/[pcre]/code/trunk/doc/pcre.txt
ViewVC logotype

Contents of /code/trunk/doc/pcre.txt

Parent Directory Parent Directory | Revision Log Revision Log


Revision 461 - (show annotations)
Mon Oct 5 10:59:35 2009 UTC (5 years, 7 months ago) by ph10
File MIME type: text/plain
File size: 335891 byte(s)
Error occurred while calculating annotation data.
Tidy up, remove trailing spaces, etc. for 8.00-RC1.
1 -----------------------------------------------------------------------------
2 This file contains a concatenation of the PCRE man pages, converted to plain
3 text format for ease of searching with a text editor, or for use on systems
4 that do not have a man page processor. The small individual files that give
5 synopses of each function in the library have not been included. Neither has
6 the pcredemo program. There are separate text files for the pcregrep and
7 pcretest commands.
8 -----------------------------------------------------------------------------
9
10
11 PCRE(3) PCRE(3)
12
13
14 NAME
15 PCRE - Perl-compatible regular expressions
16
17
18 INTRODUCTION
19
20 The PCRE library is a set of functions that implement regular expres-
21 sion pattern matching using the same syntax and semantics as Perl, with
22 just a few differences. Some features that appeared in Python and PCRE
23 before they appeared in Perl are also available using the Python syn-
24 tax, there is some support for one or two .NET and Oniguruma syntax
25 items, and there is an option for requesting some minor changes that
26 give better JavaScript compatibility.
27
28 The current implementation of PCRE corresponds approximately with Perl
29 5.10, including support for UTF-8 encoded strings and Unicode general
30 category properties. However, UTF-8 and Unicode support has to be
31 explicitly enabled; it is not the default. The Unicode tables corre-
32 spond to Unicode release 5.1.
33
34 In addition to the Perl-compatible matching function, PCRE contains an
35 alternative function that matches the same compiled patterns in a dif-
36 ferent way. In certain circumstances, the alternative function has some
37 advantages. For a discussion of the two matching algorithms, see the
38 pcrematching page.
39
40 PCRE is written in C and released as a C library. A number of people
41 have written wrappers and interfaces of various kinds. In particular,
42 Google Inc. have provided a comprehensive C++ wrapper. This is now
43 included as part of the PCRE distribution. The pcrecpp page has details
44 of this interface. Other people's contributions can be found in the
45 Contrib directory at the primary FTP site, which is:
46
47 ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre
48
49 Details of exactly which Perl regular expression features are and are
50 not supported by PCRE are given in separate documents. See the pcrepat-
51 tern and pcrecompat pages. There is a syntax summary in the pcresyntax
52 page.
53
54 Some features of PCRE can be included, excluded, or changed when the
55 library is built. The pcre_config() function makes it possible for a
56 client to discover which features are available. The features them-
57 selves are described in the pcrebuild page. Documentation about build-
58 ing PCRE for various operating systems can be found in the README and
59 NON-UNIX-USE files in the source distribution.
60
61 The library contains a number of undocumented internal functions and
62 data tables that are used by more than one of the exported external
63 functions, but which are not intended for use by external callers.
64 Their names all begin with "_pcre_", which hopefully will not provoke
65 any name clashes. In some environments, it is possible to control which
66 external symbols are exported when a shared library is built, and in
67 these cases the undocumented symbols are not exported.
68
69
70 USER DOCUMENTATION
71
72 The user documentation for PCRE comprises a number of different sec-
73 tions. In the "man" format, each of these is a separate "man page". In
74 the HTML format, each is a separate page, linked from the index page.
75 In the plain text format, all the sections, except the pcredemo sec-
76 tion, are concatenated, for ease of searching. The sections are as fol-
77 lows:
78
79 pcre this document
80 pcre-config show PCRE installation configuration information
81 pcreapi details of PCRE's native C API
82 pcrebuild options for building PCRE
83 pcrecallout details of the callout feature
84 pcrecompat discussion of Perl compatibility
85 pcrecpp details of the C++ wrapper
86 pcredemo a demonstration C program that uses PCRE
87 pcregrep description of the pcregrep command
88 pcrematching discussion of the two matching algorithms
89 pcrepartial details of the partial matching facility
90 pcrepattern syntax and semantics of supported
91 regular expressions
92 pcreperform discussion of performance issues
93 pcreposix the POSIX-compatible C API
94 pcreprecompile details of saving and re-using precompiled patterns
95 pcresample discussion of the pcredemo program
96 pcrestack discussion of stack usage
97 pcresyntax quick syntax reference
98 pcretest description of the pcretest testing command
99
100 In addition, in the "man" and HTML formats, there is a short page for
101 each C library function, listing its arguments and results.
102
103
104 LIMITATIONS
105
106 There are some size limitations in PCRE but it is hoped that they will
107 never in practice be relevant.
108
109 The maximum length of a compiled pattern is 65539 (sic) bytes if PCRE
110 is compiled with the default internal linkage size of 2. If you want to
111 process regular expressions that are truly enormous, you can compile
112 PCRE with an internal linkage size of 3 or 4 (see the README file in
113 the source distribution and the pcrebuild documentation for details).
114 In these cases the limit is substantially larger. However, the speed
115 of execution is slower.
116
117 All values in repeating quantifiers must be less than 65536.
118
119 There is no limit to the number of parenthesized subpatterns, but there
120 can be no more than 65535 capturing subpatterns.
121
122 The maximum length of name for a named subpattern is 32 characters, and
123 the maximum number of named subpatterns is 10000.
124
125 The maximum length of a subject string is the largest positive number
126 that an integer variable can hold. However, when using the traditional
127 matching function, PCRE uses recursion to handle subpatterns and indef-
128 inite repetition. This means that the available stack space may limit
129 the size of a subject string that can be processed by certain patterns.
130 For a discussion of stack issues, see the pcrestack documentation.
131
132
133 UTF-8 AND UNICODE PROPERTY SUPPORT
134
135 From release 3.3, PCRE has had some support for character strings
136 encoded in the UTF-8 format. For release 4.0 this was greatly extended
137 to cover most common requirements, and in release 5.0 additional sup-
138 port for Unicode general category properties was added.
139
140 In order process UTF-8 strings, you must build PCRE to include UTF-8
141 support in the code, and, in addition, you must call pcre_compile()
142 with the PCRE_UTF8 option flag, or the pattern must start with the
143 sequence (*UTF8). When either of these is the case, both the pattern
144 and any subject strings that are matched against it are treated as
145 UTF-8 strings instead of strings of 1-byte characters.
146
147 If you compile PCRE with UTF-8 support, but do not use it at run time,
148 the library will be a bit bigger, but the additional run time overhead
149 is limited to testing the PCRE_UTF8 flag occasionally, so should not be
150 very big.
151
152 If PCRE is built with Unicode character property support (which implies
153 UTF-8 support), the escape sequences \p{..}, \P{..}, and \X are sup-
154 ported. The available properties that can be tested are limited to the
155 general category properties such as Lu for an upper case letter or Nd
156 for a decimal number, the Unicode script names such as Arabic or Han,
157 and the derived properties Any and L&. A full list is given in the
158 pcrepattern documentation. Only the short names for properties are sup-
159 ported. For example, \p{L} matches a letter. Its Perl synonym, \p{Let-
160 ter}, is not supported. Furthermore, in Perl, many properties may
161 optionally be prefixed by "Is", for compatibility with Perl 5.6. PCRE
162 does not support this.
163
164 Validity of UTF-8 strings
165
166 When you set the PCRE_UTF8 flag, the strings passed as patterns and
167 subjects are (by default) checked for validity on entry to the relevant
168 functions. From release 7.3 of PCRE, the check is according the rules
169 of RFC 3629, which are themselves derived from the Unicode specifica-
170 tion. Earlier releases of PCRE followed the rules of RFC 2279, which
171 allows the full range of 31-bit values (0 to 0x7FFFFFFF). The current
172 check allows only values in the range U+0 to U+10FFFF, excluding U+D800
173 to U+DFFF.
174
175 The excluded code points are the "Low Surrogate Area" of Unicode, of
176 which the Unicode Standard says this: "The Low Surrogate Area does not
177 contain any character assignments, consequently no character code
178 charts or namelists are provided for this area. Surrogates are reserved
179 for use with UTF-16 and then must be used in pairs." The code points
180 that are encoded by UTF-16 pairs are available as independent code
181 points in the UTF-8 encoding. (In other words, the whole surrogate
182 thing is a fudge for UTF-16 which unfortunately messes up UTF-8.)
183
184 If an invalid UTF-8 string is passed to PCRE, an error return
185 (PCRE_ERROR_BADUTF8) is given. In some situations, you may already know
186 that your strings are valid, and therefore want to skip these checks in
187 order to improve performance. If you set the PCRE_NO_UTF8_CHECK flag at
188 compile time or at run time, PCRE assumes that the pattern or subject
189 it is given (respectively) contains only valid UTF-8 codes. In this
190 case, it does not diagnose an invalid UTF-8 string.
191
192 If you pass an invalid UTF-8 string when PCRE_NO_UTF8_CHECK is set,
193 what happens depends on why the string is invalid. If the string con-
194 forms to the "old" definition of UTF-8 (RFC 2279), it is processed as a
195 string of characters in the range 0 to 0x7FFFFFFF. In other words,
196 apart from the initial validity test, PCRE (when in UTF-8 mode) handles
197 strings according to the more liberal rules of RFC 2279. However, if
198 the string does not even conform to RFC 2279, the result is undefined.
199 Your program may crash.
200
201 If you want to process strings of values in the full range 0 to
202 0x7FFFFFFF, encoded in a UTF-8-like manner as per the old RFC, you can
203 set PCRE_NO_UTF8_CHECK to bypass the more restrictive test. However, in
204 this situation, you will have to apply your own validity check.
205
206 General comments about UTF-8 mode
207
208 1. An unbraced hexadecimal escape sequence (such as \xb3) matches a
209 two-byte UTF-8 character if the value is greater than 127.
210
211 2. Octal numbers up to \777 are recognized, and match two-byte UTF-8
212 characters for values greater than \177.
213
214 3. Repeat quantifiers apply to complete UTF-8 characters, not to indi-
215 vidual bytes, for example: \x{100}{3}.
216
217 4. The dot metacharacter matches one UTF-8 character instead of a sin-
218 gle byte.
219
220 5. The escape sequence \C can be used to match a single byte in UTF-8
221 mode, but its use can lead to some strange effects. This facility is
222 not available in the alternative matching function, pcre_dfa_exec().
223
224 6. The character escapes \b, \B, \d, \D, \s, \S, \w, and \W correctly
225 test characters of any code value, but the characters that PCRE recog-
226 nizes as digits, spaces, or word characters remain the same set as
227 before, all with values less than 256. This remains true even when PCRE
228 includes Unicode property support, because to do otherwise would slow
229 down PCRE in many common cases. If you really want to test for a wider
230 sense of, say, "digit", you must use Unicode property tests such as
231 \p{Nd}. Note that this also applies to \b, because it is defined in
232 terms of \w and \W.
233
234 7. Similarly, characters that match the POSIX named character classes
235 are all low-valued characters.
236
237 8. However, the Perl 5.10 horizontal and vertical whitespace matching
238 escapes (\h, \H, \v, and \V) do match all the appropriate Unicode char-
239 acters.
240
241 9. Case-insensitive matching applies only to characters whose values
242 are less than 128, unless PCRE is built with Unicode property support.
243 Even when Unicode property support is available, PCRE still uses its
244 own character tables when checking the case of low-valued characters,
245 so as not to degrade performance. The Unicode property information is
246 used only for characters with higher values. Even when Unicode property
247 support is available, PCRE supports case-insensitive matching only when
248 there is a one-to-one mapping between a letter's cases. There are a
249 small number of many-to-one mappings in Unicode; these are not sup-
250 ported by PCRE.
251
252
253 AUTHOR
254
255 Philip Hazel
256 University Computing Service
257 Cambridge CB2 3QH, England.
258
259 Putting an actual email address here seems to have been a spam magnet,
260 so I've taken it away. If you want to email me, use my two initials,
261 followed by the two digits 10, at the domain cam.ac.uk.
262
263
264 REVISION
265
266 Last updated: 28 September 2009
267 Copyright (c) 1997-2009 University of Cambridge.
268 ------------------------------------------------------------------------------
269
270
271 PCREBUILD(3) PCREBUILD(3)
272
273
274 NAME
275 PCRE - Perl-compatible regular expressions
276
277
278 PCRE BUILD-TIME OPTIONS
279
280 This document describes the optional features of PCRE that can be
281 selected when the library is compiled. It assumes use of the configure
282 script, where the optional features are selected or deselected by pro-
283 viding options to configure before running the make command. However,
284 the same options can be selected in both Unix-like and non-Unix-like
285 environments using the GUI facility of cmake-gui if you are using CMake
286 instead of configure to build PCRE.
287
288 There is a lot more information about building PCRE in non-Unix-like
289 environments in the file called NON_UNIX_USE, which is part of the PCRE
290 distribution. You should consult this file as well as the README file
291 if you are building in a non-Unix-like environment.
292
293 The complete list of options for configure (which includes the standard
294 ones such as the selection of the installation directory) can be
295 obtained by running
296
297 ./configure --help
298
299 The following sections include descriptions of options whose names
300 begin with --enable or --disable. These settings specify changes to the
301 defaults for the configure command. Because of the way that configure
302 works, --enable and --disable always come in pairs, so the complemen-
303 tary option always exists as well, but as it specifies the default, it
304 is not described.
305
306
307 C++ SUPPORT
308
309 By default, the configure script will search for a C++ compiler and C++
310 header files. If it finds them, it automatically builds the C++ wrapper
311 library for PCRE. You can disable this by adding
312
313 --disable-cpp
314
315 to the configure command.
316
317
318 UTF-8 SUPPORT
319
320 To build PCRE with support for UTF-8 Unicode character strings, add
321
322 --enable-utf8
323
324 to the configure command. Of itself, this does not make PCRE treat
325 strings as UTF-8. As well as compiling PCRE with this option, you also
326 have have to set the PCRE_UTF8 option when you call the pcre_compile()
327 or pcre_compile2() functions.
328
329 If you set --enable-utf8 when compiling in an EBCDIC environment, PCRE
330 expects its input to be either ASCII or UTF-8 (depending on the runtime
331 option). It is not possible to support both EBCDIC and UTF-8 codes in
332 the same version of the library. Consequently, --enable-utf8 and
333 --enable-ebcdic are mutually exclusive.
334
335
336 UNICODE CHARACTER PROPERTY SUPPORT
337
338 UTF-8 support allows PCRE to process character values greater than 255
339 in the strings that it handles. On its own, however, it does not pro-
340 vide any facilities for accessing the properties of such characters. If
341 you want to be able to use the pattern escapes \P, \p, and \X, which
342 refer to Unicode character properties, you must add
343
344 --enable-unicode-properties
345
346 to the configure command. This implies UTF-8 support, even if you have
347 not explicitly requested it.
348
349 Including Unicode property support adds around 30K of tables to the
350 PCRE library. Only the general category properties such as Lu and Nd
351 are supported. Details are given in the pcrepattern documentation.
352
353
354 CODE VALUE OF NEWLINE
355
356 By default, PCRE interprets the linefeed (LF) character as indicating
357 the end of a line. This is the normal newline character on Unix-like
358 systems. You can compile PCRE to use carriage return (CR) instead, by
359 adding
360
361 --enable-newline-is-cr
362
363 to the configure command. There is also a --enable-newline-is-lf
364 option, which explicitly specifies linefeed as the newline character.
365
366 Alternatively, you can specify that line endings are to be indicated by
367 the two character sequence CRLF. If you want this, add
368
369 --enable-newline-is-crlf
370
371 to the configure command. There is a fourth option, specified by
372
373 --enable-newline-is-anycrlf
374
375 which causes PCRE to recognize any of the three sequences CR, LF, or
376 CRLF as indicating a line ending. Finally, a fifth option, specified by
377
378 --enable-newline-is-any
379
380 causes PCRE to recognize any Unicode newline sequence.
381
382 Whatever line ending convention is selected when PCRE is built can be
383 overridden when the library functions are called. At build time it is
384 conventional to use the standard for your operating system.
385
386
387 WHAT \R MATCHES
388
389 By default, the sequence \R in a pattern matches any Unicode newline
390 sequence, whatever has been selected as the line ending sequence. If
391 you specify
392
393 --enable-bsr-anycrlf
394
395 the default is changed so that \R matches only CR, LF, or CRLF. What-
396 ever is selected when PCRE is built can be overridden when the library
397 functions are called.
398
399
400 BUILDING SHARED AND STATIC LIBRARIES
401
402 The PCRE building process uses libtool to build both shared and static
403 Unix libraries by default. You can suppress one of these by adding one
404 of
405
406 --disable-shared
407 --disable-static
408
409 to the configure command, as required.
410
411
412 POSIX MALLOC USAGE
413
414 When PCRE is called through the POSIX interface (see the pcreposix doc-
415 umentation), additional working storage is required for holding the
416 pointers to capturing substrings, because PCRE requires three integers
417 per substring, whereas the POSIX interface provides only two. If the
418 number of expected substrings is small, the wrapper function uses space
419 on the stack, because this is faster than using malloc() for each call.
420 The default threshold above which the stack is no longer used is 10; it
421 can be changed by adding a setting such as
422
423 --with-posix-malloc-threshold=20
424
425 to the configure command.
426
427
428 HANDLING VERY LARGE PATTERNS
429
430 Within a compiled pattern, offset values are used to point from one
431 part to another (for example, from an opening parenthesis to an alter-
432 nation metacharacter). By default, two-byte values are used for these
433 offsets, leading to a maximum size for a compiled pattern of around
434 64K. This is sufficient to handle all but the most gigantic patterns.
435 Nevertheless, some people do want to process truyl enormous patterns,
436 so it is possible to compile PCRE to use three-byte or four-byte off-
437 sets by adding a setting such as
438
439 --with-link-size=3
440
441 to the configure command. The value given must be 2, 3, or 4. Using
442 longer offsets slows down the operation of PCRE because it has to load
443 additional bytes when handling them.
444
445
446 AVOIDING EXCESSIVE STACK USAGE
447
448 When matching with the pcre_exec() function, PCRE implements backtrack-
449 ing by making recursive calls to an internal function called match().
450 In environments where the size of the stack is limited, this can se-
451 verely limit PCRE's operation. (The Unix environment does not usually
452 suffer from this problem, but it may sometimes be necessary to increase
453 the maximum stack size. There is a discussion in the pcrestack docu-
454 mentation.) An alternative approach to recursion that uses memory from
455 the heap to remember data, instead of using recursive function calls,
456 has been implemented to work round the problem of limited stack size.
457 If you want to build a version of PCRE that works this way, add
458
459 --disable-stack-for-recursion
460
461 to the configure command. With this configuration, PCRE will use the
462 pcre_stack_malloc and pcre_stack_free variables to call memory manage-
463 ment functions. By default these point to malloc() and free(), but you
464 can replace the pointers so that your own functions are used instead.
465
466 Separate functions are provided rather than using pcre_malloc and
467 pcre_free because the usage is very predictable: the block sizes
468 requested are always the same, and the blocks are always freed in
469 reverse order. A calling program might be able to implement optimized
470 functions that perform better than malloc() and free(). PCRE runs
471 noticeably more slowly when built in this way. This option affects only
472 the pcre_exec() function; it is not relevant for pcre_dfa_exec().
473
474
475 LIMITING PCRE RESOURCE USAGE
476
477 Internally, PCRE has a function called match(), which it calls repeat-
478 edly (sometimes recursively) when matching a pattern with the
479 pcre_exec() function. By controlling the maximum number of times this
480 function may be called during a single matching operation, a limit can
481 be placed on the resources used by a single call to pcre_exec(). The
482 limit can be changed at run time, as described in the pcreapi documen-
483 tation. The default is 10 million, but this can be changed by adding a
484 setting such as
485
486 --with-match-limit=500000
487
488 to the configure command. This setting has no effect on the
489 pcre_dfa_exec() matching function.
490
491 In some environments it is desirable to limit the depth of recursive
492 calls of match() more strictly than the total number of calls, in order
493 to restrict the maximum amount of stack (or heap, if --disable-stack-
494 for-recursion is specified) that is used. A second limit controls this;
495 it defaults to the value that is set for --with-match-limit, which
496 imposes no additional constraints. However, you can set a lower limit
497 by adding, for example,
498
499 --with-match-limit-recursion=10000
500
501 to the configure command. This value can also be overridden at run
502 time.
503
504
505 CREATING CHARACTER TABLES AT BUILD TIME
506
507 PCRE uses fixed tables for processing characters whose code values are
508 less than 256. By default, PCRE is built with a set of tables that are
509 distributed in the file pcre_chartables.c.dist. These tables are for
510 ASCII codes only. If you add
511
512 --enable-rebuild-chartables
513
514 to the configure command, the distributed tables are no longer used.
515 Instead, a program called dftables is compiled and run. This outputs
516 the source for new set of tables, created in the default locale of your
517 C runtime system. (This method of replacing the tables does not work if
518 you are cross compiling, because dftables is run on the local host. If
519 you need to create alternative tables when cross compiling, you will
520 have to do so "by hand".)
521
522
523 USING EBCDIC CODE
524
525 PCRE assumes by default that it will run in an environment where the
526 character code is ASCII (or Unicode, which is a superset of ASCII).
527 This is the case for most computer operating systems. PCRE can, how-
528 ever, be compiled to run in an EBCDIC environment by adding
529
530 --enable-ebcdic
531
532 to the configure command. This setting implies --enable-rebuild-charta-
533 bles. You should only use it if you know that you are in an EBCDIC
534 environment (for example, an IBM mainframe operating system). The
535 --enable-ebcdic option is incompatible with --enable-utf8.
536
537
538 PCREGREP OPTIONS FOR COMPRESSED FILE SUPPORT
539
540 By default, pcregrep reads all files as plain text. You can build it so
541 that it recognizes files whose names end in .gz or .bz2, and reads them
542 with libz or libbz2, respectively, by adding one or both of
543
544 --enable-pcregrep-libz
545 --enable-pcregrep-libbz2
546
547 to the configure command. These options naturally require that the rel-
548 evant libraries are installed on your system. Configuration will fail
549 if they are not.
550
551
552 PCRETEST OPTION FOR LIBREADLINE SUPPORT
553
554 If you add
555
556 --enable-pcretest-libreadline
557
558 to the configure command, pcretest is linked with the libreadline
559 library, and when its input is from a terminal, it reads it using the
560 readline() function. This provides line-editing and history facilities.
561 Note that libreadline is GPL-licensed, so if you distribute a binary of
562 pcretest linked in this way, there may be licensing issues.
563
564 Setting this option causes the -lreadline option to be added to the
565 pcretest build. In many operating environments with a sytem-installed
566 libreadline this is sufficient. However, in some environments (e.g. if
567 an unmodified distribution version of readline is in use), some extra
568 configuration may be necessary. The INSTALL file for libreadline says
569 this:
570
571 "Readline uses the termcap functions, but does not link with the
572 termcap or curses library itself, allowing applications which link
573 with readline the to choose an appropriate library."
574
575 If your environment has not been set up so that an appropriate library
576 is automatically included, you may need to add something like
577
578 LIBS="-ncurses"
579
580 immediately before the configure command.
581
582
583 SEE ALSO
584
585 pcreapi(3), pcre_config(3).
586
587
588 AUTHOR
589
590 Philip Hazel
591 University Computing Service
592 Cambridge CB2 3QH, England.
593
594
595 REVISION
596
597 Last updated: 29 September 2009
598 Copyright (c) 1997-2009 University of Cambridge.
599 ------------------------------------------------------------------------------
600
601
602 PCREMATCHING(3) PCREMATCHING(3)
603
604
605 NAME
606 PCRE - Perl-compatible regular expressions
607
608
609 PCRE MATCHING ALGORITHMS
610
611 This document describes the two different algorithms that are available
612 in PCRE for matching a compiled regular expression against a given sub-
613 ject string. The "standard" algorithm is the one provided by the
614 pcre_exec() function. This works in the same was as Perl's matching
615 function, and provides a Perl-compatible matching operation.
616
617 An alternative algorithm is provided by the pcre_dfa_exec() function;
618 this operates in a different way, and is not Perl-compatible. It has
619 advantages and disadvantages compared with the standard algorithm, and
620 these are described below.
621
622 When there is only one possible way in which a given subject string can
623 match a pattern, the two algorithms give the same answer. A difference
624 arises, however, when there are multiple possibilities. For example, if
625 the pattern
626
627 ^<.*>
628
629 is matched against the string
630
631 <something> <something else> <something further>
632
633 there are three possible answers. The standard algorithm finds only one
634 of them, whereas the alternative algorithm finds all three.
635
636
637 REGULAR EXPRESSIONS AS TREES
638
639 The set of strings that are matched by a regular expression can be rep-
640 resented as a tree structure. An unlimited repetition in the pattern
641 makes the tree of infinite size, but it is still a tree. Matching the
642 pattern to a given subject string (from a given starting point) can be
643 thought of as a search of the tree. There are two ways to search a
644 tree: depth-first and breadth-first, and these correspond to the two
645 matching algorithms provided by PCRE.
646
647
648 THE STANDARD MATCHING ALGORITHM
649
650 In the terminology of Jeffrey Friedl's book "Mastering Regular Expres-
651 sions", the standard algorithm is an "NFA algorithm". It conducts a
652 depth-first search of the pattern tree. That is, it proceeds along a
653 single path through the tree, checking that the subject matches what is
654 required. When there is a mismatch, the algorithm tries any alterna-
655 tives at the current point, and if they all fail, it backs up to the
656 previous branch point in the tree, and tries the next alternative
657 branch at that level. This often involves backing up (moving to the
658 left) in the subject string as well. The order in which repetition
659 branches are tried is controlled by the greedy or ungreedy nature of
660 the quantifier.
661
662 If a leaf node is reached, a matching string has been found, and at
663 that point the algorithm stops. Thus, if there is more than one possi-
664 ble match, this algorithm returns the first one that it finds. Whether
665 this is the shortest, the longest, or some intermediate length depends
666 on the way the greedy and ungreedy repetition quantifiers are specified
667 in the pattern.
668
669 Because it ends up with a single path through the tree, it is rela-
670 tively straightforward for this algorithm to keep track of the sub-
671 strings that are matched by portions of the pattern in parentheses.
672 This provides support for capturing parentheses and back references.
673
674
675 THE ALTERNATIVE MATCHING ALGORITHM
676
677 This algorithm conducts a breadth-first search of the tree. Starting
678 from the first matching point in the subject, it scans the subject
679 string from left to right, once, character by character, and as it does
680 this, it remembers all the paths through the tree that represent valid
681 matches. In Friedl's terminology, this is a kind of "DFA algorithm",
682 though it is not implemented as a traditional finite state machine (it
683 keeps multiple states active simultaneously).
684
685 Although the general principle of this matching algorithm is that it
686 scans the subject string only once, without backtracking, there is one
687 exception: when a lookaround assertion is encountered, the characters
688 following or preceding the current point have to be independently
689 inspected.
690
691 The scan continues until either the end of the subject is reached, or
692 there are no more unterminated paths. At this point, terminated paths
693 represent the different matching possibilities (if there are none, the
694 match has failed). Thus, if there is more than one possible match,
695 this algorithm finds all of them, and in particular, it finds the long-
696 est. There is an option to stop the algorithm after the first match
697 (which is necessarily the shortest) is found.
698
699 Note that all the matches that are found start at the same point in the
700 subject. If the pattern
701
702 cat(er(pillar)?)
703
704 is matched against the string "the caterpillar catchment", the result
705 will be the three strings "cat", "cater", and "caterpillar" that start
706 at the fourth character of the subject. The algorithm does not automat-
707 ically move on to find matches that start at later positions.
708
709 There are a number of features of PCRE regular expressions that are not
710 supported by the alternative matching algorithm. They are as follows:
711
712 1. Because the algorithm finds all possible matches, the greedy or
713 ungreedy nature of repetition quantifiers is not relevant. Greedy and
714 ungreedy quantifiers are treated in exactly the same way. However, pos-
715 sessive quantifiers can make a difference when what follows could also
716 match what is quantified, for example in a pattern like this:
717
718 ^a++\w!
719
720 This pattern matches "aaab!" but not "aaa!", which would be matched by
721 a non-possessive quantifier. Similarly, if an atomic group is present,
722 it is matched as if it were a standalone pattern at the current point,
723 and the longest match is then "locked in" for the rest of the overall
724 pattern.
725
726 2. When dealing with multiple paths through the tree simultaneously, it
727 is not straightforward to keep track of captured substrings for the
728 different matching possibilities, and PCRE's implementation of this
729 algorithm does not attempt to do this. This means that no captured sub-
730 strings are available.
731
732 3. Because no substrings are captured, back references within the pat-
733 tern are not supported, and cause errors if encountered.
734
735 4. For the same reason, conditional expressions that use a backrefer-
736 ence as the condition or test for a specific group recursion are not
737 supported.
738
739 5. Because many paths through the tree may be active, the \K escape
740 sequence, which resets the start of the match when encountered (but may
741 be on some paths and not on others), is not supported. It causes an
742 error if encountered.
743
744 6. Callouts are supported, but the value of the capture_top field is
745 always 1, and the value of the capture_last field is always -1.
746
747 7. The \C escape sequence, which (in the standard algorithm) matches a
748 single byte, even in UTF-8 mode, is not supported because the alterna-
749 tive algorithm moves through the subject string one character at a
750 time, for all active paths through the tree.
751
752 8. Except for (*FAIL), the backtracking control verbs such as (*PRUNE)
753 are not supported. (*FAIL) is supported, and behaves like a failing
754 negative assertion.
755
756
757 ADVANTAGES OF THE ALTERNATIVE ALGORITHM
758
759 Using the alternative matching algorithm provides the following advan-
760 tages:
761
762 1. All possible matches (at a single point in the subject) are automat-
763 ically found, and in particular, the longest match is found. To find
764 more than one match using the standard algorithm, you have to do kludgy
765 things with callouts.
766
767 2. Because the alternative algorithm scans the subject string just
768 once, and never needs to backtrack, it is possible to pass very long
769 subject strings to the matching function in several pieces, checking
770 for partial matching each time. The pcrepartial documentation gives
771 details of partial matching.
772
773
774 DISADVANTAGES OF THE ALTERNATIVE ALGORITHM
775
776 The alternative algorithm suffers from a number of disadvantages:
777
778 1. It is substantially slower than the standard algorithm. This is
779 partly because it has to search for all possible matches, but is also
780 because it is less susceptible to optimization.
781
782 2. Capturing parentheses and back references are not supported.
783
784 3. Although atomic groups are supported, their use does not provide the
785 performance advantage that it does for the standard algorithm.
786
787
788 AUTHOR
789
790 Philip Hazel
791 University Computing Service
792 Cambridge CB2 3QH, England.
793
794
795 REVISION
796
797 Last updated: 29 September 2009
798 Copyright (c) 1997-2009 University of Cambridge.
799 ------------------------------------------------------------------------------
800
801
802 PCREAPI(3) PCREAPI(3)
803
804
805 NAME
806 PCRE - Perl-compatible regular expressions
807
808
809 PCRE NATIVE API
810
811 #include <pcre.h>
812
813 pcre *pcre_compile(const char *pattern, int options,
814 const char **errptr, int *erroffset,
815 const unsigned char *tableptr);
816
817 pcre *pcre_compile2(const char *pattern, int options,
818 int *errorcodeptr,
819 const char **errptr, int *erroffset,
820 const unsigned char *tableptr);
821
822 pcre_extra *pcre_study(const pcre *code, int options,
823 const char **errptr);
824
825 int pcre_exec(const pcre *code, const pcre_extra *extra,
826 const char *subject, int length, int startoffset,
827 int options, int *ovector, int ovecsize);
828
829 int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
830 const char *subject, int length, int startoffset,
831 int options, int *ovector, int ovecsize,
832 int *workspace, int wscount);
833
834 int pcre_copy_named_substring(const pcre *code,
835 const char *subject, int *ovector,
836 int stringcount, const char *stringname,
837 char *buffer, int buffersize);
838
839 int pcre_copy_substring(const char *subject, int *ovector,
840 int stringcount, int stringnumber, char *buffer,
841 int buffersize);
842
843 int pcre_get_named_substring(const pcre *code,
844 const char *subject, int *ovector,
845 int stringcount, const char *stringname,
846 const char **stringptr);
847
848 int pcre_get_stringnumber(const pcre *code,
849 const char *name);
850
851 int pcre_get_stringtable_entries(const pcre *code,
852 const char *name, char **first, char **last);
853
854 int pcre_get_substring(const char *subject, int *ovector,
855 int stringcount, int stringnumber,
856 const char **stringptr);
857
858 int pcre_get_substring_list(const char *subject,
859 int *ovector, int stringcount, const char ***listptr);
860
861 void pcre_free_substring(const char *stringptr);
862
863 void pcre_free_substring_list(const char **stringptr);
864
865 const unsigned char *pcre_maketables(void);
866
867 int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
868 int what, void *where);
869
870 int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
871
872 int pcre_refcount(pcre *code, int adjust);
873
874 int pcre_config(int what, void *where);
875
876 char *pcre_version(void);
877
878 void *(*pcre_malloc)(size_t);
879
880 void (*pcre_free)(void *);
881
882 void *(*pcre_stack_malloc)(size_t);
883
884 void (*pcre_stack_free)(void *);
885
886 int (*pcre_callout)(pcre_callout_block *);
887
888
889 PCRE API OVERVIEW
890
891 PCRE has its own native API, which is described in this document. There
892 are also some wrapper functions that correspond to the POSIX regular
893 expression API. These are described in the pcreposix documentation.
894 Both of these APIs define a set of C function calls. A C++ wrapper is
895 distributed with PCRE. It is documented in the pcrecpp page.
896
897 The native API C function prototypes are defined in the header file
898 pcre.h, and on Unix systems the library itself is called libpcre. It
899 can normally be accessed by adding -lpcre to the command for linking an
900 application that uses PCRE. The header file defines the macros
901 PCRE_MAJOR and PCRE_MINOR to contain the major and minor release num-
902 bers for the library. Applications can use these to include support
903 for different releases of PCRE.
904
905 The functions pcre_compile(), pcre_compile2(), pcre_study(), and
906 pcre_exec() are used for compiling and matching regular expressions in
907 a Perl-compatible manner. A sample program that demonstrates the sim-
908 plest way of using them is provided in the file called pcredemo.c in
909 the PCRE source distribution. A listing of this program is given in the
910 pcredemo documentation, and the pcresample documentation describes how
911 to compile and run it.
912
913 A second matching function, pcre_dfa_exec(), which is not Perl-compati-
914 ble, is also provided. This uses a different algorithm for the match-
915 ing. The alternative algorithm finds all possible matches (at a given
916 point in the subject), and scans the subject just once (unless there
917 are lookbehind assertions). However, this algorithm does not return
918 captured substrings. A description of the two matching algorithms and
919 their advantages and disadvantages is given in the pcrematching docu-
920 mentation.
921
922 In addition to the main compiling and matching functions, there are
923 convenience functions for extracting captured substrings from a subject
924 string that is matched by pcre_exec(). They are:
925
926 pcre_copy_substring()
927 pcre_copy_named_substring()
928 pcre_get_substring()
929 pcre_get_named_substring()
930 pcre_get_substring_list()
931 pcre_get_stringnumber()
932 pcre_get_stringtable_entries()
933
934 pcre_free_substring() and pcre_free_substring_list() are also provided,
935 to free the memory used for extracted strings.
936
937 The function pcre_maketables() is used to build a set of character
938 tables in the current locale for passing to pcre_compile(),
939 pcre_exec(), or pcre_dfa_exec(). This is an optional facility that is
940 provided for specialist use. Most commonly, no special tables are
941 passed, in which case internal tables that are generated when PCRE is
942 built are used.
943
944 The function pcre_fullinfo() is used to find out information about a
945 compiled pattern; pcre_info() is an obsolete version that returns only
946 some of the available information, but is retained for backwards com-
947 patibility. The function pcre_version() returns a pointer to a string
948 containing the version of PCRE and its date of release.
949
950 The function pcre_refcount() maintains a reference count in a data
951 block containing a compiled pattern. This is provided for the benefit
952 of object-oriented applications.
953
954 The global variables pcre_malloc and pcre_free initially contain the
955 entry points of the standard malloc() and free() functions, respec-
956 tively. PCRE calls the memory management functions via these variables,
957 so a calling program can replace them if it wishes to intercept the
958 calls. This should be done before calling any PCRE functions.
959
960 The global variables pcre_stack_malloc and pcre_stack_free are also
961 indirections to memory management functions. These special functions
962 are used only when PCRE is compiled to use the heap for remembering
963 data, instead of recursive function calls, when running the pcre_exec()
964 function. See the pcrebuild documentation for details of how to do
965 this. It is a non-standard way of building PCRE, for use in environ-
966 ments that have limited stacks. Because of the greater use of memory
967 management, it runs more slowly. Separate functions are provided so
968 that special-purpose external code can be used for this case. When
969 used, these functions are always called in a stack-like manner (last
970 obtained, first freed), and always for memory blocks of the same size.
971 There is a discussion about PCRE's stack usage in the pcrestack docu-
972 mentation.
973
974 The global variable pcre_callout initially contains NULL. It can be set
975 by the caller to a "callout" function, which PCRE will then call at
976 specified points during a matching operation. Details are given in the
977 pcrecallout documentation.
978
979
980 NEWLINES
981
982 PCRE supports five different conventions for indicating line breaks in
983 strings: a single CR (carriage return) character, a single LF (line-
984 feed) character, the two-character sequence CRLF, any of the three pre-
985 ceding, or any Unicode newline sequence. The Unicode newline sequences
986 are the three just mentioned, plus the single characters VT (vertical
987 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
988 separator, U+2028), and PS (paragraph separator, U+2029).
989
990 Each of the first three conventions is used by at least one operating
991 system as its standard newline sequence. When PCRE is built, a default
992 can be specified. The default default is LF, which is the Unix stan-
993 dard. When PCRE is run, the default can be overridden, either when a
994 pattern is compiled, or when it is matched.
995
996 At compile time, the newline convention can be specified by the options
997 argument of pcre_compile(), or it can be specified by special text at
998 the start of the pattern itself; this overrides any other settings. See
999 the pcrepattern page for details of the special character sequences.
1000
1001 In the PCRE documentation the word "newline" is used to mean "the char-
1002 acter or pair of characters that indicate a line break". The choice of
1003 newline convention affects the handling of the dot, circumflex, and
1004 dollar metacharacters, the handling of #-comments in /x mode, and, when
1005 CRLF is a recognized line ending sequence, the match position advance-
1006 ment for a non-anchored pattern. There is more detail about this in the
1007 section on pcre_exec() options below.
1008
1009 The choice of newline convention does not affect the interpretation of
1010 the \n or \r escape sequences, nor does it affect what \R matches,
1011 which is controlled in a similar way, but by separate options.
1012
1013
1014 MULTITHREADING
1015
1016 The PCRE functions can be used in multi-threading applications, with
1017 the proviso that the memory management functions pointed to by
1018 pcre_malloc, pcre_free, pcre_stack_malloc, and pcre_stack_free, and the
1019 callout function pointed to by pcre_callout, are shared by all threads.
1020
1021 The compiled form of a regular expression is not altered during match-
1022 ing, so the same compiled pattern can safely be used by several threads
1023 at once.
1024
1025
1026 SAVING PRECOMPILED PATTERNS FOR LATER USE
1027
1028 The compiled form of a regular expression can be saved and re-used at a
1029 later time, possibly by a different program, and even on a host other
1030 than the one on which it was compiled. Details are given in the
1031 pcreprecompile documentation. However, compiling a regular expression
1032 with one version of PCRE for use with a different version is not guar-
1033 anteed to work and may cause crashes.
1034
1035
1036 CHECKING BUILD-TIME OPTIONS
1037
1038 int pcre_config(int what, void *where);
1039
1040 The function pcre_config() makes it possible for a PCRE client to dis-
1041 cover which optional features have been compiled into the PCRE library.
1042 The pcrebuild documentation has more details about these optional fea-
1043 tures.
1044
1045 The first argument for pcre_config() is an integer, specifying which
1046 information is required; the second argument is a pointer to a variable
1047 into which the information is placed. The following information is
1048 available:
1049
1050 PCRE_CONFIG_UTF8
1051
1052 The output is an integer that is set to one if UTF-8 support is avail-
1053 able; otherwise it is set to zero.
1054
1055 PCRE_CONFIG_UNICODE_PROPERTIES
1056
1057 The output is an integer that is set to one if support for Unicode
1058 character properties is available; otherwise it is set to zero.
1059
1060 PCRE_CONFIG_NEWLINE
1061
1062 The output is an integer whose value specifies the default character
1063 sequence that is recognized as meaning "newline". The four values that
1064 are supported are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF,
1065 and -1 for ANY. Though they are derived from ASCII, the same values
1066 are returned in EBCDIC environments. The default should normally corre-
1067 spond to the standard sequence for your operating system.
1068
1069 PCRE_CONFIG_BSR
1070
1071 The output is an integer whose value indicates what character sequences
1072 the \R escape sequence matches by default. A value of 0 means that \R
1073 matches any Unicode line ending sequence; a value of 1 means that \R
1074 matches only CR, LF, or CRLF. The default can be overridden when a pat-
1075 tern is compiled or matched.
1076
1077 PCRE_CONFIG_LINK_SIZE
1078
1079 The output is an integer that contains the number of bytes used for
1080 internal linkage in compiled regular expressions. The value is 2, 3, or
1081 4. Larger values allow larger regular expressions to be compiled, at
1082 the expense of slower matching. The default value of 2 is sufficient
1083 for all but the most massive patterns, since it allows the compiled
1084 pattern to be up to 64K in size.
1085
1086 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1087
1088 The output is an integer that contains the threshold above which the
1089 POSIX interface uses malloc() for output vectors. Further details are
1090 given in the pcreposix documentation.
1091
1092 PCRE_CONFIG_MATCH_LIMIT
1093
1094 The output is a long integer that gives the default limit for the num-
1095 ber of internal matching function calls in a pcre_exec() execution.
1096 Further details are given with pcre_exec() below.
1097
1098 PCRE_CONFIG_MATCH_LIMIT_RECURSION
1099
1100 The output is a long integer that gives the default limit for the depth
1101 of recursion when calling the internal matching function in a
1102 pcre_exec() execution. Further details are given with pcre_exec()
1103 below.
1104
1105 PCRE_CONFIG_STACKRECURSE
1106
1107 The output is an integer that is set to one if internal recursion when
1108 running pcre_exec() is implemented by recursive function calls that use
1109 the stack to remember their state. This is the usual way that PCRE is
1110 compiled. The output is zero if PCRE was compiled to use blocks of data
1111 on the heap instead of recursive function calls. In this case,
1112 pcre_stack_malloc and pcre_stack_free are called to manage memory
1113 blocks on the heap, thus avoiding the use of the stack.
1114
1115
1116 COMPILING A PATTERN
1117
1118 pcre *pcre_compile(const char *pattern, int options,
1119 const char **errptr, int *erroffset,
1120 const unsigned char *tableptr);
1121
1122 pcre *pcre_compile2(const char *pattern, int options,
1123 int *errorcodeptr,
1124 const char **errptr, int *erroffset,
1125 const unsigned char *tableptr);
1126
1127 Either of the functions pcre_compile() or pcre_compile2() can be called
1128 to compile a pattern into an internal form. The only difference between
1129 the two interfaces is that pcre_compile2() has an additional argument,
1130 errorcodeptr, via which a numerical error code can be returned. To
1131 avoid too much repetition, we refer just to pcre_compile() below, but
1132 the information applies equally to pcre_compile2().
1133
1134 The pattern is a C string terminated by a binary zero, and is passed in
1135 the pattern argument. A pointer to a single block of memory that is
1136 obtained via pcre_malloc is returned. This contains the compiled code
1137 and related data. The pcre type is defined for the returned block; this
1138 is a typedef for a structure whose contents are not externally defined.
1139 It is up to the caller to free the memory (via pcre_free) when it is no
1140 longer required.
1141
1142 Although the compiled code of a PCRE regex is relocatable, that is, it
1143 does not depend on memory location, the complete pcre data block is not
1144 fully relocatable, because it may contain a copy of the tableptr argu-
1145 ment, which is an address (see below).
1146
1147 The options argument contains various bit settings that affect the com-
1148 pilation. It should be zero if no options are required. The available
1149 options are described below. Some of them (in particular, those that
1150 are compatible with Perl, but some others as well) can also be set and
1151 unset from within the pattern (see the detailed description in the
1152 pcrepattern documentation). For those options that can be different in
1153 different parts of the pattern, the contents of the options argument
1154 specifies their settings at the start of compilation and execution. The
1155 PCRE_ANCHORED, PCRE_BSR_xxx, and PCRE_NEWLINE_xxx options can be set at
1156 the time of matching as well as at compile time.
1157
1158 If errptr is NULL, pcre_compile() returns NULL immediately. Otherwise,
1159 if compilation of a pattern fails, pcre_compile() returns NULL, and
1160 sets the variable pointed to by errptr to point to a textual error mes-
1161 sage. This is a static string that is part of the library. You must not
1162 try to free it. The byte offset from the start of the pattern to the
1163 character that was being processed when the error was discovered is
1164 placed in the variable pointed to by erroffset, which must not be NULL.
1165 If it is, an immediate error is given. Some errors are not detected
1166 until checks are carried out when the whole pattern has been scanned;
1167 in this case the offset is set to the end of the pattern.
1168
1169 If pcre_compile2() is used instead of pcre_compile(), and the error-
1170 codeptr argument is not NULL, a non-zero error code number is returned
1171 via this argument in the event of an error. This is in addition to the
1172 textual error message. Error codes and messages are listed below.
1173
1174 If the final argument, tableptr, is NULL, PCRE uses a default set of
1175 character tables that are built when PCRE is compiled, using the
1176 default C locale. Otherwise, tableptr must be an address that is the
1177 result of a call to pcre_maketables(). This value is stored with the
1178 compiled pattern, and used again by pcre_exec(), unless another table
1179 pointer is passed to it. For more discussion, see the section on locale
1180 support below.
1181
1182 This code fragment shows a typical straightforward call to pcre_com-
1183 pile():
1184
1185 pcre *re;
1186 const char *error;
1187 int erroffset;
1188 re = pcre_compile(
1189 "^A.*Z", /* the pattern */
1190 0, /* default options */
1191 &error, /* for error message */
1192 &erroffset, /* for error offset */
1193 NULL); /* use default character tables */
1194
1195 The following names for option bits are defined in the pcre.h header
1196 file:
1197
1198 PCRE_ANCHORED
1199
1200 If this bit is set, the pattern is forced to be "anchored", that is, it
1201 is constrained to match only at the first matching point in the string
1202 that is being searched (the "subject string"). This effect can also be
1203 achieved by appropriate constructs in the pattern itself, which is the
1204 only way to do it in Perl.
1205
1206 PCRE_AUTO_CALLOUT
1207
1208 If this bit is set, pcre_compile() automatically inserts callout items,
1209 all with number 255, before each pattern item. For discussion of the
1210 callout facility, see the pcrecallout documentation.
1211
1212 PCRE_BSR_ANYCRLF
1213 PCRE_BSR_UNICODE
1214
1215 These options (which are mutually exclusive) control what the \R escape
1216 sequence matches. The choice is either to match only CR, LF, or CRLF,
1217 or to match any Unicode newline sequence. The default is specified when
1218 PCRE is built. It can be overridden from within the pattern, or by set-
1219 ting an option when a compiled pattern is matched.
1220
1221 PCRE_CASELESS
1222
1223 If this bit is set, letters in the pattern match both upper and lower
1224 case letters. It is equivalent to Perl's /i option, and it can be
1225 changed within a pattern by a (?i) option setting. In UTF-8 mode, PCRE
1226 always understands the concept of case for characters whose values are
1227 less than 128, so caseless matching is always possible. For characters
1228 with higher values, the concept of case is supported if PCRE is com-
1229 piled with Unicode property support, but not otherwise. If you want to
1230 use caseless matching for characters 128 and above, you must ensure
1231 that PCRE is compiled with Unicode property support as well as with
1232 UTF-8 support.
1233
1234 PCRE_DOLLAR_ENDONLY
1235
1236 If this bit is set, a dollar metacharacter in the pattern matches only
1237 at the end of the subject string. Without this option, a dollar also
1238 matches immediately before a newline at the end of the string (but not
1239 before any other newlines). The PCRE_DOLLAR_ENDONLY option is ignored
1240 if PCRE_MULTILINE is set. There is no equivalent to this option in
1241 Perl, and no way to set it within a pattern.
1242
1243 PCRE_DOTALL
1244
1245 If this bit is set, a dot metacharater in the pattern matches all char-
1246 acters, including those that indicate newline. Without it, a dot does
1247 not match when the current position is at a newline. This option is
1248 equivalent to Perl's /s option, and it can be changed within a pattern
1249 by a (?s) option setting. A negative class such as [^a] always matches
1250 newline characters, independent of the setting of this option.
1251
1252 PCRE_DUPNAMES
1253
1254 If this bit is set, names used to identify capturing subpatterns need
1255 not be unique. This can be helpful for certain types of pattern when it
1256 is known that only one instance of the named subpattern can ever be
1257 matched. There are more details of named subpatterns below; see also
1258 the pcrepattern documentation.
1259
1260 PCRE_EXTENDED
1261
1262 If this bit is set, whitespace data characters in the pattern are
1263 totally ignored except when escaped or inside a character class. White-
1264 space does not include the VT character (code 11). In addition, charac-
1265 ters between an unescaped # outside a character class and the next new-
1266 line, inclusive, are also ignored. This is equivalent to Perl's /x
1267 option, and it can be changed within a pattern by a (?x) option set-
1268 ting.
1269
1270 This option makes it possible to include comments inside complicated
1271 patterns. Note, however, that this applies only to data characters.
1272 Whitespace characters may never appear within special character
1273 sequences in a pattern, for example within the sequence (?( which
1274 introduces a conditional subpattern.
1275
1276 PCRE_EXTRA
1277
1278 This option was invented in order to turn on additional functionality
1279 of PCRE that is incompatible with Perl, but it is currently of very
1280 little use. When set, any backslash in a pattern that is followed by a
1281 letter that has no special meaning causes an error, thus reserving
1282 these combinations for future expansion. By default, as in Perl, a
1283 backslash followed by a letter with no special meaning is treated as a
1284 literal. (Perl can, however, be persuaded to give a warning for this.)
1285 There are at present no other features controlled by this option. It
1286 can also be set by a (?X) option setting within a pattern.
1287
1288 PCRE_FIRSTLINE
1289
1290 If this option is set, an unanchored pattern is required to match
1291 before or at the first newline in the subject string, though the
1292 matched text may continue over the newline.
1293
1294 PCRE_JAVASCRIPT_COMPAT
1295
1296 If this option is set, PCRE's behaviour is changed in some ways so that
1297 it is compatible with JavaScript rather than Perl. The changes are as
1298 follows:
1299
1300 (1) A lone closing square bracket in a pattern causes a compile-time
1301 error, because this is illegal in JavaScript (by default it is treated
1302 as a data character). Thus, the pattern AB]CD becomes illegal when this
1303 option is set.
1304
1305 (2) At run time, a back reference to an unset subpattern group matches
1306 an empty string (by default this causes the current matching alterna-
1307 tive to fail). A pattern such as (\1)(a) succeeds when this option is
1308 set (assuming it can find an "a" in the subject), whereas it fails by
1309 default, for Perl compatibility.
1310
1311 PCRE_MULTILINE
1312
1313 By default, PCRE treats the subject string as consisting of a single
1314 line of characters (even if it actually contains newlines). The "start
1315 of line" metacharacter (^) matches only at the start of the string,
1316 while the "end of line" metacharacter ($) matches only at the end of
1317 the string, or before a terminating newline (unless PCRE_DOLLAR_ENDONLY
1318 is set). This is the same as Perl.
1319
1320 When PCRE_MULTILINE it is set, the "start of line" and "end of line"
1321 constructs match immediately following or immediately before internal
1322 newlines in the subject string, respectively, as well as at the very
1323 start and end. This is equivalent to Perl's /m option, and it can be
1324 changed within a pattern by a (?m) option setting. If there are no new-
1325 lines in a subject string, or no occurrences of ^ or $ in a pattern,
1326 setting PCRE_MULTILINE has no effect.
1327
1328 PCRE_NEWLINE_CR
1329 PCRE_NEWLINE_LF
1330 PCRE_NEWLINE_CRLF
1331 PCRE_NEWLINE_ANYCRLF
1332 PCRE_NEWLINE_ANY
1333
1334 These options override the default newline definition that was chosen
1335 when PCRE was built. Setting the first or the second specifies that a
1336 newline is indicated by a single character (CR or LF, respectively).
1337 Setting PCRE_NEWLINE_CRLF specifies that a newline is indicated by the
1338 two-character CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies
1339 that any of the three preceding sequences should be recognized. Setting
1340 PCRE_NEWLINE_ANY specifies that any Unicode newline sequence should be
1341 recognized. The Unicode newline sequences are the three just mentioned,
1342 plus the single characters VT (vertical tab, U+000B), FF (formfeed,
1343 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
1344 (paragraph separator, U+2029). The last two are recognized only in
1345 UTF-8 mode.
1346
1347 The newline setting in the options word uses three bits that are
1348 treated as a number, giving eight possibilities. Currently only six are
1349 used (default plus the five values above). This means that if you set
1350 more than one newline option, the combination may or may not be sensi-
1351 ble. For example, PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to
1352 PCRE_NEWLINE_CRLF, but other combinations may yield unused numbers and
1353 cause an error.
1354
1355 The only time that a line break is specially recognized when compiling
1356 a pattern is if PCRE_EXTENDED is set, and an unescaped # outside a
1357 character class is encountered. This indicates a comment that lasts
1358 until after the next line break sequence. In other circumstances, line
1359 break sequences are treated as literal data, except that in
1360 PCRE_EXTENDED mode, both CR and LF are treated as whitespace characters
1361 and are therefore ignored.
1362
1363 The newline option that is set at compile time becomes the default that
1364 is used for pcre_exec() and pcre_dfa_exec(), but it can be overridden.
1365
1366 PCRE_NO_AUTO_CAPTURE
1367
1368 If this option is set, it disables the use of numbered capturing paren-
1369 theses in the pattern. Any opening parenthesis that is not followed by
1370 ? behaves as if it were followed by ?: but named parentheses can still
1371 be used for capturing (and they acquire numbers in the usual way).
1372 There is no equivalent of this option in Perl.
1373
1374 PCRE_UNGREEDY
1375
1376 This option inverts the "greediness" of the quantifiers so that they
1377 are not greedy by default, but become greedy if followed by "?". It is
1378 not compatible with Perl. It can also be set by a (?U) option setting
1379 within the pattern.
1380
1381 PCRE_UTF8
1382
1383 This option causes PCRE to regard both the pattern and the subject as
1384 strings of UTF-8 characters instead of single-byte character strings.
1385 However, it is available only when PCRE is built to include UTF-8 sup-
1386 port. If not, the use of this option provokes an error. Details of how
1387 this option changes the behaviour of PCRE are given in the section on
1388 UTF-8 support in the main pcre page.
1389
1390 PCRE_NO_UTF8_CHECK
1391
1392 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
1393 automatically checked. There is a discussion about the validity of
1394 UTF-8 strings in the main pcre page. If an invalid UTF-8 sequence of
1395 bytes is found, pcre_compile() returns an error. If you already know
1396 that your pattern is valid, and you want to skip this check for perfor-
1397 mance reasons, you can set the PCRE_NO_UTF8_CHECK option. When it is
1398 set, the effect of passing an invalid UTF-8 string as a pattern is
1399 undefined. It may cause your program to crash. Note that this option
1400 can also be passed to pcre_exec() and pcre_dfa_exec(), to suppress the
1401 UTF-8 validity checking of subject strings.
1402
1403
1404 COMPILATION ERROR CODES
1405
1406 The following table lists the error codes than may be returned by
1407 pcre_compile2(), along with the error messages that may be returned by
1408 both compiling functions. As PCRE has developed, some error codes have
1409 fallen out of use. To avoid confusion, they have not been re-used.
1410
1411 0 no error
1412 1 \ at end of pattern
1413 2 \c at end of pattern
1414 3 unrecognized character follows \
1415 4 numbers out of order in {} quantifier
1416 5 number too big in {} quantifier
1417 6 missing terminating ] for character class
1418 7 invalid escape sequence in character class
1419 8 range out of order in character class
1420 9 nothing to repeat
1421 10 [this code is not in use]
1422 11 internal error: unexpected repeat
1423 12 unrecognized character after (? or (?-
1424 13 POSIX named classes are supported only within a class
1425 14 missing )
1426 15 reference to non-existent subpattern
1427 16 erroffset passed as NULL
1428 17 unknown option bit(s) set
1429 18 missing ) after comment
1430 19 [this code is not in use]
1431 20 regular expression is too large
1432 21 failed to get memory
1433 22 unmatched parentheses
1434 23 internal error: code overflow
1435 24 unrecognized character after (?<
1436 25 lookbehind assertion is not fixed length
1437 26 malformed number or name after (?(
1438 27 conditional group contains more than two branches
1439 28 assertion expected after (?(
1440 29 (?R or (?[+-]digits must be followed by )
1441 30 unknown POSIX class name
1442 31 POSIX collating elements are not supported
1443 32 this version of PCRE is not compiled with PCRE_UTF8 support
1444 33 [this code is not in use]
1445 34 character value in \x{...} sequence is too large
1446 35 invalid condition (?(0)
1447 36 \C not allowed in lookbehind assertion
1448 37 PCRE does not support \L, \l, \N, \U, or \u
1449 38 number after (?C is > 255
1450 39 closing ) for (?C expected
1451 40 recursive call could loop indefinitely
1452 41 unrecognized character after (?P
1453 42 syntax error in subpattern name (missing terminator)
1454 43 two named subpatterns have the same name
1455 44 invalid UTF-8 string
1456 45 support for \P, \p, and \X has not been compiled
1457 46 malformed \P or \p sequence
1458 47 unknown property name after \P or \p
1459 48 subpattern name is too long (maximum 32 characters)
1460 49 too many named subpatterns (maximum 10000)
1461 50 [this code is not in use]
1462 51 octal value is greater than \377 (not in UTF-8 mode)
1463 52 internal error: overran compiling workspace
1464 53 internal error: previously-checked referenced subpattern not
1465 found
1466 54 DEFINE group contains more than one branch
1467 55 repeating a DEFINE group is not allowed
1468 56 inconsistent NEWLINE options
1469 57 \g is not followed by a braced, angle-bracketed, or quoted
1470 name/number or by a plain number
1471 58 a numbered reference must not be zero
1472 59 (*VERB) with an argument is not supported
1473 60 (*VERB) not recognized
1474 61 number is too big
1475 62 subpattern name expected
1476 63 digit expected after (?+
1477 64 ] is an invalid data character in JavaScript compatibility mode
1478
1479 The numbers 32 and 10000 in errors 48 and 49 are defaults; different
1480 values may be used if the limits were changed when PCRE was built.
1481
1482
1483 STUDYING A PATTERN
1484
1485 pcre_extra *pcre_study(const pcre *code, int options
1486 const char **errptr);
1487
1488 If a compiled pattern is going to be used several times, it is worth
1489 spending more time analyzing it in order to speed up the time taken for
1490 matching. The function pcre_study() takes a pointer to a compiled pat-
1491 tern as its first argument. If studying the pattern produces additional
1492 information that will help speed up matching, pcre_study() returns a
1493 pointer to a pcre_extra block, in which the study_data field points to
1494 the results of the study.
1495
1496 The returned value from pcre_study() can be passed directly to
1497 pcre_exec() or pcre_dfa_exec(). However, a pcre_extra block also con-
1498 tains other fields that can be set by the caller before the block is
1499 passed; these are described below in the section on matching a pattern.
1500
1501 If studying the pattern does not produce any useful information,
1502 pcre_study() returns NULL. In that circumstance, if the calling program
1503 wants to pass any of the other fields to pcre_exec() or
1504 pcre_dfa_exec(), it must set up its own pcre_extra block.
1505
1506 The second argument of pcre_study() contains option bits. At present,
1507 no options are defined, and this argument should always be zero.
1508
1509 The third argument for pcre_study() is a pointer for an error message.
1510 If studying succeeds (even if no data is returned), the variable it
1511 points to is set to NULL. Otherwise it is set to point to a textual
1512 error message. This is a static string that is part of the library. You
1513 must not try to free it. You should test the error pointer for NULL
1514 after calling pcre_study(), to be sure that it has run successfully.
1515
1516 This is a typical call to pcre_study():
1517
1518 pcre_extra *pe;
1519 pe = pcre_study(
1520 re, /* result of pcre_compile() */
1521 0, /* no options exist */
1522 &error); /* set to NULL or points to a message */
1523
1524 Studying a pattern does two things: first, a lower bound for the length
1525 of subject string that is needed to match the pattern is computed. This
1526 does not mean that there are any strings of that length that match, but
1527 it does guarantee that no shorter strings match. The value is used by
1528 pcre_exec() and pcre_dfa_exec() to avoid wasting time by trying to
1529 match strings that are shorter than the lower bound. You can find out
1530 the value in a calling program via the pcre_fullinfo() function.
1531
1532 Studying a pattern is also useful for non-anchored patterns that do not
1533 have a single fixed starting character. A bitmap of possible starting
1534 bytes is created. This speeds up finding a position in the subject at
1535 which to start matching.
1536
1537
1538 LOCALE SUPPORT
1539
1540 PCRE handles caseless matching, and determines whether characters are
1541 letters, digits, or whatever, by reference to a set of tables, indexed
1542 by character value. When running in UTF-8 mode, this applies only to
1543 characters with codes less than 128. Higher-valued codes never match
1544 escapes such as \w or \d, but can be tested with \p if PCRE is built
1545 with Unicode character property support. The use of locales with Uni-
1546 code is discouraged. If you are handling characters with codes greater
1547 than 128, you should either use UTF-8 and Unicode, or use locales, but
1548 not try to mix the two.
1549
1550 PCRE contains an internal set of tables that are used when the final
1551 argument of pcre_compile() is NULL. These are sufficient for many
1552 applications. Normally, the internal tables recognize only ASCII char-
1553 acters. However, when PCRE is built, it is possible to cause the inter-
1554 nal tables to be rebuilt in the default "C" locale of the local system,
1555 which may cause them to be different.
1556
1557 The internal tables can always be overridden by tables supplied by the
1558 application that calls PCRE. These may be created in a different locale
1559 from the default. As more and more applications change to using Uni-
1560 code, the need for this locale support is expected to die away.
1561
1562 External tables are built by calling the pcre_maketables() function,
1563 which has no arguments, in the relevant locale. The result can then be
1564 passed to pcre_compile() or pcre_exec() as often as necessary. For
1565 example, to build and use tables that are appropriate for the French
1566 locale (where accented characters with values greater than 128 are
1567 treated as letters), the following code could be used:
1568
1569 setlocale(LC_CTYPE, "fr_FR");
1570 tables = pcre_maketables();
1571 re = pcre_compile(..., tables);
1572
1573 The locale name "fr_FR" is used on Linux and other Unix-like systems;
1574 if you are using Windows, the name for the French locale is "french".
1575
1576 When pcre_maketables() runs, the tables are built in memory that is
1577 obtained via pcre_malloc. It is the caller's responsibility to ensure
1578 that the memory containing the tables remains available for as long as
1579 it is needed.
1580
1581 The pointer that is passed to pcre_compile() is saved with the compiled
1582 pattern, and the same tables are used via this pointer by pcre_study()
1583 and normally also by pcre_exec(). Thus, by default, for any single pat-
1584 tern, compilation, studying and matching all happen in the same locale,
1585 but different patterns can be compiled in different locales.
1586
1587 It is possible to pass a table pointer or NULL (indicating the use of
1588 the internal tables) to pcre_exec(). Although not intended for this
1589 purpose, this facility could be used to match a pattern in a different
1590 locale from the one in which it was compiled. Passing table pointers at
1591 run time is discussed below in the section on matching a pattern.
1592
1593
1594 INFORMATION ABOUT A PATTERN
1595
1596 int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
1597 int what, void *where);
1598
1599 The pcre_fullinfo() function returns information about a compiled pat-
1600 tern. It replaces the obsolete pcre_info() function, which is neverthe-
1601 less retained for backwards compability (and is documented below).
1602
1603 The first argument for pcre_fullinfo() is a pointer to the compiled
1604 pattern. The second argument is the result of pcre_study(), or NULL if
1605 the pattern was not studied. The third argument specifies which piece
1606 of information is required, and the fourth argument is a pointer to a
1607 variable to receive the data. The yield of the function is zero for
1608 success, or one of the following negative numbers:
1609
1610 PCRE_ERROR_NULL the argument code was NULL
1611 the argument where was NULL
1612 PCRE_ERROR_BADMAGIC the "magic number" was not found
1613 PCRE_ERROR_BADOPTION the value of what was invalid
1614
1615 The "magic number" is placed at the start of each compiled pattern as
1616 an simple check against passing an arbitrary memory pointer. Here is a
1617 typical call of pcre_fullinfo(), to obtain the length of the compiled
1618 pattern:
1619
1620 int rc;
1621 size_t length;
1622 rc = pcre_fullinfo(
1623 re, /* result of pcre_compile() */
1624 pe, /* result of pcre_study(), or NULL */
1625 PCRE_INFO_SIZE, /* what is required */
1626 &length); /* where to put the data */
1627
1628 The possible values for the third argument are defined in pcre.h, and
1629 are as follows:
1630
1631 PCRE_INFO_BACKREFMAX
1632
1633 Return the number of the highest back reference in the pattern. The
1634 fourth argument should point to an int variable. Zero is returned if
1635 there are no back references.
1636
1637 PCRE_INFO_CAPTURECOUNT
1638
1639 Return the number of capturing subpatterns in the pattern. The fourth
1640 argument should point to an int variable.
1641
1642 PCRE_INFO_DEFAULT_TABLES
1643
1644 Return a pointer to the internal default character tables within PCRE.
1645 The fourth argument should point to an unsigned char * variable. This
1646 information call is provided for internal use by the pcre_study() func-
1647 tion. External callers can cause PCRE to use its internal tables by
1648 passing a NULL table pointer.
1649
1650 PCRE_INFO_FIRSTBYTE
1651
1652 Return information about the first byte of any matched string, for a
1653 non-anchored pattern. The fourth argument should point to an int vari-
1654 able. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name
1655 is still recognized for backwards compatibility.)
1656
1657 If there is a fixed first byte, for example, from a pattern such as
1658 (cat|cow|coyote), its value is returned. Otherwise, if either
1659
1660 (a) the pattern was compiled with the PCRE_MULTILINE option, and every
1661 branch starts with "^", or
1662
1663 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not
1664 set (if it were set, the pattern would be anchored),
1665
1666 -1 is returned, indicating that the pattern matches only at the start
1667 of a subject string or after any newline within the string. Otherwise
1668 -2 is returned. For anchored patterns, -2 is returned.
1669
1670 PCRE_INFO_FIRSTTABLE
1671
1672 If the pattern was studied, and this resulted in the construction of a
1673 256-bit table indicating a fixed set of bytes for the first byte in any
1674 matching string, a pointer to the table is returned. Otherwise NULL is
1675 returned. The fourth argument should point to an unsigned char * vari-
1676 able.
1677
1678 PCRE_INFO_HASCRORLF
1679
1680 Return 1 if the pattern contains any explicit matches for CR or LF
1681 characters, otherwise 0. The fourth argument should point to an int
1682 variable. An explicit match is either a literal CR or LF character, or
1683 \r or \n.
1684
1685 PCRE_INFO_JCHANGED
1686
1687 Return 1 if the (?J) or (?-J) option setting is used in the pattern,
1688 otherwise 0. The fourth argument should point to an int variable. (?J)
1689 and (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1690
1691 PCRE_INFO_LASTLITERAL
1692
1693 Return the value of the rightmost literal byte that must exist in any
1694 matched string, other than at its start, if such a byte has been
1695 recorded. The fourth argument should point to an int variable. If there
1696 is no such byte, -1 is returned. For anchored patterns, a last literal
1697 byte is recorded only if it follows something of variable length. For
1698 example, for the pattern /^a\d+z\d+/ the returned value is "z", but for
1699 /^a\dz\d/ the returned value is -1.
1700
1701 PCRE_INFO_MINLENGTH
1702
1703 If the pattern was studied and a minimum length for matching subject
1704 strings was computed, its value is returned. Otherwise the returned
1705 value is -1. The value is a number of characters, not bytes (this may
1706 be relevant in UTF-8 mode). The fourth argument should point to an int
1707 variable. A non-negative value is a lower bound to the length of any
1708 matching string. There may not be any strings of that length that do
1709 actually match, but every string that does match is at least that long.
1710
1711 PCRE_INFO_NAMECOUNT
1712 PCRE_INFO_NAMEENTRYSIZE
1713 PCRE_INFO_NAMETABLE
1714
1715 PCRE supports the use of named as well as numbered capturing parenthe-
1716 ses. The names are just an additional way of identifying the parenthe-
1717 ses, which still acquire numbers. Several convenience functions such as
1718 pcre_get_named_substring() are provided for extracting captured sub-
1719 strings by name. It is also possible to extract the data directly, by
1720 first converting the name to a number in order to access the correct
1721 pointers in the output vector (described with pcre_exec() below). To do
1722 the conversion, you need to use the name-to-number map, which is
1723 described by these three values.
1724
1725 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT
1726 gives the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size
1727 of each entry; both of these return an int value. The entry size
1728 depends on the length of the longest name. PCRE_INFO_NAMETABLE returns
1729 a pointer to the first entry of the table (a pointer to char). The
1730 first two bytes of each entry are the number of the capturing parenthe-
1731 sis, most significant byte first. The rest of the entry is the corre-
1732 sponding name, zero terminated.
1733
1734 The names are in alphabetical order. Duplicate names may appear if (?|
1735 is used to create multiple groups with the same number, as described in
1736 the section on duplicate subpattern numbers in the pcrepattern page.
1737 Duplicate names for subpatterns with different numbers are permitted
1738 only if PCRE_DUPNAMES is set. In all cases of duplicate names, they
1739 appear in the table in the order in which they were found in the pat-
1740 tern. In the absence of (?| this is the order of increasing number;
1741 when (?| is used this is not necessarily the case because later subpat-
1742 terns may have lower numbers.
1743
1744 As a simple example of the name/number table, consider the following
1745 pattern (assume PCRE_EXTENDED is set, so white space - including new-
1746 lines - is ignored):
1747
1748 (?<date> (?<year>(\d\d)?\d\d) -
1749 (?<month>\d\d) - (?<day>\d\d) )
1750
1751 There are four named subpatterns, so the table has four entries, and
1752 each entry in the table is eight bytes long. The table is as follows,
1753 with non-printing bytes shows in hexadecimal, and undefined bytes shown
1754 as ??:
1755
1756 00 01 d a t e 00 ??
1757 00 05 d a y 00 ?? ??
1758 00 04 m o n t h 00
1759 00 02 y e a r 00 ??
1760
1761 When writing code to extract data from named subpatterns using the
1762 name-to-number map, remember that the length of the entries is likely
1763 to be different for each compiled pattern.
1764
1765 PCRE_INFO_OKPARTIAL
1766
1767 Return 1 if the pattern can be used for partial matching with
1768 pcre_exec(), otherwise 0. The fourth argument should point to an int
1769 variable. From release 8.00, this always returns 1, because the
1770 restrictions that previously applied to partial matching have been
1771 lifted. The pcrepartial documentation gives details of partial match-
1772 ing.
1773
1774 PCRE_INFO_OPTIONS
1775
1776 Return a copy of the options with which the pattern was compiled. The
1777 fourth argument should point to an unsigned long int variable. These
1778 option bits are those specified in the call to pcre_compile(), modified
1779 by any top-level option settings at the start of the pattern itself. In
1780 other words, they are the options that will be in force when matching
1781 starts. For example, if the pattern /(?im)abc(?-i)d/ is compiled with
1782 the PCRE_EXTENDED option, the result is PCRE_CASELESS, PCRE_MULTILINE,
1783 and PCRE_EXTENDED.
1784
1785 A pattern is automatically anchored by PCRE if all of its top-level
1786 alternatives begin with one of the following:
1787
1788 ^ unless PCRE_MULTILINE is set
1789 \A always
1790 \G always
1791 .* if PCRE_DOTALL is set and there are no back
1792 references to the subpattern in which .* appears
1793
1794 For such patterns, the PCRE_ANCHORED bit is set in the options returned
1795 by pcre_fullinfo().
1796
1797 PCRE_INFO_SIZE
1798
1799 Return the size of the compiled pattern, that is, the value that was
1800 passed as the argument to pcre_malloc() when PCRE was getting memory in
1801 which to place the compiled data. The fourth argument should point to a
1802 size_t variable.
1803
1804 PCRE_INFO_STUDYSIZE
1805
1806 Return the size of the data block pointed to by the study_data field in
1807 a pcre_extra block. That is, it is the value that was passed to
1808 pcre_malloc() when PCRE was getting memory into which to place the data
1809 created by pcre_study(). If pcre_extra is NULL, or there is no study
1810 data, zero is returned. The fourth argument should point to a size_t
1811 variable.
1812
1813
1814 OBSOLETE INFO FUNCTION
1815
1816 int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
1817
1818 The pcre_info() function is now obsolete because its interface is too
1819 restrictive to return all the available data about a compiled pattern.
1820 New programs should use pcre_fullinfo() instead. The yield of
1821 pcre_info() is the number of capturing subpatterns, or one of the fol-
1822 lowing negative numbers:
1823
1824 PCRE_ERROR_NULL the argument code was NULL
1825 PCRE_ERROR_BADMAGIC the "magic number" was not found
1826
1827 If the optptr argument is not NULL, a copy of the options with which
1828 the pattern was compiled is placed in the integer it points to (see
1829 PCRE_INFO_OPTIONS above).
1830
1831 If the pattern is not anchored and the firstcharptr argument is not
1832 NULL, it is used to pass back information about the first character of
1833 any matched string (see PCRE_INFO_FIRSTBYTE above).
1834
1835
1836 REFERENCE COUNTS
1837
1838 int pcre_refcount(pcre *code, int adjust);
1839
1840 The pcre_refcount() function is used to maintain a reference count in
1841 the data block that contains a compiled pattern. It is provided for the
1842 benefit of applications that operate in an object-oriented manner,
1843 where different parts of the application may be using the same compiled
1844 pattern, but you want to free the block when they are all done.
1845
1846 When a pattern is compiled, the reference count field is initialized to
1847 zero. It is changed only by calling this function, whose action is to
1848 add the adjust value (which may be positive or negative) to it. The
1849 yield of the function is the new value. However, the value of the count
1850 is constrained to lie between 0 and 65535, inclusive. If the new value
1851 is outside these limits, it is forced to the appropriate limit value.
1852
1853 Except when it is zero, the reference count is not correctly preserved
1854 if a pattern is compiled on one host and then transferred to a host
1855 whose byte-order is different. (This seems a highly unlikely scenario.)
1856
1857
1858 MATCHING A PATTERN: THE TRADITIONAL FUNCTION
1859
1860 int pcre_exec(const pcre *code, const pcre_extra *extra,
1861 const char *subject, int length, int startoffset,
1862 int options, int *ovector, int ovecsize);
1863
1864 The function pcre_exec() is called to match a subject string against a
1865 compiled pattern, which is passed in the code argument. If the pattern
1866 was studied, the result of the study should be passed in the extra
1867 argument. This function is the main matching facility of the library,
1868 and it operates in a Perl-like manner. For specialist use there is also
1869 an alternative matching function, which is described below in the sec-
1870 tion about the pcre_dfa_exec() function.
1871
1872 In most applications, the pattern will have been compiled (and option-
1873 ally studied) in the same process that calls pcre_exec(). However, it
1874 is possible to save compiled patterns and study data, and then use them
1875 later in different processes, possibly even on different hosts. For a
1876 discussion about this, see the pcreprecompile documentation.
1877
1878 Here is an example of a simple call to pcre_exec():
1879
1880 int rc;
1881 int ovector[30];
1882 rc = pcre_exec(
1883 re, /* result of pcre_compile() */
1884 NULL, /* we didn't study the pattern */
1885 "some string", /* the subject string */
1886 11, /* the length of the subject string */
1887 0, /* start at offset 0 in the subject */
1888 0, /* default options */
1889 ovector, /* vector of integers for substring information */
1890 30); /* number of elements (NOT size in bytes) */
1891
1892 Extra data for pcre_exec()
1893
1894 If the extra argument is not NULL, it must point to a pcre_extra data
1895 block. The pcre_study() function returns such a block (when it doesn't
1896 return NULL), but you can also create one for yourself, and pass addi-
1897 tional information in it. The pcre_extra block contains the following
1898 fields (not necessarily in this order):
1899
1900 unsigned long int flags;
1901 void *study_data;
1902 unsigned long int match_limit;
1903 unsigned long int match_limit_recursion;
1904 void *callout_data;
1905 const unsigned char *tables;
1906
1907 The flags field is a bitmap that specifies which of the other fields
1908 are set. The flag bits are:
1909
1910 PCRE_EXTRA_STUDY_DATA
1911 PCRE_EXTRA_MATCH_LIMIT
1912 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1913 PCRE_EXTRA_CALLOUT_DATA
1914 PCRE_EXTRA_TABLES
1915
1916 Other flag bits should be set to zero. The study_data field is set in
1917 the pcre_extra block that is returned by pcre_study(), together with
1918 the appropriate flag bit. You should not set this yourself, but you may
1919 add to the block by setting the other fields and their corresponding
1920 flag bits.
1921
1922 The match_limit field provides a means of preventing PCRE from using up
1923 a vast amount of resources when running patterns that are not going to
1924 match, but which have a very large number of possibilities in their
1925 search trees. The classic example is a pattern that uses nested unlim-
1926 ited repeats.
1927
1928 Internally, PCRE uses a function called match() which it calls repeat-
1929 edly (sometimes recursively). The limit set by match_limit is imposed
1930 on the number of times this function is called during a match, which
1931 has the effect of limiting the amount of backtracking that can take
1932 place. For patterns that are not anchored, the count restarts from zero
1933 for each position in the subject string.
1934
1935 The default value for the limit can be set when PCRE is built; the
1936 default default is 10 million, which handles all but the most extreme
1937 cases. You can override the default by suppling pcre_exec() with a
1938 pcre_extra block in which match_limit is set, and
1939 PCRE_EXTRA_MATCH_LIMIT is set in the flags field. If the limit is
1940 exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1941
1942 The match_limit_recursion field is similar to match_limit, but instead
1943 of limiting the total number of times that match() is called, it limits
1944 the depth of recursion. The recursion depth is a smaller number than
1945 the total number of calls, because not all calls to match() are recur-
1946 sive. This limit is of use only if it is set smaller than match_limit.
1947
1948 Limiting the recursion depth limits the amount of stack that can be
1949 used, or, when PCRE has been compiled to use memory on the heap instead
1950 of the stack, the amount of heap memory that can be used.
1951
1952 The default value for match_limit_recursion can be set when PCRE is
1953 built; the default default is the same value as the default for
1954 match_limit. You can override the default by suppling pcre_exec() with
1955 a pcre_extra block in which match_limit_recursion is set, and
1956 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the flags field. If the
1957 limit is exceeded, pcre_exec() returns PCRE_ERROR_RECURSIONLIMIT.
1958
1959 The callout_data field is used in conjunction with the "callout" fea-
1960 ture, and is described in the pcrecallout documentation.
1961
1962 The tables field is used to pass a character tables pointer to
1963 pcre_exec(); this overrides the value that is stored with the compiled
1964 pattern. A non-NULL value is stored with the compiled pattern only if
1965 custom tables were supplied to pcre_compile() via its tableptr argu-
1966 ment. If NULL is passed to pcre_exec() using this mechanism, it forces
1967 PCRE's internal tables to be used. This facility is helpful when re-
1968 using patterns that have been saved after compiling with an external
1969 set of tables, because the external tables might be at a different
1970 address when pcre_exec() is called. See the pcreprecompile documenta-
1971 tion for a discussion of saving compiled patterns for later use.
1972
1973 Option bits for pcre_exec()
1974
1975 The unused bits of the options argument for pcre_exec() must be zero.
1976 The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_xxx,
1977 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1978 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1979 PCRE_PARTIAL_HARD.
1980
1981 PCRE_ANCHORED
1982
1983 The PCRE_ANCHORED option limits pcre_exec() to matching at the first
1984 matching position. If a pattern was compiled with PCRE_ANCHORED, or
1985 turned out to be anchored by virtue of its contents, it cannot be made
1986 unachored at matching time.
1987
1988 PCRE_BSR_ANYCRLF
1989 PCRE_BSR_UNICODE
1990
1991 These options (which are mutually exclusive) control what the \R escape
1992 sequence matches. The choice is either to match only CR, LF, or CRLF,
1993 or to match any Unicode newline sequence. These options override the
1994 choice that was made or defaulted when the pattern was compiled.
1995
1996 PCRE_NEWLINE_CR
1997 PCRE_NEWLINE_LF
1998 PCRE_NEWLINE_CRLF
1999 PCRE_NEWLINE_ANYCRLF
2000 PCRE_NEWLINE_ANY
2001
2002 These options override the newline definition that was chosen or
2003 defaulted when the pattern was compiled. For details, see the descrip-
2004 tion of pcre_compile() above. During matching, the newline choice
2005 affects the behaviour of the dot, circumflex, and dollar metacharac-
2006 ters. It may also alter the way the match position is advanced after a
2007 match failure for an unanchored pattern.
2008
2009 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is
2010 set, and a match attempt for an unanchored pattern fails when the cur-
2011 rent position is at a CRLF sequence, and the pattern contains no
2012 explicit matches for CR or LF characters, the match position is
2013 advanced by two characters instead of one, in other words, to after the
2014 CRLF.
2015
2016 The above rule is a compromise that makes the most common cases work as
2017 expected. For example, if the pattern is .+A (and the PCRE_DOTALL
2018 option is not set), it does not match the string "\r\nA" because, after
2019 failing at the start, it skips both the CR and the LF before retrying.
2020 However, the pattern [\r\n]A does match that string, because it con-
2021 tains an explicit CR or LF reference, and so advances only by one char-
2022 acter after the first failure.
2023
2024 An explicit match for CR of LF is either a literal appearance of one of
2025 those characters, or one of the \r or \n escape sequences. Implicit
2026 matches such as [^X] do not count, nor does \s (which includes CR and
2027 LF in the characters that it matches).
2028
2029 Notwithstanding the above, anomalous effects may still occur when CRLF
2030 is a valid newline sequence and explicit \r or \n escapes appear in the
2031 pattern.
2032
2033 PCRE_NOTBOL
2034
2035 This option specifies that first character of the subject string is not
2036 the beginning of a line, so the circumflex metacharacter should not
2037 match before it. Setting this without PCRE_MULTILINE (at compile time)
2038 causes circumflex never to match. This option affects only the behav-
2039 iour of the circumflex metacharacter. It does not affect \A.
2040
2041 PCRE_NOTEOL
2042
2043 This option specifies that the end of the subject string is not the end
2044 of a line, so the dollar metacharacter should not match it nor (except
2045 in multiline mode) a newline immediately before it. Setting this with-
2046 out PCRE_MULTILINE (at compile time) causes dollar never to match. This
2047 option affects only the behaviour of the dollar metacharacter. It does
2048 not affect \Z or \z.
2049
2050 PCRE_NOTEMPTY
2051
2052 An empty string is not considered to be a valid match if this option is
2053 set. If there are alternatives in the pattern, they are tried. If all
2054 the alternatives match the empty string, the entire match fails. For
2055 example, if the pattern
2056
2057 a?b?
2058
2059 is applied to a string not beginning with "a" or "b", it matches an
2060 empty string at the start of the subject. With PCRE_NOTEMPTY set, this
2061 match is not valid, so PCRE searches further into the string for occur-
2062 rences of "a" or "b".
2063
2064 PCRE_NOTEMPTY_ATSTART
2065
2066 This is like PCRE_NOTEMPTY, except that an empty string match that is
2067 not at the start of the subject is permitted. If the pattern is
2068 anchored, such a match can occur only if the pattern contains \K.
2069
2070 Perl has no direct equivalent of PCRE_NOTEMPTY or
2071 PCRE_NOTEMPTY_ATSTART, but it does make a special case of a pattern
2072 match of the empty string within its split() function, and when using
2073 the /g modifier. It is possible to emulate Perl's behaviour after
2074 matching a null string by first trying the match again at the same off-
2075 set with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then if that
2076 fails, by advancing the starting offset (see below) and trying an ordi-
2077 nary match again. There is some code that demonstrates how to do this
2078 in the pcredemo sample program.
2079
2080 PCRE_NO_START_OPTIMIZE
2081
2082 There are a number of optimizations that pcre_exec() uses at the start
2083 of a match, in order to speed up the process. For example, if it is
2084 known that a match must start with a specific character, it searches
2085 the subject for that character, and fails immediately if it cannot find
2086 it, without actually running the main matching function. When callouts
2087 are in use, these optimizations can cause them to be skipped. This
2088 option disables the "start-up" optimizations, causing performance to
2089 suffer, but ensuring that the callouts do occur.
2090
2091 PCRE_NO_UTF8_CHECK
2092
2093 When PCRE_UTF8 is set at compile time, the validity of the subject as a
2094 UTF-8 string is automatically checked when pcre_exec() is subsequently
2095 called. The value of startoffset is also checked to ensure that it
2096 points to the start of a UTF-8 character. There is a discussion about
2097 the validity of UTF-8 strings in the section on UTF-8 support in the
2098 main pcre page. If an invalid UTF-8 sequence of bytes is found,
2099 pcre_exec() returns the error PCRE_ERROR_BADUTF8. If startoffset con-
2100 tains an invalid value, PCRE_ERROR_BADUTF8_OFFSET is returned.
2101
2102 If you already know that your subject is valid, and you want to skip
2103 these checks for performance reasons, you can set the
2104 PCRE_NO_UTF8_CHECK option when calling pcre_exec(). You might want to
2105 do this for the second and subsequent calls to pcre_exec() if you are
2106 making repeated calls to find all the matches in a single subject
2107 string. However, you should be sure that the value of startoffset
2108 points to the start of a UTF-8 character. When PCRE_NO_UTF8_CHECK is
2109 set, the effect of passing an invalid UTF-8 string as a subject, or a
2110 value of startoffset that does not point to the start of a UTF-8 char-
2111 acter, is undefined. Your program may crash.
2112
2113 PCRE_PARTIAL_HARD
2114 PCRE_PARTIAL_SOFT
2115
2116 These options turn on the partial matching feature. For backwards com-
2117 patibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial
2118 match occurs if the end of the subject string is reached successfully,
2119 but there are not enough subject characters to complete the match. If
2120 this happens when PCRE_PARTIAL_HARD is set, pcre_exec() immediately
2121 returns PCRE_ERROR_PARTIAL. Otherwise, if PCRE_PARTIAL_SOFT is set,
2122 matching continues by testing any other alternatives. Only if they all
2123 fail is PCRE_ERROR_PARTIAL returned (instead of PCRE_ERROR_NOMATCH).
2124 The portion of the string that was inspected when the partial match was
2125 found is set as the first matching string. There is a more detailed
2126 discussion in the pcrepartial documentation.
2127
2128 The string to be matched by pcre_exec()
2129
2130 The subject string is passed to pcre_exec() as a pointer in subject, a
2131 length (in bytes) in length, and a starting byte offset in startoffset.
2132 In UTF-8 mode, the byte offset must point to the start of a UTF-8 char-
2133 acter. Unlike the pattern string, the subject may contain binary zero
2134 bytes. When the starting offset is zero, the search for a match starts
2135 at the beginning of the subject, and this is by far the most common
2136 case.
2137
2138 A non-zero starting offset is useful when searching for another match
2139 in the same subject by calling pcre_exec() again after a previous suc-
2140 cess. Setting startoffset differs from just passing over a shortened
2141 string and setting PCRE_NOTBOL in the case of a pattern that begins
2142 with any kind of lookbehind. For example, consider the pattern
2143
2144 \Biss\B
2145
2146 which finds occurrences of "iss" in the middle of words. (\B matches
2147 only if the current position in the subject is not a word boundary.)
2148 When applied to the string "Mississipi" the first call to pcre_exec()
2149 finds the first occurrence. If pcre_exec() is called again with just
2150 the remainder of the subject, namely "issipi", it does not match,
2151 because \B is always false at the start of the subject, which is deemed
2152 to be a word boundary. However, if pcre_exec() is passed the entire
2153 string again, but with startoffset set to 4, it finds the second occur-
2154 rence of "iss" because it is able to look behind the starting point to
2155 discover that it is preceded by a letter.
2156
2157 If a non-zero starting offset is passed when the pattern is anchored,
2158 one attempt to match at the given offset is made. This can only succeed
2159 if the pattern does not require the match to be at the start of the
2160 subject.
2161
2162 How pcre_exec() returns captured substrings
2163
2164 In general, a pattern matches a certain portion of the subject, and in
2165 addition, further substrings from the subject may be picked out by
2166 parts of the pattern. Following the usage in Jeffrey Friedl's book,
2167 this is called "capturing" in what follows, and the phrase "capturing
2168 subpattern" is used for a fragment of a pattern that picks out a sub-
2169 string. PCRE supports several other kinds of parenthesized subpattern
2170 that do not cause substrings to be captured.
2171
2172 Captured substrings are returned to the caller via a vector of integers
2173 whose address is passed in ovector. The number of elements in the vec-
2174 tor is passed in ovecsize, which must be a non-negative number. Note:
2175 this argument is NOT the size of ovector in bytes.
2176
2177 The first two-thirds of the vector is used to pass back captured sub-
2178 strings, each substring using a pair of integers. The remaining third
2179 of the vector is used as workspace by pcre_exec() while matching cap-
2180 turing subpatterns, and is not available for passing back information.
2181 The number passed in ovecsize should always be a multiple of three. If
2182 it is not, it is rounded down.
2183
2184 When a match is successful, information about captured substrings is
2185 returned in pairs of integers, starting at the beginning of ovector,
2186 and continuing up to two-thirds of its length at the most. The first
2187 element of each pair is set to the byte offset of the first character
2188 in a substring, and the second is set to the byte offset of the first
2189 character after the end of a substring. Note: these values are always
2190 byte offsets, even in UTF-8 mode. They are not character counts.
2191
2192 The first pair of integers, ovector[0] and ovector[1], identify the
2193 portion of the subject string matched by the entire pattern. The next
2194 pair is used for the first capturing subpattern, and so on. The value
2195 returned by pcre_exec() is one more than the highest numbered pair that
2196 has been set. For example, if two substrings have been captured, the
2197 returned value is 3. If there are no capturing subpatterns, the return
2198 value from a successful match is 1, indicating that just the first pair
2199 of offsets has been set.
2200
2201 If a capturing subpattern is matched repeatedly, it is the last portion
2202 of the string that it matched that is returned.
2203
2204 If the vector is too small to hold all the captured substring offsets,
2205 it is used as far as possible (up to two-thirds of its length), and the
2206 function returns a value of zero. If the substring offsets are not of
2207 interest, pcre_exec() may be called with ovector passed as NULL and
2208 ovecsize as zero. However, if the pattern contains back references and
2209 the ovector is not big enough to remember the related substrings, PCRE
2210 has to get additional memory for use during matching. Thus it is usu-
2211 ally advisable to supply an ovector.
2212
2213 The pcre_fullinfo() function can be used to find out how many capturing
2214 subpatterns there are in a compiled pattern. The smallest size for
2215 ovector that will allow for n captured substrings, in addition to the
2216 offsets of the substring matched by the whole pattern, is (n+1)*3.
2217
2218 It is possible for capturing subpattern number n+1 to match some part
2219 of the subject when subpattern n has not been used at all. For example,
2220 if the string "abc" is matched against the pattern (a|(z))(bc) the
2221 return from the function is 4, and subpatterns 1 and 3 are matched, but
2222 2 is not. When this happens, both values in the offset pairs corre-
2223 sponding to unused subpatterns are set to -1.
2224
2225 Offset values that correspond to unused subpatterns at the end of the
2226 expression are also set to -1. For example, if the string "abc" is
2227 matched against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not
2228 matched. The return from the function is 2, because the highest used
2229 capturing subpattern number is 1. However, you can refer to the offsets
2230 for the second and third capturing subpatterns if you wish (assuming
2231 the vector is large enough, of course).
2232
2233 Some convenience functions are provided for extracting the captured
2234 substrings as separate strings. These are described below.
2235
2236 Error return values from pcre_exec()
2237
2238 If pcre_exec() fails, it returns a negative number. The following are
2239 defined in the header file:
2240
2241 PCRE_ERROR_NOMATCH (-1)
2242
2243 The subject string did not match the pattern.
2244
2245 PCRE_ERROR_NULL (-2)
2246
2247 Either code or subject was passed as NULL, or ovector was NULL and
2248 ovecsize was not zero.
2249
2250 PCRE_ERROR_BADOPTION (-3)
2251
2252 An unrecognized bit was set in the options argument.
2253
2254 PCRE_ERROR_BADMAGIC (-4)
2255
2256 PCRE stores a 4-byte "magic number" at the start of the compiled code,
2257 to catch the case when it is passed a junk pointer and to detect when a
2258 pattern that was compiled in an environment of one endianness is run in
2259 an environment with the other endianness. This is the error that PCRE
2260 gives when the magic number is not present.
2261
2262 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2263
2264 While running the pattern match, an unknown item was encountered in the
2265 compiled pattern. This error could be caused by a bug in PCRE or by
2266 overwriting of the compiled pattern.
2267
2268 PCRE_ERROR_NOMEMORY (-6)
2269
2270 If a pattern contains back references, but the ovector that is passed
2271 to pcre_exec() is not big enough to remember the referenced substrings,
2272 PCRE gets a block of memory at the start of matching to use for this
2273 purpose. If the call via pcre_malloc() fails, this error is given. The
2274 memory is automatically freed at the end of matching.
2275
2276 PCRE_ERROR_NOSUBSTRING (-7)
2277
2278 This error is used by the pcre_copy_substring(), pcre_get_substring(),
2279 and pcre_get_substring_list() functions (see below). It is never
2280 returned by pcre_exec().
2281
2282 PCRE_ERROR_MATCHLIMIT (-8)
2283
2284 The backtracking limit, as specified by the match_limit field in a
2285 pcre_extra structure (or defaulted) was reached. See the description
2286 above.
2287
2288 PCRE_ERROR_CALLOUT (-9)
2289
2290 This error is never generated by pcre_exec() itself. It is provided for
2291 use by callout functions that want to yield a distinctive error code.
2292 See the pcrecallout documentation for details.
2293
2294 PCRE_ERROR_BADUTF8 (-10)
2295
2296 A string that contains an invalid UTF-8 byte sequence was passed as a
2297 subject.
2298
2299 PCRE_ERROR_BADUTF8_OFFSET (-11)
2300
2301 The UTF-8 byte sequence that was passed as a subject was valid, but the
2302 value of startoffset did not point to the beginning of a UTF-8 charac-
2303 ter.
2304
2305 PCRE_ERROR_PARTIAL (-12)
2306
2307 The subject string did not match, but it did match partially. See the
2308 pcrepartial documentation for details of partial matching.
2309
2310 PCRE_ERROR_BADPARTIAL (-13)
2311
2312 This code is no longer in use. It was formerly returned when the
2313 PCRE_PARTIAL option was used with a compiled pattern containing items
2314 that were not supported for partial matching. From release 8.00
2315 onwards, there are no restrictions on partial matching.
2316
2317 PCRE_ERROR_INTERNAL (-14)
2318
2319 An unexpected internal error has occurred. This error could be caused
2320 by a bug in PCRE or by overwriting of the compiled pattern.
2321
2322 PCRE_ERROR_BADCOUNT (-15)
2323
2324 This error is given if the value of the ovecsize argument is negative.
2325
2326 PCRE_ERROR_RECURSIONLIMIT (-21)
2327
2328 The internal recursion limit, as specified by the match_limit_recursion
2329 field in a pcre_extra structure (or defaulted) was reached. See the
2330 description above.
2331
2332 PCRE_ERROR_BADNEWLINE (-23)
2333
2334 An invalid combination of PCRE_NEWLINE_xxx options was given.
2335
2336 Error numbers -16 to -20 and -22 are not used by pcre_exec().
2337
2338
2339 EXTRACTING CAPTURED SUBSTRINGS BY NUMBER
2340
2341 int pcre_copy_substring(const char *subject, int *ovector,
2342 int stringcount, int stringnumber, char *buffer,
2343 int buffersize);
2344
2345 int pcre_get_substring(const char *subject, int *ovector,
2346 int stringcount, int stringnumber,
2347 const char **stringptr);
2348
2349 int pcre_get_substring_list(const char *subject,
2350 int *ovector, int stringcount, const char ***listptr);
2351
2352 Captured substrings can be accessed directly by using the offsets
2353 returned by pcre_exec() in ovector. For convenience, the functions
2354 pcre_copy_substring(), pcre_get_substring(), and pcre_get_sub-
2355 string_list() are provided for extracting captured substrings as new,
2356 separate, zero-terminated strings. These functions identify substrings
2357 by number. The next section describes functions for extracting named
2358 substrings.
2359
2360 A substring that contains a binary zero is correctly extracted and has
2361 a further zero added on the end, but the result is not, of course, a C
2362 string. However, you can process such a string by referring to the
2363 length that is returned by pcre_copy_substring() and pcre_get_sub-
2364 string(). Unfortunately, the interface to pcre_get_substring_list() is
2365 not adequate for handling strings containing binary zeros, because the
2366 end of the final string is not independently indicated.
2367
2368 The first three arguments are the same for all three of these func-
2369 tions: subject is the subject string that has just been successfully
2370 matched, ovector is a pointer to the vector of integer offsets that was
2371 passed to pcre_exec(), and stringcount is the number of substrings that
2372 were captured by the match, including the substring that matched the
2373 entire regular expression. This is the value returned by pcre_exec() if
2374 it is greater than zero. If pcre_exec() returned zero, indicating that
2375 it ran out of space in ovector, the value passed as stringcount should
2376 be the number of elements in the vector divided by three.
2377
2378 The functions pcre_copy_substring() and pcre_get_substring() extract a
2379 single substring, whose number is given as stringnumber. A value of
2380 zero extracts the substring that matched the entire pattern, whereas
2381 higher values extract the captured substrings. For pcre_copy_sub-
2382 string(), the string is placed in buffer, whose length is given by
2383 buffersize, while for pcre_get_substring() a new block of memory is
2384 obtained via pcre_malloc, and its address is returned via stringptr.
2385 The yield of the function is the length of the string, not including
2386 the terminating zero, or one of these error codes:
2387
2388 PCRE_ERROR_NOMEMORY (-6)
2389
2390 The buffer was too small for pcre_copy_substring(), or the attempt to
2391 get memory failed for pcre_get_substring().
2392
2393 PCRE_ERROR_NOSUBSTRING (-7)
2394
2395 There is no substring whose number is stringnumber.
2396
2397 The pcre_get_substring_list() function extracts all available sub-
2398 strings and builds a list of pointers to them. All this is done in a
2399 single block of memory that is obtained via pcre_malloc. The address of
2400 the memory block is returned via listptr, which is also the start of
2401 the list of string pointers. The end of the list is marked by a NULL
2402 pointer. The yield of the function is zero if all went well, or the
2403 error code
2404
2405 PCRE_ERROR_NOMEMORY (-6)
2406
2407 if the attempt to get the memory block failed.
2408
2409 When any of these functions encounter a substring that is unset, which
2410 can happen when capturing subpattern number n+1 matches some part of
2411 the subject, but subpattern n has not been used at all, they return an
2412 empty string. This can be distinguished from a genuine zero-length sub-
2413 string by inspecting the appropriate offset in ovector, which is nega-
2414 tive for unset substrings.
2415
2416 The two convenience functions pcre_free_substring() and pcre_free_sub-
2417 string_list() can be used to free the memory returned by a previous
2418 call of pcre_get_substring() or pcre_get_substring_list(), respec-
2419 tively. They do nothing more than call the function pointed to by
2420 pcre_free, which of course could be called directly from a C program.
2421 However, PCRE is used in some situations where it is linked via a spe-
2422 cial interface to another programming language that cannot use
2423 pcre_free directly; it is for these cases that the functions are pro-
2424 vided.
2425
2426
2427 EXTRACTING CAPTURED SUBSTRINGS BY NAME
2428
2429 int pcre_get_stringnumber(const pcre *code,
2430 const char *name);
2431
2432 int pcre_copy_named_substring(const pcre *code,
2433 const char *subject, int *ovector,
2434 int stringcount, const char *stringname,
2435 char *buffer, int buffersize);
2436
2437 int pcre_get_named_substring(const pcre *code,
2438 const char *subject, int *ovector,
2439 int stringcount, const char *stringname,
2440 const char **stringptr);
2441
2442 To extract a substring by name, you first have to find associated num-
2443 ber. For example, for this pattern
2444
2445 (a+)b(?<xxx>\d+)...
2446
2447 the number of the subpattern called "xxx" is 2. If the name is known to
2448 be unique (PCRE_DUPNAMES was not set), you can find the number from the
2449 name by calling pcre_get_stringnumber(). The first argument is the com-
2450 piled pattern, and the second is the name. The yield of the function is
2451 the subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no
2452 subpattern of that name.
2453
2454 Given the number, you can extract the substring directly, or use one of
2455 the functions described in the previous section. For convenience, there
2456 are also two functions that do the whole job.
2457
2458 Most of the arguments of pcre_copy_named_substring() and
2459 pcre_get_named_substring() are the same as those for the similarly
2460 named functions that extract by number. As these are described in the
2461 previous section, they are not re-described here. There are just two
2462 differences:
2463
2464 First, instead of a substring number, a substring name is given. Sec-
2465 ond, there is an extra argument, given at the start, which is a pointer
2466 to the compiled pattern. This is needed in order to gain access to the
2467 name-to-number translation table.
2468
2469 These functions call pcre_get_stringnumber(), and if it succeeds, they
2470 then call pcre_copy_substring() or pcre_get_substring(), as appropri-
2471 ate. NOTE: If PCRE_DUPNAMES is set and there are duplicate names, the
2472 behaviour may not be what you want (see the next section).
2473
2474 Warning: If the pattern uses the (?| feature to set up multiple subpat-
2475 terns with the same number, as described in the section on duplicate
2476 subpattern numbers in the pcrepattern page, you cannot use names to
2477 distinguish the different subpatterns, because names are not included
2478 in the compiled code. The matching process uses only numbers. For this
2479 reason, the use of different names for subpatterns of the same number
2480 causes an error at compile time.
2481
2482
2483 DUPLICATE SUBPATTERN NAMES
2484
2485 int pcre_get_stringtable_entries(const pcre *code,
2486 const char *name, char **first, char **last);
2487
2488 When a pattern is compiled with the PCRE_DUPNAMES option, names for
2489 subpatterns are not required to be unique. (Duplicate names are always
2490 allowed for subpatterns with the same number, created by using the (?|
2491 feature. Indeed, if such subpatterns are named, they are required to
2492 use the same names.)
2493
2494 Normally, patterns with duplicate names are such that in any one match,
2495 only one of the named subpatterns participates. An example is shown in
2496 the pcrepattern documentation.
2497
2498 When duplicates are present, pcre_copy_named_substring() and
2499 pcre_get_named_substring() return the first substring corresponding to
2500 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING
2501 (-7) is returned; no data is returned. The pcre_get_stringnumber()
2502 function returns one of the numbers that are associated with the name,
2503 but it is not defined which it is.
2504
2505 If you want to get full details of all captured substrings for a given
2506 name, you must use the pcre_get_stringtable_entries() function. The
2507 first argument is the compiled pattern, and the second is the name. The
2508 third and fourth are pointers to variables which are updated by the
2509 function. After it has run, they point to the first and last entries in
2510 the name-to-number table for the given name. The function itself
2511 returns the length of each entry, or PCRE_ERROR_NOSUBSTRING (-7) if
2512 there are none. The format of the table is described above in the sec-
2513 tion entitled Information about a pattern. Given all the relevant
2514 entries for the name, you can extract each of their numbers, and hence
2515 the captured data, if any.
2516
2517
2518 FINDING ALL POSSIBLE MATCHES
2519
2520 The traditional matching function uses a similar algorithm to Perl,
2521 which stops when it finds the first match, starting at a given point in
2522 the subject. If you want to find all possible matches, or the longest
2523 possible match, consider using the alternative matching function (see
2524 below) instead. If you cannot use the alternative function, but still
2525 need to find all possible matches, you can kludge it up by making use
2526 of the callout facility, which is described in the pcrecallout documen-
2527 tation.
2528
2529 What you have to do is to insert a callout right at the end of the pat-
2530 tern. When your callout function is called, extract and save the cur-
2531 rent matched substring. Then return 1, which forces pcre_exec() to
2532 backtrack and try other alternatives. Ultimately, when it runs out of
2533 matches, pcre_exec() will yield PCRE_ERROR_NOMATCH.
2534
2535
2536 MATCHING A PATTERN: THE ALTERNATIVE FUNCTION
2537
2538 int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
2539 const char *subject, int length, int startoffset,
2540 int options, int *ovector, int ovecsize,
2541 int *workspace, int wscount);
2542
2543 The function pcre_dfa_exec() is called to match a subject string
2544 against a compiled pattern, using a matching algorithm that scans the
2545 subject string just once, and does not backtrack. This has different
2546 characteristics to the normal algorithm, and is not compatible with
2547 Perl. Some of the features of PCRE patterns are not supported. Never-
2548 theless, there are times when this kind of matching can be useful. For
2549 a discussion of the two matching algorithms, and a list of features
2550 that pcre_dfa_exec() does not support, see the pcrematching documenta-
2551 tion.
2552
2553 The arguments for the pcre_dfa_exec() function are the same as for
2554 pcre_exec(), plus two extras. The ovector argument is used in a differ-
2555 ent way, and this is described below. The other common arguments are
2556 used in the same way as for pcre_exec(), so their description is not
2557 repeated here.
2558
2559 The two additional arguments provide workspace for the function. The
2560 workspace vector should contain at least 20 elements. It is used for
2561 keeping track of multiple paths through the pattern tree. More
2562 workspace will be needed for patterns and subjects where there are a
2563 lot of potential matches.
2564
2565 Here is an example of a simple call to pcre_dfa_exec():
2566
2567 int rc;
2568 int ovector[10];
2569 int wspace[20];
2570 rc = pcre_dfa_exec(
2571 re, /* result of pcre_compile() */
2572 NULL, /* we didn't study the pattern */
2573 "some string", /* the subject string */
2574 11, /* the length of the subject string */
2575 0, /* start at offset 0 in the subject */
2576 0, /* default options */
2577 ovector, /* vector of integers for substring information */
2578 10, /* number of elements (NOT size in bytes) */
2579 wspace, /* working space vector */
2580 20); /* number of elements (NOT size in bytes) */
2581
2582 Option bits for pcre_dfa_exec()
2583
2584 The unused bits of the options argument for pcre_dfa_exec() must be
2585 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEW-
2586 LINE_xxx, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
2587 PCRE_NOTEMPTY_ATSTART, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, PCRE_PAR-
2588 TIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART. All but the last
2589 four of these are exactly the same as for pcre_exec(), so their
2590 description is not repeated here.
2591
2592 PCRE_PARTIAL_HARD
2593 PCRE_PARTIAL_SOFT
2594
2595 These have the same general effect as they do for pcre_exec(), but the
2596 details are slightly different. When PCRE_PARTIAL_HARD is set for
2597 pcre_dfa_exec(), it returns PCRE_ERROR_PARTIAL if the end of the sub-
2598 ject is reached and there is still at least one matching possibility
2599 that requires additional characters. This happens even if some complete
2600 matches have also been found. When PCRE_PARTIAL_SOFT is set, the return
2601 code PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if the end
2602 of the subject is reached, there have been no complete matches, but
2603 there is still at least one matching possibility. The portion of the
2604 string that was inspected when the longest partial match was found is
2605 set as the first matching string in both cases.
2606
2607 PCRE_DFA_SHORTEST
2608
2609 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to
2610 stop as soon as it has found one match. Because of the way the alterna-
2611 tive algorithm works, this is necessarily the shortest possible match
2612 at the first possible matching point in the subject string.
2613
2614 PCRE_DFA_RESTART
2615
2616 When pcre_dfa_exec() returns a partial match, it is possible to call it
2617 again, with additional subject characters, and have it continue with
2618 the same match. The PCRE_DFA_RESTART option requests this action; when
2619 it is set, the workspace and wscount options must reference the same
2620 vector as before because data about the match so far is left in them
2621 after a partial match. There is more discussion of this facility in the
2622 pcrepartial documentation.
2623
2624 Successful returns from pcre_dfa_exec()
2625
2626 When pcre_dfa_exec() succeeds, it may have matched more than one sub-
2627 string in the subject. Note, however, that all the matches from one run
2628 of the function start at the same point in the subject. The shorter
2629 matches are all initial substrings of the longer matches. For example,
2630 if the pattern
2631
2632 <.*>
2633
2634 is matched against the string
2635
2636 This is <something> <something else> <something further> no more
2637
2638 the three matched strings are
2639
2640 <something>
2641 <something> <something else>
2642 <something> <something else> <something further>
2643
2644 On success, the yield of the function is a number greater than zero,
2645 which is the number of matched substrings. The substrings themselves
2646 are returned in ovector. Each string uses two elements; the first is
2647 the offset to the start, and the second is the offset to the end. In
2648 fact, all the strings have the same start offset. (Space could have
2649 been saved by giving this only once, but it was decided to retain some
2650 compatibility with the way pcre_exec() returns data, even though the
2651 meaning of the strings is different.)
2652
2653 The strings are returned in reverse order of length; that is, the long-
2654 est matching string is given first. If there were too many matches to
2655 fit into ovector, the yield of the function is zero, and the vector is
2656 filled with the longest matches.
2657
2658 Error returns from pcre_dfa_exec()
2659
2660 The pcre_dfa_exec() function returns a negative number when it fails.
2661 Many of the errors are the same as for pcre_exec(), and these are
2662 described above. There are in addition the following errors that are
2663 specific to pcre_dfa_exec():
2664
2665 PCRE_ERROR_DFA_UITEM (-16)
2666
2667 This return is given if pcre_dfa_exec() encounters an item in the pat-
2668 tern that it does not support, for instance, the use of \C or a back
2669 reference.
2670
2671 PCRE_ERROR_DFA_UCOND (-17)
2672
2673 This return is given if pcre_dfa_exec() encounters a condition item
2674 that uses a back reference for the condition, or a test for recursion
2675 in a specific group. These are not supported.
2676
2677 PCRE_ERROR_DFA_UMLIMIT (-18)
2678
2679 This return is given if pcre_dfa_exec() is called with an extra block
2680 that contains a setting of the match_limit field. This is not supported
2681 (it is meaningless).
2682
2683 PCRE_ERROR_DFA_WSSIZE (-19)
2684
2685 This return is given if pcre_dfa_exec() runs out of space in the
2686 workspace vector.
2687
2688 PCRE_ERROR_DFA_RECURSE (-20)
2689
2690 When a recursive subpattern is processed, the matching function calls
2691 itself recursively, using private vectors for ovector and workspace.
2692 This error is given if the output vector is not large enough. This
2693 should be extremely rare, as a vector of size 1000 is used.
2694
2695
2696 SEE ALSO
2697
2698 pcrebuild(3), pcrecallout(3), pcrecpp(3)(3), pcrematching(3), pcrepar-
2699 tial(3), pcreposix(3), pcreprecompile(3), pcresample(3), pcrestack(3).
2700
2701
2702 AUTHOR
2703
2704 Philip Hazel
2705 University Computing Service
2706 Cambridge CB2 3QH, England.
2707
2708
2709 REVISION
2710
2711 Last updated: 03 October 2009
2712 Copyright (c) 1997-2009 University of Cambridge.
2713 ------------------------------------------------------------------------------
2714
2715
2716 PCRECALLOUT(3) PCRECALLOUT(3)
2717
2718
2719 NAME
2720 PCRE - Perl-compatible regular expressions
2721
2722
2723 PCRE CALLOUTS
2724
2725 int (*pcre_callout)(pcre_callout_block *);
2726
2727 PCRE provides a feature called "callout", which is a means of temporar-
2728 ily passing control to the caller of PCRE in the middle of pattern
2729 matching. The caller of PCRE provides an external function by putting
2730 its entry point in the global variable pcre_callout. By default, this
2731 variable contains NULL, which disables all calling out.
2732
2733 Within a regular expression, (?C) indicates the points at which the
2734 external function is to be called. Different callout points can be
2735 identified by putting a number less than 256 after the letter C. The
2736 default value is zero. For example, this pattern has two callout
2737 points:
2738
2739 (?C1)abc(?C2)def
2740
2741 If the PCRE_AUTO_CALLOUT option bit is set when pcre_compile() or
2742 pcre_compile2() is called, PCRE automatically inserts callouts, all
2743 with number 255, before each item in the pattern. For example, if
2744 PCRE_AUTO_CALLOUT is used with the pattern
2745
2746 A(\d{2}|--)
2747
2748 it is processed as if it were
2749
2750 (?C255)A(?C255)((?C255)\d{2}(?C255)|(?C255)-(?C255)-(?C255))(?C255)
2751
2752 Notice that there is a callout before and after each parenthesis and
2753 alternation bar. Automatic callouts can be used for tracking the
2754 progress of pattern matching. The pcretest command has an option that
2755 sets automatic callouts; when it is used, the output indicates how the
2756 pattern is matched. This is useful information when you are trying to
2757 optimize the performance of a particular pattern.
2758
2759
2760 MISSING CALLOUTS
2761
2762 You should be aware that, because of optimizations in the way PCRE
2763 matches patterns by default, callouts sometimes do not happen. For
2764 example, if the pattern is
2765
2766 ab(?C4)cd
2767
2768 PCRE knows that any matching string must contain the letter "d". If the
2769 subject string is "abyz", the lack of "d" means that matching doesn't
2770 ever start, and the callout is never reached. However, with "abyd",
2771 though the result is still no match, the callout is obeyed.
2772
2773 If the pattern is studied, PCRE knows the minimum length of a matching
2774 string, and will immediately give a "no match" return without actually
2775 running a match if the subject is not long enough, or, for unanchored
2776 patterns, if it has been scanned far enough.
2777
2778 You can disable these optimizations by passing the PCRE_NO_START_OPTI-
2779 MIZE option to pcre_exec() or pcre_dfa_exec(). This slows down the
2780 matching process, but does ensure that callouts such as the example
2781 above are obeyed.
2782
2783
2784 THE CALLOUT INTERFACE
2785
2786 During matching, when PCRE reaches a callout point, the external func-
2787 tion defined by pcre_callout is called (if it is set). This applies to
2788 both the pcre_exec() and the pcre_dfa_exec() matching functions. The
2789 only argument to the callout function is a pointer to a pcre_callout
2790 block. This structure contains the following fields:
2791
2792 int version;
2793 int callout_number;
2794 int *offset_vector;
2795 const char *subject;
2796 int subject_length;
2797 int start_match;
2798 int current_position;
2799 int capture_top;
2800 int capture_last;
2801 void *callout_data;
2802 int pattern_position;
2803 int next_item_length;
2804
2805 The version field is an integer containing the version number of the
2806 block format. The initial version was 0; the current version is 1. The
2807 version number will change again in future if additional fields are
2808 added, but the intention is never to remove any of the existing fields.
2809
2810 The callout_number field contains the number of the callout, as com-
2811 piled into the pattern (that is, the number after ?C for manual call-
2812 outs, and 255 for automatically generated callouts).
2813
2814 The offset_vector field is a pointer to the vector of offsets that was
2815 passed by the caller to pcre_exec() or pcre_dfa_exec(). When
2816 pcre_exec() is used, the contents can be inspected in order to extract
2817 substrings that have been matched so far, in the same way as for
2818 extracting substrings after a match has completed. For pcre_dfa_exec()
2819 this field is not useful.
2820
2821 The subject and subject_length fields contain copies of the values that
2822 were passed to pcre_exec().
2823
2824 The start_match field normally contains the offset within the subject
2825 at which the current match attempt started. However, if the escape
2826 sequence \K has been encountered, this value is changed to reflect the
2827 modified starting point. If the pattern is not anchored, the callout
2828 function may be called several times from the same point in the pattern
2829 for different starting points in the subject.
2830
2831 The current_position field contains the offset within the subject of
2832 the current match pointer.
2833
2834 When the pcre_exec() function is used, the capture_top field contains
2835 one more than the number of the highest numbered captured substring so
2836 far. If no substrings have been captured, the value of capture_top is
2837 one. This is always the case when pcre_dfa_exec() is used, because it
2838 does not support captured substrings.
2839
2840 The capture_last field contains the number of the most recently cap-
2841 tured substring. If no substrings have been captured, its value is -1.
2842 This is always the case when pcre_dfa_exec() is used.
2843
2844 The callout_data field contains a value that is passed to pcre_exec()
2845 or pcre_dfa_exec() specifically so that it can be passed back in call-
2846 outs. It is passed in the pcre_callout field of the pcre_extra data
2847 structure. If no such data was passed, the value of callout_data in a
2848 pcre_callout block is NULL. There is a description of the pcre_extra
2849 structure in the pcreapi documentation.
2850
2851 The pattern_position field is present from version 1 of the pcre_call-
2852 out structure. It contains the offset to the next item to be matched in
2853 the pattern string.
2854
2855 The next_item_length field is present from version 1 of the pcre_call-
2856 out structure. It contains the length of the next item to be matched in
2857 the pattern string. When the callout immediately precedes an alterna-
2858 tion bar, a closing parenthesis, or the end of the pattern, the length
2859 is zero. When the callout precedes an opening parenthesis, the length
2860 is that of the entire subpattern.
2861
2862 The pattern_position and next_item_length fields are intended to help
2863 in distinguishing between different automatic callouts, which all have
2864 the same callout number. However, they are set for all callouts.
2865
2866
2867 RETURN VALUES
2868
2869 The external callout function returns an integer to PCRE. If the value
2870 is zero, matching proceeds as normal. If the value is greater than
2871 zero, matching fails at the current point, but the testing of other
2872 matching possibilities goes ahead, just as if a lookahead assertion had
2873 failed. If the value is less than zero, the match is abandoned, and
2874 pcre_exec() or pcre_dfa_exec() returns the negative value.
2875
2876 Negative values should normally be chosen from the set of
2877 PCRE_ERROR_xxx values. In particular, PCRE_ERROR_NOMATCH forces a stan-
2878 dard "no match" failure. The error number PCRE_ERROR_CALLOUT is
2879 reserved for use by callout functions; it will never be used by PCRE
2880 itself.
2881
2882
2883 AUTHOR
2884
2885 Philip Hazel
2886 University Computing Service
2887 Cambridge CB2 3QH, England.
2888
2889
2890 REVISION
2891
2892 Last updated: 29 September 2009
2893 Copyright (c) 1997-2009 University of Cambridge.
2894 ------------------------------------------------------------------------------
2895
2896
2897 PCRECOMPAT(3) PCRECOMPAT(3)
2898
2899
2900 NAME
2901 PCRE - Perl-compatible regular expressions
2902
2903
2904 DIFFERENCES BETWEEN PCRE AND PERL
2905
2906 This document describes the differences in the ways that PCRE and Perl
2907 handle regular expressions. The differences described here are with
2908 respect to Perl 5.10.
2909
2910 1. PCRE has only a subset of Perl's UTF-8 and Unicode support. Details
2911 of what it does have are given in the section on UTF-8 support in the
2912 main pcre page.
2913
2914 2. PCRE does not allow repeat quantifiers on lookahead assertions. Perl
2915 permits them, but they do not mean what you might think. For example,
2916 (?!a){3} does not assert that the next three characters are not "a". It
2917 just asserts that the next character is not "a" three times.
2918
2919 3. Capturing subpatterns that occur inside negative lookahead asser-
2920 tions are counted, but their entries in the offsets vector are never
2921 set. Perl sets its numerical variables from any such patterns that are
2922 matched before the assertion fails to match something (thereby succeed-
2923 ing), but only if the negative lookahead assertion contains just one
2924 branch.
2925
2926 4. Though binary zero characters are supported in the subject string,
2927 they are not allowed in a pattern string because it is passed as a nor-
2928 mal C string, terminated by zero. The escape sequence \0 can be used in
2929 the pattern to represent a binary zero.
2930
2931 5. The following Perl escape sequences are not supported: \l, \u, \L,
2932 \U, and \N. In fact these are implemented by Perl's general string-han-
2933 dling and are not part of its pattern matching engine. If any of these
2934 are encountered by PCRE, an error is generated.
2935
2936 6. The Perl escape sequences \p, \P, and \X are supported only if PCRE
2937 is built with Unicode character property support. The properties that
2938 can be tested with \p and \P are limited to the general category prop-
2939 erties such as Lu and Nd, script names such as Greek or Han, and the
2940 derived properties Any and L&. PCRE does support the Cs (surrogate)
2941 property, which Perl does not; the Perl documentation says "Because
2942 Perl hides the need for the user to understand the internal representa-
2943 tion of Unicode characters, there is no need to implement the somewhat
2944 messy concept of surrogates."
2945
2946 7. PCRE does support the \Q...\E escape for quoting substrings. Charac-
2947 ters in between are treated as literals. This is slightly different
2948 from Perl in that $ and @ are also handled as literals inside the
2949 quotes. In Perl, they cause variable interpolation (but of course PCRE
2950 does not have variables). Note the following examples:
2951
2952 Pattern PCRE matches Perl matches
2953
2954 \Qabc$xyz\E abc$xyz abc followed by the
2955 contents of $xyz
2956 \Qabc\$xyz\E abc\$xyz abc\$xyz
2957 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2958
2959 The \Q...\E sequence is recognized both inside and outside character
2960 classes.
2961
2962 8. Fairly obviously, PCRE does not support the (?{code}) and (??{code})
2963 constructions. However, there is support for recursive patterns. This
2964 is not available in Perl 5.8, but it is in Perl 5.10. Also, the PCRE
2965 "callout" feature allows an external function to be called during pat-
2966 tern matching. See the pcrecallout documentation for details.
2967
2968 9. Subpatterns that are called recursively or as "subroutines" are
2969 always treated as atomic groups in PCRE. This is like Python, but
2970 unlike Perl. There is a discussion of an example that explains this in
2971 more detail in the section on recursion differences from Perl in the
2972 pcrepattern page.
2973
2974 10. There are some differences that are concerned with the settings of
2975 captured strings when part of a pattern is repeated. For example,
2976 matching "aba" against the pattern /^(a(b)?)+$/ in Perl leaves $2
2977 unset, but in PCRE it is set to "b".
2978
2979 11. PCRE does support Perl 5.10's backtracking verbs (*ACCEPT),
2980 (*FAIL), (*F), (*COMMIT), (*PRUNE), (*SKIP), and (*THEN), but only in
2981 the forms without an argument. PCRE does not support (*MARK).
2982
2983 12. PCRE's handling of duplicate subpattern numbers and duplicate sub-
2984 pattern names is not as general as Perl's. This is a consequence of the
2985 fact the PCRE works internally just with numbers, using an external ta-
2986 ble to translate between numbers and names. In particular, a pattern
2987 such as (?|(?<a>A)|(?<b)B), where the two capturing parentheses have
2988 the same number but different names, is not supported, and causes an
2989 error at compile time. If it were allowed, it would not be possible to
2990 distinguish which parentheses matched, because both names map to cap-
2991 turing subpattern number 1. To avoid this confusing situation, an error
2992 is given at compile time.
2993
2994 13. PCRE provides some extensions to the Perl regular expression facil-
2995 ities. Perl 5.10 includes new features that are not in earlier ver-
2996 sions of Perl, some of which (such as named parentheses) have been in
2997 PCRE for some time. This list is with respect to Perl 5.10:
2998
2999 (a) Although lookbehind assertions in PCRE must match fixed length
3000 strings, each alternative branch of a lookbehind assertion can match a
3001 different length of string. Perl requires them all to have the same
3002 length.
3003
3004 (b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $
3005 meta-character matches only at the very end of the string.
3006
3007 (c) If PCRE_EXTRA is set, a backslash followed by a letter with no spe-
3008 cial meaning is faulted. Otherwise, like Perl, the backslash is quietly
3009 ignored. (Perl can be made to issue a warning.)
3010
3011 (d) If PCRE_UNGREEDY is set, the greediness of the repetition quanti-
3012 fiers is inverted, that is, by default they are not greedy, but if fol-
3013 lowed by a question mark they are.
3014
3015 (e) PCRE_ANCHORED can be used at matching time to force a pattern to be
3016 tried only at the first matching position in the subject string.
3017
3018 (f) The PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
3019 and PCRE_NO_AUTO_CAPTURE options for pcre_exec() have no Perl equiva-
3020 lents.
3021
3022 (g) The \R escape sequence can be restricted to match only CR, LF, or
3023 CRLF by the PCRE_BSR_ANYCRLF option.
3024
3025 (h) The callout facility is PCRE-specific.
3026
3027 (i) The partial matching facility is PCRE-specific.
3028
3029 (j) Patterns compiled by PCRE can be saved and re-used at a later time,
3030 even on different hosts that have the other endianness.
3031
3032 (k) The alternative matching function (pcre_dfa_exec()) matches in a
3033 different way and is not Perl-compatible.
3034
3035 (l) PCRE recognizes some special sequences such as (*CR) at the start
3036 of a pattern that set overall options that cannot be changed within the
3037 pattern.
3038
3039
3040 AUTHOR
3041
3042 Philip Hazel
3043 University Computing Service
3044 Cambridge CB2 3QH, England.
3045
3046
3047 REVISION
3048
3049 Last updated: 04 October 2009
3050 Copyright (c) 1997-2009 University of Cambridge.
3051 ------------------------------------------------------------------------------
3052
3053
3054 PCREPATTERN(3) PCREPATTERN(3)
3055
3056
3057 NAME
3058 PCRE - Perl-compatible regular expressions
3059
3060
3061 PCRE REGULAR EXPRESSION DETAILS
3062
3063 The syntax and semantics of the regular expressions that are supported
3064 by PCRE are described in detail below. There is a quick-reference syn-
3065 tax summary in the pcresyntax page. PCRE tries to match Perl syntax and
3066 semantics as closely as it can. PCRE also supports some alternative
3067 regular expression syntax (which does not conflict with the Perl syn-
3068 tax) in order to provide some compatibility with regular expressions in
3069 Python, .NET, and Oniguruma.
3070
3071 Perl's regular expressions are described in its own documentation, and
3072 regular expressions in general are covered in a number of books, some
3073 of which have copious examples. Jeffrey Friedl's "Mastering Regular
3074 Expressions", published by O'Reilly, covers regular expressions in
3075 great detail. This description of PCRE's regular expressions is
3076 intended as reference material.
3077
3078 The original operation of PCRE was on strings of one-byte characters.
3079 However, there is now also support for UTF-8 character strings. To use
3080 this, PCRE must be built to include UTF-8 support, and you must call
3081 pcre_compile() or pcre_compile2() with the PCRE_UTF8 option. There is
3082 also a special sequence that can be given at the start of a pattern:
3083
3084 (*UTF8)
3085
3086 Starting a pattern with this sequence is equivalent to setting the
3087 PCRE_UTF8 option. This feature is not Perl-compatible. How setting
3088 UTF-8 mode affects pattern matching is mentioned in several places
3089 below. There is also a summary of UTF-8 features in the section on
3090 UTF-8 support in the main pcre page.
3091
3092 The remainder of this document discusses the patterns that are sup-
3093 ported by PCRE when its main matching function, pcre_exec(), is used.
3094 From release 6.0, PCRE offers a second matching function,
3095 pcre_dfa_exec(), which matches using a different algorithm that is not
3096 Perl-compatible. Some of the features discussed below are not available
3097 when pcre_dfa_exec() is used. The advantages and disadvantages of the
3098 alternative function, and how it differs from the normal function, are
3099 discussed in the pcrematching page.
3100
3101
3102 NEWLINE CONVENTIONS
3103
3104 PCRE supports five different conventions for indicating line breaks in
3105 strings: a single CR (carriage return) character, a single LF (line-
3106 feed) character, the two-character sequence CRLF, any of the three pre-
3107 ceding, or any Unicode newline sequence. The pcreapi page has further
3108 discussion about newlines, and shows how to set the newline convention
3109 in the options arguments for the compiling and matching functions.
3110
3111 It is also possible to specify a newline convention by starting a pat-
3112 tern string with one of the following five sequences:
3113
3114 (*CR) carriage return
3115 (*LF) linefeed
3116 (*CRLF) carriage return, followed by linefeed
3117 (*ANYCRLF) any of the three above
3118 (*ANY) all Unicode newline sequences
3119
3120 These override the default and the options given to pcre_compile() or
3121 pcre_compile2(). For example, on a Unix system where LF is the default
3122 newline sequence, the pattern
3123
3124 (*CR)a.b
3125
3126 changes the convention to CR. That pattern matches "a\nb" because LF is
3127 no longer a newline. Note that these special settings, which are not
3128 Perl-compatible, are recognized only at the very start of a pattern,
3129 and that they must be in upper case. If more than one of them is
3130 present, the last one is used.
3131
3132 The newline convention does not affect what the \R escape sequence
3133 matches. By default, this is any Unicode newline sequence, for Perl
3134 compatibility. However, this can be changed; see the description of \R
3135 in the section entitled "Newline sequences" below. A change of \R set-
3136 ting can be combined with a change of newline convention.
3137
3138
3139 CHARACTERS AND METACHARACTERS
3140
3141 A regular expression is a pattern that is matched against a subject
3142 string from left to right. Most characters stand for themselves in a
3143 pattern, and match the corresponding characters in the subject. As a
3144 trivial example, the pattern
3145
3146 The quick brown fox
3147
3148 matches a portion of a subject string that is identical to itself. When
3149 caseless matching is specified (the PCRE_CASELESS option), letters are
3150 matched independently of case. In UTF-8 mode, PCRE always understands
3151 the concept of case for characters whose values are less than 128, so
3152 caseless matching is always possible. For characters with higher val-
3153 ues, the concept of case is supported if PCRE is compiled with Unicode
3154 property support, but not otherwise. If you want to use caseless
3155 matching for characters 128 and above, you must ensure that PCRE is
3156 compiled with Unicode property support as well as with UTF-8 support.
3157
3158 The power of regular expressions comes from the ability to include
3159 alternatives and repetitions in the pattern. These are encoded in the
3160 pattern by the use of metacharacters, which do not stand for themselves
3161 but instead are interpreted in some special way.
3162
3163 There are two different sets of metacharacters: those that are recog-
3164 nized anywhere in the pattern except within square brackets, and those
3165 that are recognized within square brackets. Outside square brackets,
3166 the metacharacters are as follows:
3167
3168 \ general escape character with several uses
3169 ^ assert start of string (or line, in multiline mode)
3170 $ assert end of string (or line, in multiline mode)
3171 . match any character except newline (by default)
3172 [ start character class definition
3173 | start of alternative branch
3174 ( start subpattern
3175 ) end subpattern
3176 ? extends the meaning of (
3177 also 0 or 1 quantifier
3178 also quantifier minimizer
3179 * 0 or more quantifier
3180 + 1 or more quantifier
3181 also "possessive quantifier"
3182 { start min/max quantifier
3183
3184 Part of a pattern that is in square brackets is called a "character
3185 class". In a character class the only metacharacters are:
3186
3187 \ general escape character
3188 ^ negate the class, but only if the first character
3189 - indicates character range
3190 [ POSIX character class (only if followed by POSIX
3191 syntax)
3192 ] terminates the character class
3193
3194 The following sections describe the use of each of the metacharacters.
3195
3196
3197 BACKSLASH
3198
3199 The backslash character has several uses. Firstly, if it is followed by
3200 a non-alphanumeric character, it takes away any special meaning that
3201 character may have. This use of backslash as an escape character
3202 applies both inside and outside character classes.
3203
3204 For example, if you want to match a * character, you write \* in the
3205 pattern. This escaping action applies whether or not the following
3206 character would otherwise be interpreted as a metacharacter, so it is
3207 always safe to precede a non-alphanumeric with backslash to specify
3208 that it stands for itself. In particular, if you want to match a back-
3209 slash, you write \\.
3210
3211 If a pattern is compiled with the PCRE_EXTENDED option, whitespace in
3212 the pattern (other than in a character class) and characters between a
3213 # outside a character class and the next newline are ignored. An escap-
3214 ing backslash can be used to include a whitespace or # character as
3215 part of the pattern.
3216
3217 If you want to remove the special meaning from a sequence of charac-
3218 ters, you can do so by putting them between \Q and \E. This is differ-
3219 ent from Perl in that $ and @ are handled as literals in \Q...\E
3220 sequences in PCRE, whereas in Perl, $ and @ cause variable interpola-
3221 tion. Note the following examples:
3222
3223 Pattern PCRE matches Perl matches
3224
3225 \Qabc$xyz\E abc$xyz abc followed by the
3226 contents of $xyz
3227 \Qabc\$xyz\E abc\$xyz abc\$xyz
3228 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3229
3230 The \Q...\E sequence is recognized both inside and outside character
3231 classes.
3232
3233 Non-printing characters
3234
3235 A second use of backslash provides a way of encoding non-printing char-
3236 acters in patterns in a visible manner. There is no restriction on the
3237 appearance of non-printing characters, apart from the binary zero that
3238 terminates a pattern, but when a pattern is being prepared by text
3239 editing, it is often easier to use one of the following escape
3240 sequences than the binary character it represents:
3241
3242 \a alarm, that is, the BEL character (hex 07)
3243 \cx "control-x", where x is any character
3244 \e escape (hex 1B)
3245 \f formfeed (hex 0C)
3246 \n linefeed (hex 0A)
3247 \r carriage return (hex 0D)
3248 \t tab (hex 09)
3249 \ddd character with octal code ddd, or backreference
3250 \xhh character with hex code hh
3251 \x{hhh..} character with hex code hhh..
3252
3253 The precise effect of \cx is as follows: if x is a lower case letter,
3254 it is converted to upper case. Then bit 6 of the character (hex 40) is
3255 inverted. Thus \cz becomes hex 1A, but \c{ becomes hex 3B, while \c;
3256 becomes hex 7B.
3257
3258 After \x, from zero to two hexadecimal digits are read (letters can be
3259 in upper or lower case). Any number of hexadecimal digits may appear
3260 between \x{ and }, but the value of the character code must be less
3261 than 256 in non-UTF-8 mode, and less than 2**31 in UTF-8 mode. That is,
3262 the maximum value in hexadecimal is 7FFFFFFF. Note that this is bigger
3263 than the largest Unicode code point, which is 10FFFF.
3264
3265 If characters other than hexadecimal digits appear between \x{ and },
3266 or if there is no terminating }, this form of escape is not recognized.
3267 Instead, the initial \x will be interpreted as a basic hexadecimal
3268 escape, with no following digits, giving a character whose value is
3269 zero.
3270
3271 Characters whose value is less than 256 can be defined by either of the
3272 two syntaxes for \x. There is no difference in the way they are han-
3273 dled. For example, \xdc is exactly the same as \x{dc}.
3274
3275 After \0 up to two further octal digits are read. If there are fewer
3276 than two digits, just those that are present are used. Thus the
3277 sequence \0\x\07 specifies two binary zeros followed by a BEL character
3278 (code value 7). Make sure you supply two digits after the initial zero
3279 if the pattern character that follows is itself an octal digit.
3280
3281 The handling of a backslash followed by a digit other than 0 is compli-
3282 cated. Outside a character class, PCRE reads it and any following dig-
3283 its as a decimal number. If the number is less than 10, or if there
3284 have been at least that many previous capturing left parentheses in the
3285 expression, the entire sequence is taken as a back reference. A
3286 description of how this works is given later, following the discussion
3287 of parenthesized subpatterns.
3288
3289 Inside a character class, or if the decimal number is greater than 9
3290 and there have not been that many capturing subpatterns, PCRE re-reads
3291 up to three octal digits following the backslash, and uses them to gen-
3292 erate a data character. Any subsequent digits stand for themselves. In
3293 non-UTF-8 mode, the value of a character specified in octal must be
3294 less than \400. In UTF-8 mode, values up to \777 are permitted. For
3295 example:
3296
3297 \040 is another way of writing a space
3298 \40 is the same, provided there are fewer than 40
3299 previous capturing subpatterns
3300 \7 is always a back reference
3301 \11 might be a back reference, or another way of
3302 writing a tab
3303 \011 is always a tab
3304 \0113 is a tab followed by the character "3"
3305 \113 might be a back reference, otherwise the
3306 character with octal code 113
3307 \377 might be a back reference, otherwise
3308 the byte consisting entirely of 1 bits
3309 \81 is either a back reference, or a binary zero
3310 followed by the two characters "8" and "1"
3311
3312 Note that octal values of 100 or greater must not be introduced by a
3313 leading zero, because no more than three octal digits are ever read.
3314
3315 All the sequences that define a single character value can be used both
3316 inside and outside character classes. In addition, inside a character
3317 class, the sequence \b is interpreted as the backspace character (hex
3318 08), and the sequences \R and \X are interpreted as the characters "R"
3319 and "X", respectively. Outside a character class, these sequences have
3320 different meanings (see below).
3321
3322 Absolute and relative back references
3323
3324 The sequence \g followed by an unsigned or a negative number, option-
3325 ally enclosed in braces, is an absolute or relative back reference. A
3326 named back reference can be coded as \g{name}. Back references are dis-
3327 cussed later, following the discussion of parenthesized subpatterns.
3328
3329 Absolute and relative subroutine calls
3330
3331 For compatibility with Oniguruma, the non-Perl syntax \g followed by a
3332 name or a number enclosed either in angle brackets or single quotes, is
3333 an alternative syntax for referencing a subpattern as a "subroutine".
3334 Details are discussed later. Note that \g{...} (Perl syntax) and
3335 \g<...> (Oniguruma syntax) are not synonymous. The former is a back
3336 reference; the latter is a subroutine call.
3337
3338 Generic character types
3339
3340 Another use of backslash is for specifying generic character types. The
3341 following are always recognized:
3342
3343 \d any decimal digit
3344 \D any character that is not a decimal digit
3345 \h any horizontal whitespace character
3346 \H any character that is not a horizontal whitespace character
3347 \s any whitespace character
3348 \S any character that is not a whitespace character
3349 \v any vertical whitespace character
3350 \V any character that is not a vertical whitespace character
3351 \w any "word" character
3352 \W any "non-word" character
3353
3354 Each pair of escape sequences partitions the complete set of characters
3355 into two disjoint sets. Any given character matches one, and only one,
3356 of each pair.
3357
3358 These character type sequences can appear both inside and outside char-
3359 acter classes. They each match one character of the appropriate type.
3360 If the current matching point is at the end of the subject string, all
3361 of them fail, since there is no character to match.
3362
3363 For compatibility with Perl, \s does not match the VT character (code
3364 11). This makes it different from the the POSIX "space" class. The \s
3365 characters are HT (9), LF (10), FF (12), CR (13), and space (32). If
3366 "use locale;" is included in a Perl script, \s may match the VT charac-
3367 ter. In PCRE, it never does.
3368
3369 In UTF-8 mode, characters with values greater than 128 never match \d,
3370 \s, or \w, and always match \D, \S, and \W. This is true even when Uni-
3371 code character property support is available. These sequences retain
3372 their original meanings from before UTF-8 support was available, mainly
3373 for efficiency reasons. Note that this also affects \b, because it is
3374 defined in terms of \w and \W.
3375
3376 The sequences \h, \H, \v, and \V are Perl 5.10 features. In contrast to
3377 the other sequences, these do match certain high-valued codepoints in
3378 UTF-8 mode. The horizontal space characters are:
3379
3380 U+0009 Horizontal tab
3381 U+0020 Space
3382 U+00A0 Non-break space
3383 U+1680 Ogham space mark
3384 U+180E Mongolian vowel separator
3385 U+2000 En quad
3386 U+2001 Em quad
3387 U+2002 En space
3388 U+2003 Em space
3389 U+2004 Three-per-em space
3390 U+2005 Four-per-em space
3391 U+2006 Six-per-em space
3392 U+2007 Figure space
3393 U+2008 Punctuation space
3394 U+2009 Thin space
3395 U+200A Hair space
3396 U+202F Narrow no-break space
3397 U+205F Medium mathematical space
3398 U+3000 Ideographic space
3399
3400 The vertical space characters are:
3401
3402 U+000A Linefeed
3403 U+000B Vertical tab
3404 U+000C Formfeed
3405 U+000D Carriage return
3406 U+0085 Next line
3407 U+2028 Line separator
3408 U+2029 Paragraph separator
3409
3410 A "word" character is an underscore or any character less than 256 that
3411 is a letter or digit. The definition of letters and digits is con-
3412 trolled by PCRE's low-valued character tables, and may vary if locale-
3413 specific matching is taking place (see "Locale support" in the pcreapi
3414 page). For example, in a French locale such as "fr_FR" in Unix-like
3415 systems, or "french" in Windows, some character codes greater than 128
3416 are used for accented letters, and these are matched by \w. The use of
3417 locales with Unicode is discouraged.
3418
3419 Newline sequences
3420
3421 Outside a character class, by default, the escape sequence \R matches
3422 any Unicode newline sequence. This is a Perl 5.10 feature. In non-UTF-8
3423 mode \R is equivalent to the following:
3424
3425 (?>\r\n|\n|\x0b|\f|\r|\x85)
3426
3427 This is an example of an "atomic group", details of which are given
3428 below. This particular group matches either the two-character sequence
3429 CR followed by LF, or one of the single characters LF (linefeed,
3430 U+000A), VT (vertical tab, U+000B), FF (formfeed, U+000C), CR (carriage
3431 return, U+000D), or NEL (next line, U+0085). The two-character sequence
3432 is treated as a single unit that cannot be split.
3433
3434 In UTF-8 mode, two additional characters whose codepoints are greater
3435 than 255 are added: LS (line separator, U+2028) and PS (paragraph sepa-
3436 rator, U+2029). Unicode character property support is not needed for
3437 these characters to be recognized.
3438
3439 It is possible to restrict \R to match only CR, LF, or CRLF (instead of
3440 the complete set of Unicode line endings) by setting the option
3441 PCRE_BSR_ANYCRLF either at compile time or when the pattern is matched.
3442 (BSR is an abbrevation for "backslash R".) This can be made the default
3443 when PCRE is built; if this is the case, the other behaviour can be
3444 requested via the PCRE_BSR_UNICODE option. It is also possible to
3445 specify these settings by starting a pattern string with one of the
3446 following sequences:
3447
3448 (*BSR_ANYCRLF) CR, LF, or CRLF only
3449 (*BSR_UNICODE) any Unicode newline sequence
3450
3451 These override the default and the options given to pcre_compile() or
3452 pcre_compile2(), but they can be overridden by options given to
3453 pcre_exec() or pcre_dfa_exec(). Note that these special settings, which
3454 are not Perl-compatible, are recognized only at the very start of a
3455 pattern, and that they must be in upper case. If more than one of them
3456 is present, the last one is used. They can be combined with a change of
3457 newline convention, for example, a pattern can start with:
3458
3459 (*ANY)(*BSR_ANYCRLF)
3460
3461 Inside a character class, \R matches the letter "R".
3462
3463 Unicode character properties
3464
3465 When PCRE is built with Unicode character property support, three addi-
3466 tional escape sequences that match characters with specific properties
3467 are available. When not in UTF-8 mode, these sequences are of course
3468 limited to testing characters whose codepoints are less than 256, but
3469 they do work in this mode. The extra escape sequences are:
3470
3471 \p{xx} a character with the xx property
3472 \P{xx} a character without the xx property
3473 \X an extended Unicode sequence
3474
3475 The property names represented by xx above are limited to the Unicode
3476 script names, the general category properties, and "Any", which matches
3477 any character (including newline). Other properties such as "InMusical-
3478 Symbols" are not currently supported by PCRE. Note that \P{Any} does
3479 not match any characters, so always causes a match failure.
3480
3481 Sets of Unicode characters are defined as belonging to certain scripts.
3482 A character from one of these sets can be matched using a script name.
3483 For example:
3484
3485 \p{Greek}
3486 \P{Han}
3487
3488 Those that are not part of an identified script are lumped together as
3489 "Common". The current list of scripts is:
3490
3491 Arabic, Armenian, Balinese, Bengali, Bopomofo, Braille, Buginese,
3492 Buhid, Canadian_Aboriginal, Cherokee, Common, Coptic, Cuneiform,
3493 Cypriot, Cyrillic, Deseret, Devanagari, Ethiopic, Georgian, Glagolitic,
3494 Gothic, Greek, Gujarati, Gurmukhi, Han, Hangul, Hanunoo, Hebrew, Hira-
3495 gana, Inherited, Kannada, Katakana, Kharoshthi, Khmer, Lao, Latin,
3496 Limbu, Linear_B, Malayalam, Mongolian, Myanmar, New_Tai_Lue, Nko,
3497 Ogham, Old_Italic, Old_Persian, Oriya, Osmanya, Phags_Pa, Phoenician,
3498 Runic, Shavian, Sinhala, Syloti_Nagri, Syriac, Tagalog, Tagbanwa,
3499 Tai_Le, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh, Ugaritic, Yi.
3500
3501 Each character has exactly one general category property, specified by
3502 a two-letter abbreviation. For compatibility with Perl, negation can be
3503 specified by including a circumflex between the opening brace and the
3504 property name. For example, \p{^Lu} is the same as \P{Lu}.
3505
3506 If only one letter is specified with \p or \P, it includes all the gen-
3507 eral category properties that start with that letter. In this case, in
3508 the absence of negation, the curly brackets in the escape sequence are
3509 optional; these two examples have the same effect:
3510
3511 \p{L}
3512 \pL
3513
3514 The following general category property codes are supported:
3515
3516 C Other
3517 Cc Control
3518 Cf Format
3519 Cn Unassigned
3520 Co Private use
3521 Cs Surrogate
3522
3523 L Letter
3524 Ll Lower case letter
3525 Lm Modifier letter
3526 Lo Other letter
3527 Lt Title case letter
3528 Lu Upper case letter
3529
3530 M Mark
3531 Mc Spacing mark
3532 Me Enclosing mark
3533 Mn Non-spacing mark
3534
3535 N Number
3536 Nd Decimal number
3537 Nl Letter number
3538 No Other number
3539
3540 P Punctuation
3541 Pc Connector punctuation
3542 Pd Dash punctuation
3543 Pe Close punctuation
3544 Pf Final punctuation
3545 Pi Initial punctuation
3546 Po Other punctuation
3547 Ps Open punctuation
3548
3549 S Symbol
3550 Sc Currency symbol
3551 Sk Modifier symbol
3552 Sm Mathematical symbol
3553 So Other symbol
3554
3555 Z Separator
3556 Zl Line separator
3557 Zp Paragraph separator
3558 Zs Space separator
3559
3560 The special property L& is also supported: it matches a character that
3561 has the Lu, Ll, or Lt property, in other words, a letter that is not
3562 classified as a modifier or "other".
3563
3564 The Cs (Surrogate) property applies only to characters in the range
3565 U+D800 to U+DFFF. Such characters are not valid in UTF-8 strings (see
3566 RFC 3629) and so cannot be tested by PCRE, unless UTF-8 validity check-
3567 ing has been turned off (see the discussion of PCRE_NO_UTF8_CHECK in
3568 the pcreapi page). Perl does not support the Cs property.
3569
3570 The long synonyms for property names that Perl supports (such as
3571 \p{Letter}) are not supported by PCRE, nor is it permitted to prefix
3572 any of these properties with "Is".
3573
3574 No character that is in the Unicode table has the Cn (unassigned) prop-
3575 erty. Instead, this property is assumed for any code point that is not
3576 in the Unicode table.
3577
3578 Specifying caseless matching does not affect these escape sequences.
3579 For example, \p{Lu} always matches only upper case letters.
3580
3581 The \X escape matches any number of Unicode characters that form an
3582 extended Unicode sequence. \X is equivalent to
3583
3584 (?>\PM\pM*)
3585
3586 That is, it matches a character without the "mark" property, followed
3587 by zero or more characters with the "mark" property, and treats the
3588 sequence as an atomic group (see below). Characters with the "mark"
3589 property are typically accents that affect the preceding character.
3590 None of them have codepoints less than 256, so in non-UTF-8 mode \X
3591 matches any one character.
3592
3593 Matching characters by Unicode property is not fast, because PCRE has
3594 to search a structure that contains data for over fifteen thousand
3595 characters. That is why the traditional escape sequences such as \d and
3596 \w do not use Unicode properties in PCRE.
3597
3598 Resetting the match start
3599
3600 The escape sequence \K, which is a Perl 5.10 feature, causes any previ-
3601 ously matched characters not to be included in the final matched
3602 sequence. For example, the pattern:
3603
3604 foo\Kbar
3605
3606 matches "foobar", but reports that it has matched "bar". This feature
3607 is similar to a lookbehind assertion (described below). However, in
3608 this case, the part of the subject before the real match does not have
3609 to be of fixed length, as lookbehind assertions do. The use of \K does
3610 not interfere with the setting of captured substrings. For example,
3611 when the pattern
3612
3613 (foo)\Kbar
3614
3615 matches "foobar", the first substring is still set to "foo".
3616
3617 Simple assertions
3618
3619 The final use of backslash is for certain simple assertions. An asser-
3620 tion specifies a condition that has to be met at a particular point in
3621 a match, without consuming any characters from the subject string. The
3622 use of subpatterns for more complicated assertions is described below.
3623 The backslashed assertions are:
3624
3625 \b matches at a word boundary
3626 \B matches when not at a word boundary
3627 \A matches at the start of the subject
3628 \Z matches at the end of the subject
3629 also matches before a newline at the end of the subject
3630 \z matches only at the end of the subject
3631 \G matches at the first matching position in the subject
3632
3633 These assertions may not appear in character classes (but note that \b
3634 has a different meaning, namely the backspace character, inside a char-
3635 acter class).
3636
3637 A word boundary is a position in the subject string where the current
3638 character and the previous character do not both match \w or \W (i.e.
3639 one matches \w and the other matches \W), or the start or end of the
3640 string if the first or last character matches \w, respectively. Neither
3641 PCRE nor Perl has a separte "start of word" or "end of word" metase-
3642 quence. However, whatever follows \b normally determines which it is.
3643 For example, the fragment \ba matches "a" at the start of a word.
3644
3645 The \A, \Z, and \z assertions differ from the traditional circumflex
3646 and dollar (described in the next section) in that they only ever match
3647 at the very start and end of the subject string, whatever options are
3648 set. Thus, they are independent of multiline mode. These three asser-
3649 tions are not affected by the PCRE_NOTBOL or PCRE_NOTEOL options, which
3650 affect only the behaviour of the circumflex and dollar metacharacters.
3651 However, if the startoffset argument of pcre_exec() is non-zero, indi-
3652 cating that matching is to start at a point other than the beginning of
3653 the subject, \A can never match. The difference between \Z and \z is
3654 that \Z matches before a newline at the end of the string as well as at
3655 the very end, whereas \z matches only at the end.
3656
3657 The \G assertion is true only when the current matching position is at
3658 the start point of the match, as specified by the startoffset argument
3659 of pcre_exec(). It differs from \A when the value of startoffset is
3660 non-zero. By calling pcre_exec() multiple times with appropriate argu-
3661 ments, you can mimic Perl's /g option, and it is in this kind of imple-
3662 mentation where \G can be useful.
3663
3664 Note, however, that PCRE's interpretation of \G, as the start of the
3665 current match, is subtly different from Perl's, which defines it as the
3666 end of the previous match. In Perl, these can be different when the
3667 previously matched string was empty. Because PCRE does just one match
3668 at a time, it cannot reproduce this behaviour.
3669
3670 If all the alternatives of a pattern begin with \G, the expression is
3671 anchored to the starting match position, and the "anchored" flag is set
3672 in the compiled regular expression.
3673
3674
3675 CIRCUMFLEX AND DOLLAR
3676
3677 Outside a character class, in the default matching mode, the circumflex
3678 character is an assertion that is true only if the current matching
3679 point is at the start of the subject string. If the startoffset argu-
3680 ment of pcre_exec() is non-zero, circumflex can never match if the
3681 PCRE_MULTILINE option is unset. Inside a character class, circumflex
3682 has an entirely different meaning (see below).
3683
3684 Circumflex need not be the first character of the pattern if a number
3685 of alternatives are involved, but it should be the first thing in each
3686 alternative in which it appears if the pattern is ever to match that
3687 branch. If all possible alternatives start with a circumflex, that is,
3688 if the pattern is constrained to match only at the start of the sub-
3689 ject, it is said to be an "anchored" pattern. (There are also other
3690 constructs that can cause a pattern to be anchored.)
3691
3692 A dollar character is an assertion that is true only if the current
3693 matching point is at the end of the subject string, or immediately
3694 before a newline at the end of the string (by default). Dollar need not
3695 be the last character of the pattern if a number of alternatives are
3696 involved, but it should be the last item in any branch in which it
3697 appears. Dollar has no special meaning in a character class.
3698
3699 The meaning of dollar can be changed so that it matches only at the
3700 very end of the string, by setting the PCRE_DOLLAR_ENDONLY option at
3701 compile time. This does not affect the \Z assertion.
3702
3703 The meanings of the circumflex and dollar characters are changed if the
3704 PCRE_MULTILINE option is set. When this is the case, a circumflex
3705 matches immediately after internal newlines as well as at the start of
3706 the subject string. It does not match after a newline that ends the
3707 string. A dollar matches before any newlines in the string, as well as
3708 at the very end, when PCRE_MULTILINE is set. When newline is specified
3709 as the two-character sequence CRLF, isolated CR and LF characters do
3710 not indicate newlines.
3711
3712 For example, the pattern /^abc$/ matches the subject string "def\nabc"
3713 (where \n represents a newline) in multiline mode, but not otherwise.
3714 Consequently, patterns that are anchored in single line mode because
3715 all branches start with ^ are not anchored in multiline mode, and a
3716 match for circumflex is possible when the startoffset argument of
3717 pcre_exec() is non-zero. The PCRE_DOLLAR_ENDONLY option is ignored if
3718 PCRE_MULTILINE is set.
3719
3720 Note that the sequences \A, \Z, and \z can be used to match the start
3721 and end of the subject in both modes, and if all branches of a pattern
3722 start with \A it is always anchored, whether or not PCRE_MULTILINE is
3723 set.
3724
3725
3726 FULL STOP (PERIOD, DOT)
3727
3728 Outside a character class, a dot in the pattern matches any one charac-
3729 ter in the subject string except (by default) a character that signi-
3730 fies the end of a line. In UTF-8 mode, the matched character may be
3731 more than one byte long.
3732
3733 When a line ending is defined as a single character, dot never matches
3734 that character; when the two-character sequence CRLF is used, dot does
3735 not match CR if it is immediately followed by LF, but otherwise it
3736 matches all characters (including isolated CRs and LFs). When any Uni-
3737 code line endings are being recognized, dot does not match CR or LF or
3738 any of the other line ending characters.
3739
3740 The behaviour of dot with regard to newlines can be changed. If the
3741 PCRE_DOTALL option is set, a dot matches any one character, without
3742 exception. If the two-character sequence CRLF is present in the subject
3743 string, it takes two dots to match it.
3744
3745 The handling of dot is entirely independent of the handling of circum-
3746 flex and dollar, the only relationship being that they both involve
3747 newlines. Dot has no special meaning in a character class.
3748
3749
3750 MATCHING A SINGLE BYTE
3751
3752 Outside a character class, the escape sequence \C matches any one byte,
3753 both in and out of UTF-8 mode. Unlike a dot, it always matches any
3754 line-ending characters. The feature is provided in Perl in order to
3755 match individual bytes in UTF-8 mode. Because it breaks up UTF-8 char-
3756 acters into individual bytes, what remains in the string may be a mal-
3757 formed UTF-8 string. For this reason, the \C escape sequence is best
3758 avoided.
3759
3760 PCRE does not allow \C to appear in lookbehind assertions (described
3761 below), because in UTF-8 mode this would make it impossible to calcu-
3762 late the length of the lookbehind.
3763
3764
3765 SQUARE BRACKETS AND CHARACTER CLASSES
3766
3767 An opening square bracket introduces a character class, terminated by a
3768 closing square bracket. A closing square bracket on its own is not spe-
3769 cial by default. However, if the PCRE_JAVASCRIPT_COMPAT option is set,
3770 a lone closing square bracket causes a compile-time error. If a closing
3771 square bracket is required as a member of the class, it should be the
3772 first data character in the class (after an initial circumflex, if
3773 present) or escaped with a backslash.
3774
3775 A character class matches a single character in the subject. In UTF-8
3776 mode, the character may be more than one byte long. A matched character
3777 must be in the set of characters defined by the class, unless the first
3778 character in the class definition is a circumflex, in which case the
3779 subject character must not be in the set defined by the class. If a
3780 circumflex is actually required as a member of the class, ensure it is
3781 not the first character, or escape it with a backslash.
3782
3783 For example, the character class [aeiou] matches any lower case vowel,
3784 while [^aeiou] matches any character that is not a lower case vowel.
3785 Note that a circumflex is just a convenient notation for specifying the
3786 characters that are in the class by enumerating those that are not. A
3787 class that starts with a circumflex is not an assertion; it still con-
3788 sumes a character from the subject string, and therefore it fails if
3789 the current pointer is at the end of the string.
3790
3791 In UTF-8 mode, characters with values greater than 255 can be included
3792 in a class as a literal string of bytes, or by using the \x{ escaping
3793 mechanism.
3794
3795 When caseless matching is set, any letters in a class represent both
3796 their upper case and lower case versions, so for example, a caseless
3797 [aeiou] matches "A" as well as "a", and a caseless [^aeiou] does not
3798 match "A", whereas a caseful version would. In UTF-8 mode, PCRE always
3799 understands the concept of case for characters whose values are less
3800 than 128, so caseless matching is always possible. For characters with
3801 higher values, the concept of case is supported if PCRE is compiled
3802 with Unicode property support, but not otherwise. If you want to use
3803 caseless matching in UTF8-mode for characters 128 and above, you must
3804 ensure that PCRE is compiled with Unicode property support as well as
3805 with UTF-8 support.
3806
3807 Characters that might indicate line breaks are never treated in any
3808 special way when matching character classes, whatever line-ending
3809 sequence is in use, and whatever setting of the PCRE_DOTALL and
3810 PCRE_MULTILINE options is used. A class such as [^a] always matches one
3811 of these characters.
3812
3813 The minus (hyphen) character can be used to specify a range of charac-
3814 ters in a character class. For example, [d-m] matches any letter
3815 between d and m, inclusive. If a minus character is required in a
3816 class, it must be escaped with a backslash or appear in a position
3817 where it cannot be interpreted as indicating a range, typically as the
3818 first or last character in the class.
3819
3820 It is not possible to have the literal character "]" as the end charac-
3821 ter of a range. A pattern such as [W-]46] is interpreted as a class of
3822 two characters ("W" and "-") followed by a literal string "46]", so it
3823 would match "W46]" or "-46]". However, if the "]" is escaped with a
3824 backslash it is interpreted as the end of range, so [W-\]46] is inter-
3825 preted as a class containing a range followed by two other characters.
3826 The octal or hexadecimal representation of "]" can also be used to end
3827 a range.
3828
3829 Ranges operate in the collating sequence of character values. They can
3830 also be used for characters specified numerically, for example
3831 [\000-\037]. In UTF-8 mode, ranges can include characters whose values
3832 are greater than 255, for example [\x{100}-\x{2ff}].
3833
3834 If a range that includes letters is used when caseless matching is set,
3835 it matches the letters in either case. For example, [W-c] is equivalent
3836 to [][\\^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if
3837 character tables for a French locale are in use, [\xc8-\xcb] matches
3838 accented E characters in both cases. In UTF-8 mode, PCRE supports the
3839 concept of case for characters with values greater than 128 only when
3840 it is compiled with Unicode property support.
3841
3842 The character types \d, \D, \p, \P, \s, \S, \w, and \W may also appear
3843 in a character class, and add the characters that they match to the
3844 class. For example, [\dABCDEF] matches any hexadecimal digit. A circum-
3845 flex can conveniently be used with the upper case character types to
3846 specify a more restricted set of characters than the matching lower
3847 case type. For example, the class [^\W_] matches any letter or digit,
3848 but not underscore.
3849
3850 The only metacharacters that are recognized in character classes are
3851 backslash, hyphen (only where it can be interpreted as specifying a
3852 range), circumflex (only at the start), opening square bracket (only
3853 when it can be interpreted as introducing a POSIX class name - see the
3854 next section), and the terminating closing square bracket. However,
3855 escaping other non-alphanumeric characters does no harm.
3856
3857
3858 POSIX CHARACTER CLASSES
3859
3860 Perl supports the POSIX notation for character classes. This uses names
3861 enclosed by [: and :] within the enclosing square brackets. PCRE also
3862 supports this notation. For example,
3863
3864 [01[:alpha:]%]
3865
3866 matches "0", "1", any alphabetic character, or "%". The supported class
3867 names are
3868
3869 alnum letters and digits
3870 alpha letters
3871 ascii character codes 0 - 127
3872 blank space or tab only
3873 cntrl control characters
3874 digit decimal digits (same as \d)
3875 graph printing characters, excluding space
3876 lower lower case letters
3877 print printing characters, including space
3878 punct printing characters, excluding letters and digits
3879 space white space (not quite the same as \s)
3880 upper upper case letters
3881 word "word" characters (same as \w)
3882 xdigit hexadecimal digits
3883
3884 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13),
3885 and space (32). Notice that this list includes the VT character (code
3886 11). This makes "space" different to \s, which does not include VT (for
3887 Perl compatibility).
3888
3889 The name "word" is a Perl extension, and "blank" is a GNU extension
3890 from Perl 5.8. Another Perl extension is negation, which is indicated
3891 by a ^ character after the colon. For example,
3892
3893 [12[:^digit:]]
3894
3895 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the
3896 POSIX syntax [.ch.] and [=ch=] where "ch" is a "collating element", but
3897 these are not supported, and an error is given if they are encountered.
3898
3899 In UTF-8 mode, characters with values greater than 128 do not match any
3900 of the POSIX character classes.
3901
3902
3903 VERTICAL BAR
3904
3905 Vertical bar characters are used to separate alternative patterns. For
3906 example, the pattern
3907
3908 gilbert|sullivan
3909
3910 matches either "gilbert" or "sullivan". Any number of alternatives may
3911 appear, and an empty alternative is permitted (matching the empty
3912 string). The matching process tries each alternative in turn, from left
3913 to right, and the first one that succeeds is used. If the alternatives
3914 are within a subpattern (defined below), "succeeds" means matching the
3915 rest of the main pattern as well as the alternative in the subpattern.
3916
3917
3918 INTERNAL OPTION SETTING
3919
3920 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
3921 PCRE_EXTENDED options (which are Perl-compatible) can be changed from
3922 within the pattern by a sequence of Perl option letters enclosed
3923 between "(?" and ")". The option letters are
3924
3925 i for PCRE_CASELESS
3926 m for PCRE_MULTILINE
3927 s for PCRE_DOTALL
3928 x for PCRE_EXTENDED
3929
3930 For example, (?im) sets caseless, multiline matching. It is also possi-
3931 ble to unset these options by preceding the letter with a hyphen, and a
3932 combined setting and unsetting such as (?im-sx), which sets PCRE_CASE-
3933 LESS and PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED,
3934 is also permitted. If a letter appears both before and after the
3935 hyphen, the option is unset.
3936
3937 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA
3938 can be changed in the same way as the Perl-compatible options by using
3939 the characters J, U and X respectively.
3940
3941 When one of these option changes occurs at top level (that is, not
3942 inside subpattern parentheses), the change applies to the remainder of
3943 the pattern that follows. If the change is placed right at the start of
3944 a pattern, PCRE extracts it into the global options (and it will there-
3945 fore show up in data extracted by the pcre_fullinfo() function).
3946
3947 An option change within a subpattern (see below for a description of
3948 subpatterns) affects only that part of the current pattern that follows
3949 it, so
3950
3951 (a(?i)b)c
3952
3953 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not
3954 used). By this means, options can be made to have different settings
3955 in different parts of the pattern. Any changes made in one alternative
3956 do carry on into subsequent branches within the same subpattern. For
3957 example,
3958
3959 (a(?i)b|c)
3960
3961 matches "ab", "aB", "c", and "C", even though when matching "C" the
3962 first branch is abandoned before the option setting. This is because
3963 the effects of option settings happen at compile time. There would be
3964 some very weird behaviour otherwise.
3965
3966 Note: There are other PCRE-specific options that can be set by the
3967 application when the compile or match functions are called. In some
3968 cases the pattern can contain special leading sequences such as (*CRLF)
3969 to override what the application has set or what has been defaulted.
3970 Details are given in the section entitled "Newline sequences" above.
3971 There is also the (*UTF8) leading sequence that can be used to set
3972 UTF-8 mode; this is equivalent to setting the PCRE_UTF8 option.
3973
3974
3975 SUBPATTERNS
3976
3977 Subpatterns are delimited by parentheses (round brackets), which can be
3978 nested. Turning part of a pattern into a subpattern does two things:
3979
3980 1. It localizes a set of alternatives. For example, the pattern
3981
3982 cat(aract|erpillar|)
3983
3984 matches one of the words "cat", "cataract", or "caterpillar". Without
3985 the parentheses, it would match "cataract", "erpillar" or an empty
3986 string.
3987
3988 2. It sets up the subpattern as a capturing subpattern. This means
3989 that, when the whole pattern matches, that portion of the subject
3990 string that matched the subpattern is passed back to the caller via the
3991 ovector argument of pcre_exec(). Opening parentheses are counted from
3992 left to right (starting from 1) to obtain numbers for the capturing
3993 subpatterns.
3994
3995 For example, if the string "the red king" is matched against the pat-
3996 tern
3997
3998 the ((red|white) (king|queen))
3999
4000 the captured substrings are "red king", "red", and "king", and are num-
4001 bered 1, 2, and 3, respectively.
4002
4003 The fact that plain parentheses fulfil two functions is not always
4004 helpful. There are often times when a grouping subpattern is required
4005 without a capturing requirement. If an opening parenthesis is followed
4006 by a question mark and a colon, the subpattern does not do any captur-
4007 ing, and is not counted when computing the number of any subsequent
4008 capturing subpatterns. For example, if the string "the white queen" is
4009 matched against the pattern
4010
4011 the ((?:red|white) (king|queen))
4012
4013 the captured substrings are "white queen" and "queen", and are numbered
4014 1 and 2. The maximum number of capturing subpatterns is 65535.
4015
4016 As a convenient shorthand, if any option settings are required at the
4017 start of a non-capturing subpattern, the option letters may appear
4018 between the "?" and the ":". Thus the two patterns
4019
4020 (?i:saturday|sunday)
4021 (?:(?i)saturday|sunday)
4022
4023 match exactly the same set of strings. Because alternative branches are
4024 tried from left to right, and options are not reset until the end of
4025 the subpattern is reached, an option setting in one branch does affect
4026 subsequent branches, so the above patterns match "SUNDAY" as well as
4027 "Saturday".
4028
4029
4030 DUPLICATE SUBPATTERN NUMBERS
4031
4032 Perl 5.10 introduced a feature whereby each alternative in a subpattern
4033 uses the same numbers for its capturing parentheses. Such a subpattern
4034 starts with (?| and is itself a non-capturing subpattern. For example,
4035 consider this pattern:
4036
4037 (?|(Sat)ur|(Sun))day
4038
4039 Because the two alternatives are inside a (?| group, both sets of cap-
4040 turing parentheses are numbered one. Thus, when the pattern matches,
4041 you can look at captured substring number one, whichever alternative
4042 matched. This construct is useful when you want to capture part, but
4043 not all, of one of a number of alternatives. Inside a (?| group, paren-
4044 theses are numbered as usual, but the number is reset at the start of
4045 each branch. The numbers of any capturing buffers that follow the sub-
4046 pattern start after the highest number used in any branch. The follow-
4047 ing example is taken from the Perl documentation. The numbers under-
4048 neath show in which buffer the captured content will be stored.
4049
4050 # before ---------------branch-reset----------- after
4051 / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
4052 # 1 2 2 3 2 3 4
4053
4054 A backreference to a numbered subpattern uses the most recent value
4055 that is set for that number by any subpattern. The following pattern
4056 matches "abcabc" or "defdef":
4057
4058 /(?|(abc)|(def))\1/
4059
4060 In contrast, a recursive or "subroutine" call to a numbered subpattern
4061 always refers to the first one in the pattern with the given number.
4062 The following pattern matches "abcabc" or "defabc":
4063
4064 /(?|(abc)|(def))(?1)/
4065
4066 If a condition test for a subpattern's having matched refers to a non-
4067 unique number, the test is true if any of the subpatterns of that num-
4068 ber have matched.
4069
4070 An alternative approach to using this "branch reset" feature is to use
4071 duplicate named subpatterns, as described in the next section.
4072
4073
4074 NAMED SUBPATTERNS
4075
4076 Identifying capturing parentheses by number is simple, but it can be
4077 very hard to keep track of the numbers in complicated regular expres-
4078 sions. Furthermore, if an expression is modified, the numbers may
4079 change. To help with this difficulty, PCRE supports the naming of sub-
4080 patterns. This feature was not added to Perl until release 5.10. Python
4081 had the feature earlier, and PCRE introduced it at release 4.0, using
4082 the Python syntax. PCRE now supports both the Perl and the Python syn-
4083 tax. Perl allows identically numbered subpatterns to have different
4084 names, but PCRE does not.
4085
4086 In PCRE, a subpattern can be named in one of three ways: (?<name>...)
4087 or (?'name'...) as in Perl, or (?P<name>...) as in Python. References
4088 to capturing parentheses from other parts of the pattern, such as back-
4089 references, recursion, and conditions, can be made by name as well as
4090 by number.
4091
4092 Names consist of up to 32 alphanumeric characters and underscores.
4093 Named capturing parentheses are still allocated numbers as well as
4094 names, exactly as if the names were not present. The PCRE API provides
4095 function calls for extracting the name-to-number translation table from
4096 a compiled pattern. There is also a convenience function for extracting
4097 a captured substring by name.
4098
4099 By default, a name must be unique within a pattern, but it is possible
4100 to relax this constraint by setting the PCRE_DUPNAMES option at compile
4101 time. (Duplicate names are also always permitted for subpatterns with
4102 the same number, set up as described in the previous section.) Dupli-
4103 cate names can be useful for patterns where only one instance of the
4104 named parentheses can match. Suppose you want to match the name of a
4105 weekday, either as a 3-letter abbreviation or as the full name, and in
4106 both cases you want to extract the abbreviation. This pattern (ignoring
4107 the line breaks) does the job:
4108
4109 (?<DN>Mon|Fri|Sun)(?:day)?|
4110 (?<DN>Tue)(?:sday)?|
4111 (?<DN>Wed)(?:nesday)?|
4112 (?<DN>Thu)(?:rsday)?|
4113 (?<DN>Sat)(?:urday)?
4114
4115 There are five capturing substrings, but only one is ever set after a
4116 match. (An alternative way of solving this problem is to use a "branch
4117 reset" subpattern, as described in the previous section.)
4118
4119 The convenience function for extracting the data by name returns the
4120 substring for the first (and in this example, the only) subpattern of
4121 that name that matched. This saves searching to find which numbered
4122 subpattern it was.
4123
4124 If you make a backreference to a non-unique named subpattern from else-
4125 where in the pattern, the one that corresponds to the first occurrence
4126 of the name is used. In the absence of duplicate numbers (see the pre-
4127 vious section) this is the one with the lowest number. If you use a
4128 named reference in a condition test (see the section about conditions
4129 below), either to check whether a subpattern has matched, or to check
4130 for recursion, all subpatterns with the same name are tested. If the
4131 condition is true for any one of them, the overall condition is true.
4132 This is the same behaviour as testing by number. For further details of
4133 the interfaces for handling named subpatterns, see the pcreapi documen-
4134 tation.
4135
4136 Warning: You cannot use different names to distinguish between two sub-
4137 patterns with the same number because PCRE uses only the numbers when
4138 matching. For this reason, an error is given at compile time if differ-
4139 ent names are given to subpatterns with the same number. However, you
4140 can give the same name to subpatterns with the same number, even when
4141 PCRE_DUPNAMES is not set.
4142
4143
4144 REPETITION
4145
4146 Repetition is specified by quantifiers, which can follow any of the
4147 following items:
4148
4149 a literal data character
4150 the dot metacharacter
4151 the \C escape sequence
4152 the \X escape sequence (in UTF-8 mode with Unicode properties)
4153 the \R escape sequence
4154 an escape such as \d that matches a single character
4155 a character class
4156 a back reference (see next section)
4157 a parenthesized subpattern (unless it is an assertion)
4158 a recursive or "subroutine" call to a subpattern
4159
4160 The general repetition quantifier specifies a minimum and maximum num-
4161 ber of permitted matches, by giving the two numbers in curly brackets
4162 (braces), separated by a comma. The numbers must be less than 65536,
4163 and the first must be less than or equal to the second. For example:
4164
4165 z{2,4}
4166
4167 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a
4168 special character. If the second number is omitted, but the comma is
4169 present, there is no upper limit; if the second number and the comma
4170 are both omitted, the quantifier specifies an exact number of required
4171 matches. Thus
4172
4173 [aeiou]{3,}
4174
4175 matches at least 3 successive vowels, but may match many more, while
4176
4177 \d{8}
4178
4179 matches exactly 8 digits. An opening curly bracket that appears in a
4180 position where a quantifier is not allowed, or one that does not match
4181 the syntax of a quantifier, is taken as a literal character. For exam-
4182 ple, {,6} is not a quantifier, but a literal string of four characters.
4183
4184 In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to
4185 individual bytes. Thus, for example, \x{100}{2} matches two UTF-8 char-
4186 acters, each of which is represented by a two-byte sequence. Similarly,
4187 when Unicode property support is available, \X{3} matches three Unicode
4188 extended sequences, each of which may be several bytes long (and they
4189 may be of different lengths).
4190
4191 The quantifier {0} is permitted, causing the expression to behave as if
4192 the previous item and the quantifier were not present. This may be use-
4193 ful for subpatterns that are referenced as subroutines from elsewhere
4194 in the pattern. Items other than subpatterns that have a {0} quantifier
4195 are omitted from the compiled pattern.
4196
4197 For convenience, the three most common quantifiers have single-charac-
4198 ter abbreviations:
4199
4200 * is equivalent to {0,}
4201 + is equivalent to {1,}
4202 ? is equivalent to {0,1}
4203
4204 It is possible to construct infinite loops by following a subpattern
4205 that can match no characters with a quantifier that has no upper limit,
4206 for example:
4207
4208 (a?)*
4209
4210 Earlier versions of Perl and PCRE used to give an error at compile time
4211 for such patterns. However, because there are cases where this can be
4212 useful, such patterns are now accepted, but if any repetition of the
4213 subpattern does in fact match no characters, the loop is forcibly bro-
4214 ken.
4215
4216 By default, the quantifiers are "greedy", that is, they match as much
4217 as possible (up to the maximum number of permitted times), without
4218 causing the rest of the pattern to fail. The classic example of where
4219 this gives problems is in trying to match comments in C programs. These
4220 appear between /* and */ and within the comment, individual * and /
4221 characters may appear. An attempt to match C comments by applying the
4222 pattern
4223
4224 /\*.*\*/
4225
4226 to the string
4227
4228 /* first comment */ not comment /* second comment */
4229
4230 fails, because it matches the entire string owing to the greediness of
4231 the .* item.
4232
4233 However, if a quantifier is followed by a question mark, it ceases to
4234 be greedy, and instead matches the minimum number of times possible, so
4235 the pattern
4236
4237 /\*.*?\*/
4238
4239 does the right thing with the C comments. The meaning of the various
4240 quantifiers is not otherwise changed, just the preferred number of
4241 matches. Do not confuse this use of question mark with its use as a
4242 quantifier in its own right. Because it has two uses, it can sometimes
4243 appear doubled, as in
4244
4245 \d??\d
4246
4247 which matches one digit by preference, but can match two if that is the
4248 only way the rest of the pattern matches.
4249
4250 If the PCRE_UNGREEDY option is set (an option that is not available in
4251 Perl), the quantifiers are not greedy by default, but individual ones
4252 can be made greedy by following them with a question mark. In other
4253 words, it inverts the default behaviour.
4254
4255 When a parenthesized subpattern is quantified with a minimum repeat
4256 count that is greater than 1 or with a limited maximum, more memory is
4257 required for the compiled pattern, in proportion to the size of the
4258 minimum or maximum.
4259
4260 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equiv-
4261 alent to Perl's /s) is set, thus allowing the dot to match newlines,
4262 the pattern is implicitly anchored, because whatever follows will be
4263 tried against every character position in the subject string, so there
4264 is no point in retrying the overall match at any position after the
4265 first. PCRE normally treats such a pattern as though it were preceded
4266 by \A.
4267
4268 In cases where it is known that the subject string contains no new-
4269 lines, it is worth setting PCRE_DOTALL in order to obtain this opti-
4270 mization, or alternatively using ^ to indicate anchoring explicitly.
4271
4272 However, there is one situation where the optimization cannot be used.
4273 When .* is inside capturing parentheses that are the subject of a
4274 backreference elsewhere in the pattern, a match at the start may fail
4275 where a later one succeeds. Consider, for example:
4276
4277 (.*)abc\1
4278
4279 If the subject is "xyz123abc123" the match point is the fourth charac-
4280 ter. For this reason, such a pattern is not implicitly anchored.
4281
4282 When a capturing subpattern is repeated, the value captured is the sub-
4283 string that matched the final iteration. For example, after
4284
4285 (tweedle[dume]{3}\s*)+
4286
4287 has matched "tweedledum tweedledee" the value of the captured substring
4288 is "tweedledee". However, if there are nested capturing subpatterns,
4289 the corresponding captured values may have been set in previous itera-
4290 tions. For example, after
4291
4292 /(a|(b))+/
4293
4294 matches "aba" the value of the second captured substring is "b".
4295
4296
4297 ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS
4298
4299 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
4300 repetition, failure of what follows normally causes the repeated item
4301 to be re-evaluated to see if a different number of repeats allows the
4302 rest of the pattern to match. Sometimes it is useful to prevent this,
4303 either to change the nature of the match, or to cause it fail earlier
4304 than it otherwise might, when the author of the pattern knows there is
4305 no point in carrying on.
4306
4307 Consider, for example, the pattern \d+foo when applied to the subject
4308 line
4309
4310 123456bar
4311
4312 After matching all 6 digits and then failing to match "foo", the normal
4313 action of the matcher is to try again with only 5 digits matching the
4314 \d+ item, and then with 4, and so on, before ultimately failing.
4315 "Atomic grouping" (a term taken from Jeffrey Friedl's book) provides
4316 the means for specifying that once a subpattern has matched, it is not
4317 to be re-evaluated in this way.
4318
4319 If we use atomic grouping for the previous example, the matcher gives
4320 up immediately on failing to match "foo" the first time. The notation
4321 is a kind of special parenthesis, starting with (?> as in this example:
4322
4323 (?>\d+)foo
4324
4325 This kind of parenthesis "locks up" the part of the pattern it con-
4326 tains once it has matched, and a failure further into the pattern is
4327 prevented from backtracking into it. Backtracking past it to previous
4328 items, however, works as normal.
4329
4330 An alternative description is that a subpattern of this type matches
4331 the string of characters that an identical standalone pattern would
4332 match, if anchored at the current point in the subject string.
4333
4334 Atomic grouping subpatterns are not capturing subpatterns. Simple cases
4335 such as the above example can be thought of as a maximizing repeat that
4336 must swallow everything it can. So, while both \d+ and \d+? are pre-
4337 pared to adjust the number of digits they match in order to make the
4338 rest of the pattern match, (?>\d+) can only match an entire sequence of
4339 digits.
4340
4341 Atomic groups in general can of course contain arbitrarily complicated
4342 subpatterns, and can be nested. However, when the subpattern for an
4343 atomic group is just a single repeated item, as in the example above, a
4344 simpler notation, called a "possessive quantifier" can be used. This
4345 consists of an additional + character following a quantifier. Using
4346 this notation, the previous example can be rewritten as
4347
4348 \d++foo
4349
4350 Note that a possessive quantifier can be used with an entire group, for
4351 example:
4352
4353 (abc|xyz){2,3}+
4354
4355 Possessive quantifiers are always greedy; the setting of the
4356 PCRE_UNGREEDY option is ignored. They are a convenient notation for the
4357 simpler forms of atomic group. However, there is no difference in the
4358 meaning of a possessive quantifier and the equivalent atomic group,
4359 though there may be a performance difference; possessive quantifiers
4360 should be slightly faster.
4361
4362 The possessive quantifier syntax is an extension to the Perl 5.8 syn-
4363 tax. Jeffrey Friedl originated the idea (and the name) in the first
4364 edition of his book. Mike McCloskey liked it, so implemented it when he
4365 built Sun's Java package, and PCRE copied it from there. It ultimately
4366 found its way into Perl at release 5.10.
4367
4368 PCRE has an optimization that automatically "possessifies" certain sim-
4369 ple pattern constructs. For example, the sequence A+B is treated as
4370 A++B because there is no point in backtracking into a sequence of A's
4371 when B must follow.
4372
4373 When a pattern contains an unlimited repeat inside a subpattern that
4374 can itself be repeated an unlimited number of times, the use of an
4375 atomic group is the only way to avoid some failing matches taking a
4376 very long time indeed. The pattern
4377
4378 (\D+|<\d+>)*[!?]
4379
4380 matches an unlimited number of substrings that either consist of non-
4381 digits, or digits enclosed in <>, followed by either ! or ?. When it
4382 matches, it runs quickly. However, if it is applied to
4383
4384 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
4385
4386 it takes a long time before reporting failure. This is because the
4387 string can be divided between the internal \D+ repeat and the external
4388 * repeat in a large number of ways, and all have to be tried. (The
4389 example uses [!?] rather than a single character at the end, because
4390 both PCRE and Perl have an optimization that allows for fast failure
4391 when a single character is used. They remember the last single charac-
4392 ter that is required for a match, and fail early if it is not present
4393 in the string.) If the pattern is changed so that it uses an atomic
4394 group, like this:
4395
4396 ((?>\D+)|<\d+>)*[!?]
4397
4398 sequences of non-digits cannot be broken, and failure happens quickly.
4399
4400
4401 BACK REFERENCES
4402
4403 Outside a character class, a backslash followed by a digit greater than
4404 0 (and possibly further digits) is a back reference to a capturing sub-
4405 pattern earlier (that is, to its left) in the pattern, provided there
4406 have been that many previous capturing left parentheses.
4407
4408 However, if the decimal number following the backslash is less than 10,
4409 it is always taken as a back reference, and causes an error only if
4410 there are not that many capturing left parentheses in the entire pat-
4411 tern. In other words, the parentheses that are referenced need not be
4412 to the left of the reference for numbers less than 10. A "forward back
4413 reference" of this type can make sense when a repetition is involved
4414 and the subpattern to the right has participated in an earlier itera-
4415 tion.
4416
4417 It is not possible to have a numerical "forward back reference" to a
4418 subpattern whose number is 10 or more using this syntax because a
4419 sequence such as \50 is interpreted as a character defined in octal.
4420 See the subsection entitled "Non-printing characters" above for further
4421 details of the handling of digits following a backslash. There is no
4422 such problem when named parentheses are used. A back reference to any
4423 subpattern is possible using named parentheses (see below).
4424
4425 Another way of avoiding the ambiguity inherent in the use of digits
4426 following a backslash is to use the \g escape sequence, which is a fea-
4427 ture introduced in Perl 5.10. This escape must be followed by an
4428 unsigned number or a negative number, optionally enclosed in braces.
4429 These examples are all identical:
4430
4431 (ring), \1
4432 (ring), \g1
4433 (ring), \g{1}
4434
4435 An unsigned number specifies an absolute reference without the ambigu-
4436 ity that is present in the older syntax. It is also useful when literal
4437 digits follow the reference. A negative number is a relative reference.
4438 Consider this example:
4439
4440 (abc(def)ghi)\g{-1}
4441
4442 The sequence \g{-1} is a reference to the most recently started captur-
4443 ing subpattern before \g, that is, is it equivalent to \2. Similarly,
4444 \g{-2} would be equivalent to \1. The use of relative references can be
4445 helpful in long patterns, and also in patterns that are created by
4446 joining together fragments that contain references within themselves.
4447
4448 A back reference matches whatever actually matched the capturing sub-
4449 pattern in the current subject string, rather than anything matching
4450 the subpattern itself (see "Subpatterns as subroutines" below for a way
4451 of doing that). So the pattern
4452
4453 (sens|respons)e and \1ibility
4454
4455 matches "sense and sensibility" and "response and responsibility", but
4456 not "sense and responsibility". If caseful matching is in force at the
4457 time of the back reference, the case of letters is relevant. For exam-
4458 ple,
4459
4460 ((?i)rah)\s+\1
4461
4462 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the
4463 original capturing subpattern is matched caselessly.
4464
4465 There are several different ways of writing back references to named
4466 subpatterns. The .NET syntax \k{name} and the Perl syntax \k<name> or
4467 \k'name' are supported, as is the Python syntax (?P=name). Perl 5.10's
4468 unified back reference syntax, in which \g can be used for both numeric
4469 and named references, is also supported. We could rewrite the above
4470 example in any of the following ways:
4471
4472 (?<p1>(?i)rah)\s+\k<p1>
4473 (?'p1'(?i)rah)\s+\k{p1}
4474 (?P<p1>(?i)rah)\s+(?P=p1)
4475 (?<p1>(?i)rah)\s+\g{p1}
4476
4477 A subpattern that is referenced by name may appear in the pattern
4478 before or after the reference.
4479
4480 There may be more than one back reference to the same subpattern. If a
4481 subpattern has not actually been used in a particular match, any back
4482 references to it always fail by default. For example, the pattern
4483
4484 (a|(bc))\2
4485
4486 always fails if it starts to match "a" rather than "bc". However, if
4487 the PCRE_JAVASCRIPT_COMPAT option is set at compile time, a back refer-
4488 ence to an unset value matches an empty string.
4489
4490 Because there may be many capturing parentheses in a pattern, all dig-
4491 its following a backslash are taken as part of a potential back refer-
4492 ence number. If the pattern continues with a digit character, some
4493 delimiter must be used to terminate the back reference. If the
4494 PCRE_EXTENDED option is set, this can be whitespace. Otherwise, the \g{
4495 syntax or an empty comment (see "Comments" below) can be used.
4496
4497 A back reference that occurs inside the parentheses to which it refers
4498 fails when the subpattern is first used, so, for example, (a\1) never
4499 matches. However, such references can be useful inside repeated sub-
4500 patterns. For example, the pattern
4501
4502 (a|b\1)+
4503
4504 matches any number of "a"s and also "aba", "ababbaa" etc. At each iter-
4505 ation of the subpattern, the back reference matches the character
4506 string corresponding to the previous iteration. In order for this to
4507 work, the pattern must be such that the first iteration does not need
4508 to match the back reference. This can be done using alternation, as in
4509 the example above, or by a quantifier with a minimum of zero.
4510
4511
4512 ASSERTIONS
4513
4514 An assertion is a test on the characters following or preceding the
4515 current matching point that does not actually consume any characters.
4516 The simple assertions coded as \b, \B, \A, \G, \Z, \z, ^ and $ are
4517 described above.
4518
4519 More complicated assertions are coded as subpatterns. There are two
4520 kinds: those that look ahead of the current position in the subject
4521 string, and those that look behind it. An assertion subpattern is
4522 matched in the normal way, except that it does not cause the current
4523 matching position to be changed.
4524
4525 Assertion subpatterns are not capturing subpatterns, and may not be
4526 repeated, because it makes no sense to assert the same thing several
4527 times. If any kind of assertion contains capturing subpatterns within
4528 it, these are counted for the purposes of numbering the capturing sub-
4529 patterns in the whole pattern. However, substring capturing is carried
4530 out only for positive assertions, because it does not make sense for
4531 negative assertions.
4532
4533 Lookahead assertions
4534
4535 Lookahead assertions start with (?= for positive assertions and (?! for
4536 negative assertions. For example,
4537
4538 \w+(?=;)
4539
4540 matches a word followed by a semicolon, but does not include the semi-
4541 colon in the match, and
4542
4543 foo(?!bar)
4544
4545 matches any occurrence of "foo" that is not followed by "bar". Note
4546 that the apparently similar pattern
4547
4548 (?!foo)bar
4549
4550 does not find an occurrence of "bar" that is preceded by something
4551 other than "foo"; it finds any occurrence of "bar" whatsoever, because
4552 the assertion (?!foo) is always true when the next three characters are
4553 "bar". A lookbehind assertion is needed to achieve the other effect.
4554
4555 If you want to force a matching failure at some point in a pattern, the
4556 most convenient way to do it is with (?!) because an empty string
4557 always matches, so an assertion that requires there not to be an empty
4558 string must always fail. The Perl 5.10 backtracking control verb
4559 (*FAIL) or (*F) is essentially a synonym for (?!).
4560
4561 Lookbehind assertions
4562
4563 Lookbehind assertions start with (?<= for positive assertions and (?<!
4564 for negative assertions. For example,
4565
4566 (?<!foo)bar
4567
4568 does find an occurrence of "bar" that is not preceded by "foo". The
4569 contents of a lookbehind assertion are restricted such that all the
4570 strings it matches must have a fixed length. However, if there are sev-
4571 eral top-level alternatives, they do not all have to have the same
4572 fixed length. Thus
4573
4574 (?<=bullock|donkey)
4575
4576 is permitted, but
4577
4578 (?<!dogs?|cats?)
4579
4580 causes an error at compile time. Branches that match different length
4581 strings are permitted only at the top level of a lookbehind assertion.
4582 This is an extension compared with Perl (5.8 and 5.10), which requires
4583 all branches to match the same length of string. An assertion such as
4584
4585 (?<=ab(c|de))
4586
4587 is not permitted, because its single top-level branch can match two
4588 different lengths, but it is acceptable to PCRE if rewritten to use two
4589 top-level branches:
4590
4591 (?<=abc|abde)
4592
4593 In some cases, the Perl 5.10 escape sequence \K (see above) can be used
4594 instead of a lookbehind assertion to get round the fixed-length
4595 restriction.
4596
4597 The implementation of lookbehind assertions is, for each alternative,
4598 to temporarily move the current position back by the fixed length and
4599 then try to match. If there are insufficient characters before the cur-
4600 rent position, the assertion fails.
4601
4602 PCRE does not allow the \C escape (which matches a single byte in UTF-8
4603 mode) to appear in lookbehind assertions, because it makes it impossi-
4604 ble to calculate the length of the lookbehind. The \X and \R escapes,
4605 which can match different numbers of bytes, are also not permitted.
4606
4607 "Subroutine" calls (see below) such as (?2) or (?&X) are permitted in
4608 lookbehinds, as long as the subpattern matches a fixed-length string.
4609 Recursion, however, is not supported.
4610
4611 Possessive quantifiers can be used in conjunction with lookbehind
4612 assertions to specify efficient matching of fixed-length strings at the
4613 end of subject strings. Consider a simple pattern such as
4614
4615 abcd$
4616
4617 when applied to a long string that does not match. Because matching
4618 proceeds from left to right, PCRE will look for each "a" in the subject
4619 and then see if what follows matches the rest of the pattern. If the
4620 pattern is specified as
4621
4622 ^.*abcd$
4623
4624 the initial .* matches the entire string at first, but when this fails
4625 (because there is no following "a"), it backtracks to match all but the
4626 last character, then all but the last two characters, and so on. Once
4627 again the search for "a" covers the entire string, from right to left,
4628 so we are no better off. However, if the pattern is written as
4629
4630 ^.*+(?<=abcd)
4631
4632 there can be no backtracking for the .*+ item; it can match only the
4633 entire string. The subsequent lookbehind assertion does a single test
4634 on the last four characters. If it fails, the match fails immediately.
4635 For long strings, this approach makes a significant difference to the
4636 processing time.
4637
4638 Using multiple assertions
4639
4640 Several assertions (of any sort) may occur in succession. For example,
4641
4642 (?<=\d{3})(?<!999)foo
4643
4644 matches "foo" preceded by three digits that are not "999". Notice that
4645 each of the assertions is applied independently at the same point in
4646 the subject string. First there is a check that the previous three
4647 characters are all digits, and then there is a check that the same
4648 three characters are not "999". This pattern does not match "foo" pre-
4649 ceded by six characters, the first of which are digits and the last
4650 three of which are not "999". For example, it doesn't match "123abc-
4651 foo". A pattern to do that is
4652
4653 (?<=\d{3}...)(?<!999)foo
4654
4655 This time the first assertion looks at the preceding six characters,
4656 checking that the first three are digits, and then the second assertion
4657 checks that the preceding three characters are not "999".
4658
4659 Assertions can be nested in any combination. For example,
4660
4661 (?<=(?<!foo)bar)baz
4662
4663 matches an occurrence of "baz" that is preceded by "bar" which in turn
4664 is not preceded by "foo", while
4665
4666 (?<=\d{3}(?!999)...)foo
4667
4668 is another pattern that matches "foo" preceded by three digits and any
4669 three characters that are not "999".
4670
4671
4672 CONDITIONAL SUBPATTERNS
4673
4674 It is possible to cause the matching process to obey a subpattern con-
4675 ditionally or to choose between two alternative subpatterns, depending
4676 on the result of an assertion, or whether a specific capturing subpat-
4677 tern has already been matched. The two possible forms of conditional
4678 subpattern are:
4679
4680 (?(condition)yes-pattern)
4681 (?(condition)yes-pattern|no-pattern)
4682
4683 If the condition is satisfied, the yes-pattern is used; otherwise the
4684 no-pattern (if present) is used. If there are more than two alterna-
4685 tives in the subpattern, a compile-time error occurs.
4686
4687 There are four kinds of condition: references to subpatterns, refer-
4688 ences to recursion, a pseudo-condition called DEFINE, and assertions.
4689
4690 Checking for a used subpattern by number
4691
4692 If the text between the parentheses consists of a sequence of digits,
4693 the condition is true if a capturing subpattern of that number has pre-
4694 viously matched. If there is more than one capturing subpattern with
4695 the same number (see the earlier section about duplicate subpattern
4696 numbers), the condition is true if any of them have been set. An alter-
4697 native notation is to precede the digits with a plus or minus sign. In
4698 this case, the subpattern number is relative rather than absolute. The
4699 most recently opened parentheses can be referenced by (?(-1), the next
4700 most recent by (?(-2), and so on. In looping constructs it can also
4701 make sense to refer to subsequent groups with constructs such as
4702 (?(+2).
4703
4704 Consider the following pattern, which contains non-significant white
4705 space to make it more readable (assume the PCRE_EXTENDED option) and to
4706 divide it into three parts for ease of discussion:
4707
4708 ( \( )? [^()]+ (?(1) \) )
4709
4710 The first part matches an optional opening parenthesis, and if that
4711 character is present, sets it as the first captured substring. The sec-
4712 ond part matches one or more characters that are not parentheses. The
4713 third part is a conditional subpattern that tests whether the first set
4714 of parentheses matched or not. If they did, that is, if subject started
4715 with an opening parenthesis, the condition is true, and so the yes-pat-
4716 tern is executed and a closing parenthesis is required. Otherwise,
4717 since no-pattern is not present, the subpattern matches nothing. In
4718 other words, this pattern matches a sequence of non-parentheses,
4719 optionally enclosed in parentheses.
4720
4721 If you were embedding this pattern in a larger one, you could use a
4722 relative reference:
4723
4724 ...other stuff... ( \( )? [^()]+ (?(-1) \) ) ...
4725
4726 This makes the fragment independent of the parentheses in the larger
4727 pattern.
4728
4729 Checking for a used subpattern by name
4730
4731 Perl uses the syntax (?(<name>)...) or (?('name')...) to test for a
4732 used subpattern by name. For compatibility with earlier versions of
4733 PCRE, which had this facility before Perl, the syntax (?(name)...) is
4734 also recognized. However, there is a possible ambiguity with this syn-
4735 tax, because subpattern names may consist entirely of digits. PCRE
4736 looks first for a named subpattern; if it cannot find one and the name
4737 consists entirely of digits, PCRE looks for a subpattern of that num-
4738 ber, which must be greater than zero. Using subpattern names that con-
4739 sist entirely of digits is not recommended.
4740
4741 Rewriting the above example to use a named subpattern gives this:
4742
4743 (?<OPEN> \( )? [^()]+ (?(<OPEN>) \) )
4744
4745 If the name used in a condition of this kind is a duplicate, the test
4746 is applied to all subpatterns of the same name, and is true if any one
4747 of them has matched.
4748
4749 Checking for pattern recursion
4750
4751 If the condition is the string (R), and there is no subpattern with the
4752 name R, the condition is true if a recursive call to the whole pattern
4753 or any subpattern has been made. If digits or a name preceded by amper-
4754 sand follow the letter R, for example:
4755
4756 (?(R3)...) or (?(R&name)...)
4757
4758 the condition is true if the most recent recursion is into a subpattern
4759 whose number or name is given. This condition does not check the entire
4760 recursion stack. If the name used in a condition of this kind is a
4761 duplicate, the test is applied to all subpatterns of the same name, and
4762 is true if any one of them is the most recent recursion.
4763
4764 At "top level", all these recursion test conditions are false. The
4765 syntax for recursive patterns is described below.
4766
4767 Defining subpatterns for use by reference only
4768
4769 If the condition is the string (DEFINE), and there is no subpattern
4770 with the name DEFINE, the condition is always false. In this case,
4771 there may be only one alternative in the subpattern. It is always
4772 skipped if control reaches this point in the pattern; the idea of
4773 DEFINE is that it can be used to define "subroutines" that can be ref-
4774 erenced from elsewhere. (The use of "subroutines" is described below.)
4775 For example, a pattern to match an IPv4 address could be written like
4776 this (ignore whitespace and line breaks):
4777
4778 (?(DEFINE) (?<byte> 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
4779 \b (?&byte) (\.(?&byte)){3} \b
4780
4781 The first part of the pattern is a DEFINE group inside which a another
4782 group named "byte" is defined. This matches an individual component of
4783 an IPv4 address (a number less than 256). When matching takes place,
4784 this part of the pattern is skipped because DEFINE acts like a false
4785 condition. The rest of the pattern uses references to the named group
4786 to match the four dot-separated components of an IPv4 address, insist-
4787 ing on a word boundary at each end.
4788
4789 Assertion conditions
4790
4791 If the condition is not in any of the above formats, it must be an
4792 assertion. This may be a positive or negative lookahead or lookbehind
4793 assertion. Consider this pattern, again containing non-significant
4794 white space, and with the two alternatives on the second line:
4795
4796 (?(?=[^a-z]*[a-z])
4797 \d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} )
4798
4799 The condition is a positive lookahead assertion that matches an
4800 optional sequence of non-letters followed by a letter. In other words,
4801 it tests for the presence of at least one letter in the subject. If a
4802 letter is found, the subject is matched against the first alternative;
4803 otherwise it is matched against the second. This pattern matches
4804 strings in one of the two forms dd-aaa-dd or dd-dd-dd, where aaa are
4805 letters and dd are digits.
4806
4807
4808 COMMENTS
4809
4810 The sequence (?# marks the start of a comment that continues up to the
4811 next closing parenthesis. Nested parentheses are not permitted. The
4812 characters that make up a comment play no part in the pattern matching
4813 at all.
4814
4815 If the PCRE_EXTENDED option is set, an unescaped # character outside a
4816 character class introduces a comment that continues to immediately
4817 after the next newline in the pattern.
4818
4819
4820 RECURSIVE PATTERNS
4821
4822 Consider the problem of matching a string in parentheses, allowing for
4823 unlimited nested parentheses. Without the use of recursion, the best
4824 that can be done is to use a pattern that matches up to some fixed
4825 depth of nesting. It is not possible to handle an arbitrary nesting
4826 depth.
4827
4828 For some time, Perl has provided a facility that allows regular expres-
4829 sions to recurse (amongst other things). It does this by interpolating
4830 Perl code in the expression at run time, and the code can refer to the
4831 expression itself. A Perl pattern using code interpolation to solve the
4832 parentheses problem can be created like this:
4833
4834 $re = qr{\( (?: (?>[^()]+) | (?p{$re}) )* \)}x;
4835
4836 The (?p{...}) item interpolates Perl code at run time, and in this case
4837 refers recursively to the pattern in which it appears.
4838
4839 Obviously, PCRE cannot support the interpolation of Perl code. Instead,
4840 it supports special syntax for recursion of the entire pattern, and
4841 also for individual subpattern recursion. After its introduction in
4842 PCRE and Python, this kind of recursion was subsequently introduced
4843 into Perl at release 5.10.
4844
4845 A special item that consists of (? followed by a number greater than
4846 zero and a closing parenthesis is a recursive call of the subpattern of
4847 the given number, provided that it occurs inside that subpattern. (If
4848 not, it is a "subroutine" call, which is described in the next sec-
4849 tion.) The special item (?R) or (?0) is a recursive call of the entire
4850 regular expression.
4851
4852 This PCRE pattern solves the nested parentheses problem (assume the
4853 PCRE_EXTENDED option is set so that white space is ignored):
4854
4855 \( ( [^()]++ | (?R) )* \)
4856
4857 First it matches an opening parenthesis. Then it matches any number of
4858 substrings which can either be a sequence of non-parentheses, or a
4859 recursive match of the pattern itself (that is, a correctly parenthe-
4860 sized substring). Finally there is a closing parenthesis. Note the use
4861 of a possessive quantifier to avoid backtracking into sequences of non-
4862 parentheses.
4863
4864 If this were part of a larger pattern, you would not want to recurse
4865 the entire pattern, so instead you could use this:
4866
4867 ( \( ( [^()]++ | (?1) )* \) )
4868
4869 We have put the pattern into parentheses, and caused the recursion to
4870 refer to them instead of the whole pattern.
4871
4872 In a larger pattern, keeping track of parenthesis numbers can be
4873 tricky. This is made easier by the use of relative references (a Perl
4874 5.10 feature). Instead of (?1) in the pattern above you can write
4875 (?-2) to refer to the second most recently opened parentheses preceding
4876 the recursion. In other words, a negative number counts capturing
4877 parentheses leftwards from the point at which it is encountered.
4878
4879 It is also possible to refer to subsequently opened parentheses, by
4880 writing references such as (?+2). However, these cannot be recursive
4881 because the reference is not inside the parentheses that are refer-
4882 enced. They are always "subroutine" calls, as described in the next
4883 section.
4884
4885 An alternative approach is to use named parentheses instead. The Perl
4886 syntax for this is (?&name); PCRE's earlier syntax (?P>name) is also
4887 supported. We could rewrite the above example as follows:
4888
4889 (?<pn> \( ( [^()]++ | (?&pn) )* \) )
4890
4891 If there is more than one subpattern with the same name, the earliest
4892 one is used.
4893
4894 This particular example pattern that we have been looking at contains
4895 nested unlimited repeats, and so the use of a possessive quantifier for
4896 matching strings of non-parentheses is important when applying the pat-
4897 tern to strings that do not match. For example, when this pattern is
4898 applied to
4899
4900 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
4901
4902 it yields "no match" quickly. However, if a possessive quantifier is
4903 not used, the match runs for a very long time indeed because there are
4904 so many different ways the + and * repeats can carve up the subject,
4905 and all have to be tested before failure can be reported.
4906
4907 At the end of a match, the values set for any capturing subpatterns are
4908 those from the outermost level of the recursion at which the subpattern
4909 value is set. If you want to obtain intermediate values, a callout
4910 function can be used (see below and the pcrecallout documentation). If
4911 the pattern above is matched against
4912
4913 (ab(cd)ef)
4914
4915 the value for the capturing parentheses is "ef", which is the last
4916 value taken on at the top level. If additional parentheses are added,
4917 giving
4918
4919 \( ( ( [^()]++ | (?R) )* ) \)
4920 ^ ^
4921 ^ ^
4922
4923 the string they capture is "ab(cd)ef", the contents of the top level
4924 parentheses. If there are more than 15 capturing parentheses in a pat-
4925 tern, PCRE has to obtain extra memory to store data during a recursion,
4926 which it does by using pcre_malloc, freeing it via pcre_free after-
4927 wards. If no memory can be obtained, the match fails with the
4928 PCRE_ERROR_NOMEMORY error.
4929
4930 Do not confuse the (?R) item with the condition (R), which tests for
4931 recursion. Consider this pattern, which matches text in angle brack-
4932 ets, allowing for arbitrary nesting. Only digits are allowed in nested
4933 brackets (that is, when recursing), whereas any characters are permit-
4934 ted at the outer level.
4935
4936 < (?: (?(R) \d++ | [^<>]*+) | (?R)) * >
4937
4938 In this pattern, (?(R) is the start of a conditional subpattern, with
4939 two different alternatives for the recursive and non-recursive cases.
4940 The (?R) item is the actual recursive call.
4941
4942 Recursion difference from Perl
4943
4944 In PCRE (like Python, but unlike Perl), a recursive subpattern call is
4945 always treated as an atomic group. That is, once it has matched some of
4946 the subject string, it is never re-entered, even if it contains untried
4947 alternatives and there is a subsequent matching failure. This can be
4948 illustrated by the following pattern, which purports to match a palin-
4949 dromic string that contains an odd number of characters (for example,
4950 "a", "aba", "abcba", "abcdcba"):
4951
4952 ^(.|(.)(?1)\2)$
4953
4954 The idea is that it either matches a single character, or two identical
4955 characters surrounding a sub-palindrome. In Perl, this pattern works;
4956 in PCRE it does not if the pattern is longer than three characters.
4957 Consider the subject string "abcba":
4958
4959 At the top level, the first character is matched, but as it is not at
4960 the end of the string, the first alternative fails; the second alterna-
4961 tive is taken and the recursion kicks in. The recursive call to subpat-
4962 tern 1 successfully matches the next character ("b"). (Note that the
4963 beginning and end of line tests are not part of the recursion).
4964
4965 Back at the top level, the next character ("c") is compared with what
4966 subpattern 2 matched, which was "a". This fails. Because the recursion
4967 is treated as an atomic group, there are now no backtracking points,
4968 and so the entire match fails. (Perl is able, at this point, to re-
4969 enter the recursion and try the second alternative.) However, if the
4970 pattern is written with the alternatives in the other order, things are
4971 different:
4972
4973 ^((.)(?1)\2|.)$
4974
4975 This time, the recursing alternative is tried first, and continues to
4976 recurse until it runs out of characters, at which point the recursion
4977 fails. But this time we do have another alternative to try at the
4978 higher level. That is the big difference: in the previous case the
4979 remaining alternative is at a deeper recursion level, which PCRE cannot
4980 use.
4981
4982 To change the pattern so that matches all palindromic strings, not just
4983 those with an odd number of characters, it is tempting to change the
4984 pattern to this:
4985
4986 ^((.)(?1)\2|.?)$
4987
4988 Again, this works in Perl, but not in PCRE, and for the same reason.
4989 When a deeper recursion has matched a single character, it cannot be
4990 entered again in order to match an empty string. The solution is to
4991 separate the two cases, and write out the odd and even cases as alter-
4992 natives at the higher level:
4993
4994 ^(?:((.)(?1)\2|)|((.)(?3)\4|.))
4995
4996 If you want to match typical palindromic phrases, the pattern has to
4997 ignore all non-word characters, which can be done like this:
4998
4999 ^\W*+(?:((.)\W*+(?1)\W*+\2|)|((.)\W*+(?3)\W*+\4|\W*+.\W*+))\W*+$
5000
5001 If run with the PCRE_CASELESS option, this pattern matches phrases such
5002 as "A man, a plan, a canal: Panama!" and it works well in both PCRE and
5003 Perl. Note the use of the possessive quantifier *+ to avoid backtrack-
5004 ing into sequences of non-word characters. Without this, PCRE takes a
5005 great deal longer (ten times or more) to match typical phrases, and
5006 Perl takes so long that you think it has gone into a loop.
5007
5008 WARNING: The palindrome-matching patterns above work only if the sub-
5009 ject string does not start with a palindrome that is shorter than the
5010 entire string. For example, although "abcba" is correctly matched, if
5011 the subject is "ababa", PCRE finds the palindrome "aba" at the start,
5012 then fails at top level because the end of the string does not follow.
5013 Once again, it cannot jump back into the recursion to try other alter-
5014 natives, so the entire match fails.
5015
5016
5017 SUBPATTERNS AS SUBROUTINES
5018
5019 If the syntax for a recursive subpattern reference (either by number or
5020 by name) is used outside the parentheses to which it refers, it oper-
5021 ates like a subroutine in a programming language. The "called" subpat-
5022 tern may be defined before or after the reference. A numbered reference
5023 can be absolute or relative, as in these examples:
5024
5025 (...(absolute)...)...(?2)...
5026 (...(relative)...)...(?-1)...
5027 (...(?+1)...(relative)...
5028
5029 An earlier example pointed out that the pattern
5030
5031 (sens|respons)e and \1ibility
5032
5033 matches "sense and sensibility" and "response and responsibility", but
5034 not "sense and responsibility". If instead the pattern
5035
5036 (sens|respons)e and (?1)ibility
5037
5038 is used, it does match "sense and responsibility" as well as the other
5039 two strings. Another example is given in the discussion of DEFINE
5040 above.
5041
5042 Like recursive subpatterns, a "subroutine" call is always treated as an
5043 atomic group. That is, once it has matched some of the subject string,
5044 it is never re-entered, even if it contains untried alternatives and
5045 there is a subsequent matching failure.
5046
5047 When a subpattern is used as a subroutine, processing options such as
5048 case-independence are fixed when the subpattern is defined. They cannot
5049 be changed for different calls. For example, consider this pattern:
5050
5051 (abc)(?i:(?-1))
5052
5053 It matches "abcabc". It does not match "abcABC" because the change of
5054 processing option does not affect the called subpattern.
5055
5056
5057 ONIGURUMA SUBROUTINE SYNTAX
5058
5059 For compatibility with Oniguruma, the non-Perl syntax \g followed by a
5060 name or a number enclosed either in angle brackets or single quotes, is
5061 an alternative syntax for referencing a subpattern as a subroutine,
5062 possibly recursively. Here are two of the examples used above, rewrit-
5063 ten using this syntax:
5064
5065 (?<pn> \( ( (?>[^()]+) | \g<pn> )* \) )
5066 (sens|respons)e and \g'1'ibility
5067
5068 PCRE supports an extension to Oniguruma: if a number is preceded by a
5069 plus or a minus sign it is taken as a relative reference. For example:
5070
5071 (abc)(?i:\g<-1>)
5072
5073 Note that \g{...} (Perl syntax) and \g<...> (Oniguruma syntax) are not
5074 synonymous. The former is a back reference; the latter is a subroutine
5075 call.
5076
5077
5078 CALLOUTS
5079
5080 Perl has a feature whereby using the sequence (?{...}) causes arbitrary
5081 Perl code to be obeyed in the middle of matching a regular expression.
5082 This makes it possible, amongst other things, to extract different sub-
5083 strings that match the same pair of parentheses when there is a repeti-
5084 tion.
5085
5086 PCRE provides a similar feature, but of course it cannot obey arbitrary
5087 Perl code. The feature is called "callout". The caller of PCRE provides
5088 an external function by putting its entry point in the global variable
5089 pcre_callout. By default, this variable contains NULL, which disables
5090 all calling out.
5091
5092 Within a regular expression, (?C) indicates the points at which the
5093 external function is to be called. If you want to identify different
5094 callout points, you can put a number less than 256 after the letter C.
5095 The default value is zero. For example, this pattern has two callout
5096 points:
5097
5098 (?C1)abc(?C2)def
5099
5100 If the PCRE_AUTO_CALLOUT flag is passed to pcre_compile(), callouts are
5101 automatically installed before each item in the pattern. They are all
5102 numbered 255.
5103
5104 During matching, when PCRE reaches a callout point (and pcre_callout is
5105 set), the external function is called. It is provided with the number
5106 of the callout, the position in the pattern, and, optionally, one item
5107 of data originally supplied by the caller of pcre_exec(). The callout
5108 function may cause matching to proceed, to backtrack, or to fail alto-
5109 gether. A complete description of the interface to the callout function
5110 is given in the pcrecallout documentation.
5111
5112
5113 BACKTRACKING CONTROL
5114
5115 Perl 5.10 introduced a number of "Special Backtracking Control Verbs",
5116 which are described in the Perl documentation as "experimental and sub-
5117 ject to change or removal in a future version of Perl". It goes on to
5118 say: "Their usage in production code should be noted to avoid problems
5119 during upgrades." The same remarks apply to the PCRE features described
5120 in this section.
5121
5122 Since these verbs are specifically related to backtracking, most of
5123 them can be used only when the pattern is to be matched using
5124 pcre_exec(), which uses a backtracking algorithm. With the exception of
5125 (*FAIL), which behaves like a failing negative assertion, they cause an
5126 error if encountered by pcre_dfa_exec().
5127
5128 If any of these verbs are used in an assertion subpattern, their effect
5129 is confined to that subpattern; it does not extend to the surrounding
5130 pattern. Note that assertion subpatterns are processed as anchored at
5131 the point where they are tested.
5132
5133 The new verbs make use of what was previously invalid syntax: an open-
5134 ing parenthesis followed by an asterisk. In Perl, they are generally of
5135 the form (*VERB:ARG) but PCRE does not support the use of arguments, so
5136 its general form is just (*VERB). Any number of these verbs may occur
5137 in a pattern. There are two kinds:
5138
5139 Verbs that act immediately
5140
5141 The following verbs act as soon as they are encountered:
5142
5143 (*ACCEPT)
5144
5145 This verb causes the match to end successfully, skipping the remainder
5146 of the pattern. When inside a recursion, only the innermost pattern is
5147 ended immediately. If (*ACCEPT) is inside capturing parentheses, the
5148 data so far is captured. (This feature was added to PCRE at release
5149 8.00.) For example:
5150
5151 A((?:A|B(*ACCEPT)|C)D)
5152
5153 This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is cap-
5154 tured by the outer parentheses.
5155
5156 (*FAIL) or (*F)
5157
5158 This verb causes the match to fail, forcing backtracking to occur. It
5159 is equivalent to (?!) but easier to read. The Perl documentation notes
5160 that it is probably useful only when combined with (?{}) or (??{}).
5161 Those are, of course, Perl features that are not present in PCRE. The
5162 nearest equivalent is the callout feature, as for example in this pat-
5163 tern:
5164
5165 a+(?C)(*FAIL)
5166
5167 A match with the string "aaaa" always fails, but the callout is taken
5168 before each backtrack happens (in this example, 10 times).
5169
5170 Verbs that act after backtracking
5171
5172 The following verbs do nothing when they are encountered. Matching con-
5173 tinues with what follows, but if there is no subsequent match, a fail-
5174 ure is forced. The verbs differ in exactly what kind of failure
5175 occurs.
5176
5177 (*COMMIT)
5178
5179 This verb causes the whole match to fail outright if the rest of the
5180 pattern does not match. Even if the pattern is unanchored, no further
5181 attempts to find a match by advancing the starting point take place.
5182 Once (*COMMIT) has been passed, pcre_exec() is committed to finding a
5183 match at the current starting point, or not at all. For example:
5184
5185 a+(*COMMIT)b
5186
5187 This matches "xxaab" but not "aacaab". It can be thought of as a kind
5188 of dynamic anchor, or "I've started, so I must finish."
5189
5190 (*PRUNE)
5191
5192 This verb causes the match to fail at the current position if the rest
5193 of the pattern does not match. If the pattern is unanchored, the normal
5194 "bumpalong" advance to the next starting character then happens. Back-
5195 tracking can occur as usual to the left of (*PRUNE), or when matching
5196 to the right of (*PRUNE), but if there is no match to the right, back-
5197 tracking cannot cross (*PRUNE). In simple cases, the use of (*PRUNE)
5198 is just an alternative to an atomic group or possessive quantifier, but
5199 there are some uses of (*PRUNE) that cannot be expressed in any other
5200 way.
5201
5202 (*SKIP)
5203
5204 This verb is like (*PRUNE), except that if the pattern is unanchored,
5205 the "bumpalong" advance is not to the next character, but to the posi-
5206 tion in the subject where (*SKIP) was encountered. (*SKIP) signifies
5207 that whatever text was matched leading up to it cannot be part of a
5208 successful match. Consider:
5209
5210 a+(*SKIP)b
5211
5212 If the subject is "aaaac...", after the first match attempt fails
5213 (starting at the first character in the string), the starting point
5214 skips on to start the next attempt at "c". Note that a possessive quan-
5215 tifer does not have the same effect as this example; although it would
5216 suppress backtracking during the first match attempt, the second
5217 attempt would start at the second character instead of skipping on to
5218 "c".
5219
5220 (*THEN)
5221
5222 This verb causes a skip to the next alternation if the rest of the pat-
5223 tern does not match. That is, it cancels pending backtracking, but only
5224 within the current alternation. Its name comes from the observation
5225 that it can be used for a pattern-based if-then-else block:
5226
5227 ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
5228
5229 If the COND1 pattern matches, FOO is tried (and possibly further items
5230 after the end of the group if FOO succeeds); on failure the matcher
5231 skips to the second alternative and tries COND2, without backtracking
5232 into COND1. If (*THEN) is used outside of any alternation, it acts
5233 exactly like (*PRUNE).
5234
5235
5236 SEE ALSO
5237
5238 pcreapi(3), pcrecallout(3), pcrematching(3), pcresyntax(3), pcre(3).
5239
5240
5241 AUTHOR
5242
5243 Philip Hazel
5244 University Computing Service
5245 Cambridge CB2 3QH, England.
5246
5247
5248 REVISION
5249
5250 Last updated: 04 October 2009
5251 Copyright (c) 1997-2009 University of Cambridge.
5252 ------------------------------------------------------------------------------
5253
5254
5255 PCRESYNTAX(3) PCRESYNTAX(3)
5256
5257
5258 NAME
5259 PCRE - Perl-compatible regular expressions
5260
5261
5262 PCRE REGULAR EXPRESSION SYNTAX SUMMARY
5263
5264 The full syntax and semantics of the regular expressions that are sup-
5265 ported by PCRE are described in the pcrepattern documentation. This
5266 document contains just a quick-reference summary of the syntax.
5267
5268
5269 QUOTING
5270
5271 \x where x is non-alphanumeric is a literal x
5272 \Q...\E treat enclosed characters as literal
5273
5274
5275 CHARACTERS
5276
5277 \a alarm, that is, the BEL character (hex 07)
5278 \cx "control-x", where x is any character
5279 \e escape (hex 1B)
5280 \f formfeed (hex 0C)
5281 \n newline (hex 0A)
5282 \r carriage return (hex 0D)
5283 \t tab (hex 09)
5284 \ddd character with octal code ddd, or backreference
5285 \xhh character with hex code hh
5286 \x{hhh..} character with hex code hhh..
5287
5288
5289 CHARACTER TYPES
5290
5291 . any character except newline;
5292 in dotall mode, any character whatsoever
5293 \C one byte, even in UTF-8 mode (best avoided)
5294 \d a decimal digit
5295 \D a character that is not a decimal digit
5296 \h a horizontal whitespace character
5297 \H a character that is not a horizontal whitespace character
5298 \p{xx} a character with the xx property
5299 \P{xx} a character without the xx property
5300 \R a newline sequence
5301 \s a whitespace character
5302 \S a character that is not a whitespace character
5303 \v a vertical whitespace character
5304 \V a character that is not a vertical whitespace character
5305 \w a "word" character
5306 \W a "non-word" character
5307 \X an extended Unicode sequence
5308
5309 In PCRE, \d, \D, \s, \S, \w, and \W recognize only ASCII characters.
5310
5311
5312 GENERAL CATEGORY PROPERTY CODES FOR \p and \P
5313
5314 C Other
5315 Cc Control
5316 Cf Format
5317 Cn Unassigned
5318 Co Private use
5319 Cs Surrogate
5320
5321 L Letter
5322 Ll Lower case letter
5323 Lm Modifier letter
5324 Lo Other letter
5325 Lt Title case letter
5326 Lu Upper case letter
5327 L& Ll, Lu, or Lt
5328
5329 M Mark
5330 Mc Spacing mark
5331 Me Enclosing mark
5332 Mn Non-spacing mark
5333
5334 N Number
5335 Nd Decimal number
5336 Nl Letter number
5337 No Other number
5338
5339 P Punctuation
5340 Pc Connector punctuation
5341 Pd Dash punctuation
5342 Pe Close punctuation
5343 Pf Final punctuation
5344 Pi Initial punctuation
5345 Po Other punctuation
5346 Ps Open punctuation
5347
5348 S Symbol
5349 Sc Currency symbol
5350 Sk Modifier symbol
5351 Sm Mathematical symbol
5352 So Other symbol
5353
5354 Z Separator
5355 Zl Line separator
5356 Zp Paragraph separator
5357 Zs Space separator
5358
5359
5360 SCRIPT NAMES FOR \p AND \P
5361
5362 Arabic, Armenian, Balinese, Bengali, Bopomofo, Braille, Buginese,
5363 Buhid, Canadian_Aboriginal, Carian, Cham, Cherokee, Common, Coptic, Cu-
5364 neiform, Cypriot, Cyrillic, Deseret, Devanagari, Ethiopic, Georgian,
5365 Glagolitic, Gothic, Greek, Gujarati, Gurmukhi, Han, Hangul, Hanunoo,
5366 Hebrew, Hiragana, Inherited, Kannada, Katakana, Kayah_Li, Kharoshthi,
5367 Khmer, Lao, Latin, Lepcha, Limbu, Linear_B, Lycian, Lydian, Malayalam,
5368 Mongolian, Myanmar, New_Tai_Lue, Nko, Ogham, Old_Italic, Old_Persian,
5369 Ol_Chiki, Oriya, Osmanya, Phags_Pa, Phoenician, Rejang, Runic, Saurash-
5370 tra, Shavian, Sinhala, Sudanese, Syloti_Nagri, Syriac, Tagalog, Tag-
5371 banwa, Tai_Le, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh,
5372 Ugaritic, Vai, Yi.
5373
5374
5375 CHARACTER CLASSES
5376
5377 [...] positive character class
5378 [^...] negative character class
5379 [x-y] range (can be used for hex characters)
5380 [[:xxx:]] positive POSIX named set
5381 [[:^xxx:]] negative POSIX named set
5382
5383 alnum alphanumeric
5384 alpha alphabetic
5385 ascii 0-127
5386 blank space or tab
5387 cntrl control character
5388 digit decimal digit
5389 graph printing, excluding space
5390 lower lower case letter
5391 print printing, including space
5392 punct printing, excluding alphanumeric
5393 space whitespace
5394 upper upper case letter
5395 word same as \w
5396 xdigit hexadecimal digit
5397
5398 In PCRE, POSIX character set names recognize only ASCII characters. You
5399 can use \Q...\E inside a character class.
5400
5401
5402 QUANTIFIERS
5403
5404 ? 0 or 1, greedy
5405 ?+ 0 or 1, possessive
5406 ?? 0 or 1, lazy
5407 * 0 or more, greedy
5408 *+ 0 or more, possessive
5409 *? 0 or more, lazy
5410 + 1 or more, greedy
5411 ++ 1 or more, possessive
5412 +? 1 or more, lazy
5413 {n} exactly n
5414 {n,m} at least n, no more than m, greedy
5415 {n,m}+ at least n, no more than m, possessive
5416 {n,m}? at least n, no more than m, lazy
5417 {n,} n or more, greedy
5418 {n,}+ n or more, possessive
5419 {n,}? n or more, lazy
5420
5421
5422 ANCHORS AND SIMPLE ASSERTIONS
5423
5424 \b word boundary (only ASCII letters recognized)
5425 \B not a word boundary
5426 ^ start of subject
5427 also after internal newline in multiline mode
5428 \A start of subject
5429 $ end of subject
5430 also before newline at end of subject
5431 also before internal newline in multiline mode
5432 \Z end of subject
5433 also before newline at end of subject
5434 \z end of subject
5435 \G first matching position in subject
5436
5437
5438 MATCH POINT RESET
5439
5440 \K reset start of match
5441
5442
5443 ALTERNATION
5444
5445 expr|expr|expr...
5446
5447
5448 CAPTURING
5449
5450 (...) capturing group
5451 (?<name>...) named capturing group (Perl)
5452 (?'name'...) named capturing group (Perl)
5453 (?P<name>...) named capturing group (Python)
5454 (?:...) non-capturing group
5455 (?|...) non-capturing group; reset group numbers for
5456 capturing groups in each alternative
5457
5458
5459 ATOMIC GROUPS
5460
5461 (?>...) atomic, non-capturing group
5462
5463
5464 COMMENT
5465
5466 (?#....) comment (not nestable)
5467
5468
5469 OPTION SETTING
5470
5471 (?i) caseless
5472 (?J) allow duplicate names
5473 (?m) multiline
5474 (?s) single line (dotall)
5475 (?U) default ungreedy (lazy)
5476 (?x) extended (ignore white space)
5477 (?-...) unset option(s)
5478
5479 The following is recognized only at the start of a pattern or after one
5480 of the newline-setting options with similar syntax:
5481
5482 (*UTF8) set UTF-8 mode
5483
5484
5485 LOOKAHEAD AND LOOKBEHIND ASSERTIONS
5486
5487 (?=...) positive look ahead
5488 (?!...) negative look ahead
5489 (?<=...) positive look behind
5490 (?<!...) negative look behind
5491
5492 Each top-level branch of a look behind must be of a fixed length.
5493
5494
5495 BACKREFERENCES
5496
5497 \n reference by number (can be ambiguous)
5498 \gn reference by number
5499 \g{n} reference by number
5500 \g{-n} relative reference by number
5501 \k<name> reference by name (Perl)
5502 \k'name' reference by name (Perl)
5503 \g{name} reference by name (Perl)
5504 \k{name} reference by name (.NET)
5505 (?P=name) reference by name (Python)
5506
5507
5508 SUBROUTINE REFERENCES (POSSIBLY RECURSIVE)
5509
5510 (?R) recurse whole pattern
5511 (?n) call subpattern by absolute number
5512 (?+n) call subpattern by relative number
5513 (?-n) call subpattern by relative number
5514 (?&name) call subpattern by name (Perl)
5515 (?P>name) call subpattern by name (Python)
5516 \g<name> call subpattern by name (Oniguruma)
5517 \g'name' call subpattern by name (Oniguruma)
5518 \g<n> call subpattern by absolute number (Oniguruma)
5519 \g'n' call subpattern by absolute number (Oniguruma)
5520 \g<+n> call subpattern by relative number (PCRE extension)
5521 \g'+n' call subpattern by relative number (PCRE extension)
5522 \g<-n> call subpattern by relative number (PCRE extension)
5523 \g'-n' call subpattern by relative number (PCRE extension)
5524
5525
5526 CONDITIONAL PATTERNS
5527
5528 (?(condition)yes-pattern)
5529 (?(condition)yes-pattern|no-pattern)
5530
5531 (?(n)... absolute reference condition
5532 (?(+n)... relative reference condition
5533 (?(-n)... relative reference condition
5534 (?(<name>)... named reference condition (Perl)
5535 (?('name')... named reference condition (Perl)
5536 (?(name)... named reference condition (PCRE)
5537 (?(R)... overall recursion condition
5538 (?(Rn)... specific group recursion condition
5539 (?(R&name)... specific recursion condition
5540 (?(DEFINE)... define subpattern for reference
5541 (?(assert)... assertion condition
5542
5543
5544 BACKTRACKING CONTROL
5545
5546 The following act immediately they are reached:
5547
5548 (*ACCEPT) force successful match
5549 (*FAIL) force backtrack; synonym (*F)
5550
5551 The following act only when a subsequent match failure causes a back-
5552 track to reach them. They all force a match failure, but they differ in
5553 what happens afterwards. Those that advance the start-of-match point do
5554 so only if the pattern is not anchored.
5555
5556 (*COMMIT) overall failure, no advance of starting point
5557 (*PRUNE) advance to next starting character
5558 (*SKIP) advance start to current matching position
5559 (*THEN) local failure, backtrack to next alternation
5560
5561
5562 NEWLINE CONVENTIONS
5563
5564 These are recognized only at the very start of the pattern or after a
5565 (*BSR_...) or (*UTF8) option.
5566
5567 (*CR) carriage return only
5568 (*LF) linefeed only
5569 (*CRLF) carriage return followed by linefeed
5570 (*ANYCRLF) all three of the above
5571 (*ANY) any Unicode newline sequence
5572
5573
5574 WHAT \R MATCHES
5575
5576 These are recognized only at the very start of the pattern or after a
5577 (*...) option that sets the newline convention or UTF-8 mode.
5578
5579 (*BSR_ANYCRLF) CR, LF, or CRLF
5580 (*BSR_UNICODE) any Unicode newline sequence
5581
5582
5583 CALLOUTS
5584
5585 (?C) callout
5586 (?Cn) callout with data n
5587
5588
5589 SEE ALSO
5590
5591 pcrepattern(3), pcreapi(3), pcrecallout(3), pcrematching(3), pcre(3).
5592
5593
5594 AUTHOR
5595
5596 Philip Hazel
5597 University Computing Service
5598 Cambridge CB2 3QH, England.
5599
5600
5601 REVISION
5602
5603 Last updated: 11 April 2009
5604 Copyright (c) 1997-2009 University of Cambridge.
5605 ------------------------------------------------------------------------------
5606
5607
5608 PCREPARTIAL(3) PCREPARTIAL(3)
5609
5610
5611 NAME
5612 PCRE - Perl-compatible regular expressions
5613
5614
5615 PARTIAL MATCHING IN PCRE
5616
5617 In normal use of PCRE, if the subject string that is passed to
5618 pcre_exec() or pcre_dfa_exec() matches as far as it goes, but is too
5619 short to match the entire pattern, PCRE_ERROR_NOMATCH is returned.
5620 There are circumstances where it might be helpful to distinguish this
5621 case from other cases in which there is no match.
5622
5623 Consider, for example, an application where a human is required to type
5624 in data for a field with specific formatting requirements. An example
5625 might be a date in the form ddmmmyy, defined by this pattern:
5626
5627 ^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$
5628
5629 If the application sees the user's keystrokes one by one, and can check
5630 that what has been typed so far is potentially valid, it is able to
5631 raise an error as soon as a mistake is made, by beeping and not
5632 reflecting the character that has been typed, for example. This immedi-
5633 ate feedback is likely to be a better user interface than a check that
5634 is delayed until the entire string has been entered. Partial matching
5635 can also sometimes be useful when the subject string is very long and
5636 is not all available at once.
5637
5638 PCRE supports partial matching by means of the PCRE_PARTIAL_SOFT and
5639 PCRE_PARTIAL_HARD options, which can be set when calling pcre_exec() or
5640 pcre_dfa_exec(). For backwards compatibility, PCRE_PARTIAL is a synonym
5641 for PCRE_PARTIAL_SOFT. The essential difference between the two options
5642 is whether or not a partial match is preferred to an alternative com-
5643 plete match, though the details differ between the two matching func-
5644 tions. If both options are set, PCRE_PARTIAL_HARD takes precedence.
5645
5646 Setting a partial matching option disables two of PCRE's optimizations.
5647 PCRE remembers the last literal byte in a pattern, and abandons match-
5648 ing immediately if such a byte is not present in the subject string.
5649 This optimization cannot be used for a subject string that might match
5650 only partially. If the pattern was studied, PCRE knows the minimum
5651 length of a matching string, and does not bother to run the matching
5652 function on shorter strings. This optimization is also disabled for
5653 partial matching.
5654
5655
5656 PARTIAL MATCHING USING pcre_exec()
5657
5658 A partial match occurs during a call to pcre_exec() whenever the end of
5659 the subject string is reached successfully, but matching cannot con-
5660 tinue because more characters are needed. However, at least one charac-
5661 ter must have been matched. (In other words, a partial match can never
5662 be an empty string.)
5663
5664 If PCRE_PARTIAL_SOFT is set, the partial match is remembered, but
5665 matching continues as normal, and other alternatives in the pattern are
5666 tried. If no complete match can be found, pcre_exec() returns
5667 PCRE_ERROR_PARTIAL instead of PCRE_ERROR_NOMATCH. If there are at least
5668 two slots in the offsets vector, the first of them is set to the offset
5669 of the earliest character that was inspected when the partial match was
5670 found. For convenience, the second offset points to the end of the
5671 string so that a substring can easily be identified.
5672
5673 For the majority of patterns, the first offset identifies the start of
5674 the partially matched string. However, for patterns that contain look-
5675 behind assertions, or \K, or begin with \b or \B, earlier characters
5676 have been inspected while carrying out the match. For example:
5677
5678 /(?<=abc)123/
5679
5680 This pattern matches "123", but only if it is preceded by "abc". If the
5681 subject string is "xyzabc12", the offsets after a partial match are for
5682 the substring "abc12", because all these characters are needed if
5683 another match is tried with extra characters added.
5684
5685 If there is more than one partial match, the first one that was found
5686 provides the data that is returned. Consider this pattern:
5687
5688 /123\w+X|dogY/
5689
5690 If this is matched against the subject string "abc123dog", both alter-
5691 natives fail to match, but the end of the subject is reached during
5692 matching, so PCRE_ERROR_PARTIAL is returned instead of
5693 PCRE_ERROR_NOMATCH. The offsets are set to 3 and 9, identifying
5694 "123dog" as the first partial match that was found. (In this example,
5695 there are two partial matches, because "dog" on its own partially
5696 matches the second alternative.)
5697
5698 If PCRE_PARTIAL_HARD is set for pcre_exec(), it returns PCRE_ERROR_PAR-
5699 TIAL as soon as a partial match is found, without continuing to search
5700 for possible complete matches. The difference between the two options
5701 can be illustrated by a pattern such as:
5702
5703 /dog(sbody)?/
5704
5705 This matches either "dog" or "dogsbody", greedily (that is, it prefers
5706 the longer string if possible). If it is matched against the string
5707 "dog" with PCRE_PARTIAL_SOFT, it yields a complete match for "dog".
5708 However, if PCRE_PARTIAL_HARD is set, the result is PCRE_ERROR_PARTIAL.
5709 On the other hand, if the pattern is made ungreedy the result is dif-
5710 ferent:
5711
5712 /dog(sbody)??/
5713
5714 In this case the result is always a complete match because pcre_exec()
5715 finds that first, and it never continues after finding a match. It
5716 might be easier to follow this explanation by thinking of the two pat-
5717 terns like this:
5718
5719 /dog(sbody)?/ is the same as /dogsbody|dog/
5720 /dog(sbody)??/ is the same as /dog|dogsbody/
5721
5722 The second pattern will never match "dogsbody" when pcre_exec() is
5723 used, because it will always find the shorter match first.
5724
5725
5726 PARTIAL MATCHING USING pcre_dfa_exec()
5727
5728 The pcre_dfa_exec() function moves along the subject string character
5729 by character, without backtracking, searching for all possible matches
5730 simultaneously. If the end of the subject is reached before the end of
5731 the pattern, there is the possibility of a partial match, again pro-
5732 vided that at least one character has matched.
5733
5734 When PCRE_PARTIAL_SOFT is set, PCRE_ERROR_PARTIAL is returned only if
5735 there have been no complete matches. Otherwise, the complete matches
5736 are returned. However, if PCRE_PARTIAL_HARD is set, a partial match
5737 takes precedence over any complete matches. The portion of the string
5738 that was inspected when the longest partial match was found is set as
5739 the first matching string, provided there are at least two slots in the
5740 offsets vector.
5741
5742 Because pcre_dfa_exec() always searches for all possible matches, and
5743 there is no difference between greedy and ungreedy repetition, its be-
5744 haviour is different from pcre_exec when PCRE_PARTIAL_HARD is set. Con-
5745 sider the string "dog" matched against the ungreedy pattern shown
5746 above:
5747
5748 /dog(sbody)??/
5749
5750 Whereas pcre_exec() stops as soon as it finds the complete match for
5751 "dog", pcre_dfa_exec() also finds the partial match for "dogsbody", and
5752 so returns that when PCRE_PARTIAL_HARD is set.
5753
5754
5755 PARTIAL MATCHING AND WORD BOUNDARIES
5756
5757 If a pattern ends with one of sequences \w or \W, which test for word
5758 boundaries, partial matching with PCRE_PARTIAL_SOFT can give counter-
5759 intuitive results. Consider this pattern:
5760
5761 /\bcat\b/
5762
5763 This matches "cat", provided there is a word boundary at either end. If
5764 the subject string is "the cat", the comparison of the final "t" with a
5765 following character cannot take place, so a partial match is found.
5766 However, pcre_exec() carries on with normal matching, which matches \b
5767 at the end of the subject when the last character is a letter, thus
5768 finding a complete match. The result, therefore, is not PCRE_ERROR_PAR-
5769 TIAL. The same thing happens with pcre_dfa_exec(), because it also
5770 finds the complete match.
5771
5772 Using PCRE_PARTIAL_HARD in this case does yield PCRE_ERROR_PARTIAL,
5773 because then the partial match takes precedence.
5774
5775
5776 FORMERLY RESTRICTED PATTERNS
5777
5778 For releases of PCRE prior to 8.00, because of the way certain internal
5779 optimizations were implemented in the pcre_exec() function, the
5780 PCRE_PARTIAL option (predecessor of PCRE_PARTIAL_SOFT) could not be
5781 used with all patterns. From release 8.00 onwards, the restrictions no
5782 longer apply, and partial matching with pcre_exec() can be requested
5783 for any pattern.
5784
5785 Items that were formerly restricted were repeated single characters and
5786 repeated metasequences. If PCRE_PARTIAL was set for a pattern that did
5787 not conform to the restrictions, pcre_exec() returned the error code
5788 PCRE_ERROR_BADPARTIAL (-13). This error code is no longer in use. The
5789 PCRE_INFO_OKPARTIAL call to pcre_fullinfo() to find out if a compiled
5790 pattern can be used for partial matching now always returns 1.
5791
5792
5793 EXAMPLE OF PARTIAL MATCHING USING PCRETEST
5794
5795 If the escape sequence \P is present in a pcretest data line, the
5796 PCRE_PARTIAL_SOFT option is used for the match. Here is a run of
5797 pcretest that uses the date example quoted above:
5798
5799 re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/
5800 data> 25jun04\P
5801 0: 25jun04
5802 1: jun
5803 data> 25dec3\P
5804 Partial match: 23dec3
5805 data> 3ju\P
5806 Partial match: 3ju
5807 data> 3juj\P
5808 No match
5809 data> j\P
5810 No match
5811
5812 The first data string is matched completely, so pcretest shows the
5813 matched substrings. The remaining four strings do not match the com-
5814 plete pattern, but the first two are partial matches. Similar output is
5815 obtained when pcre_dfa_exec() is used.
5816
5817 If the escape sequence \P is present more than once in a pcretest data
5818 line, the PCRE_PARTIAL_HARD option is set for the match.
5819
5820
5821 MULTI-SEGMENT MATCHING WITH pcre_dfa_exec()
5822
5823 When a partial match has been found using pcre_dfa_exec(), it is possi-
5824 ble to continue the match by providing additional subject data and
5825 calling pcre_dfa_exec() again with the same compiled regular expres-
5826 sion, this time setting the PCRE_DFA_RESTART option. You must pass the
5827 same working space as before, because this is where details of the pre-
5828 vious partial match are stored. Here is an example using pcretest,
5829 using the \R escape sequence to set the PCRE_DFA_RESTART option (\D
5830 specifies the use of pcre_dfa_exec()):
5831
5832 re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/
5833 data> 23ja\P\D
5834 Partial match: 23ja
5835 data> n05\R\D
5836 0: n05
5837
5838 The first call has "23ja" as the subject, and requests partial match-
5839 ing; the second call has "n05" as the subject for the continued
5840 (restarted) match. Notice that when the match is complete, only the
5841 last part is shown; PCRE does not retain the previously partially-
5842 matched string. It is up to the calling program to do that if it needs
5843 to.
5844
5845 You can set the PCRE_PARTIAL_SOFT or PCRE_PARTIAL_HARD options with
5846 PCRE_DFA_RESTART to continue partial matching over multiple segments.
5847 This facility can be used to pass very long subject strings to
5848 pcre_dfa_exec().
5849
5850
5851 MULTI-SEGMENT MATCHING WITH pcre_exec()
5852
5853 From release 8.00, pcre_exec() can also be used to do multi-segment
5854 matching. Unlike pcre_dfa_exec(), it is not possible to restart the
5855 previous match with a new segment of data. Instead, new data must be
5856 added to the previous subject string, and the entire match re-run,
5857 starting from the point where the partial match occurred. Earlier data
5858 can be discarded. Consider an unanchored pattern that matches dates:
5859
5860 re> /\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d/
5861 data> The date is 23ja\P
5862 Partial match: 23ja
5863
5864 The this stage, an application could discard the text preceding "23ja",
5865 add on text from the next segment, and call pcre_exec() again. Unlike
5866 pcre_dfa_exec(), the entire matching string must always be available,
5867 and the complete matching process occurs for each call, so more memory
5868 and more processing time is needed.
5869
5870 Note: If the pattern contains lookbehind assertions, or \K, or starts
5871 with \b or \B, the string that is returned for a partial match will
5872 include characters that precede the partially matched string itself,
5873 because these must be retained when adding on more characters for a
5874 subsequent matching attempt.
5875
5876
5877 ISSUES WITH MULTI-SEGMENT MATCHING
5878
5879 Certain types of pattern may give problems with multi-segment matching,
5880 whichever matching function is used.
5881
5882 1. If the pattern contains tests for the beginning or end of a line,
5883 you need to pass the PCRE_NOTBOL or PCRE_NOTEOL options, as appropri-
5884 ate, when the subject string for any call does not contain the begin-
5885 ning or end of a line.
5886
5887 2. Lookbehind assertions at the start of a pattern are catered for in
5888 the offsets that are returned for a partial match. However, in theory,
5889 a lookbehind assertion later in the pattern could require even earlier
5890 characters to be inspected, and it might not have been reached when a
5891 partial match occurs. This is probably an extremely unlikely case; you
5892 could guard against it to a certain extent by always including extra
5893 characters at the start.
5894
5895 3. Matching a subject string that is split into multiple segments may
5896 not always produce exactly the same result as matching over one single
5897 long string, especially when PCRE_PARTIAL_SOFT is used. The section
5898 "Partial Matching and Word Boundaries" above describes an issue that
5899 arises if the pattern ends with \b or \B. Another kind of difference
5900 may occur when there are multiple matching possibilities, because a
5901 partial match result is given only when there are no completed matches.
5902 This means that as soon as the shortest match has been found, continua-
5903 tion to a new subject segment is no longer possible. Consider again
5904 this pcretest example:
5905
5906 re> /dog(sbody)?/
5907 data> dogsb\P
5908 0: dog
5909 data> do\P\D
5910 Partial match: do
5911 data> gsb\R\P\D
5912 0: g
5913 data> dogsbody\D
5914 0: dogsbody
5915 1: dog
5916
5917 The first data line passes the string "dogsb" to pcre_exec(), setting
5918 the PCRE_PARTIAL_SOFT option. Although the string is a partial match
5919 for "dogsbody", the result is not PCRE_ERROR_PARTIAL, because the
5920 shorter string "dog" is a complete match. Similarly, when the subject
5921 is presented to pcre_dfa_exec() in several parts ("do" and "gsb" being
5922 the first two) the match stops when "dog" has been found, and it is not
5923 possible to continue. On the other hand, if "dogsbody" is presented as
5924 a single string, pcre_dfa_exec() finds both matches.
5925
5926 Because of these problems, it is probably best to use PCRE_PARTIAL_HARD
5927 when matching multi-segment data. The example above then behaves dif-
5928 ferently:
5929
5930 re> /dog(sbody)?/
5931 data> dogsb\P\P
5932 Partial match: dogsb
5933 data> do\P\D
5934 Partial match: do
5935 data> gsb\R\P\P\D
5936 Partial match: gsb
5937
5938
5939 4. Patterns that contain alternatives at the top level which do not all
5940 start with the same pattern item may not work as expected when
5941 pcre_dfa_exec() is used. For example, consider this pattern:
5942
5943 1234|3789
5944
5945 If the first part of the subject is "ABC123", a partial match of the
5946 first alternative is found at offset 3. There is no partial match for
5947 the second alternative, because such a match does not start at the same
5948 point in the subject string. Attempting to continue with the string
5949 "7890" does not yield a match because only those alternatives that
5950 match at one point in the subject are remembered. The problem arises
5951 because the start of the second alternative matches within the first
5952 alternative. There is no problem with anchored patterns or patterns
5953 such as:
5954
5955 1234|ABCD
5956
5957 where no string can be a partial match for both alternatives. This is
5958 not a problem if pcre_exec() is used, because the entire match has to
5959 be rerun each time:
5960
5961 re> /1234|3789/
5962 data> ABC123\P
5963 Partial match: 123
5964 data> 1237890
5965 0: 3789
5966
5967
5968 AUTHOR
5969
5970 Philip Hazel
5971 University Computing Service
5972 Cambridge CB2 3QH, England.
5973
5974
5975 REVISION
5976
5977 Last updated: 29 September 2009
5978 Copyright (c) 1997-2009 University of Cambridge.
5979 ------------------------------------------------------------------------------
5980
5981
5982 PCREPRECOMPILE(3) PCREPRECOMPILE(3)
5983
5984
5985 NAME
5986 PCRE - Perl-compatible regular expressions
5987
5988
5989 SAVING AND RE-USING PRECOMPILED PCRE PATTERNS
5990
5991 If you are running an application that uses a large number of regular
5992 expression patterns, it may be useful to store them in a precompiled
5993 form instead of having to compile them every time the application is
5994 run. If you are not using any private character tables (see the
5995 pcre_maketables() documentation), this is relatively straightforward.
5996 If you are using private tables, it is a little bit more complicated.
5997
5998 If you save compiled patterns to a file, you can copy them to a differ-
5999 ent host and run them there. This works even if the new host has the
6000 opposite endianness to the one on which the patterns were compiled.
6001 There may be a small performance penalty, but it should be insignifi-
6002 cant. However, compiling regular expressions with one version of PCRE
6003 for use with a different version is not guaranteed to work and may
6004 cause crashes.
6005
6006
6007 SAVING A COMPILED PATTERN
6008 The value returned by pcre_compile() points to a single block of memory
6009 that holds the compiled pattern and associated data. You can find the
6010 length of this block in bytes by calling pcre_fullinfo() with an argu-
6011 ment of PCRE_INFO_SIZE. You can then save the data in any appropriate
6012 manner. Here is sample code that compiles a pattern and writes it to a
6013 file. It assumes that the variable fd refers to a file that is open for
6014 output:
6015
6016 int erroroffset, rc, size;
6017 char *error;
6018 pcre *re;
6019
6020 re = pcre_compile("my pattern", 0, &error, &erroroffset, NULL);
6021 if (re == NULL) { ... handle errors ... }
6022 rc = pcre_fullinfo(re, NULL, PCRE_INFO_SIZE, &size);
6023 if (rc < 0) { ... handle errors ... }
6024 rc = fwrite(re, 1, size, fd);
6025 if (rc != size) { ... handle errors ... }
6026
6027 In this example, the bytes that comprise the compiled pattern are
6028 copied exactly. Note that this is binary data that may contain any of
6029 the 256 possible byte values. On systems that make a distinction
6030 between binary and non-binary data, be sure that the file is opened for
6031 binary output.
6032
6033 If you want to write more than one pattern to a file, you will have to
6034 devise a way of separating them. For binary data, preceding each pat-
6035 tern with its length is probably the most straightforward approach.
6036 Another possibility is to write out the data in hexadecimal instead of
6037 binary, one pattern to a line.
6038
6039 Saving compiled patterns in a file is only one possible way of storing
6040 them for later use. They could equally well be saved in a database, or
6041 in the memory of some daemon process that passes them via sockets to
6042 the processes that want them.
6043
6044 If the pattern has been studied, it is also possible to save the study
6045 data in a similar way to the compiled pattern itself. When studying
6046 generates additional information, pcre_study() returns a pointer to a
6047 pcre_extra data block. Its format is defined in the section on matching
6048 a pattern in the pcreapi documentation. The study_data field points to
6049 the binary study data, and this is what you must save (not the
6050 pcre_extra block itself). The length of the study data can be obtained
6051 by calling pcre_fullinfo() with an argument of PCRE_INFO_STUDYSIZE.