (op == '+' && (flags&SIMPLE)) {
8409 reginsert(pRExC_state, PLUS, ret, depth+1);
8413 else if (op == '+') {
8417 else if (op == '?') {
8422 if (!SIZE_ONLY && !(flags&(HASWIDTH|POSTPONED)) && max > REG_INFTY/3) {
8423 ckWARN3reg(RExC_parse,
8424 "%.*s matches null string many times",
8425 (int)(RExC_parse >= origparse ? RExC_parse - origparse : 0),
8429 if (RExC_parse < RExC_end && *RExC_parse == '?') {
8430 nextchar(pRExC_state);
8431 reginsert(pRExC_state, MINMOD, ret, depth+1);
8432 REGTAIL(pRExC_state, ret, ret + NODE_STEP_REGNODE);
8434 #ifndef REG_ALLOW_MINMOD_SUSPEND
8437 if (RExC_parse < RExC_end && *RExC_parse == '+') {
8439 nextchar(pRExC_state);
8440 ender = reg_node(pRExC_state, SUCCEED);
8441 REGTAIL(pRExC_state, ret, ender);
8442 reginsert(pRExC_state, SUSPEND, ret, depth+1);
8444 ender = reg_node(pRExC_state, TAIL);
8445 REGTAIL(pRExC_state, ret, ender);
8449 if (RExC_parse < RExC_end && ISMULT2(RExC_parse)) {
8451 vFAIL("Nested quantifiers");
8458 /* reg_namedseq(pRExC_state,UVp, UV depth)
8460 This is expected to be called by a parser routine that has
8461 recognized '\N' and needs to handle the rest. RExC_parse is
8462 expected to point at the first char following the N at the time
8465 The \N may be inside (indicated by valuep not being NULL) or outside a
8468 \N may begin either a named sequence, or if outside a character class, mean
8469 to match a non-newline. For non single-quoted regexes, the tokenizer has
8470 attempted to decide which, and in the case of a named sequence converted it
8471 into one of the forms: \N{} (if the sequence is null), or \N{U+c1.c2...},
8472 where c1... are the characters in the sequence. For single-quoted regexes,
8473 the tokenizer passes the \N sequence through unchanged; this code will not
8474 attempt to determine this nor expand those. The net effect is that if the
8475 beginning of the passed-in pattern isn't '{U+' or there is no '}', it
8476 signals that this \N occurrence means to match a non-newline.
8478 Only the \N{U+...} form should occur in a character class, for the same
8479 reason that '.' inside a character class means to just match a period: it
8480 just doesn't make sense.
8482 If valuep is non-null then it is assumed that we are parsing inside
8483 of a charclass definition and the first codepoint in the resolved
8484 string is returned via *valuep and the routine will return NULL.
8485 In this mode if a multichar string is returned from the charnames
8486 handler, a warning will be issued, and only the first char in the
8487 sequence will be examined. If the string returned is zero length
8488 then the value of *valuep is undefined and NON-NULL will
8489 be returned to indicate failure. (This will NOT be a valid pointer
8492 If valuep is null then it is assumed that we are parsing normal text and a
8493 new EXACT node is inserted into the program containing the resolved string,
8494 and a pointer to the new node is returned. But if the string is zero length
8495 a NOTHING node is emitted instead.
8497 On success RExC_parse is set to the char following the endbrace.
8498 Parsing failures will generate a fatal error via vFAIL(...)
8501 S_reg_namedseq(pTHX_ RExC_state_t *pRExC_state, UV *valuep, I32 *flagp, U32 depth)
8503 char * endbrace; /* '}' following the name */
8504 regnode *ret = NULL;
8507 GET_RE_DEBUG_FLAGS_DECL;
8509 PERL_ARGS_ASSERT_REG_NAMEDSEQ;
8513 /* The [^\n] meaning of \N ignores spaces and comments under the /x
8514 * modifier. The other meaning does not */
8515 p = (RExC_flags & RXf_PMf_EXTENDED)
8516 ? regwhite( pRExC_state, RExC_parse )
8519 /* Disambiguate between \N meaning a named character versus \N meaning
8520 * [^\n]. The former is assumed when it can't be the latter. */
8521 if (*p != '{' || regcurly(p)) {
8524 /* no bare \N in a charclass */
8525 vFAIL("\\N in a character class must be a named character: \\N{...}");
8527 nextchar(pRExC_state);
8528 ret = reg_node(pRExC_state, REG_ANY);
8529 *flagp |= HASWIDTH|SIMPLE;
8532 Set_Node_Length(ret, 1); /* MJD */
8536 /* Here, we have decided it should be a named sequence */
8538 /* The test above made sure that the next real character is a '{', but
8539 * under the /x modifier, it could be separated by space (or a comment and
8540 * \n) and this is not allowed (for consistency with \x{...} and the
8541 * tokenizer handling of \N{NAME}). */
8542 if (*RExC_parse != '{') {
8543 vFAIL("Missing braces on \\N{}");
8546 RExC_parse++; /* Skip past the '{' */
8548 if (! (endbrace = strchr(RExC_parse, '}')) /* no trailing brace */
8549 || ! (endbrace == RExC_parse /* nothing between the {} */
8550 || (endbrace - RExC_parse >= 2 /* U+ (bad hex is checked below */
8551 && strnEQ(RExC_parse, "U+", 2)))) /* for a better error msg) */
8553 if (endbrace) RExC_parse = endbrace; /* position msg's '<--HERE' */
8554 vFAIL("\\N{NAME} must be resolved by the lexer");
8557 if (endbrace == RExC_parse) { /* empty: \N{} */
8559 RExC_parse = endbrace + 1;
8560 return reg_node(pRExC_state,NOTHING);
8564 ckWARNreg(RExC_parse,
8565 "Ignoring zero length \\N{} in character class"
8567 RExC_parse = endbrace + 1;
8570 return (regnode *) &RExC_parse; /* Invalid regnode pointer */
8573 REQUIRE_UTF8; /* named sequences imply Unicode semantics */
8574 RExC_parse += 2; /* Skip past the 'U+' */
8576 if (valuep) { /* In a bracketed char class */
8577 /* We only pay attention to the first char of
8578 multichar strings being returned. I kinda wonder
8579 if this makes sense as it does change the behaviour
8580 from earlier versions, OTOH that behaviour was broken
8581 as well. XXX Solution is to recharacterize as
8582 [rest-of-class]|multi1|multi2... */
8584 STRLEN length_of_hex;
8585 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
8586 | PERL_SCAN_DISALLOW_PREFIX
8587 | (SIZE_ONLY ? PERL_SCAN_SILENT_ILLDIGIT : 0);
8589 char * endchar = RExC_parse + strcspn(RExC_parse, ".}");
8590 if (endchar < endbrace) {
8591 ckWARNreg(endchar, "Using just the first character returned by \\N{} in character class");
8594 length_of_hex = (STRLEN)(endchar - RExC_parse);
8595 *valuep = grok_hex(RExC_parse, &length_of_hex, &flags, NULL);
8597 /* The tokenizer should have guaranteed validity, but it's possible to
8598 * bypass it by using single quoting, so check */
8599 if (length_of_hex == 0
8600 || length_of_hex != (STRLEN)(endchar - RExC_parse) )
8602 RExC_parse += length_of_hex; /* Includes all the valid */
8603 RExC_parse += (RExC_orig_utf8) /* point to after 1st invalid */
8604 ? UTF8SKIP(RExC_parse)
8606 /* Guard against malformed utf8 */
8607 if (RExC_parse >= endchar) RExC_parse = endchar;
8608 vFAIL("Invalid hexadecimal number in \\N{U+...}");
8611 RExC_parse = endbrace + 1;
8612 if (endchar == endbrace) return NULL;
8614 ret = (regnode *) &RExC_parse; /* Invalid regnode pointer */
8616 else { /* Not a char class */
8618 /* What is done here is to convert this to a sub-pattern of the form
8619 * (?:\x{char1}\x{char2}...)
8620 * and then call reg recursively. That way, it retains its atomicness,
8621 * while not having to worry about special handling that some code
8622 * points may have. toke.c has converted the original Unicode values
8623 * to native, so that we can just pass on the hex values unchanged. We
8624 * do have to set a flag to keep recoding from happening in the
8627 SV * substitute_parse = newSVpvn_flags("?:", 2, SVf_UTF8|SVs_TEMP);
8629 char *endchar; /* Points to '.' or '}' ending cur char in the input
8631 char *orig_end = RExC_end;
8633 while (RExC_parse < endbrace) {
8635 /* Code points are separated by dots. If none, there is only one
8636 * code point, and is terminated by the brace */
8637 endchar = RExC_parse + strcspn(RExC_parse, ".}");
8639 /* Convert to notation the rest of the code understands */
8640 sv_catpv(substitute_parse, "\\x{");
8641 sv_catpvn(substitute_parse, RExC_parse, endchar - RExC_parse);
8642 sv_catpv(substitute_parse, "}");
8644 /* Point to the beginning of the next character in the sequence. */
8645 RExC_parse = endchar + 1;
8647 sv_catpv(substitute_parse, ")");
8649 RExC_parse = SvPV(substitute_parse, len);
8651 /* Don't allow empty number */
8653 vFAIL("Invalid hexadecimal number in \\N{U+...}");
8655 RExC_end = RExC_parse + len;
8657 /* The values are Unicode, and therefore not subject to recoding */
8658 RExC_override_recoding = 1;
8660 ret = reg(pRExC_state, 1, flagp, depth+1);
8662 RExC_parse = endbrace;
8663 RExC_end = orig_end;
8664 RExC_override_recoding = 0;
8666 nextchar(pRExC_state);
8676 * It returns the code point in utf8 for the value in *encp.
8677 * value: a code value in the source encoding
8678 * encp: a pointer to an Encode object
8680 * If the result from Encode is not a single character,
8681 * it returns U+FFFD (Replacement character) and sets *encp to NULL.
8684 S_reg_recode(pTHX_ const char value, SV **encp)
8687 SV * const sv = newSVpvn_flags(&value, numlen, SVs_TEMP);
8688 const char * const s = *encp ? sv_recode_to_utf8(sv, *encp) : SvPVX(sv);
8689 const STRLEN newlen = SvCUR(sv);
8690 UV uv = UNICODE_REPLACEMENT;
8692 PERL_ARGS_ASSERT_REG_RECODE;
8696 ? utf8n_to_uvchr((U8*)s, newlen, &numlen, UTF8_ALLOW_DEFAULT)
8699 if (!newlen || numlen != newlen) {
8700 uv = UNICODE_REPLACEMENT;
8708 - regatom - the lowest level
8710 Try to identify anything special at the start of the pattern. If there
8711 is, then handle it as required. This may involve generating a single regop,
8712 such as for an assertion; or it may involve recursing, such as to
8713 handle a () structure.
8715 If the string doesn't start with something special then we gobble up
8716 as much literal text as we can.
8718 Once we have been able to handle whatever type of thing started the
8719 sequence, we return.
8721 Note: we have to be careful with escapes, as they can be both literal
8722 and special, and in the case of \10 and friends can either, depending
8723 on context. Specifically there are two separate switches for handling
8724 escape sequences, with the one for handling literal escapes requiring
8725 a dummy entry for all of the special escapes that are actually handled
8730 S_regatom(pTHX_ RExC_state_t *pRExC_state, I32 *flagp, U32 depth)
8733 register regnode *ret = NULL;
8735 char *parse_start = RExC_parse;
8737 GET_RE_DEBUG_FLAGS_DECL;
8738 DEBUG_PARSE("atom");
8739 *flagp = WORST; /* Tentatively. */
8741 PERL_ARGS_ASSERT_REGATOM;
8744 switch ((U8)*RExC_parse) {
8746 RExC_seen_zerolen++;
8747 nextchar(pRExC_state);
8748 if (RExC_flags & RXf_PMf_MULTILINE)
8749 ret = reg_node(pRExC_state, MBOL);
8750 else if (RExC_flags & RXf_PMf_SINGLELINE)
8751 ret = reg_node(pRExC_state, SBOL);
8753 ret = reg_node(pRExC_state, BOL);
8754 Set_Node_Length(ret, 1); /* MJD */
8757 nextchar(pRExC_state);
8759 RExC_seen_zerolen++;
8760 if (RExC_flags & RXf_PMf_MULTILINE)
8761 ret = reg_node(pRExC_state, MEOL);
8762 else if (RExC_flags & RXf_PMf_SINGLELINE)
8763 ret = reg_node(pRExC_state, SEOL);
8765 ret = reg_node(pRExC_state, EOL);
8766 Set_Node_Length(ret, 1); /* MJD */
8769 nextchar(pRExC_state);
8770 if (RExC_flags & RXf_PMf_SINGLELINE)
8771 ret = reg_node(pRExC_state, SANY);
8773 ret = reg_node(pRExC_state, REG_ANY);
8774 *flagp |= HASWIDTH|SIMPLE;
8776 Set_Node_Length(ret, 1); /* MJD */
8780 char * const oregcomp_parse = ++RExC_parse;
8781 ret = regclass(pRExC_state,depth+1);
8782 if (*RExC_parse != ']') {
8783 RExC_parse = oregcomp_parse;
8784 vFAIL("Unmatched [");
8786 nextchar(pRExC_state);
8787 *flagp |= HASWIDTH|SIMPLE;
8788 Set_Node_Length(ret, RExC_parse - oregcomp_parse + 1); /* MJD */
8792 nextchar(pRExC_state);
8793 ret = reg(pRExC_state, 1, &flags,depth+1);
8795 if (flags & TRYAGAIN) {
8796 if (RExC_parse == RExC_end) {
8797 /* Make parent create an empty node if needed. */
8805 *flagp |= flags&(HASWIDTH|SPSTART|SIMPLE|POSTPONED);
8809 if (flags & TRYAGAIN) {
8813 vFAIL("Internal urp");
8814 /* Supposed to be caught earlier. */
8817 if (!regcurly(RExC_parse)) {
8826 vFAIL("Quantifier follows nothing");
8831 This switch handles escape sequences that resolve to some kind
8832 of special regop and not to literal text. Escape sequnces that
8833 resolve to literal text are handled below in the switch marked
8836 Every entry in this switch *must* have a corresponding entry
8837 in the literal escape switch. However, the opposite is not
8838 required, as the default for this switch is to jump to the
8839 literal text handling code.
8841 switch ((U8)*++RExC_parse) {
8842 /* Special Escapes */
8844 RExC_seen_zerolen++;
8845 ret = reg_node(pRExC_state, SBOL);
8847 goto finish_meta_pat;
8849 ret = reg_node(pRExC_state, GPOS);
8850 RExC_seen |= REG_SEEN_GPOS;
8852 goto finish_meta_pat;
8854 RExC_seen_zerolen++;
8855 ret = reg_node(pRExC_state, KEEPS);
8857 /* XXX:dmq : disabling in-place substitution seems to
8858 * be necessary here to avoid cases of memory corruption, as
8859 * with: C<$_="x" x 80; s/x\K/y/> -- rgs
8861 RExC_seen |= REG_SEEN_LOOKBEHIND;
8862 goto finish_meta_pat;
8864 ret = reg_node(pRExC_state, SEOL);
8866 RExC_seen_zerolen++; /* Do not optimize RE away */
8867 goto finish_meta_pat;
8869 ret = reg_node(pRExC_state, EOS);
8871 RExC_seen_zerolen++; /* Do not optimize RE away */
8872 goto finish_meta_pat;
8874 ret = reg_node(pRExC_state, CANY);
8875 RExC_seen |= REG_SEEN_CANY;
8876 *flagp |= HASWIDTH|SIMPLE;
8877 goto finish_meta_pat;
8879 ret = reg_node(pRExC_state, CLUMP);
8881 goto finish_meta_pat;
8883 switch (get_regex_charset(RExC_flags)) {
8884 case REGEX_LOCALE_CHARSET:
8887 case REGEX_UNICODE_CHARSET:
8890 case REGEX_ASCII_RESTRICTED_CHARSET:
8891 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
8894 case REGEX_DEPENDS_CHARSET:
8900 ret = reg_node(pRExC_state, op);
8901 *flagp |= HASWIDTH|SIMPLE;
8902 goto finish_meta_pat;
8904 switch (get_regex_charset(RExC_flags)) {
8905 case REGEX_LOCALE_CHARSET:
8908 case REGEX_UNICODE_CHARSET:
8911 case REGEX_ASCII_RESTRICTED_CHARSET:
8912 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
8915 case REGEX_DEPENDS_CHARSET:
8921 ret = reg_node(pRExC_state, op);
8922 *flagp |= HASWIDTH|SIMPLE;
8923 goto finish_meta_pat;
8925 RExC_seen_zerolen++;
8926 RExC_seen |= REG_SEEN_LOOKBEHIND;
8927 switch (get_regex_charset(RExC_flags)) {
8928 case REGEX_LOCALE_CHARSET:
8931 case REGEX_UNICODE_CHARSET:
8934 case REGEX_ASCII_RESTRICTED_CHARSET:
8935 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
8938 case REGEX_DEPENDS_CHARSET:
8944 ret = reg_node(pRExC_state, op);
8945 FLAGS(ret) = get_regex_charset(RExC_flags);
8947 if (! SIZE_ONLY && (U8) *(RExC_parse + 1) == '{') {
8948 ckWARNregdep(RExC_parse, "\"\\b{\" is deprecated; use \"\\b\\{\" instead");
8950 goto finish_meta_pat;
8952 RExC_seen_zerolen++;
8953 RExC_seen |= REG_SEEN_LOOKBEHIND;
8954 switch (get_regex_charset(RExC_flags)) {
8955 case REGEX_LOCALE_CHARSET:
8958 case REGEX_UNICODE_CHARSET:
8961 case REGEX_ASCII_RESTRICTED_CHARSET:
8962 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
8965 case REGEX_DEPENDS_CHARSET:
8971 ret = reg_node(pRExC_state, op);
8972 FLAGS(ret) = get_regex_charset(RExC_flags);
8974 if (! SIZE_ONLY && (U8) *(RExC_parse + 1) == '{') {
8975 ckWARNregdep(RExC_parse, "\"\\B{\" is deprecated; use \"\\B\\{\" instead");
8977 goto finish_meta_pat;
8979 switch (get_regex_charset(RExC_flags)) {
8980 case REGEX_LOCALE_CHARSET:
8983 case REGEX_UNICODE_CHARSET:
8986 case REGEX_ASCII_RESTRICTED_CHARSET:
8987 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
8990 case REGEX_DEPENDS_CHARSET:
8996 ret = reg_node(pRExC_state, op);
8997 *flagp |= HASWIDTH|SIMPLE;
8998 goto finish_meta_pat;
9000 switch (get_regex_charset(RExC_flags)) {
9001 case REGEX_LOCALE_CHARSET:
9004 case REGEX_UNICODE_CHARSET:
9007 case REGEX_ASCII_RESTRICTED_CHARSET:
9008 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
9011 case REGEX_DEPENDS_CHARSET:
9017 ret = reg_node(pRExC_state, op);
9018 *flagp |= HASWIDTH|SIMPLE;
9019 goto finish_meta_pat;
9021 switch (get_regex_charset(RExC_flags)) {
9022 case REGEX_LOCALE_CHARSET:
9025 case REGEX_ASCII_RESTRICTED_CHARSET:
9026 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
9029 case REGEX_DEPENDS_CHARSET: /* No difference between these */
9030 case REGEX_UNICODE_CHARSET:
9036 ret = reg_node(pRExC_state, op);
9037 *flagp |= HASWIDTH|SIMPLE;
9038 goto finish_meta_pat;
9040 switch (get_regex_charset(RExC_flags)) {
9041 case REGEX_LOCALE_CHARSET:
9044 case REGEX_ASCII_RESTRICTED_CHARSET:
9045 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
9048 case REGEX_DEPENDS_CHARSET: /* No difference between these */
9049 case REGEX_UNICODE_CHARSET:
9055 ret = reg_node(pRExC_state, op);
9056 *flagp |= HASWIDTH|SIMPLE;
9057 goto finish_meta_pat;
9059 ret = reg_node(pRExC_state, LNBREAK);
9060 *flagp |= HASWIDTH|SIMPLE;
9061 goto finish_meta_pat;
9063 ret = reg_node(pRExC_state, HORIZWS);
9064 *flagp |= HASWIDTH|SIMPLE;
9065 goto finish_meta_pat;
9067 ret = reg_node(pRExC_state, NHORIZWS);
9068 *flagp |= HASWIDTH|SIMPLE;
9069 goto finish_meta_pat;
9071 ret = reg_node(pRExC_state, VERTWS);
9072 *flagp |= HASWIDTH|SIMPLE;
9073 goto finish_meta_pat;
9075 ret = reg_node(pRExC_state, NVERTWS);
9076 *flagp |= HASWIDTH|SIMPLE;
9078 nextchar(pRExC_state);
9079 Set_Node_Length(ret, 2); /* MJD */
9084 char* const oldregxend = RExC_end;
9086 char* parse_start = RExC_parse - 2;
9089 if (RExC_parse[1] == '{') {
9090 /* a lovely hack--pretend we saw [\pX] instead */
9091 RExC_end = strchr(RExC_parse, '}');
9093 const U8 c = (U8)*RExC_parse;
9095 RExC_end = oldregxend;
9096 vFAIL2("Missing right brace on \\%c{}", c);
9101 RExC_end = RExC_parse + 2;
9102 if (RExC_end > oldregxend)
9103 RExC_end = oldregxend;
9107 ret = regclass(pRExC_state,depth+1);
9109 RExC_end = oldregxend;
9112 Set_Node_Offset(ret, parse_start + 2);
9113 Set_Node_Cur_Length(ret);
9114 nextchar(pRExC_state);
9115 *flagp |= HASWIDTH|SIMPLE;
9119 /* Handle \N and \N{NAME} here and not below because it can be
9120 multicharacter. join_exact() will join them up later on.
9121 Also this makes sure that things like /\N{BLAH}+/ and
9122 \N{BLAH} being multi char Just Happen. dmq*/
9124 ret= reg_namedseq(pRExC_state, NULL, flagp, depth);
9126 case 'k': /* Handle \k<NAME> and \k'NAME' */
9129 char ch= RExC_parse[1];
9130 if (ch != '<' && ch != '\'' && ch != '{') {
9132 vFAIL2("Sequence %.2s... not terminated",parse_start);
9134 /* this pretty much dupes the code for (?P=...) in reg(), if
9135 you change this make sure you change that */
9136 char* name_start = (RExC_parse += 2);
9138 SV *sv_dat = reg_scan_name(pRExC_state,
9139 SIZE_ONLY ? REG_RSN_RETURN_NULL : REG_RSN_RETURN_DATA);
9140 ch= (ch == '<') ? '>' : (ch == '{') ? '}' : '\'';
9141 if (RExC_parse == name_start || *RExC_parse != ch)
9142 vFAIL2("Sequence %.3s... not terminated",parse_start);
9145 num = add_data( pRExC_state, 1, "S" );
9146 RExC_rxi->data->data[num]=(void*)sv_dat;
9147 SvREFCNT_inc_simple_void(sv_dat);
9151 ret = reganode(pRExC_state,
9154 : (MORE_ASCII_RESTRICTED)
9156 : (AT_LEAST_UNI_SEMANTICS)
9164 /* override incorrect value set in reganode MJD */
9165 Set_Node_Offset(ret, parse_start+1);
9166 Set_Node_Cur_Length(ret); /* MJD */
9167 nextchar(pRExC_state);
9173 case '1': case '2': case '3': case '4':
9174 case '5': case '6': case '7': case '8': case '9':
9177 bool isg = *RExC_parse == 'g';
9182 if (*RExC_parse == '{') {
9186 if (*RExC_parse == '-') {
9190 if (hasbrace && !isDIGIT(*RExC_parse)) {
9191 if (isrel) RExC_parse--;
9193 goto parse_named_seq;
9195 num = atoi(RExC_parse);
9196 if (isg && num == 0)
9197 vFAIL("Reference to invalid group 0");
9199 num = RExC_npar - num;
9201 vFAIL("Reference to nonexistent or unclosed group");
9203 if (!isg && num > 9 && num >= RExC_npar)
9206 char * const parse_start = RExC_parse - 1; /* MJD */
9207 while (isDIGIT(*RExC_parse))
9209 if (parse_start == RExC_parse - 1)
9210 vFAIL("Unterminated \\g... pattern");
9212 if (*RExC_parse != '}')
9213 vFAIL("Unterminated \\g{...} pattern");
9217 if (num > (I32)RExC_rx->nparens)
9218 vFAIL("Reference to nonexistent group");
9221 ret = reganode(pRExC_state,
9224 : (MORE_ASCII_RESTRICTED)
9226 : (AT_LEAST_UNI_SEMANTICS)
9234 /* override incorrect value set in reganode MJD */
9235 Set_Node_Offset(ret, parse_start+1);
9236 Set_Node_Cur_Length(ret); /* MJD */
9238 nextchar(pRExC_state);
9243 if (RExC_parse >= RExC_end)
9244 FAIL("Trailing \\");
9247 /* Do not generate "unrecognized" warnings here, we fall
9248 back into the quick-grab loop below */
9255 if (RExC_flags & RXf_PMf_EXTENDED) {
9256 if ( reg_skipcomment( pRExC_state ) )
9263 parse_start = RExC_parse - 1;
9268 register STRLEN len;
9273 U8 tmpbuf[UTF8_MAXBYTES_CASE+1], *foldbuf;
9274 regnode * orig_emit;
9277 /* Is this a LATIN LOWER CASE SHARP S in an EXACTFU node? If so,
9278 * it is folded to 'ss' even if not utf8 */
9279 bool is_exactfu_sharp_s;
9282 orig_emit = RExC_emit; /* Save the original output node position in
9283 case we need to output a different node
9285 node_type = ((! FOLD) ? EXACT
9288 : (MORE_ASCII_RESTRICTED)
9290 : (AT_LEAST_UNI_SEMANTICS)
9293 ret = reg_node(pRExC_state, node_type);
9296 /* XXX The node can hold up to 255 bytes, yet this only goes to
9297 * 127. I (khw) do not know why. Keeping it somewhat less than
9298 * 255 allows us to not have to worry about overflow due to
9299 * converting to utf8 and fold expansion, but that value is
9300 * 255-UTF8_MAXBYTES_CASE. join_exact() may join adjacent nodes
9301 * split up by this limit into a single one using the real max of
9302 * 255. Even at 127, this breaks under rare circumstances. If
9303 * folding, we do not want to split a node at a character that is a
9304 * non-final in a multi-char fold, as an input string could just
9305 * happen to want to match across the node boundary. The join
9306 * would solve that problem if the join actually happens. But a
9307 * series of more than two nodes in a row each of 127 would cause
9308 * the first join to succeed to get to 254, but then there wouldn't
9309 * be room for the next one, which could at be one of those split
9310 * multi-char folds. I don't know of any fool-proof solution. One
9311 * could back off to end with only a code point that isn't such a
9312 * non-final, but it is possible for there not to be any in the
9314 for (len = 0, p = RExC_parse - 1;
9315 len < 127 && p < RExC_end;
9318 char * const oldp = p;
9320 if (RExC_flags & RXf_PMf_EXTENDED)
9321 p = regwhite( pRExC_state, p );
9332 /* Literal Escapes Switch
9334 This switch is meant to handle escape sequences that
9335 resolve to a literal character.
9337 Every escape sequence that represents something
9338 else, like an assertion or a char class, is handled
9339 in the switch marked 'Special Escapes' above in this
9340 routine, but also has an entry here as anything that
9341 isn't explicitly mentioned here will be treated as
9342 an unescaped equivalent literal.
9346 /* These are all the special escapes. */
9347 case 'A': /* Start assertion */
9348 case 'b': case 'B': /* Word-boundary assertion*/
9349 case 'C': /* Single char !DANGEROUS! */
9350 case 'd': case 'D': /* digit class */
9351 case 'g': case 'G': /* generic-backref, pos assertion */
9352 case 'h': case 'H': /* HORIZWS */
9353 case 'k': case 'K': /* named backref, keep marker */
9354 case 'N': /* named char sequence */
9355 case 'p': case 'P': /* Unicode property */
9356 case 'R': /* LNBREAK */
9357 case 's': case 'S': /* space class */
9358 case 'v': case 'V': /* VERTWS */
9359 case 'w': case 'W': /* word class */
9360 case 'X': /* eXtended Unicode "combining character sequence" */
9361 case 'z': case 'Z': /* End of line/string assertion */
9365 /* Anything after here is an escape that resolves to a
9366 literal. (Except digits, which may or may not)
9385 ender = ASCII_TO_NATIVE('\033');
9389 ender = ASCII_TO_NATIVE('\007');
9394 STRLEN brace_len = len;
9396 const char* error_msg;
9398 bool valid = grok_bslash_o(p,
9405 RExC_parse = p; /* going to die anyway; point
9406 to exact spot of failure */
9413 if (PL_encoding && ender < 0x100) {
9414 goto recode_encoding;
9423 char* const e = strchr(p, '}');
9427 vFAIL("Missing right brace on \\x{}");
9430 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
9431 | PERL_SCAN_DISALLOW_PREFIX;
9432 STRLEN numlen = e - p - 1;
9433 ender = grok_hex(p + 1, &numlen, &flags, NULL);
9440 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
9442 ender = grok_hex(p, &numlen, &flags, NULL);
9445 if (PL_encoding && ender < 0x100)
9446 goto recode_encoding;
9450 ender = grok_bslash_c(*p++, UTF, SIZE_ONLY);
9452 case '0': case '1': case '2': case '3':case '4':
9453 case '5': case '6': case '7': case '8':case '9':
9455 (isDIGIT(p[1]) && atoi(p) >= RExC_npar))
9457 I32 flags = PERL_SCAN_SILENT_ILLDIGIT;
9459 ender = grok_oct(p, &numlen, &flags, NULL);
9469 if (PL_encoding && ender < 0x100)
9470 goto recode_encoding;
9473 if (! RExC_override_recoding) {
9474 SV* enc = PL_encoding;
9475 ender = reg_recode((const char)(U8)ender, &enc);
9476 if (!enc && SIZE_ONLY)
9477 ckWARNreg(p, "Invalid escape in the specified encoding");
9483 FAIL("Trailing \\");
9486 if (!SIZE_ONLY&& isALPHA(*p)) {
9487 /* Include any { following the alpha to emphasize
9488 * that it could be part of an escape at some point
9490 int len = (*(p + 1) == '{') ? 2 : 1;
9491 ckWARN3reg(p + len, "Unrecognized escape \\%.*s passed through", len, p);
9493 goto normal_default;
9498 if (UTF8_IS_START(*p) && UTF) {
9500 ender = utf8n_to_uvchr((U8*)p, RExC_end - p,
9501 &numlen, UTF8_ALLOW_DEFAULT);
9507 } /* End of switch on the literal */
9509 is_exactfu_sharp_s = (node_type == EXACTFU
9510 && ender == LATIN_SMALL_LETTER_SHARP_S);
9511 if ( RExC_flags & RXf_PMf_EXTENDED)
9512 p = regwhite( pRExC_state, p );
9513 if ((UTF && FOLD) || is_exactfu_sharp_s) {
9514 /* Prime the casefolded buffer. Locale rules, which apply
9515 * only to code points < 256, aren't known until execution,
9516 * so for them, just output the original character using
9517 * utf8. If we start to fold non-UTF patterns, be sure to
9518 * update join_exact() */
9519 if (LOC && ender < 256) {
9520 if (UNI_IS_INVARIANT(ender)) {
9521 *tmpbuf = (U8) ender;
9524 *tmpbuf = UTF8_TWO_BYTE_HI(ender);
9525 *(tmpbuf + 1) = UTF8_TWO_BYTE_LO(ender);
9529 else if (isASCII(ender)) { /* Note: Here can't also be LOC
9531 ender = toLOWER(ender);
9532 *tmpbuf = (U8) ender;
9535 else if (! MORE_ASCII_RESTRICTED && ! LOC) {
9537 /* Locale and /aa require more selectivity about the
9538 * fold, so are handled below. Otherwise, here, just
9540 ender = toFOLD_uni(ender, tmpbuf, &foldlen);
9543 /* Under locale rules or /aa we are not to mix,
9544 * respectively, ords < 256 or ASCII with non-. So
9545 * reject folds that mix them, using only the
9546 * non-folded code point. So do the fold to a
9547 * temporary, and inspect each character in it. */
9548 U8 trialbuf[UTF8_MAXBYTES_CASE+1];
9550 UV tmpender = toFOLD_uni(ender, trialbuf, &foldlen);
9551 U8* e = s + foldlen;
9552 bool fold_ok = TRUE;
9556 || (LOC && (UTF8_IS_INVARIANT(*s)
9557 || UTF8_IS_DOWNGRADEABLE_START(*s))))
9565 Copy(trialbuf, tmpbuf, foldlen, U8);
9569 uvuni_to_utf8(tmpbuf, ender);
9570 foldlen = UNISKIP(ender);
9574 if (p < RExC_end && ISMULT2(p)) { /* Back off on ?+*. */
9577 else if (UTF || is_exactfu_sharp_s) {
9579 /* Emit all the Unicode characters. */
9581 for (foldbuf = tmpbuf;
9583 foldlen -= numlen) {
9584 ender = utf8_to_uvchr(foldbuf, &numlen);
9586 const STRLEN unilen = reguni(pRExC_state, ender, s);
9589 /* In EBCDIC the numlen
9590 * and unilen can differ. */
9592 if (numlen >= foldlen)
9596 break; /* "Can't happen." */
9600 const STRLEN unilen = reguni(pRExC_state, ender, s);
9609 REGC((char)ender, s++);
9613 if (UTF || is_exactfu_sharp_s) {
9615 /* Emit all the Unicode characters. */
9617 for (foldbuf = tmpbuf;
9619 foldlen -= numlen) {
9620 ender = utf8_to_uvchr(foldbuf, &numlen);
9622 const STRLEN unilen = reguni(pRExC_state, ender, s);
9625 /* In EBCDIC the numlen
9626 * and unilen can differ. */
9628 if (numlen >= foldlen)
9636 const STRLEN unilen = reguni(pRExC_state, ender, s);
9645 REGC((char)ender, s++);
9648 loopdone: /* Jumped to when encounters something that shouldn't be in
9651 Set_Node_Cur_Length(ret); /* MJD */
9652 nextchar(pRExC_state);
9654 /* len is STRLEN which is unsigned, need to copy to signed */
9657 vFAIL("Internal disaster");
9661 if (len == 1 && UNI_IS_INVARIANT(ender))
9665 RExC_size += STR_SZ(len);
9668 RExC_emit += STR_SZ(len);
9676 /* Jumped to when an unrecognized character set is encountered */
9678 Perl_croak(aTHX_ "panic: Unknown regex character set encoding: %u", get_regex_charset(RExC_flags));
9683 S_regwhite( RExC_state_t *pRExC_state, char *p )
9685 const char *e = RExC_end;
9687 PERL_ARGS_ASSERT_REGWHITE;
9692 else if (*p == '#') {
9701 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
9709 /* Parse POSIX character classes: [[:foo:]], [[=foo=]], [[.foo.]].
9710 Character classes ([:foo:]) can also be negated ([:^foo:]).
9711 Returns a named class id (ANYOF_XXX) if successful, -1 otherwise.
9712 Equivalence classes ([=foo=]) and composites ([.foo.]) are parsed,
9713 but trigger failures because they are currently unimplemented. */
9715 #define POSIXCC_DONE(c) ((c) == ':')
9716 #define POSIXCC_NOTYET(c) ((c) == '=' || (c) == '.')
9717 #define POSIXCC(c) (POSIXCC_DONE(c) || POSIXCC_NOTYET(c))
9720 S_regpposixcc(pTHX_ RExC_state_t *pRExC_state, I32 value)
9723 I32 namedclass = OOB_NAMEDCLASS;
9725 PERL_ARGS_ASSERT_REGPPOSIXCC;
9727 if (value == '[' && RExC_parse + 1 < RExC_end &&
9728 /* I smell either [: or [= or [. -- POSIX has been here, right? */
9729 POSIXCC(UCHARAT(RExC_parse))) {
9730 const char c = UCHARAT(RExC_parse);
9731 char* const s = RExC_parse++;
9733 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != c)
9735 if (RExC_parse == RExC_end)
9736 /* Grandfather lone [:, [=, [. */
9739 const char* const t = RExC_parse++; /* skip over the c */
9742 if (UCHARAT(RExC_parse) == ']') {
9743 const char *posixcc = s + 1;
9744 RExC_parse++; /* skip over the ending ] */
9747 const I32 complement = *posixcc == '^' ? *posixcc++ : 0;
9748 const I32 skip = t - posixcc;
9750 /* Initially switch on the length of the name. */
9753 if (memEQ(posixcc, "word", 4)) /* this is not POSIX, this is the Perl \w */
9754 namedclass = complement ? ANYOF_NALNUM : ANYOF_ALNUM;
9757 /* Names all of length 5. */
9758 /* alnum alpha ascii blank cntrl digit graph lower
9759 print punct space upper */
9760 /* Offset 4 gives the best switch position. */
9761 switch (posixcc[4]) {
9763 if (memEQ(posixcc, "alph", 4)) /* alpha */
9764 namedclass = complement ? ANYOF_NALPHA : ANYOF_ALPHA;
9767 if (memEQ(posixcc, "spac", 4)) /* space */
9768 namedclass = complement ? ANYOF_NPSXSPC : ANYOF_PSXSPC;
9771 if (memEQ(posixcc, "grap", 4)) /* graph */
9772 namedclass = complement ? ANYOF_NGRAPH : ANYOF_GRAPH;
9775 if (memEQ(posixcc, "asci", 4)) /* ascii */
9776 namedclass = complement ? ANYOF_NASCII : ANYOF_ASCII;
9779 if (memEQ(posixcc, "blan", 4)) /* blank */
9780 namedclass = complement ? ANYOF_NBLANK : ANYOF_BLANK;
9783 if (memEQ(posixcc, "cntr", 4)) /* cntrl */
9784 namedclass = complement ? ANYOF_NCNTRL : ANYOF_CNTRL;
9787 if (memEQ(posixcc, "alnu", 4)) /* alnum */
9788 namedclass = complement ? ANYOF_NALNUMC : ANYOF_ALNUMC;
9791 if (memEQ(posixcc, "lowe", 4)) /* lower */
9792 namedclass = complement ? ANYOF_NLOWER : ANYOF_LOWER;
9793 else if (memEQ(posixcc, "uppe", 4)) /* upper */
9794 namedclass = complement ? ANYOF_NUPPER : ANYOF_UPPER;
9797 if (memEQ(posixcc, "digi", 4)) /* digit */
9798 namedclass = complement ? ANYOF_NDIGIT : ANYOF_DIGIT;
9799 else if (memEQ(posixcc, "prin", 4)) /* print */
9800 namedclass = complement ? ANYOF_NPRINT : ANYOF_PRINT;
9801 else if (memEQ(posixcc, "punc", 4)) /* punct */
9802 namedclass = complement ? ANYOF_NPUNCT : ANYOF_PUNCT;
9807 if (memEQ(posixcc, "xdigit", 6))
9808 namedclass = complement ? ANYOF_NXDIGIT : ANYOF_XDIGIT;
9812 if (namedclass == OOB_NAMEDCLASS)
9813 Simple_vFAIL3("POSIX class [:%.*s:] unknown",
9815 assert (posixcc[skip] == ':');
9816 assert (posixcc[skip+1] == ']');
9817 } else if (!SIZE_ONLY) {
9818 /* [[=foo=]] and [[.foo.]] are still future. */
9820 /* adjust RExC_parse so the warning shows after
9822 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse) != ']')
9824 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
9827 /* Maternal grandfather:
9828 * "[:" ending in ":" but not in ":]" */
9838 S_checkposixcc(pTHX_ RExC_state_t *pRExC_state)
9842 PERL_ARGS_ASSERT_CHECKPOSIXCC;
9844 if (POSIXCC(UCHARAT(RExC_parse))) {
9845 const char *s = RExC_parse;
9846 const char c = *s++;
9850 if (*s && c == *s && s[1] == ']') {
9852 "POSIX syntax [%c %c] belongs inside character classes",
9855 /* [[=foo=]] and [[.foo.]] are still future. */
9856 if (POSIXCC_NOTYET(c)) {
9857 /* adjust RExC_parse so the error shows after
9859 while (UCHARAT(RExC_parse) && UCHARAT(RExC_parse++) != ']')
9861 Simple_vFAIL3("POSIX syntax [%c %c] is reserved for future extensions", c, c);
9867 /* No locale test, and always Unicode semantics, no ignore-case differences */
9868 #define _C_C_T_NOLOC_(NAME,TEST,WORD) \
9870 for (value = 0; value < 256; value++) \
9872 stored += set_regclass_bit(pRExC_state, ret, (U8) value, &l1_fold_invlist, &unicode_alternate); \
9876 case ANYOF_N##NAME: \
9877 for (value = 0; value < 256; value++) \
9879 stored += set_regclass_bit(pRExC_state, ret, (U8) value, &l1_fold_invlist, &unicode_alternate); \
9884 /* Like the above, but there are differences if we are in uni-8-bit or not, so
9885 * there are two tests passed in, to use depending on that. There aren't any
9886 * cases where the label is different from the name, so no need for that
9888 * Sets 'what' to WORD which is the property name for non-bitmap code points;
9889 * But, uses FOLD_WORD instead if /i has been selected, to allow a different
9891 #define _C_C_T_(NAME, TEST_8, TEST_7, WORD, FOLD_WORD) \
9893 if (LOC) ANYOF_CLASS_SET(ret, ANYOF_##NAME); \
9894 else if (UNI_SEMANTICS) { \
9895 for (value = 0; value < 256; value++) { \
9896 if (TEST_8(value)) stored += \
9897 set_regclass_bit(pRExC_state, ret, (U8) value, &l1_fold_invlist, &unicode_alternate); \
9901 for (value = 0; value < 128; value++) { \
9902 if (TEST_7(UNI_TO_NATIVE(value))) stored += \
9903 set_regclass_bit(pRExC_state, ret, \
9904 (U8) UNI_TO_NATIVE(value), &l1_fold_invlist, &unicode_alternate); \
9915 case ANYOF_N##NAME: \
9916 if (LOC) ANYOF_CLASS_SET(ret, ANYOF_N##NAME); \
9917 else if (UNI_SEMANTICS) { \
9918 for (value = 0; value < 256; value++) { \
9919 if (! TEST_8(value)) stored += \
9920 set_regclass_bit(pRExC_state, ret, (U8) value, &l1_fold_invlist, &unicode_alternate); \
9924 for (value = 0; value < 128; value++) { \
9925 if (! TEST_7(UNI_TO_NATIVE(value))) stored += set_regclass_bit( \
9926 pRExC_state, ret, (U8) UNI_TO_NATIVE(value), &l1_fold_invlist, &unicode_alternate); \
9928 if (AT_LEAST_ASCII_RESTRICTED) { \
9929 for (value = 128; value < 256; value++) { \
9930 stored += set_regclass_bit( \
9931 pRExC_state, ret, (U8) UNI_TO_NATIVE(value), &l1_fold_invlist, &unicode_alternate); \
9933 ANYOF_FLAGS(ret) |= ANYOF_UNICODE_ALL; \
9936 /* For a non-ut8 target string with DEPENDS semantics, all above \
9937 * ASCII Latin1 code points match the complement of any of the \
9938 * classes. But in utf8, they have their Unicode semantics, so \
9939 * can't just set them in the bitmap, or else regexec.c will think \
9940 * they matched when they shouldn't. */ \
9941 ANYOF_FLAGS(ret) |= ANYOF_NON_UTF8_LATIN1_ALL; \
9954 S_set_regclass_bit_fold(pTHX_ RExC_state_t *pRExC_state, regnode* node, const U8 value, SV** invlist_ptr, AV** alternate_ptr)
9957 /* Handle the setting of folds in the bitmap for non-locale ANYOF nodes.
9958 * Locale folding is done at run-time, so this function should not be
9959 * called for nodes that are for locales.
9961 * This function sets the bit corresponding to the fold of the input
9962 * 'value', if not already set. The fold of 'f' is 'F', and the fold of
9965 * It also knows about the characters that are in the bitmap that have
9966 * folds that are matchable only outside it, and sets the appropriate lists
9969 * It returns the number of bits that actually changed from 0 to 1 */
9974 PERL_ARGS_ASSERT_SET_REGCLASS_BIT_FOLD;
9976 fold = (AT_LEAST_UNI_SEMANTICS) ? PL_fold_latin1[value]
9979 /* It assumes the bit for 'value' has already been set */
9980 if (fold != value && ! ANYOF_BITMAP_TEST(node, fold)) {
9981 ANYOF_BITMAP_SET(node, fold);
9984 if (_HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(value) && (! isASCII(value) || ! MORE_ASCII_RESTRICTED)) {
9985 /* Certain Latin1 characters have matches outside the bitmap. To get
9986 * here, 'value' is one of those characters. None of these matches is
9987 * valid for ASCII characters under /aa, which have been excluded by
9988 * the 'if' above. The matches fall into three categories:
9989 * 1) They are singly folded-to or -from an above 255 character, as
9990 * LATIN SMALL LETTER Y WITH DIAERESIS and LATIN CAPITAL LETTER Y
9992 * 2) They are part of a multi-char fold with another character in the
9993 * bitmap, only LATIN SMALL LETTER SHARP S => "ss" fits that bill;
9994 * 3) They are part of a multi-char fold with a character not in the
9995 * bitmap, such as various ligatures.
9996 * We aren't dealing fully with multi-char folds, except we do deal
9997 * with the pattern containing a character that has a multi-char fold
9998 * (not so much the inverse).
9999 * For types 1) and 3), the matches only happen when the target string
10000 * is utf8; that's not true for 2), and we set a flag for it.
10002 * The code below adds to the passed in inversion list the single fold
10003 * closures for 'value'. The values are hard-coded here so that an
10004 * innocent-looking character class, like /[ks]/i won't have to go out
10005 * to disk to find the possible matches. XXX It would be better to
10006 * generate these via regen, in case a new version of the Unicode
10007 * standard adds new mappings, though that is not really likely. */
10012 *invlist_ptr = add_cp_to_invlist(*invlist_ptr, 0x212A);
10016 /* LATIN SMALL LETTER LONG S */
10017 *invlist_ptr = add_cp_to_invlist(*invlist_ptr, 0x017F);
10020 *invlist_ptr = add_cp_to_invlist(*invlist_ptr,
10021 GREEK_SMALL_LETTER_MU);
10022 *invlist_ptr = add_cp_to_invlist(*invlist_ptr,
10023 GREEK_CAPITAL_LETTER_MU);
10025 case LATIN_CAPITAL_LETTER_A_WITH_RING_ABOVE:
10026 case LATIN_SMALL_LETTER_A_WITH_RING_ABOVE:
10027 /* ANGSTROM SIGN */
10028 *invlist_ptr = add_cp_to_invlist(*invlist_ptr, 0x212B);
10029 if (DEPENDS_SEMANTICS) { /* See DEPENDS comment below */
10030 *invlist_ptr = add_cp_to_invlist(*invlist_ptr,
10031 PL_fold_latin1[value]);
10034 case LATIN_SMALL_LETTER_Y_WITH_DIAERESIS:
10035 *invlist_ptr = add_cp_to_invlist(*invlist_ptr,
10036 LATIN_CAPITAL_LETTER_Y_WITH_DIAERESIS);
10038 case LATIN_SMALL_LETTER_SHARP_S:
10039 *invlist_ptr = add_cp_to_invlist(*invlist_ptr,
10040 LATIN_CAPITAL_LETTER_SHARP_S);
10042 /* Under /a, /d, and /u, this can match the two chars "ss" */
10043 if (! MORE_ASCII_RESTRICTED) {
10044 add_alternate(alternate_ptr, (U8 *) "ss", 2);
10046 /* And under /u or /a, it can match even if the target is
10048 if (AT_LEAST_UNI_SEMANTICS) {
10049 ANYOF_FLAGS(node) |= ANYOF_NONBITMAP_NON_UTF8;
10053 case 'F': case 'f':
10054 case 'I': case 'i':
10055 case 'L': case 'l':
10056 case 'T': case 't':
10057 case 'A': case 'a':
10058 case 'H': case 'h':
10059 case 'J': case 'j':
10060 case 'N': case 'n':
10061 case 'W': case 'w':
10062 case 'Y': case 'y':
10063 /* These all are targets of multi-character folds from code
10064 * points that require UTF8 to express, so they can't match
10065 * unless the target string is in UTF-8, so no action here is
10066 * necessary, as regexec.c properly handles the general case
10067 * for UTF-8 matching */
10070 /* Use deprecated warning to increase the chances of this
10072 ckWARN2regdep(RExC_parse, "Perl folding rules are not up-to-date for 0x%x; please use the perlbug utility to report;", value);
10076 else if (DEPENDS_SEMANTICS
10077 && ! isASCII(value)
10078 && PL_fold_latin1[value] != value)
10080 /* Under DEPENDS rules, non-ASCII Latin1 characters match their
10081 * folds only when the target string is in UTF-8. We add the fold
10082 * here to the list of things to match outside the bitmap, which
10083 * won't be looked at unless it is UTF8 (or else if something else
10084 * says to look even if not utf8, but those things better not happen
10085 * under DEPENDS semantics. */
10086 *invlist_ptr = add_cp_to_invlist(*invlist_ptr, PL_fold_latin1[value]);
10093 PERL_STATIC_INLINE U8
10094 S_set_regclass_bit(pTHX_ RExC_state_t *pRExC_state, regnode* node, const U8 value, SV** invlist_ptr, AV** alternate_ptr)
10096 /* This inline function sets a bit in the bitmap if not already set, and if
10097 * appropriate, its fold, returning the number of bits that actually
10098 * changed from 0 to 1 */
10102 PERL_ARGS_ASSERT_SET_REGCLASS_BIT;
10104 if (ANYOF_BITMAP_TEST(node, value)) { /* Already set */
10108 ANYOF_BITMAP_SET(node, value);
10111 if (FOLD && ! LOC) { /* Locale folds aren't known until runtime */
10112 stored += set_regclass_bit_fold(pRExC_state, node, value, invlist_ptr, alternate_ptr);
10119 S_add_alternate(pTHX_ AV** alternate_ptr, U8* string, STRLEN len)
10121 /* Adds input 'string' with length 'len' to the ANYOF node's unicode
10122 * alternate list, pointed to by 'alternate_ptr'. This is an array of
10123 * the multi-character folds of characters in the node */
10126 PERL_ARGS_ASSERT_ADD_ALTERNATE;
10128 if (! *alternate_ptr) {
10129 *alternate_ptr = newAV();
10131 sv = newSVpvn_utf8((char*)string, len, TRUE);
10132 av_push(*alternate_ptr, sv);
10137 parse a class specification and produce either an ANYOF node that
10138 matches the pattern or perhaps will be optimized into an EXACTish node
10139 instead. The node contains a bit map for the first 256 characters, with the
10140 corresponding bit set if that character is in the list. For characters
10141 above 255, a range list is used */
10144 S_regclass(pTHX_ RExC_state_t *pRExC_state, U32 depth)
10147 register UV nextvalue;
10148 register IV prevvalue = OOB_UNICODE;
10149 register IV range = 0;
10150 UV value = 0; /* XXX:dmq: needs to be referenceable (unfortunately) */
10151 register regnode *ret;
10154 char *rangebegin = NULL;
10155 bool need_class = 0;
10156 bool allow_full_fold = TRUE; /* Assume wants multi-char folding */
10158 STRLEN initial_listsv_len = 0; /* Kind of a kludge to see if it is more
10159 than just initialized. */
10160 SV* properties = NULL; /* Code points that match \p{} \P{} */
10161 UV element_count = 0; /* Number of distinct elements in the class.
10162 Optimizations may be possible if this is tiny */
10165 /* Unicode properties are stored in a swash; this holds the current one
10166 * being parsed. If this swash is the only above-latin1 component of the
10167 * character class, an optimization is to pass it directly on to the
10168 * execution engine. Otherwise, it is set to NULL to indicate that there
10169 * are other things in the class that have to be dealt with at execution
10171 SV* swash = NULL; /* Code points that match \p{} \P{} */
10173 /* Set if a component of this character class is user-defined; just passed
10174 * on to the engine */
10175 UV has_user_defined_property = 0;
10177 /* code points this node matches that can't be stored in the bitmap */
10178 SV* nonbitmap = NULL;
10180 /* The items that are to match that aren't stored in the bitmap, but are a
10181 * result of things that are stored there. This is the fold closure of
10182 * such a character, either because it has DEPENDS semantics and shouldn't
10183 * be matched unless the target string is utf8, or is a code point that is
10184 * too large for the bit map, as for example, the fold of the MICRO SIGN is
10185 * above 255. This all is solely for performance reasons. By having this
10186 * code know the outside-the-bitmap folds that the bitmapped characters are
10187 * involved with, we don't have to go out to disk to find the list of
10188 * matches, unless the character class includes code points that aren't
10189 * storable in the bit map. That means that a character class with an 's'
10190 * in it, for example, doesn't need to go out to disk to find everything
10191 * that matches. A 2nd list is used so that the 'nonbitmap' list is kept
10192 * empty unless there is something whose fold we don't know about, and will
10193 * have to go out to the disk to find. */
10194 SV* l1_fold_invlist = NULL;
10196 /* List of multi-character folds that are matched by this node */
10197 AV* unicode_alternate = NULL;
10199 UV literal_endpoint = 0;
10201 UV stored = 0; /* how many chars stored in the bitmap */
10203 regnode * const orig_emit = RExC_emit; /* Save the original RExC_emit in
10204 case we need to change the emitted regop to an EXACT. */
10205 const char * orig_parse = RExC_parse;
10206 GET_RE_DEBUG_FLAGS_DECL;
10208 PERL_ARGS_ASSERT_REGCLASS;
10210 PERL_UNUSED_ARG(depth);
10213 DEBUG_PARSE("clas");
10215 /* Assume we are going to generate an ANYOF node. */
10216 ret = reganode(pRExC_state, ANYOF, 0);
10220 ANYOF_FLAGS(ret) = 0;
10223 if (UCHARAT(RExC_parse) == '^') { /* Complement of range. */
10227 ANYOF_FLAGS(ret) |= ANYOF_INVERT;
10229 /* We have decided to not allow multi-char folds in inverted character
10230 * classes, due to the confusion that can happen, especially with
10231 * classes that are designed for a non-Unicode world: You have the
10232 * peculiar case that:
10233 "s s" =~ /^[^\xDF]+$/i => Y
10234 "ss" =~ /^[^\xDF]+$/i => N
10236 * See [perl #89750] */
10237 allow_full_fold = FALSE;
10241 RExC_size += ANYOF_SKIP;
10242 listsv = &PL_sv_undef; /* For code scanners: listsv always non-NULL. */
10245 RExC_emit += ANYOF_SKIP;
10247 ANYOF_FLAGS(ret) |= ANYOF_LOCALE;
10249 ANYOF_BITMAP_ZERO(ret);
10250 listsv = newSVpvs("# comment\n");
10251 initial_listsv_len = SvCUR(listsv);
10254 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
10256 if (!SIZE_ONLY && POSIXCC(nextvalue))
10257 checkposixcc(pRExC_state);
10259 /* allow 1st char to be ] (allowing it to be - is dealt with later) */
10260 if (UCHARAT(RExC_parse) == ']')
10261 goto charclassloop;
10264 while (RExC_parse < RExC_end && UCHARAT(RExC_parse) != ']') {
10268 namedclass = OOB_NAMEDCLASS; /* initialize as illegal */
10271 rangebegin = RExC_parse;
10275 value = utf8n_to_uvchr((U8*)RExC_parse,
10276 RExC_end - RExC_parse,
10277 &numlen, UTF8_ALLOW_DEFAULT);
10278 RExC_parse += numlen;
10281 value = UCHARAT(RExC_parse++);
10283 nextvalue = RExC_parse < RExC_end ? UCHARAT(RExC_parse) : 0;
10284 if (value == '[' && POSIXCC(nextvalue))
10285 namedclass = regpposixcc(pRExC_state, value);
10286 else if (value == '\\') {
10288 value = utf8n_to_uvchr((U8*)RExC_parse,
10289 RExC_end - RExC_parse,
10290 &numlen, UTF8_ALLOW_DEFAULT);
10291 RExC_parse += numlen;
10294 value = UCHARAT(RExC_parse++);
10295 /* Some compilers cannot handle switching on 64-bit integer
10296 * values, therefore value cannot be an UV. Yes, this will
10297 * be a problem later if we want switch on Unicode.
10298 * A similar issue a little bit later when switching on
10299 * namedclass. --jhi */
10300 switch ((I32)value) {
10301 case 'w': namedclass = ANYOF_ALNUM; break;
10302 case 'W': namedclass = ANYOF_NALNUM; break;
10303 case 's': namedclass = ANYOF_SPACE; break;
10304 case 'S': namedclass = ANYOF_NSPACE; break;
10305 case 'd': namedclass = ANYOF_DIGIT; break;
10306 case 'D': namedclass = ANYOF_NDIGIT; break;
10307 case 'v': namedclass = ANYOF_VERTWS; break;
10308 case 'V': namedclass = ANYOF_NVERTWS; break;
10309 case 'h': namedclass = ANYOF_HORIZWS; break;
10310 case 'H': namedclass = ANYOF_NHORIZWS; break;
10311 case 'N': /* Handle \N{NAME} in class */
10313 /* We only pay attention to the first char of
10314 multichar strings being returned. I kinda wonder
10315 if this makes sense as it does change the behaviour
10316 from earlier versions, OTOH that behaviour was broken
10318 UV v; /* value is register so we cant & it /grrr */
10319 if (reg_namedseq(pRExC_state, &v, NULL, depth)) {
10329 if (RExC_parse >= RExC_end)
10330 vFAIL2("Empty \\%c{}", (U8)value);
10331 if (*RExC_parse == '{') {
10332 const U8 c = (U8)value;
10333 e = strchr(RExC_parse++, '}');
10335 vFAIL2("Missing right brace on \\%c{}", c);
10336 while (isSPACE(UCHARAT(RExC_parse)))
10338 if (e == RExC_parse)
10339 vFAIL2("Empty \\%c{}", c);
10340 n = e - RExC_parse;
10341 while (isSPACE(UCHARAT(RExC_parse + n - 1)))
10352 if (UCHARAT(RExC_parse) == '^') {
10355 value = value == 'p' ? 'P' : 'p'; /* toggle */
10356 while (isSPACE(UCHARAT(RExC_parse))) {
10361 /* Try to get the definition of the property into
10362 * <invlist>. If /i is in effect, the effective property
10363 * will have its name be <__NAME_i>. The design is
10364 * discussed in commit
10365 * 2f833f5208e26b208886e51e09e2c072b5eabb46 */
10366 Newx(name, n + sizeof("_i__\n"), char);
10368 sprintf(name, "%s%.*s%s\n",
10369 (FOLD) ? "__" : "",
10375 /* Look up the property name, and get its swash and
10376 * inversion list, if the property is found */
10378 SvREFCNT_dec(swash);
10380 swash = _core_swash_init("utf8", name, &PL_sv_undef,
10383 TRUE, /* this routine will handle
10384 undefined properties */
10385 NULL, FALSE /* No inversion list */
10389 || ! SvTYPE(SvRV(swash)) == SVt_PVHV
10391 hv_fetchs(MUTABLE_HV(SvRV(swash)),
10393 || ! (invlist = *invlistsvp))
10396 SvREFCNT_dec(swash);
10400 /* Here didn't find it. It could be a user-defined
10401 * property that will be available at run-time. Add it
10402 * to the list to look up then */
10403 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::%s\n",
10404 (value == 'p' ? '+' : '!'),
10406 has_user_defined_property = 1;
10408 /* We don't know yet, so have to assume that the
10409 * property could match something in the Latin1 range,
10410 * hence something that isn't utf8 */
10411 ANYOF_FLAGS(ret) |= ANYOF_NONBITMAP_NON_UTF8;
10415 /* Here, did get the swash and its inversion list. If
10416 * the swash is from a user-defined property, then this
10417 * whole character class should be regarded as such */
10418 SV** user_defined_svp =
10419 hv_fetchs(MUTABLE_HV(SvRV(swash)),
10420 "USER_DEFINED", FALSE);
10421 if (user_defined_svp) {
10422 has_user_defined_property
10423 |= SvUV(*user_defined_svp);
10426 /* Invert if asking for the complement */
10427 if (value == 'P') {
10429 /* Add to any existing list */
10430 if (! properties) {
10431 properties = invlist_clone(invlist);
10432 _invlist_invert(properties);
10435 invlist = invlist_clone(invlist);
10436 _invlist_invert(invlist);
10437 _invlist_union(properties, invlist, &properties);
10438 SvREFCNT_dec(invlist);
10441 /* The swash can't be used as-is, because we've
10442 * inverted things; delay removing it to here after
10443 * have copied its invlist above */
10444 SvREFCNT_dec(swash);
10448 if (! properties) {
10449 properties = invlist_clone(invlist);
10452 _invlist_union(properties, invlist, &properties);
10458 RExC_parse = e + 1;
10459 namedclass = ANYOF_MAX; /* no official name, but it's named */
10461 /* \p means they want Unicode semantics */
10462 RExC_uni_semantics = 1;
10465 case 'n': value = '\n'; break;
10466 case 'r': value = '\r'; break;
10467 case 't': value = '\t'; break;
10468 case 'f': value = '\f'; break;
10469 case 'b': value = '\b'; break;
10470 case 'e': value = ASCII_TO_NATIVE('\033');break;
10471 case 'a': value = ASCII_TO_NATIVE('\007');break;
10473 RExC_parse--; /* function expects to be pointed at the 'o' */
10475 const char* error_msg;
10476 bool valid = grok_bslash_o(RExC_parse,
10481 RExC_parse += numlen;
10486 if (PL_encoding && value < 0x100) {
10487 goto recode_encoding;
10491 if (*RExC_parse == '{') {
10492 I32 flags = PERL_SCAN_ALLOW_UNDERSCORES
10493 | PERL_SCAN_DISALLOW_PREFIX;
10494 char * const e = strchr(RExC_parse++, '}');
10496 vFAIL("Missing right brace on \\x{}");
10498 numlen = e - RExC_parse;
10499 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
10500 RExC_parse = e + 1;
10503 I32 flags = PERL_SCAN_DISALLOW_PREFIX;
10505 value = grok_hex(RExC_parse, &numlen, &flags, NULL);
10506 RExC_parse += numlen;
10508 if (PL_encoding && value < 0x100)
10509 goto recode_encoding;
10512 value = grok_bslash_c(*RExC_parse++, UTF, SIZE_ONLY);
10514 case '0': case '1': case '2': case '3': case '4':
10515 case '5': case '6': case '7':
10517 /* Take 1-3 octal digits */
10518 I32 flags = PERL_SCAN_SILENT_ILLDIGIT;
10520 value = grok_oct(--RExC_parse, &numlen, &flags, NULL);
10521 RExC_parse += numlen;
10522 if (PL_encoding && value < 0x100)
10523 goto recode_encoding;
10527 if (! RExC_override_recoding) {
10528 SV* enc = PL_encoding;
10529 value = reg_recode((const char)(U8)value, &enc);
10530 if (!enc && SIZE_ONLY)
10531 ckWARNreg(RExC_parse,
10532 "Invalid escape in the specified encoding");
10536 /* Allow \_ to not give an error */
10537 if (!SIZE_ONLY && isALNUM(value) && value != '_') {
10538 ckWARN2reg(RExC_parse,
10539 "Unrecognized escape \\%c in character class passed through",
10544 } /* end of \blah */
10547 literal_endpoint++;
10550 if (namedclass > OOB_NAMEDCLASS) { /* this is a named class \blah */
10552 /* What matches in a locale is not known until runtime, so need to
10553 * (one time per class) allocate extra space to pass to regexec.
10554 * The space will contain a bit for each named class that is to be
10555 * matched against. This isn't needed for \p{} and pseudo-classes,
10556 * as they are not affected by locale, and hence are dealt with
10558 if (LOC && namedclass < ANYOF_MAX && ! need_class) {
10561 RExC_size += ANYOF_CLASS_SKIP - ANYOF_SKIP;
10564 RExC_emit += ANYOF_CLASS_SKIP - ANYOF_SKIP;
10565 ANYOF_CLASS_ZERO(ret);
10567 ANYOF_FLAGS(ret) |= ANYOF_CLASS;
10570 /* a bad range like a-\d, a-[:digit:]. The '-' is taken as a
10571 * literal, as is the character that began the false range, i.e.
10572 * the 'a' in the examples */
10576 RExC_parse >= rangebegin ?
10577 RExC_parse - rangebegin : 0;
10578 ckWARN4reg(RExC_parse,
10579 "False [] range \"%*.*s\"",
10583 set_regclass_bit(pRExC_state, ret, '-', &l1_fold_invlist, &unicode_alternate);
10584 if (prevvalue < 256) {
10586 set_regclass_bit(pRExC_state, ret, (U8) prevvalue, &l1_fold_invlist, &unicode_alternate);
10589 nonbitmap = add_cp_to_invlist(nonbitmap, prevvalue);
10593 range = 0; /* this was not a true range */
10597 const char *what = NULL;
10600 /* Possible truncation here but in some 64-bit environments
10601 * the compiler gets heartburn about switch on 64-bit values.
10602 * A similar issue a little earlier when switching on value.
10604 switch ((I32)namedclass) {
10606 case _C_C_T_(ALNUMC, isALNUMC_L1, isALNUMC, "XPosixAlnum", "XPosixAlnum");
10607 case _C_C_T_(ALPHA, isALPHA_L1, isALPHA, "XPosixAlpha", "XPosixAlpha");
10608 case _C_C_T_(BLANK, isBLANK_L1, isBLANK, "XPosixBlank", "XPosixBlank");
10609 case _C_C_T_(CNTRL, isCNTRL_L1, isCNTRL, "XPosixCntrl", "XPosixCntrl");
10610 case _C_C_T_(GRAPH, isGRAPH_L1, isGRAPH, "XPosixGraph", "XPosixGraph");
10611 case _C_C_T_(LOWER, isLOWER_L1, isLOWER, "XPosixLower", "__XPosixLower_i");
10612 case _C_C_T_(PRINT, isPRINT_L1, isPRINT, "XPosixPrint", "XPosixPrint");
10613 case _C_C_T_(PSXSPC, isPSXSPC_L1, isPSXSPC, "XPosixSpace", "XPosixSpace");
10614 case _C_C_T_(PUNCT, isPUNCT_L1, isPUNCT, "XPosixPunct", "XPosixPunct");
10615 case _C_C_T_(UPPER, isUPPER_L1, isUPPER, "XPosixUpper", "__XPosixUpper_i");
10616 /* \s, \w match all unicode if utf8. */
10617 case _C_C_T_(SPACE, isSPACE_L1, isSPACE, "SpacePerl", "SpacePerl");
10618 case _C_C_T_(ALNUM, isWORDCHAR_L1, isALNUM, "Word", "Word");
10619 case _C_C_T_(XDIGIT, isXDIGIT_L1, isXDIGIT, "XPosixXDigit", "XPosixXDigit");
10620 case _C_C_T_NOLOC_(VERTWS, is_VERTWS_latin1(&value), "VertSpace");
10621 case _C_C_T_NOLOC_(HORIZWS, is_HORIZWS_latin1(&value), "HorizSpace");
10624 ANYOF_CLASS_SET(ret, ANYOF_ASCII);
10626 for (value = 0; value < 128; value++)
10628 set_regclass_bit(pRExC_state, ret, (U8) ASCII_TO_NATIVE(value), &l1_fold_invlist, &unicode_alternate);
10631 what = NULL; /* Doesn't match outside ascii, so
10632 don't want to add +utf8:: */
10636 ANYOF_CLASS_SET(ret, ANYOF_NASCII);
10638 for (value = 128; value < 256; value++)
10640 set_regclass_bit(pRExC_state, ret, (U8) ASCII_TO_NATIVE(value), &l1_fold_invlist, &unicode_alternate);
10642 ANYOF_FLAGS(ret) |= ANYOF_UNICODE_ALL;
10648 ANYOF_CLASS_SET(ret, ANYOF_DIGIT);
10650 /* consecutive digits assumed */
10651 for (value = '0'; value <= '9'; value++)
10653 set_regclass_bit(pRExC_state, ret, (U8) value, &l1_fold_invlist, &unicode_alternate);
10660 ANYOF_CLASS_SET(ret, ANYOF_NDIGIT);
10662 /* consecutive digits assumed */
10663 for (value = 0; value < '0'; value++)
10665 set_regclass_bit(pRExC_state, ret, (U8) value, &l1_fold_invlist, &unicode_alternate);
10666 for (value = '9' + 1; value < 256; value++)
10668 set_regclass_bit(pRExC_state, ret, (U8) value, &l1_fold_invlist, &unicode_alternate);
10672 if (AT_LEAST_ASCII_RESTRICTED ) {
10673 ANYOF_FLAGS(ret) |= ANYOF_UNICODE_ALL;
10677 /* this is to handle \p and \P */
10680 vFAIL("Invalid [::] class");
10683 if (what && ! (AT_LEAST_ASCII_RESTRICTED)) {
10684 /* Strings such as "+utf8::isWord\n" */
10685 Perl_sv_catpvf(aTHX_ listsv, "%cutf8::%s\n", yesno, what);
10690 } /* end of namedclass \blah */
10693 if (prevvalue > (IV)value) /* b-a */ {
10694 const int w = RExC_parse - rangebegin;
10695 Simple_vFAIL4("Invalid [] range \"%*.*s\"", w, w, rangebegin);
10696 range = 0; /* not a valid range */
10700 prevvalue = value; /* save the beginning of the range */
10701 if (RExC_parse+1 < RExC_end
10702 && *RExC_parse == '-'
10703 && RExC_parse[1] != ']')
10707 /* a bad range like \w-, [:word:]- ? */
10708 if (namedclass > OOB_NAMEDCLASS) {
10709 if (ckWARN(WARN_REGEXP)) {
10711 RExC_parse >= rangebegin ?
10712 RExC_parse - rangebegin : 0;
10714 "False [] range \"%*.*s\"",
10719 set_regclass_bit(pRExC_state, ret, '-', &l1_fold_invlist, &unicode_alternate);
10721 range = 1; /* yeah, it's a range! */
10722 continue; /* but do it the next time */
10726 /* non-Latin1 code point implies unicode semantics. Must be set in
10727 * pass1 so is there for the whole of pass 2 */
10729 RExC_uni_semantics = 1;
10732 /* now is the next time */
10734 if (prevvalue < 256) {
10735 const IV ceilvalue = value < 256 ? value : 255;
10738 /* In EBCDIC [\x89-\x91] should include
10739 * the \x8e but [i-j] should not. */
10740 if (literal_endpoint == 2 &&
10741 ((isLOWER(prevvalue) && isLOWER(ceilvalue)) ||
10742 (isUPPER(prevvalue) && isUPPER(ceilvalue))))
10744 if (isLOWER(prevvalue)) {
10745 for (i = prevvalue; i <= ceilvalue; i++)
10746 if (isLOWER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
10748 set_regclass_bit(pRExC_state, ret, (U8) i, &l1_fold_invlist, &unicode_alternate);
10751 for (i = prevvalue; i <= ceilvalue; i++)
10752 if (isUPPER(i) && !ANYOF_BITMAP_TEST(ret,i)) {
10754 set_regclass_bit(pRExC_state, ret, (U8) i, &l1_fold_invlist, &unicode_alternate);
10760 for (i = prevvalue; i <= ceilvalue; i++) {
10761 stored += set_regclass_bit(pRExC_state, ret, (U8) i, &l1_fold_invlist, &unicode_alternate);
10765 const UV prevnatvalue = NATIVE_TO_UNI(prevvalue);
10766 const UV natvalue = NATIVE_TO_UNI(value);
10767 nonbitmap = add_range_to_invlist(nonbitmap, prevnatvalue, natvalue);
10770 literal_endpoint = 0;
10774 range = 0; /* this range (if it was one) is done now */
10781 /****** !SIZE_ONLY AFTER HERE *********/
10783 /* If folding and there are code points above 255, we calculate all
10784 * characters that could fold to or from the ones already on the list */
10785 if (FOLD && nonbitmap) {
10786 UV start, end; /* End points of code point ranges */
10788 SV* fold_intersection = NULL;
10790 /* This is a list of all the characters that participate in folds
10791 * (except marks, etc in multi-char folds */
10792 if (! PL_utf8_foldable) {
10793 SV* swash = swash_init("utf8", "Cased", &PL_sv_undef, 1, 0);
10794 PL_utf8_foldable = _swash_to_invlist(swash);
10795 SvREFCNT_dec(swash);
10798 /* This is a hash that for a particular fold gives all characters
10799 * that are involved in it */
10800 if (! PL_utf8_foldclosures) {
10802 /* If we were unable to find any folds, then we likely won't be
10803 * able to find the closures. So just create an empty list.
10804 * Folding will effectively be restricted to the non-Unicode rules
10805 * hard-coded into Perl. (This case happens legitimately during
10806 * compilation of Perl itself before the Unicode tables are
10808 if (invlist_len(PL_utf8_foldable) == 0) {
10809 PL_utf8_foldclosures = newHV();
10811 /* If the folds haven't been read in, call a fold function
10813 if (! PL_utf8_tofold) {
10814 U8 dummy[UTF8_MAXBYTES+1];
10817 /* This particular string is above \xff in both UTF-8 and
10819 to_utf8_fold((U8*) "\xC8\x80", dummy, &dummy_len);
10820 assert(PL_utf8_tofold); /* Verify that worked */
10822 PL_utf8_foldclosures = _swash_inversion_hash(PL_utf8_tofold);
10826 /* Only the characters in this class that participate in folds need be
10827 * checked. Get the intersection of this class and all the possible
10828 * characters that are foldable. This can quickly narrow down a large
10830 _invlist_intersection(PL_utf8_foldable, nonbitmap, &fold_intersection);
10832 /* Now look at the foldable characters in this class individually */
10833 invlist_iterinit(fold_intersection);
10834 while (invlist_iternext(fold_intersection, &start, &end)) {
10837 /* Look at every character in the range */
10838 for (j = start; j <= end; j++) {
10841 U8 foldbuf[UTF8_MAXBYTES_CASE+1];
10844 _to_uni_fold_flags(j, foldbuf, &foldlen, allow_full_fold);
10846 if (foldlen > (STRLEN)UNISKIP(f)) {
10848 /* Any multicharacter foldings (disallowed in lookbehind
10849 * patterns) require the following transform: [ABCDEF] ->
10850 * (?:[ABCabcDEFd]|pq|rst) where E folds into "pq" and F
10851 * folds into "rst", all other characters fold to single
10852 * characters. We save away these multicharacter foldings,
10853 * to be later saved as part of the additional "s" data. */
10854 if (! RExC_in_lookbehind) {
10856 U8* e = foldbuf + foldlen;
10858 /* If any of the folded characters of this are in the
10859 * Latin1 range, tell the regex engine that this can
10860 * match a non-utf8 target string. The only multi-byte
10861 * fold whose source is in the Latin1 range (U+00DF)
10862 * applies only when the target string is utf8, or
10863 * under unicode rules */
10864 if (j > 255 || AT_LEAST_UNI_SEMANTICS) {
10867 /* Can't mix ascii with non- under /aa */
10868 if (MORE_ASCII_RESTRICTED
10869 && (isASCII(*loc) != isASCII(j)))
10871 goto end_multi_fold;
10873 if (UTF8_IS_INVARIANT(*loc)
10874 || UTF8_IS_DOWNGRADEABLE_START(*loc))
10876 /* Can't mix above and below 256 under LOC
10879 goto end_multi_fold;
10882 |= ANYOF_NONBITMAP_NON_UTF8;
10885 loc += UTF8SKIP(loc);
10889 add_alternate(&unicode_alternate, foldbuf, foldlen);
10893 /* This is special-cased, as it is the only letter which
10894 * has both a multi-fold and single-fold in Latin1. All
10895 * the other chars that have single and multi-folds are
10896 * always in utf8, and the utf8 folding algorithm catches
10898 if (! LOC && j == LATIN_CAPITAL_LETTER_SHARP_S) {
10899 stored += set_regclass_bit(pRExC_state,
10901 LATIN_SMALL_LETTER_SHARP_S,
10902 &l1_fold_invlist, &unicode_alternate);
10906 /* Single character fold. Add everything in its fold
10907 * closure to the list that this node should match */
10910 /* The fold closures data structure is a hash with the keys
10911 * being every character that is folded to, like 'k', and
10912 * the values each an array of everything that folds to its
10913 * key. e.g. [ 'k', 'K', KELVIN_SIGN ] */
10914 if ((listp = hv_fetch(PL_utf8_foldclosures,
10915 (char *) foldbuf, foldlen, FALSE)))
10917 AV* list = (AV*) *listp;
10919 for (k = 0; k <= av_len(list); k++) {
10920 SV** c_p = av_fetch(list, k, FALSE);
10923 Perl_croak(aTHX_ "panic: invalid PL_utf8_foldclosures structure");
10927 /* /aa doesn't allow folds between ASCII and non-;
10928 * /l doesn't allow them between above and below
10930 if ((MORE_ASCII_RESTRICTED
10931 && (isASCII(c) != isASCII(j)))
10932 || (LOC && ((c < 256) != (j < 256))))
10937 if (c < 256 && AT_LEAST_UNI_SEMANTICS) {
10938 stored += set_regclass_bit(pRExC_state,
10941 &l1_fold_invlist, &unicode_alternate);
10943 /* It may be that the code point is already in
10944 * this range or already in the bitmap, in
10945 * which case we need do nothing */
10946 else if ((c < start || c > end)
10948 || ! ANYOF_BITMAP_TEST(ret, c)))
10950 nonbitmap = add_cp_to_invlist(nonbitmap, c);
10957 SvREFCNT_dec(fold_intersection);
10960 /* Combine the two lists into one. */
10961 if (l1_fold_invlist) {
10963 _invlist_union(nonbitmap, l1_fold_invlist, &nonbitmap);
10964 SvREFCNT_dec(l1_fold_invlist);
10967 nonbitmap = l1_fold_invlist;
10971 /* And combine the result (if any) with any inversion list from properties.
10972 * The lists are kept separate up to now because we don't want to fold the
10976 _invlist_union(nonbitmap, properties, &nonbitmap);
10977 SvREFCNT_dec(properties);
10980 nonbitmap = properties;
10984 /* Here, <nonbitmap> contains all the code points we can determine at
10985 * compile time that we haven't put into the bitmap. Go through it, and
10986 * for things that belong in the bitmap, put them there, and delete from
10990 /* Above-ASCII code points in /d have to stay in <nonbitmap>, as they
10991 * possibly only should match when the target string is UTF-8 */
10992 UV max_cp_to_set = (DEPENDS_SEMANTICS) ? 127 : 255;
10994 /* This gets set if we actually need to modify things */
10995 bool change_invlist = FALSE;
10999 /* Start looking through <nonbitmap> */
11000 invlist_iterinit(nonbitmap);
11001 while (invlist_iternext(nonbitmap, &start, &end)) {
11005 /* Quit if are above what we should change */
11006 if (start > max_cp_to_set) {
11010 change_invlist = TRUE;
11012 /* Set all the bits in the range, up to the max that we are doing */
11013 high = (end < max_cp_to_set) ? end : max_cp_to_set;
11014 for (i = start; i <= (int) high; i++) {
11015 if (! ANYOF_BITMAP_TEST(ret, i)) {
11016 ANYOF_BITMAP_SET(ret, i);
11024 /* Done with loop; set <nonbitmap> to not include any code points that
11025 * are in the bitmap */
11026 if (change_invlist) {
11027 SV* keep_list = _new_invlist(2);
11028 _append_range_to_invlist(keep_list, max_cp_to_set + 1, UV_MAX);
11029 _invlist_intersection(nonbitmap, keep_list, &nonbitmap);
11030 SvREFCNT_dec(keep_list);
11033 /* If have completely emptied it, remove it completely */
11034 if (invlist_len(nonbitmap) == 0) {
11035 SvREFCNT_dec(nonbitmap);
11040 /* Here, we have calculated what code points should be in the character
11041 * class. <nonbitmap> does not overlap the bitmap except possibly in the
11042 * case of DEPENDS rules.
11044 * Now we can see about various optimizations. Fold calculation (which we
11045 * did above) needs to take place before inversion. Otherwise /[^k]/i
11046 * would invert to include K, which under /i would match k, which it
11049 /* Optimize inverted simple patterns (e.g. [^a-z]). Note that we haven't
11050 * set the FOLD flag yet, so this does optimize those. It doesn't
11051 * optimize locale. Doing so perhaps could be done as long as there is
11052 * nothing like \w in it; some thought also would have to be given to the
11053 * interaction with above 0x100 chars */
11054 if ((ANYOF_FLAGS(ret) & ANYOF_INVERT)
11056 && ! unicode_alternate
11057 /* In case of /d, there are some things that should match only when in
11058 * not in the bitmap, i.e., they require UTF8 to match. These are
11059 * listed in nonbitmap, but if ANYOF_NONBITMAP_NON_UTF8 is set in this
11060 * case, they don't require UTF8, so can invert here */
11062 || ! DEPENDS_SEMANTICS
11063 || (ANYOF_FLAGS(ret) & ANYOF_NONBITMAP_NON_UTF8))
11064 && SvCUR(listsv) == initial_listsv_len)
11068 for (i = 0; i < 256; ++i) {
11069 if (ANYOF_BITMAP_TEST(ret, i)) {
11070 ANYOF_BITMAP_CLEAR(ret, i);
11073 ANYOF_BITMAP_SET(ret, i);
11078 /* The inversion means that everything above 255 is matched */
11079 ANYOF_FLAGS(ret) |= ANYOF_UNICODE_ALL;
11082 /* Here, also has things outside the bitmap that may overlap with
11083 * the bitmap. We have to sync them up, so that they get inverted
11084 * in both places. Earlier, we removed all overlaps except in the
11085 * case of /d rules, so no syncing is needed except for this case
11087 SV *remove_list = NULL;
11089 if (DEPENDS_SEMANTICS) {
11092 /* Set the bits that correspond to the ones that aren't in the
11093 * bitmap. Otherwise, when we invert, we'll miss these.
11094 * Earlier, we removed from the nonbitmap all code points
11095 * < 128, so there is no extra work here */
11096 invlist_iterinit(nonbitmap);
11097 while (invlist_iternext(nonbitmap, &start, &end)) {
11098 if (start > 255) { /* The bit map goes to 255 */
11104 for (i = start; i <= (int) end; ++i) {
11105 ANYOF_BITMAP_SET(ret, i);
11112 /* Now invert both the bitmap and the nonbitmap. Anything in the
11113 * bitmap has to also be removed from the non-bitmap, but again,
11114 * there should not be overlap unless is /d rules. */
11115 _invlist_invert(nonbitmap);
11117 for (i = 0; i < 256; ++i) {
11118 if (ANYOF_BITMAP_TEST(ret, i)) {
11119 ANYOF_BITMAP_CLEAR(ret, i);
11120 if (DEPENDS_SEMANTICS) {
11121 if (! remove_list) {
11122 remove_list = _new_invlist(2);
11124 remove_list = add_cp_to_invlist(remove_list, i);
11128 ANYOF_BITMAP_SET(ret, i);
11134 /* And do the removal */
11135 if (DEPENDS_SEMANTICS) {
11137 _invlist_subtract(nonbitmap, remove_list, &nonbitmap);
11138 SvREFCNT_dec(remove_list);
11142 /* There is no overlap for non-/d, so just delete anything
11144 SV* keep_list = _new_invlist(2);
11145 _append_range_to_invlist(keep_list, 256, UV_MAX);
11146 _invlist_intersection(nonbitmap, keep_list, &nonbitmap);
11147 SvREFCNT_dec(keep_list);
11151 stored = 256 - stored;
11153 /* Clear the invert flag since have just done it here */
11154 ANYOF_FLAGS(ret) &= ~ANYOF_INVERT;
11157 /* Folding in the bitmap is taken care of above, but not for locale (for
11158 * which we have to wait to see what folding is in effect at runtime), and
11159 * for some things not in the bitmap (only the upper latin folds in this
11160 * case, as all other single-char folding has been set above). Set
11161 * run-time fold flag for these */
11163 || (DEPENDS_SEMANTICS
11165 && ! (ANYOF_FLAGS(ret) & ANYOF_NONBITMAP_NON_UTF8))
11166 || unicode_alternate))
11168 ANYOF_FLAGS(ret) |= ANYOF_LOC_NONBITMAP_FOLD;
11171 /* A single character class can be "optimized" into an EXACTish node.
11172 * Note that since we don't currently count how many characters there are
11173 * outside the bitmap, we are XXX missing optimization possibilities for
11174 * them. This optimization can't happen unless this is a truly single
11175 * character class, which means that it can't be an inversion into a
11176 * many-character class, and there must be no possibility of there being
11177 * things outside the bitmap. 'stored' (only) for locales doesn't include
11178 * \w, etc, so have to make a special test that they aren't present
11180 * Similarly A 2-character class of the very special form like [bB] can be
11181 * optimized into an EXACTFish node, but only for non-locales, and for
11182 * characters which only have the two folds; so things like 'fF' and 'Ii'
11183 * wouldn't work because they are part of the fold of 'LATIN SMALL LIGATURE
11186 && ! unicode_alternate
11187 && SvCUR(listsv) == initial_listsv_len
11188 && ! (ANYOF_FLAGS(ret) & (ANYOF_INVERT|ANYOF_UNICODE_ALL))
11189 && (((stored == 1 && ((! (ANYOF_FLAGS(ret) & ANYOF_LOCALE))
11190 || (! ANYOF_CLASS_TEST_ANY_SET(ret)))))
11191 || (stored == 2 && ((! (ANYOF_FLAGS(ret) & ANYOF_LOCALE))
11192 && (! _HAS_NONLATIN1_FOLD_CLOSURE_ONLY_FOR_USE_BY_REGCOMP_DOT_C_AND_REGEXEC_DOT_C(value))
11193 /* If the latest code point has a fold whose
11194 * bit is set, it must be the only other one */
11195 && ((prevvalue = PL_fold_latin1[value]) != (IV)value)
11196 && ANYOF_BITMAP_TEST(ret, prevvalue)))))
11198 /* Note that the information needed to decide to do this optimization
11199 * is not currently available until the 2nd pass, and that the actually
11200 * used EXACTish node takes less space than the calculated ANYOF node,
11201 * and hence the amount of space calculated in the first pass is larger
11202 * than actually used, so this optimization doesn't gain us any space.
11203 * But an EXACT node is faster than an ANYOF node, and can be combined
11204 * with any adjacent EXACT nodes later by the optimizer for further
11205 * gains. The speed of executing an EXACTF is similar to an ANYOF
11206 * node, so the optimization advantage comes from the ability to join
11207 * it to adjacent EXACT nodes */
11209 const char * cur_parse= RExC_parse;
11211 RExC_emit = (regnode *)orig_emit;
11212 RExC_parse = (char *)orig_parse;
11216 /* A locale node with one point can be folded; all the other cases
11217 * with folding will have two points, since we calculate them above
11219 if (ANYOF_FLAGS(ret) & ANYOF_LOC_NONBITMAP_FOLD) {
11226 else { /* else 2 chars in the bit map: the folds of each other */
11228 /* Use the folded value, which for the cases where we get here,
11229 * is just the lower case of the current one (which may resolve to
11230 * itself, or to the other one */
11231 value = toLOWER_LATIN1(value);
11233 /* To join adjacent nodes, they must be the exact EXACTish type.
11234 * Try to use the most likely type, by using EXACTFA if possible,
11235 * then EXACTFU if the regex calls for it, or is required because
11236 * the character is non-ASCII. (If <value> is ASCII, its fold is
11237 * also ASCII for the cases where we get here.) */
11238 if (MORE_ASCII_RESTRICTED && isASCII(value)) {
11241 else if (AT_LEAST_UNI_SEMANTICS || !isASCII(value)) {
11244 else { /* Otherwise, more likely to be EXACTF type */
11249 ret = reg_node(pRExC_state, op);
11250 RExC_parse = (char *)cur_parse;
11251 if (UTF && ! NATIVE_IS_INVARIANT(value)) {
11252 *STRING(ret)= UTF8_EIGHT_BIT_HI((U8) value);
11253 *(STRING(ret) + 1)= UTF8_EIGHT_BIT_LO((U8) value);
11255 RExC_emit += STR_SZ(2);
11258 *STRING(ret)= (char)value;
11260 RExC_emit += STR_SZ(1);
11262 SvREFCNT_dec(listsv);
11266 /* If there is a swash and more than one element, we can't use the swash in
11267 * the optimization below. */
11268 if (swash && element_count > 1) {
11269 SvREFCNT_dec(swash);
11273 && SvCUR(listsv) == initial_listsv_len
11274 && ! unicode_alternate)
11276 ARG_SET(ret, ANYOF_NONBITMAP_EMPTY);
11277 SvREFCNT_dec(listsv);
11278 SvREFCNT_dec(unicode_alternate);
11281 /* av[0] stores the character class description in its textual form:
11282 * used later (regexec.c:Perl_regclass_swash()) to initialize the
11283 * appropriate swash, and is also useful for dumping the regnode.
11284 * av[1] if NULL, is a placeholder to later contain the swash computed
11285 * from av[0]. But if no further computation need be done, the
11286 * swash is stored there now.
11287 * av[2] stores the multicharacter foldings, used later in
11288 * regexec.c:S_reginclass().
11289 * av[3] stores the nonbitmap inversion list for use in addition or
11290 * instead of av[0]; not used if av[1] isn't NULL
11291 * av[4] is set if any component of the class is from a user-defined
11292 * property; not used if av[1] isn't NULL */
11293 AV * const av = newAV();
11296 av_store(av, 0, (SvCUR(listsv) == initial_listsv_len)
11300 av_store(av, 1, swash);
11301 SvREFCNT_dec(nonbitmap);
11304 av_store(av, 1, NULL);
11306 av_store(av, 3, nonbitmap);
11307 av_store(av, 4, newSVuv(has_user_defined_property));
11311 /* Store any computed multi-char folds only if we are allowing
11313 if (allow_full_fold) {
11314 av_store(av, 2, MUTABLE_SV(unicode_alternate));
11315 if (unicode_alternate) { /* This node is variable length */
11320 av_store(av, 2, NULL);
11322 rv = newRV_noinc(MUTABLE_SV(av));
11323 n = add_data(pRExC_state, 1, "s");
11324 RExC_rxi->data->data[n] = (void*)rv;
11332 /* reg_skipcomment()
11334 Absorbs an /x style # comments from the input stream.
11335 Returns true if there is more text remaining in the stream.
11336 Will set the REG_SEEN_RUN_ON_COMMENT flag if the comment
11337 terminates the pattern without including a newline.
11339 Note its the callers responsibility to ensure that we are
11340 actually in /x mode
11345 S_reg_skipcomment(pTHX_ RExC_state_t *pRExC_state)
11349 PERL_ARGS_ASSERT_REG_SKIPCOMMENT;
11351 while (RExC_parse < RExC_end)
11352 if (*RExC_parse++ == '\n') {
11357 /* we ran off the end of the pattern without ending
11358 the comment, so we have to add an \n when wrapping */
11359 RExC_seen |= REG_SEEN_RUN_ON_COMMENT;
11367 Advances the parse position, and optionally absorbs
11368 "whitespace" from the inputstream.
11370 Without /x "whitespace" means (?#...) style comments only,
11371 with /x this means (?#...) and # comments and whitespace proper.
11373 Returns the RExC_parse point from BEFORE the scan occurs.
11375 This is the /x friendly way of saying RExC_parse++.
11379 S_nextchar(pTHX_ RExC_state_t *pRExC_state)
11381 char* const retval = RExC_parse++;
11383 PERL_ARGS_ASSERT_NEXTCHAR;
11386 if (RExC_end - RExC_parse >= 3
11387 && *RExC_parse == '('
11388 && RExC_parse[1] == '?'
11389 && RExC_parse[2] == '#')
11391 while (*RExC_parse != ')') {
11392 if (RExC_parse == RExC_end)
11393 FAIL("Sequence (?#... not terminated");
11399 if (RExC_flags & RXf_PMf_EXTENDED) {
11400 if (isSPACE(*RExC_parse)) {
11404 else if (*RExC_parse == '#') {
11405 if ( reg_skipcomment( pRExC_state ) )
11414 - reg_node - emit a node
11416 STATIC regnode * /* Location. */
11417 S_reg_node(pTHX_ RExC_state_t *pRExC_state, U8 op)
11420 register regnode *ptr;
11421 regnode * const ret = RExC_emit;
11422 GET_RE_DEBUG_FLAGS_DECL;
11424 PERL_ARGS_ASSERT_REG_NODE;
11427 SIZE_ALIGN(RExC_size);
11431 if (RExC_emit >= RExC_emit_bound)
11432 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d, %p>=%p",
11433 op, RExC_emit, RExC_emit_bound);
11435 NODE_ALIGN_FILL(ret);
11437 FILL_ADVANCE_NODE(ptr, op);
11438 #ifdef RE_TRACK_PATTERN_OFFSETS
11439 if (RExC_offsets) { /* MJD */
11440 MJD_OFFSET_DEBUG(("%s:%d: (op %s) %s %"UVuf" (len %"UVuf") (max %"UVuf").\n",
11441 "reg_node", __LINE__,
11443 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0]
11444 ? "Overwriting end of array!\n" : "OK",
11445 (UV)(RExC_emit - RExC_emit_start),
11446 (UV)(RExC_parse - RExC_start),
11447 (UV)RExC_offsets[0]));
11448 Set_Node_Offset(RExC_emit, RExC_parse + (op == END));
11456 - reganode - emit a node with an argument
11458 STATIC regnode * /* Location. */
11459 S_reganode(pTHX_ RExC_state_t *pRExC_state, U8 op, U32 arg)
11462 register regnode *ptr;
11463 regnode * const ret = RExC_emit;
11464 GET_RE_DEBUG_FLAGS_DECL;
11466 PERL_ARGS_ASSERT_REGANODE;
11469 SIZE_ALIGN(RExC_size);
11474 assert(2==regarglen[op]+1);
11476 Anything larger than this has to allocate the extra amount.
11477 If we changed this to be:
11479 RExC_size += (1 + regarglen[op]);
11481 then it wouldn't matter. Its not clear what side effect
11482 might come from that so its not done so far.
11487 if (RExC_emit >= RExC_emit_bound)
11488 Perl_croak(aTHX_ "panic: reg_node overrun trying to emit %d, %p>=%p",
11489 op, RExC_emit, RExC_emit_bound);
11491 NODE_ALIGN_FILL(ret);
11493 FILL_ADVANCE_NODE_ARG(ptr, op, arg);
11494 #ifdef RE_TRACK_PATTERN_OFFSETS
11495 if (RExC_offsets) { /* MJD */
11496 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
11500 (UV)(RExC_emit - RExC_emit_start) > RExC_offsets[0] ?
11501 "Overwriting end of array!\n" : "OK",
11502 (UV)(RExC_emit - RExC_emit_start),
11503 (UV)(RExC_parse - RExC_start),
11504 (UV)RExC_offsets[0]));
11505 Set_Cur_Node_Offset;
11513 - reguni - emit (if appropriate) a Unicode character
11516 S_reguni(pTHX_ const RExC_state_t *pRExC_state, UV uv, char* s)
11520 PERL_ARGS_ASSERT_REGUNI;
11522 return SIZE_ONLY ? UNISKIP(uv) : (uvchr_to_utf8((U8*)s, uv) - (U8*)s);
11526 - reginsert - insert an operator in front of already-emitted operand
11528 * Means relocating the operand.
11531 S_reginsert(pTHX_ RExC_state_t *pRExC_state, U8 op, regnode *opnd, U32 depth)
11534 register regnode *src;
11535 register regnode *dst;
11536 register regnode *place;
11537 const int offset = regarglen[(U8)op];
11538 const int size = NODE_STEP_REGNODE + offset;
11539 GET_RE_DEBUG_FLAGS_DECL;
11541 PERL_ARGS_ASSERT_REGINSERT;
11542 PERL_UNUSED_ARG(depth);
11543 /* (PL_regkind[(U8)op] == CURLY ? EXTRA_STEP_2ARGS : 0); */
11544 DEBUG_PARSE_FMT("inst"," - %s",PL_reg_name[op]);
11553 if (RExC_open_parens) {
11555 /*DEBUG_PARSE_FMT("inst"," - %"IVdf, (IV)RExC_npar);*/
11556 for ( paren=0 ; paren < RExC_npar ; paren++ ) {
11557 if ( RExC_open_parens[paren] >= opnd ) {
11558 /*DEBUG_PARSE_FMT("open"," - %d",size);*/
11559 RExC_open_parens[paren] += size;
11561 /*DEBUG_PARSE_FMT("open"," - %s","ok");*/
11563 if ( RExC_close_parens[paren] >= opnd ) {
11564 /*DEBUG_PARSE_FMT("close"," - %d",size);*/
11565 RExC_close_parens[paren] += size;
11567 /*DEBUG_PARSE_FMT("close"," - %s","ok");*/
11572 while (src > opnd) {
11573 StructCopy(--src, --dst, regnode);
11574 #ifdef RE_TRACK_PATTERN_OFFSETS
11575 if (RExC_offsets) { /* MJD 20010112 */
11576 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s copy %"UVuf" -> %"UVuf" (max %"UVuf").\n",
11580 (UV)(dst - RExC_emit_start) > RExC_offsets[0]
11581 ? "Overwriting end of array!\n" : "OK",
11582 (UV)(src - RExC_emit_start),
11583 (UV)(dst - RExC_emit_start),
11584 (UV)RExC_offsets[0]));
11585 Set_Node_Offset_To_R(dst-RExC_emit_start, Node_Offset(src));
11586 Set_Node_Length_To_R(dst-RExC_emit_start, Node_Length(src));
11592 place = opnd; /* Op node, where operand used to be. */
11593 #ifdef RE_TRACK_PATTERN_OFFSETS
11594 if (RExC_offsets) { /* MJD */
11595 MJD_OFFSET_DEBUG(("%s(%d): (op %s) %s %"UVuf" <- %"UVuf" (max %"UVuf").\n",
11599 (UV)(place - RExC_emit_start) > RExC_offsets[0]
11600 ? "Overwriting end of array!\n" : "OK",
11601 (UV)(place - RExC_emit_start),
11602 (UV)(RExC_parse - RExC_start),
11603 (UV)RExC_offsets[0]));
11604 Set_Node_Offset(place, RExC_parse);
11605 Set_Node_Length(place, 1);
11608 src = NEXTOPER(place);
11609 FILL_ADVANCE_NODE(place, op);
11610 Zero(src, offset, regnode);
11614 - regtail - set the next-pointer at the end of a node chain of p to val.
11615 - SEE ALSO: regtail_study
11617 /* TODO: All three parms should be const */
11619 S_regtail(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
11622 register regnode *scan;
11623 GET_RE_DEBUG_FLAGS_DECL;
11625 PERL_ARGS_ASSERT_REGTAIL;
11627 PERL_UNUSED_ARG(depth);
11633 /* Find last node. */
11636 regnode * const temp = regnext(scan);
11638 SV * const mysv=sv_newmortal();
11639 DEBUG_PARSE_MSG((scan==p ? "tail" : ""));
11640 regprop(RExC_rx, mysv, scan);
11641 PerlIO_printf(Perl_debug_log, "~ %s (%d) %s %s\n",
11642 SvPV_nolen_const(mysv), REG_NODE_NUM(scan),
11643 (temp == NULL ? "->" : ""),
11644 (temp == NULL ? PL_reg_name[OP(val)] : "")
11652 if (reg_off_by_arg[OP(scan)]) {
11653 ARG_SET(scan, val - scan);
11656 NEXT_OFF(scan) = val - scan;
11662 - regtail_study - set the next-pointer at the end of a node chain of p to val.
11663 - Look for optimizable sequences at the same time.
11664 - currently only looks for EXACT chains.
11666 This is experimental code. The idea is to use this routine to perform
11667 in place optimizations on branches and groups as they are constructed,
11668 with the long term intention of removing optimization from study_chunk so
11669 that it is purely analytical.
11671 Currently only used when in DEBUG mode. The macro REGTAIL_STUDY() is used
11672 to control which is which.
11675 /* TODO: All four parms should be const */
11678 S_regtail_study(pTHX_ RExC_state_t *pRExC_state, regnode *p, const regnode *val,U32 depth)
11681 register regnode *scan;
11683 #ifdef EXPERIMENTAL_INPLACESCAN
11686 GET_RE_DEBUG_FLAGS_DECL;
11688 PERL_ARGS_ASSERT_REGTAIL_STUDY;
11694 /* Find last node. */
11698 regnode * const temp = regnext(scan);
11699 #ifdef EXPERIMENTAL_INPLACESCAN
11700 if (PL_regkind[OP(scan)] == EXACT) {
11701 bool has_exactf_sharp_s; /* Unexamined in this routine */
11702 if (join_exact(pRExC_state,scan,&min, &has_exactf_sharp_s, 1,val,depth+1))
11707 switch (OP(scan)) {
11713 case EXACTFU_NO_TRIE:
11715 if( exact == PSEUDO )
11717 else if ( exact != OP(scan) )
11726 SV * const mysv=sv_newmortal();
11727 DEBUG_PARSE_MSG((scan==p ? "tsdy" : ""));
11728 regprop(RExC_rx, mysv, scan);
11729 PerlIO_printf(Perl_debug_log, "~ %s (%d) -> %s\n",
11730 SvPV_nolen_const(mysv),
11731 REG_NODE_NUM(scan),
11732 PL_reg_name[exact]);
11739 SV * const mysv_val=sv_newmortal();
11740 DEBUG_PARSE_MSG("");
11741 regprop(RExC_rx, mysv_val, val);
11742 PerlIO_printf(Perl_debug_log, "~ attach to %s (%"IVdf") offset to %"IVdf"\n",
11743 SvPV_nolen_const(mysv_val),
11744 (IV)REG_NODE_NUM(val),
11748 if (reg_off_by_arg[OP(scan)]) {
11749 ARG_SET(scan, val - scan);
11752 NEXT_OFF(scan) = val - scan;
11760 - regdump - dump a regexp onto Perl_debug_log in vaguely comprehensible form
11764 S_regdump_extflags(pTHX_ const char *lead, const U32 flags)
11770 for (bit=0; bit<32; bit++) {
11771 if (flags & (1<<bit)) {
11772 if ((1<<bit) & RXf_PMf_CHARSET) { /* Output separately, below */
11775 if (!set++ && lead)
11776 PerlIO_printf(Perl_debug_log, "%s",lead);
11777 PerlIO_printf(Perl_debug_log, "%s ",PL_reg_extflags_name[bit]);
11780 if ((cs = get_regex_charset(flags)) != REGEX_DEPENDS_CHARSET) {
11781 if (!set++ && lead) {
11782 PerlIO_printf(Perl_debug_log, "%s",lead);
11785 case REGEX_UNICODE_CHARSET:
11786 PerlIO_printf(Perl_debug_log, "UNICODE");
11788 case REGEX_LOCALE_CHARSET:
11789 PerlIO_printf(Perl_debug_log, "LOCALE");
11791 case REGEX_ASCII_RESTRICTED_CHARSET:
11792 PerlIO_printf(Perl_debug_log, "ASCII-RESTRICTED");
11794 case REGEX_ASCII_MORE_RESTRICTED_CHARSET:
11795 PerlIO_printf(Perl_debug_log, "ASCII-MORE_RESTRICTED");
11798 PerlIO_printf(Perl_debug_log, "UNKNOWN CHARACTER SET");
11804 PerlIO_printf(Perl_debug_log, "\n");
11806 PerlIO_printf(Perl_debug_log, "%s[none-set]\n",lead);
11812 Perl_regdump(pTHX_ const regexp *r)
11816 SV * const sv = sv_newmortal();
11817 SV *dsv= sv_newmortal();
11818 RXi_GET_DECL(r,ri);
11819 GET_RE_DEBUG_FLAGS_DECL;
11821 PERL_ARGS_ASSERT_REGDUMP;
11823 (void)dumpuntil(r, ri->program, ri->program + 1, NULL, NULL, sv, 0, 0);
11825 /* Header fields of interest. */
11826 if (r->anchored_substr) {
11827 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->anchored_substr),
11828 RE_SV_DUMPLEN(r->anchored_substr), 30);
11829 PerlIO_printf(Perl_debug_log,
11830 "anchored %s%s at %"IVdf" ",
11831 s, RE_SV_TAIL(r->anchored_substr),
11832 (IV)r->anchored_offset);
11833 } else if (r->anchored_utf8) {
11834 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->anchored_utf8),
11835 RE_SV_DUMPLEN(r->anchored_utf8), 30);
11836 PerlIO_printf(Perl_debug_log,
11837 "anchored utf8 %s%s at %"IVdf" ",
11838 s, RE_SV_TAIL(r->anchored_utf8),
11839 (IV)r->anchored_offset);
11841 if (r->float_substr) {
11842 RE_PV_QUOTED_DECL(s, 0, dsv, SvPVX_const(r->float_substr),
11843 RE_SV_DUMPLEN(r->float_substr), 30);
11844 PerlIO_printf(Perl_debug_log,
11845 "floating %s%s at %"IVdf"..%"UVuf" ",
11846 s, RE_SV_TAIL(r->float_substr),
11847 (IV)r->float_min_offset, (UV)r->float_max_offset);
11848 } else if (r->float_utf8) {
11849 RE_PV_QUOTED_DECL(s, 1, dsv, SvPVX_const(r->float_utf8),
11850 RE_SV_DUMPLEN(r->float_utf8), 30);
11851 PerlIO_printf(Perl_debug_log,
11852 "floating utf8 %s%s at %"IVdf"..%"UVuf" ",
11853 s, RE_SV_TAIL(r->float_utf8),
11854 (IV)r->float_min_offset, (UV)r->float_max_offset);
11856 if (r->check_substr || r->check_utf8)
11857 PerlIO_printf(Perl_debug_log,
11859 (r->check_substr == r->float_substr
11860 && r->check_utf8 == r->float_utf8
11861 ? "(checking floating" : "(checking anchored"));
11862 if (r->extflags & RXf_NOSCAN)
11863 PerlIO_printf(Perl_debug_log, " noscan");
11864 if (r->extflags & RXf_CHECK_ALL)
11865 PerlIO_printf(Perl_debug_log, " isall");
11866 if (r->check_substr || r->check_utf8)
11867 PerlIO_printf(Perl_debug_log, ") ");
11869 if (ri->regstclass) {
11870 regprop(r, sv, ri->regstclass);
11871 PerlIO_printf(Perl_debug_log, "stclass %s ", SvPVX_const(sv));
11873 if (r->extflags & RXf_ANCH) {
11874 PerlIO_printf(Perl_debug_log, "anchored");
11875 if (r->extflags & RXf_ANCH_BOL)
11876 PerlIO_printf(Perl_debug_log, "(BOL)");
11877 if (r->extflags & RXf_ANCH_MBOL)
11878 PerlIO_printf(Perl_debug_log, "(MBOL)");
11879 if (r->extflags & RXf_ANCH_SBOL)
11880 PerlIO_printf(Perl_debug_log, "(SBOL)");
11881 if (r->extflags & RXf_ANCH_GPOS)
11882 PerlIO_printf(Perl_debug_log, "(GPOS)");
11883 PerlIO_putc(Perl_debug_log, ' ');
11885 if (r->extflags & RXf_GPOS_SEEN)
11886 PerlIO_printf(Perl_debug_log, "GPOS:%"UVuf" ", (UV)r->gofs);
11887 if (r->intflags & PREGf_SKIP)
11888 PerlIO_printf(Perl_debug_log, "plus ");
11889 if (r->intflags & PREGf_IMPLICIT)
11890 PerlIO_printf(Perl_debug_log, "implicit ");
11891 PerlIO_printf(Perl_debug_log, "minlen %"IVdf" ", (IV)r->minlen);
11892 if (r->extflags & RXf_EVAL_SEEN)
11893 PerlIO_printf(Perl_debug_log, "with eval ");
11894 PerlIO_printf(Perl_debug_log, "\n");
11895 DEBUG_FLAGS_r(regdump_extflags("r->extflags: ",r->extflags));
11897 PERL_ARGS_ASSERT_REGDUMP;
11898 PERL_UNUSED_CONTEXT;
11899 PERL_UNUSED_ARG(r);
11900 #endif /* DEBUGGING */
11904 - regprop - printable representation of opcode
11906 #define EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags) \
11909 Perl_sv_catpvf(aTHX_ sv,"%s][%s",PL_colors[1],PL_colors[0]); \
11910 if (flags & ANYOF_INVERT) \
11911 /*make sure the invert info is in each */ \
11912 sv_catpvs(sv, "^"); \
11918 Perl_regprop(pTHX_ const regexp *prog, SV *sv, const regnode *o)
11923 RXi_GET_DECL(prog,progi);
11924 GET_RE_DEBUG_FLAGS_DECL;
11926 PERL_ARGS_ASSERT_REGPROP;
11930 if (OP(o) > REGNODE_MAX) /* regnode.type is unsigned */
11931 /* It would be nice to FAIL() here, but this may be called from
11932 regexec.c, and it would be hard to supply pRExC_state. */
11933 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(o), (int)REGNODE_MAX);
11934 sv_catpv(sv, PL_reg_name[OP(o)]); /* Take off const! */
11936 k = PL_regkind[OP(o)];
11939 sv_catpvs(sv, " ");
11940 /* Using is_utf8_string() (via PERL_PV_UNI_DETECT)
11941 * is a crude hack but it may be the best for now since
11942 * we have no flag "this EXACTish node was UTF-8"
11944 pv_pretty(sv, STRING(o), STR_LEN(o), 60, PL_colors[0], PL_colors[1],
11945 PERL_PV_ESCAPE_UNI_DETECT |
11946 PERL_PV_ESCAPE_NONASCII |
11947 PERL_PV_PRETTY_ELLIPSES |
11948 PERL_PV_PRETTY_LTGT |
11949 PERL_PV_PRETTY_NOCLEAR
11951 } else if (k == TRIE) {
11952 /* print the details of the trie in dumpuntil instead, as
11953 * progi->data isn't available here */
11954 const char op = OP(o);
11955 const U32 n = ARG(o);
11956 const reg_ac_data * const ac = IS_TRIE_AC(op) ?
11957 (reg_ac_data *)progi->data->data[n] :
11959 const reg_trie_data * const trie
11960 = (reg_trie_data*)progi->data->data[!IS_TRIE_AC(op) ? n : ac->trie];
11962 Perl_sv_catpvf(aTHX_ sv, "-%s",PL_reg_name[o->flags]);
11963 DEBUG_TRIE_COMPILE_r(
11964 Perl_sv_catpvf(aTHX_ sv,
11965 "<S:%"UVuf"/%"IVdf" W:%"UVuf" L:%"UVuf"/%"UVuf" C:%"UVuf"/%"UVuf">",
11966 (UV)trie->startstate,
11967 (IV)trie->statecount-1, /* -1 because of the unused 0 element */
11968 (UV)trie->wordcount,
11971 (UV)TRIE_CHARCOUNT(trie),
11972 (UV)trie->uniquecharcount
11975 if ( IS_ANYOF_TRIE(op) || trie->bitmap ) {
11977 int rangestart = -1;
11978 U8* bitmap = IS_ANYOF_TRIE(op) ? (U8*)ANYOF_BITMAP(o) : (U8*)TRIE_BITMAP(trie);
11979 sv_catpvs(sv, "[");
11980 for (i = 0; i <= 256; i++) {
11981 if (i < 256 && BITMAP_TEST(bitmap,i)) {
11982 if (rangestart == -1)
11984 } else if (rangestart != -1) {
11985 if (i <= rangestart + 3)
11986 for (; rangestart < i; rangestart++)
11987 put_byte(sv, rangestart);
11989 put_byte(sv, rangestart);
11990 sv_catpvs(sv, "-");
11991 put_byte(sv, i - 1);
11996 sv_catpvs(sv, "]");
11999 } else if (k == CURLY) {
12000 if (OP(o) == CURLYM || OP(o) == CURLYN || OP(o) == CURLYX)
12001 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* Parenth number */
12002 Perl_sv_catpvf(aTHX_ sv, " {%d,%d}", ARG1(o), ARG2(o));
12004 else if (k == WHILEM && o->flags) /* Ordinal/of */
12005 Perl_sv_catpvf(aTHX_ sv, "[%d/%d]", o->flags & 0xf, o->flags>>4);
12006 else if (k == REF || k == OPEN || k == CLOSE || k == GROUPP || OP(o)==ACCEPT) {
12007 Perl_sv_catpvf(aTHX_ sv, "%d", (int)ARG(o)); /* Parenth number */
12008 if ( RXp_PAREN_NAMES(prog) ) {
12009 if ( k != REF || (OP(o) < NREF)) {
12010 AV *list= MUTABLE_AV(progi->data->data[progi->name_list_idx]);
12011 SV **name= av_fetch(list, ARG(o), 0 );
12013 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
12016 AV *list= MUTABLE_AV(progi->data->data[ progi->name_list_idx ]);
12017 SV *sv_dat= MUTABLE_SV(progi->data->data[ ARG( o ) ]);
12018 I32 *nums=(I32*)SvPVX(sv_dat);
12019 SV **name= av_fetch(list, nums[0], 0 );
12022 for ( n=0; n<SvIVX(sv_dat); n++ ) {
12023 Perl_sv_catpvf(aTHX_ sv, "%s%"IVdf,
12024 (n ? "," : ""), (IV)nums[n]);
12026 Perl_sv_catpvf(aTHX_ sv, " '%"SVf"'", SVfARG(*name));
12030 } else if (k == GOSUB)
12031 Perl_sv_catpvf(aTHX_ sv, "%d[%+d]", (int)ARG(o),(int)ARG2L(o)); /* Paren and offset */
12032 else if (k == VERB) {
12034 Perl_sv_catpvf(aTHX_ sv, ":%"SVf,
12035 SVfARG((MUTABLE_SV(progi->data->data[ ARG( o ) ]))));
12036 } else if (k == LOGICAL)
12037 Perl_sv_catpvf(aTHX_ sv, "[%d]", o->flags); /* 2: embedded, otherwise 1 */
12038 else if (k == ANYOF) {
12039 int i, rangestart = -1;
12040 const U8 flags = ANYOF_FLAGS(o);
12043 /* Should be synchronized with * ANYOF_ #xdefines in regcomp.h */
12044 static const char * const anyofs[] = {
12077 if (flags & ANYOF_LOCALE)
12078 sv_catpvs(sv, "{loc}");
12079 if (flags & ANYOF_LOC_NONBITMAP_FOLD)
12080 sv_catpvs(sv, "{i}");
12081 Perl_sv_catpvf(aTHX_ sv, "[%s", PL_colors[0]);
12082 if (flags & ANYOF_INVERT)
12083 sv_catpvs(sv, "^");
12085 /* output what the standard cp 0-255 bitmap matches */
12086 for (i = 0; i <= 256; i++) {
12087 if (i < 256 && ANYOF_BITMAP_TEST(o,i)) {
12088 if (rangestart == -1)
12090 } else if (rangestart != -1) {
12091 if (i <= rangestart + 3)
12092 for (; rangestart < i; rangestart++)
12093 put_byte(sv, rangestart);
12095 put_byte(sv, rangestart);
12096 sv_catpvs(sv, "-");
12097 put_byte(sv, i - 1);
12104 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
12105 /* output any special charclass tests (used entirely under use locale) */
12106 if (ANYOF_CLASS_TEST_ANY_SET(o))
12107 for (i = 0; i < (int)(sizeof(anyofs)/sizeof(char*)); i++)
12108 if (ANYOF_CLASS_TEST(o,i)) {
12109 sv_catpv(sv, anyofs[i]);
12113 EMIT_ANYOF_TEST_SEPARATOR(do_sep,sv,flags);
12115 if (flags & ANYOF_NON_UTF8_LATIN1_ALL) {
12116 sv_catpvs(sv, "{non-utf8-latin1-all}");
12119 /* output information about the unicode matching */
12120 if (flags & ANYOF_UNICODE_ALL)
12121 sv_catpvs(sv, "{unicode_all}");
12122 else if (ANYOF_NONBITMAP(o))
12123 sv_catpvs(sv, "{unicode}");
12124 if (flags & ANYOF_NONBITMAP_NON_UTF8)
12125 sv_catpvs(sv, "{outside bitmap}");
12127 if (ANYOF_NONBITMAP(o)) {
12128 SV *lv; /* Set if there is something outside the bit map */
12129 SV * const sw = regclass_swash(prog, o, FALSE, &lv, 0);
12130 bool byte_output = FALSE; /* If something in the bitmap has been
12133 if (lv && lv != &PL_sv_undef) {
12135 U8 s[UTF8_MAXBYTES_CASE+1];
12137 for (i = 0; i <= 256; i++) { /* Look at chars in bitmap */
12138 uvchr_to_utf8(s, i);
12141 && ! ANYOF_BITMAP_TEST(o, i) /* Don't duplicate
12145 && swash_fetch(sw, s, TRUE))
12147 if (rangestart == -1)
12149 } else if (rangestart != -1) {
12150 byte_output = TRUE;
12151 if (i <= rangestart + 3)
12152 for (; rangestart < i; rangestart++) {
12153 put_byte(sv, rangestart);
12156 put_byte(sv, rangestart);
12157 sv_catpvs(sv, "-");
12166 char *s = savesvpv(lv);
12167 char * const origs = s;
12169 while (*s && *s != '\n')
12173 const char * const t = ++s;
12176 sv_catpvs(sv, " ");
12182 /* Truncate very long output */
12183 if (s - origs > 256) {
12184 Perl_sv_catpvf(aTHX_ sv,
12186 (int) (s - origs - 1),
12192 else if (*s == '\t') {
12211 Perl_sv_catpvf(aTHX_ sv, "%s]", PL_colors[1]);
12213 else if (k == BRANCHJ && (OP(o) == UNLESSM || OP(o) == IFMATCH))
12214 Perl_sv_catpvf(aTHX_ sv, "[%d]", -(o->flags));
12216 PERL_UNUSED_CONTEXT;
12217 PERL_UNUSED_ARG(sv);
12218 PERL_UNUSED_ARG(o);
12219 PERL_UNUSED_ARG(prog);
12220 #endif /* DEBUGGING */
12224 Perl_re_intuit_string(pTHX_ REGEXP * const r)
12225 { /* Assume that RE_INTUIT is set */
12227 struct regexp *const prog = (struct regexp *)SvANY(r);
12228 GET_RE_DEBUG_FLAGS_DECL;
12230 PERL_ARGS_ASSERT_RE_INTUIT_STRING;
12231 PERL_UNUSED_CONTEXT;
12235 const char * const s = SvPV_nolen_const(prog->check_substr
12236 ? prog->check_substr : prog->check_utf8);
12238 if (!PL_colorset) reginitcolors();
12239 PerlIO_printf(Perl_debug_log,
12240 "%sUsing REx %ssubstr:%s \"%s%.60s%s%s\"\n",
12242 prog->check_substr ? "" : "utf8 ",
12243 PL_colors[5],PL_colors[0],
12246 (strlen(s) > 60 ? "..." : ""));
12249 return prog->check_substr ? prog->check_substr : prog->check_utf8;
12255 handles refcounting and freeing the perl core regexp structure. When
12256 it is necessary to actually free the structure the first thing it
12257 does is call the 'free' method of the regexp_engine associated to
12258 the regexp, allowing the handling of the void *pprivate; member
12259 first. (This routine is not overridable by extensions, which is why
12260 the extensions free is called first.)
12262 See regdupe and regdupe_internal if you change anything here.
12264 #ifndef PERL_IN_XSUB_RE
12266 Perl_pregfree(pTHX_ REGEXP *r)
12272 Perl_pregfree2(pTHX_ REGEXP *rx)
12275 struct regexp *const r = (struct regexp *)SvANY(rx);
12276 GET_RE_DEBUG_FLAGS_DECL;
12278 PERL_ARGS_ASSERT_PREGFREE2;
12280 if (r->mother_re) {
12281 ReREFCNT_dec(r->mother_re);
12283 CALLREGFREE_PVT(rx); /* free the private data */
12284 SvREFCNT_dec(RXp_PAREN_NAMES(r));
12287 SvREFCNT_dec(r->anchored_substr);
12288 SvREFCNT_dec(r->anchored_utf8);
12289 SvREFCNT_dec(r->float_substr);
12290 SvREFCNT_dec(r->float_utf8);
12291 Safefree(r->substrs);
12293 RX_MATCH_COPY_FREE(rx);
12294 #ifdef PERL_OLD_COPY_ON_WRITE
12295 SvREFCNT_dec(r->saved_copy);
12302 This is a hacky workaround to the structural issue of match results
12303 being stored in the regexp structure which is in turn stored in
12304 PL_curpm/PL_reg_curpm. The problem is that due to qr// the pattern
12305 could be PL_curpm in multiple contexts, and could require multiple
12306 result sets being associated with the pattern simultaneously, such
12307 as when doing a recursive match with (??{$qr})
12309 The solution is to make a lightweight copy of the regexp structure
12310 when a qr// is returned from the code executed by (??{$qr}) this
12311 lightweight copy doesn't actually own any of its data except for
12312 the starp/end and the actual regexp structure itself.
12318 Perl_reg_temp_copy (pTHX_ REGEXP *ret_x, REGEXP *rx)
12320 struct regexp *ret;
12321 struct regexp *const r = (struct regexp *)SvANY(rx);
12322 register const I32 npar = r->nparens+1;
12324 PERL_ARGS_ASSERT_REG_TEMP_COPY;
12327 ret_x = (REGEXP*) newSV_type(SVt_REGEXP);
12328 ret = (struct regexp *)SvANY(ret_x);
12330 (void)ReREFCNT_inc(rx);
12331 /* We can take advantage of the existing "copied buffer" mechanism in SVs
12332 by pointing directly at the buffer, but flagging that the allocated
12333 space in the copy is zero. As we've just done a struct copy, it's now
12334 a case of zero-ing that, rather than copying the current length. */
12335 SvPV_set(ret_x, RX_WRAPPED(rx));
12336 SvFLAGS(ret_x) |= SvFLAGS(rx) & (SVf_POK|SVp_POK|SVf_UTF8);
12337 memcpy(&(ret->xpv_cur), &(r->xpv_cur),
12338 sizeof(regexp) - STRUCT_OFFSET(regexp, xpv_cur));
12339 SvLEN_set(ret_x, 0);
12340 SvSTASH_set(ret_x, NULL);
12341 SvMAGIC_set(ret_x, NULL);
12342 Newx(ret->offs, npar, regexp_paren_pair);
12343 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
12345 Newx(ret->substrs, 1, struct reg_substr_data);
12346 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
12348 SvREFCNT_inc_void(ret->anchored_substr);
12349 SvREFCNT_inc_void(ret->anchored_utf8);
12350 SvREFCNT_inc_void(ret->float_substr);
12351 SvREFCNT_inc_void(ret->float_utf8);
12353 /* check_substr and check_utf8, if non-NULL, point to either their
12354 anchored or float namesakes, and don't hold a second reference. */
12356 RX_MATCH_COPIED_off(ret_x);
12357 #ifdef PERL_OLD_COPY_ON_WRITE
12358 ret->saved_copy = NULL;
12360 ret->mother_re = rx;
12366 /* regfree_internal()
12368 Free the private data in a regexp. This is overloadable by
12369 extensions. Perl takes care of the regexp structure in pregfree(),
12370 this covers the *pprivate pointer which technically perl doesn't
12371 know about, however of course we have to handle the
12372 regexp_internal structure when no extension is in use.
12374 Note this is called before freeing anything in the regexp
12379 Perl_regfree_internal(pTHX_ REGEXP * const rx)
12382 struct regexp *const r = (struct regexp *)SvANY(rx);
12383 RXi_GET_DECL(r,ri);
12384 GET_RE_DEBUG_FLAGS_DECL;
12386 PERL_ARGS_ASSERT_REGFREE_INTERNAL;
12392 SV *dsv= sv_newmortal();
12393 RE_PV_QUOTED_DECL(s, RX_UTF8(rx),
12394 dsv, RX_PRECOMP(rx), RX_PRELEN(rx), 60);
12395 PerlIO_printf(Perl_debug_log,"%sFreeing REx:%s %s\n",
12396 PL_colors[4],PL_colors[5],s);
12399 #ifdef RE_TRACK_PATTERN_OFFSETS
12401 Safefree(ri->u.offsets); /* 20010421 MJD */
12404 int n = ri->data->count;
12405 PAD* new_comppad = NULL;
12410 /* If you add a ->what type here, update the comment in regcomp.h */
12411 switch (ri->data->what[n]) {
12416 SvREFCNT_dec(MUTABLE_SV(ri->data->data[n]));
12419 Safefree(ri->data->data[n]);
12422 new_comppad = MUTABLE_AV(ri->data->data[n]);
12425 if (new_comppad == NULL)
12426 Perl_croak(aTHX_ "panic: pregfree comppad");
12427 PAD_SAVE_LOCAL(old_comppad,
12428 /* Watch out for global destruction's random ordering. */
12429 (SvTYPE(new_comppad) == SVt_PVAV) ? new_comppad : NULL
12432 refcnt = OpREFCNT_dec((OP_4tree*)ri->data->data[n]);
12435 op_free((OP_4tree*)ri->data->data[n]);
12437 PAD_RESTORE_LOCAL(old_comppad);
12438 SvREFCNT_dec(MUTABLE_SV(new_comppad));
12439 new_comppad = NULL;
12444 { /* Aho Corasick add-on structure for a trie node.
12445 Used in stclass optimization only */
12447 reg_ac_data *aho=(reg_ac_data*)ri->data->data[n];
12449 refcount = --aho->refcount;
12452 PerlMemShared_free(aho->states);
12453 PerlMemShared_free(aho->fail);
12454 /* do this last!!!! */
12455 PerlMemShared_free(ri->data->data[n]);
12456 PerlMemShared_free(ri->regstclass);
12462 /* trie structure. */
12464 reg_trie_data *trie=(reg_trie_data*)ri->data->data[n];
12466 refcount = --trie->refcount;
12469 PerlMemShared_free(trie->charmap);
12470 PerlMemShared_free(trie->states);
12471 PerlMemShared_free(trie->trans);
12473 PerlMemShared_free(trie->bitmap);
12475 PerlMemShared_free(trie->jump);
12476 PerlMemShared_free(trie->wordinfo);
12477 /* do this last!!!! */
12478 PerlMemShared_free(ri->data->data[n]);
12483 Perl_croak(aTHX_ "panic: regfree data code '%c'", ri->data->what[n]);
12486 Safefree(ri->data->what);
12487 Safefree(ri->data);
12493 #define av_dup_inc(s,t) MUTABLE_AV(sv_dup_inc((const SV *)s,t))
12494 #define hv_dup_inc(s,t) MUTABLE_HV(sv_dup_inc((const SV *)s,t))
12495 #define SAVEPVN(p,n) ((p) ? savepvn(p,n) : NULL)
12498 re_dup - duplicate a regexp.
12500 This routine is expected to clone a given regexp structure. It is only
12501 compiled under USE_ITHREADS.
12503 After all of the core data stored in struct regexp is duplicated
12504 the regexp_engine.dupe method is used to copy any private data
12505 stored in the *pprivate pointer. This allows extensions to handle
12506 any duplication it needs to do.
12508 See pregfree() and regfree_internal() if you change anything here.
12510 #if defined(USE_ITHREADS)
12511 #ifndef PERL_IN_XSUB_RE
12513 Perl_re_dup_guts(pTHX_ const REGEXP *sstr, REGEXP *dstr, CLONE_PARAMS *param)
12517 const struct regexp *r = (const struct regexp *)SvANY(sstr);
12518 struct regexp *ret = (struct regexp *)SvANY(dstr);
12520 PERL_ARGS_ASSERT_RE_DUP_GUTS;
12522 npar = r->nparens+1;
12523 Newx(ret->offs, npar, regexp_paren_pair);
12524 Copy(r->offs, ret->offs, npar, regexp_paren_pair);
12526 /* no need to copy these */
12527 Newx(ret->swap, npar, regexp_paren_pair);
12530 if (ret->substrs) {
12531 /* Do it this way to avoid reading from *r after the StructCopy().
12532 That way, if any of the sv_dup_inc()s dislodge *r from the L1
12533 cache, it doesn't matter. */
12534 const bool anchored = r->check_substr
12535 ? r->check_substr == r->anchored_substr
12536 : r->check_utf8 == r->anchored_utf8;
12537 Newx(ret->substrs, 1, struct reg_substr_data);
12538 StructCopy(r->substrs, ret->substrs, struct reg_substr_data);
12540 ret->anchored_substr = sv_dup_inc(ret->anchored_substr, param);
12541 ret->anchored_utf8 = sv_dup_inc(ret->anchored_utf8, param);
12542 ret->float_substr = sv_dup_inc(ret->float_substr, param);
12543 ret->float_utf8 = sv_dup_inc(ret->float_utf8, param);
12545 /* check_substr and check_utf8, if non-NULL, point to either their
12546 anchored or float namesakes, and don't hold a second reference. */
12548 if (ret->check_substr) {
12550 assert(r->check_utf8 == r->anchored_utf8);
12551 ret->check_substr = ret->anchored_substr;
12552 ret->check_utf8 = ret->anchored_utf8;
12554 assert(r->check_substr == r->float_substr);
12555 assert(r->check_utf8 == r->float_utf8);
12556 ret->check_substr = ret->float_substr;
12557 ret->check_utf8 = ret->float_utf8;
12559 } else if (ret->check_utf8) {
12561 ret->check_utf8 = ret->anchored_utf8;
12563 ret->check_utf8 = ret->float_utf8;
12568 RXp_PAREN_NAMES(ret) = hv_dup_inc(RXp_PAREN_NAMES(ret), param);
12571 RXi_SET(ret,CALLREGDUPE_PVT(dstr,param));
12573 if (RX_MATCH_COPIED(dstr))
12574 ret->subbeg = SAVEPVN(ret->subbeg, ret->sublen);
12576 ret->subbeg = NULL;
12577 #ifdef PERL_OLD_COPY_ON_WRITE
12578 ret->saved_copy = NULL;
12581 if (ret->mother_re) {
12582 if (SvPVX_const(dstr) == SvPVX_const(ret->mother_re)) {
12583 /* Our storage points directly to our mother regexp, but that's
12584 1: a buffer in a different thread
12585 2: something we no longer hold a reference on
12586 so we need to copy it locally. */
12587 /* Note we need to use SvCUR(), rather than
12588 SvLEN(), on our mother_re, because it, in
12589 turn, may well be pointing to its own mother_re. */
12590 SvPV_set(dstr, SAVEPVN(SvPVX_const(ret->mother_re),
12591 SvCUR(ret->mother_re)+1));
12592 SvLEN_set(dstr, SvCUR(ret->mother_re)+1);
12594 ret->mother_re = NULL;
12598 #endif /* PERL_IN_XSUB_RE */
12603 This is the internal complement to regdupe() which is used to copy
12604 the structure pointed to by the *pprivate pointer in the regexp.
12605 This is the core version of the extension overridable cloning hook.
12606 The regexp structure being duplicated will be copied by perl prior
12607 to this and will be provided as the regexp *r argument, however
12608 with the /old/ structures pprivate pointer value. Thus this routine
12609 may override any copying normally done by perl.
12611 It returns a pointer to the new regexp_internal structure.
12615 Perl_regdupe_internal(pTHX_ REGEXP * const rx, CLONE_PARAMS *param)
12618 struct regexp *const r = (struct regexp *)SvANY(rx);
12619 regexp_internal *reti;
12621 RXi_GET_DECL(r,ri);
12623 PERL_ARGS_ASSERT_REGDUPE_INTERNAL;
12627 Newxc(reti, sizeof(regexp_internal) + len*sizeof(regnode), char, regexp_internal);
12628 Copy(ri->program, reti->program, len+1, regnode);
12631 reti->regstclass = NULL;
12634 struct reg_data *d;
12635 const int count = ri->data->count;
12638 Newxc(d, sizeof(struct reg_data) + count*sizeof(void *),
12639 char, struct reg_data);
12640 Newx(d->what, count, U8);
12643 for (i = 0; i < count; i++) {
12644 d->what[i] = ri->data->what[i];
12645 switch (d->what[i]) {
12646 /* legal options are one of: sSfpontTua
12647 see also regcomp.h and pregfree() */
12648 case 'a': /* actually an AV, but the dup function is identical. */
12651 case 'p': /* actually an AV, but the dup function is identical. */
12652 case 'u': /* actually an HV, but the dup function is identical. */
12653 d->data[i] = sv_dup_inc((const SV *)ri->data->data[i], param);
12656 /* This is cheating. */
12657 Newx(d->data[i], 1, struct regnode_charclass_class);
12658 StructCopy(ri->data->data[i], d->data[i],
12659 struct regnode_charclass_class);
12660 reti->regstclass = (regnode*)d->data[i];
12663 /* Compiled op trees are readonly and in shared memory,
12664 and can thus be shared without duplication. */
12666 d->data[i] = (void*)OpREFCNT_inc((OP*)ri->data->data[i]);
12670 /* Trie stclasses are readonly and can thus be shared
12671 * without duplication. We free the stclass in pregfree
12672 * when the corresponding reg_ac_data struct is freed.
12674 reti->regstclass= ri->regstclass;
12678 ((reg_trie_data*)ri->data->data[i])->refcount++;
12682 d->data[i] = ri->data->data[i];
12685 Perl_croak(aTHX_ "panic: re_dup unknown data code '%c'", ri->data->what[i]);
12694 reti->name_list_idx = ri->name_list_idx;
12696 #ifdef RE_TRACK_PATTERN_OFFSETS
12697 if (ri->u.offsets) {
12698 Newx(reti->u.offsets, 2*len+1, U32);
12699 Copy(ri->u.offsets, reti->u.offsets, 2*len+1, U32);
12702 SetProgLen(reti,len);
12705 return (void*)reti;
12708 #endif /* USE_ITHREADS */
12710 #ifndef PERL_IN_XSUB_RE
12713 - regnext - dig the "next" pointer out of a node
12716 Perl_regnext(pTHX_ register regnode *p)
12719 register I32 offset;
12724 if (OP(p) > REGNODE_MAX) { /* regnode.type is unsigned */
12725 Perl_croak(aTHX_ "Corrupted regexp opcode %d > %d", (int)OP(p), (int)REGNODE_MAX);
12728 offset = (reg_off_by_arg[OP(p)] ? ARG(p) : NEXT_OFF(p));
12737 S_re_croak2(pTHX_ const char* pat1,const char* pat2,...)
12740 STRLEN l1 = strlen(pat1);
12741 STRLEN l2 = strlen(pat2);
12744 const char *message;
12746 PERL_ARGS_ASSERT_RE_CROAK2;
12752 Copy(pat1, buf, l1 , char);
12753 Copy(pat2, buf + l1, l2 , char);
12754 buf[l1 + l2] = '\n';
12755 buf[l1 + l2 + 1] = '\0';
12757 /* ANSI variant takes additional second argument */
12758 va_start(args, pat2);
12762 msv = vmess(buf, &args);
12764 message = SvPV_const(msv,l1);
12767 Copy(message, buf, l1 , char);
12768 buf[l1-1] = '\0'; /* Overwrite \n */
12769 Perl_croak(aTHX_ "%s", buf);
12772 /* XXX Here's a total kludge. But we need to re-enter for swash routines. */
12774 #ifndef PERL_IN_XSUB_RE
12776 Perl_save_re_context(pTHX)
12780 struct re_save_state *state;
12782 SAVEVPTR(PL_curcop);
12783 SSGROW(SAVESTACK_ALLOC_FOR_RE_SAVE_STATE + 1);
12785 state = (struct re_save_state *)(PL_savestack + PL_savestack_ix);
12786 PL_savestack_ix += SAVESTACK_ALLOC_FOR_RE_SAVE_STATE;
12787 SSPUSHUV(SAVEt_RE_STATE);
12789 Copy(&PL_reg_state, state, 1, struct re_save_state);
12791 PL_reg_start_tmp = 0;
12792 PL_reg_start_tmpl = 0;
12793 PL_reg_oldsaved = NULL;
12794 PL_reg_oldsavedlen = 0;
12795 PL_reg_maxiter = 0;
12796 PL_reg_leftiter = 0;
12797 PL_reg_poscache = NULL;
12798 PL_reg_poscache_size = 0;
12799 #ifdef PERL_OLD_COPY_ON_WRITE
12803 /* Save $1..$n (#18107: UTF-8 s/(\w+)/uc($1)/e); AMS 20021106. */
12805 const REGEXP * const rx = PM_GETRE(PL_curpm);
12808 for (i = 1; i <= RX_NPARENS(rx); i++) {
12809 char digits[TYPE_CHARS(long)];
12810 const STRLEN len = my_snprintf(digits, sizeof(digits), "%lu", (long)i);
12811 GV *const *const gvp
12812 = (GV**)hv_fetch(PL_defstash, digits, len, 0);
12815 GV * const gv = *gvp;
12816 if (SvTYPE(gv) == SVt_PVGV && GvSV(gv))
12826 clear_re(pTHX_ void *r)
12829 ReREFCNT_dec((REGEXP *)r);
12835 S_put_byte(pTHX_ SV *sv, int c)
12837 PERL_ARGS_ASSERT_PUT_BYTE;
12839 /* Our definition of isPRINT() ignores locales, so only bytes that are
12840 not part of UTF-8 are considered printable. I assume that the same
12841 holds for UTF-EBCDIC.
12842 Also, code point 255 is not printable in either (it's E0 in EBCDIC,
12843 which Wikipedia says:
12845 EO, or Eight Ones, is an 8-bit EBCDIC character code represented as all
12846 ones (binary 1111 1111, hexadecimal FF). It is similar, but not
12847 identical, to the ASCII delete (DEL) or rubout control character.
12848 ) So the old condition can be simplified to !isPRINT(c) */
12851 Perl_sv_catpvf(aTHX_ sv, "\\x%02x", c);
12854 Perl_sv_catpvf(aTHX_ sv, "\\x{%x}", c);
12858 const char string = c;
12859 if (c == '-' || c == ']' || c == '\\' || c == '^')
12860 sv_catpvs(sv, "\\");
12861 sv_catpvn(sv, &string, 1);
12866 #define CLEAR_OPTSTART \
12867 if (optstart) STMT_START { \
12868 DEBUG_OPTIMISE_r(PerlIO_printf(Perl_debug_log, " (%"IVdf" nodes)\n", (IV)(node - optstart))); \
12872 #define DUMPUNTIL(b,e) CLEAR_OPTSTART; node=dumpuntil(r,start,(b),(e),last,sv,indent+1,depth+1);
12874 STATIC const regnode *
12875 S_dumpuntil(pTHX_ const regexp *r, const regnode *start, const regnode *node,
12876 const regnode *last, const regnode *plast,
12877 SV* sv, I32 indent, U32 depth)
12880 register U8 op = PSEUDO; /* Arbitrary non-END op. */
12881 register const regnode *next;
12882 const regnode *optstart= NULL;
12884 RXi_GET_DECL(r,ri);
12885 GET_RE_DEBUG_FLAGS_DECL;
12887 PERL_ARGS_ASSERT_DUMPUNTIL;
12889 #ifdef DEBUG_DUMPUNTIL
12890 PerlIO_printf(Perl_debug_log, "--- %d : %d - %d - %d\n",indent,node-start,
12891 last ? last-start : 0,plast ? plast-start : 0);
12894 if (plast && plast < last)
12897 while (PL_regkind[op] != END && (!last || node < last)) {
12898 /* While that wasn't END last time... */
12901 if (op == CLOSE || op == WHILEM)
12903 next = regnext((regnode *)node);
12906 if (OP(node) == OPTIMIZED) {
12907 if (!optstart && RE_DEBUG_FLAG(RE_DEBUG_COMPILE_OPTIMISE))
12914 regprop(r, sv, node);
12915 PerlIO_printf(Perl_debug_log, "%4"IVdf":%*s%s", (IV)(node - start),
12916 (int)(2*indent + 1), "", SvPVX_const(sv));
12918 if (OP(node) != OPTIMIZED) {
12919 if (next == NULL) /* Next ptr. */
12920 PerlIO_printf(Perl_debug_log, " (0)");
12921 else if (PL_regkind[(U8)op] == BRANCH && PL_regkind[OP(next)] != BRANCH )
12922 PerlIO_printf(Perl_debug_log, " (FAIL)");
12924 PerlIO_printf(Perl_debug_log, " (%"IVdf")", (IV)(next - start));
12925 (void)PerlIO_putc(Perl_debug_log, '\n');
12929 if (PL_regkind[(U8)op] == BRANCHJ) {
12932 register const regnode *nnode = (OP(next) == LONGJMP
12933 ? regnext((regnode *)next)
12935 if (last && nnode > last)
12937 DUMPUNTIL(NEXTOPER(NEXTOPER(node)), nnode);
12940 else if (PL_regkind[(U8)op] == BRANCH) {
12942 DUMPUNTIL(NEXTOPER(node), next);
12944 else if ( PL_regkind[(U8)op] == TRIE ) {
12945 const regnode *this_trie = node;
12946 const char op = OP(node);
12947 const U32 n = ARG(node);
12948 const reg_ac_data * const ac = op>=AHOCORASICK ?
12949 (reg_ac_data *)ri->data->data[n] :
12951 const reg_trie_data * const trie =
12952 (reg_trie_data*)ri->data->data[op<AHOCORASICK ? n : ac->trie];
12954 AV *const trie_words = MUTABLE_AV(ri->data->data[n + TRIE_WORDS_OFFSET]);
12956 const regnode *nextbranch= NULL;
12959 for (word_idx= 0; word_idx < (I32)trie->wordcount; word_idx++) {
12960 SV ** const elem_ptr = av_fetch(trie_words,word_idx,0);
12962 PerlIO_printf(Perl_debug_log, "%*s%s ",
12963 (int)(2*(indent+3)), "",
12964 elem_ptr ? pv_pretty(sv, SvPV_nolen_const(*elem_ptr), SvCUR(*elem_ptr), 60,
12965 PL_colors[0], PL_colors[1],
12966 (SvUTF8(*elem_ptr) ? PERL_PV_ESCAPE_UNI : 0) |
12967 PERL_PV_PRETTY_ELLIPSES |
12968 PERL_PV_PRETTY_LTGT
12973 U16 dist= trie->jump[word_idx+1];
12974 PerlIO_printf(Perl_debug_log, "(%"UVuf")\n",
12975 (UV)((dist ? this_trie + dist : next) - start));
12978 nextbranch= this_trie + trie->jump[0];
12979 DUMPUNTIL(this_trie + dist, nextbranch);
12981 if (nextbranch && PL_regkind[OP(nextbranch)]==BRANCH)
12982 nextbranch= regnext((regnode *)nextbranch);
12984 PerlIO_printf(Perl_debug_log, "\n");
12987 if (last && next > last)
12992 else if ( op == CURLY ) { /* "next" might be very big: optimizer */
12993 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS,
12994 NEXTOPER(node) + EXTRA_STEP_2ARGS + 1);
12996 else if (PL_regkind[(U8)op] == CURLY && op != CURLYX) {
12998 DUMPUNTIL(NEXTOPER(node) + EXTRA_STEP_2ARGS, next);
13000 else if ( op == PLUS || op == STAR) {
13001 DUMPUNTIL(NEXTOPER(node), NEXTOPER(node) + 1);
13003 else if (PL_regkind[(U8)op] == ANYOF) {
13004 /* arglen 1 + class block */
13005 node += 1 + ((ANYOF_FLAGS(node) & ANYOF_CLASS)
13006 ? ANYOF_CLASS_SKIP : ANYOF_SKIP);
13007 node = NEXTOPER(node);
13009 else if (PL_regkind[(U8)op] == EXACT) {
13010 /* Literal string, where present. */
13011 node += NODE_SZ_STR(node) - 1;
13012 node = NEXTOPER(node);
13015 node = NEXTOPER(node);
13016 node += regarglen[(U8)op];
13018 if (op == CURLYX || op == OPEN)
13022 #ifdef DEBUG_DUMPUNTIL
13023 PerlIO_printf(Perl_debug_log, "--- %d\n", (int)indent);
13028 #endif /* DEBUGGING */
13032 * c-indentation-style: bsd
13033 * c-basic-offset: 4
13034 * indent-tabs-mode: t
13037 * ex: set ts=8 sts=4 sw=4 noet:
https://perl5.git.perl.org / perl5.git / blob