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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 454 - (show annotations)
Tue Sep 22 09:42:11 2009 UTC (10 years ago) by ph10
File MIME type: text/plain
File size: 328434 byte(s)
Error occurred while calculating annotation data.
Allow fixed-length subroutine calls in lookbehinds.
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. Certain features that appeared in Python and
23 PCRE before they appeared in Perl are also available using the Python
24 syntax. There is also some support for certain .NET and Oniguruma syn-
25 tax items, and there is an option for requesting some minor changes
26 that give better JavaScript compatibility.
27
28 The current implementation of PCRE (release 8.xx) corresponds approxi-
29 mately with Perl 5.10, including support for UTF-8 encoded strings and
30 Unicode general category properties. However, UTF-8 and Unicode support
31 has to be explicitly enabled; it is not the default. The Unicode tables
32 correspond to Unicode release 5.1.
33
34 In addition to the Perl-compatible matching function, PCRE contains an
35 alternative matching function that matches the same compiled patterns
36 in a different way. In certain circumstances, the alternative function
37 has some advantages. For a discussion of the two matching algorithms,
38 see the 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 file
59 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 pcresyntax quick syntax reference
93 pcreperform discussion of performance issues
94 pcreposix the POSIX-compatible C API
95 pcreprecompile details of saving and re-using precompiled patterns
96 pcresample discussion of the pcredemo program
97 pcrestack discussion of stack usage
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 just strings of bytes.
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: 01 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 function.
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 enormous patterns, so it
436 is possible to compile PCRE to use three-byte or four-byte offsets by
437 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.
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 the the
473 pcre_dfa_exec() function.
474
475
476 LIMITING PCRE RESOURCE USAGE
477
478 Internally, PCRE has a function called match(), which it calls repeat-
479 edly (sometimes recursively) when matching a pattern with the
480 pcre_exec() function. By controlling the maximum number of times this
481 function may be called during a single matching operation, a limit can
482 be placed on the resources used by a single call to pcre_exec(). The
483 limit can be changed at run time, as described in the pcreapi documen-
484 tation. The default is 10 million, but this can be changed by adding a
485 setting such as
486
487 --with-match-limit=500000
488
489 to the configure command. This setting has no effect on the
490 pcre_dfa_exec() matching function.
491
492 In some environments it is desirable to limit the depth of recursive
493 calls of match() more strictly than the total number of calls, in order
494 to restrict the maximum amount of stack (or heap, if --disable-stack-
495 for-recursion is specified) that is used. A second limit controls this;
496 it defaults to the value that is set for --with-match-limit, which
497 imposes no additional constraints. However, you can set a lower limit
498 by adding, for example,
499
500 --with-match-limit-recursion=10000
501
502 to the configure command. This value can also be overridden at run
503 time.
504
505
506 CREATING CHARACTER TABLES AT BUILD TIME
507
508 PCRE uses fixed tables for processing characters whose code values are
509 less than 256. By default, PCRE is built with a set of tables that are
510 distributed in the file pcre_chartables.c.dist. These tables are for
511 ASCII codes only. If you add
512
513 --enable-rebuild-chartables
514
515 to the configure command, the distributed tables are no longer used.
516 Instead, a program called dftables is compiled and run. This outputs
517 the source for new set of tables, created in the default locale of your
518 C runtime system. (This method of replacing the tables does not work if
519 you are cross compiling, because dftables is run on the local host. If
520 you need to create alternative tables when cross compiling, you will
521 have to do so "by hand".)
522
523
524 USING EBCDIC CODE
525
526 PCRE assumes by default that it will run in an environment where the
527 character code is ASCII (or Unicode, which is a superset of ASCII).
528 This is the case for most computer operating systems. PCRE can, how-
529 ever, be compiled to run in an EBCDIC environment by adding
530
531 --enable-ebcdic
532
533 to the configure command. This setting implies --enable-rebuild-charta-
534 bles. You should only use it if you know that you are in an EBCDIC
535 environment (for example, an IBM mainframe operating system). The
536 --enable-ebcdic option is incompatible with --enable-utf8.
537
538
539 PCREGREP OPTIONS FOR COMPRESSED FILE SUPPORT
540
541 By default, pcregrep reads all files as plain text. You can build it so
542 that it recognizes files whose names end in .gz or .bz2, and reads them
543 with libz or libbz2, respectively, by adding one or both of
544
545 --enable-pcregrep-libz
546 --enable-pcregrep-libbz2
547
548 to the configure command. These options naturally require that the rel-
549 evant libraries are installed on your system. Configuration will fail
550 if they are not.
551
552
553 PCRETEST OPTION FOR LIBREADLINE SUPPORT
554
555 If you add
556
557 --enable-pcretest-libreadline
558
559 to the configure command, pcretest is linked with the libreadline
560 library, and when its input is from a terminal, it reads it using the
561 readline() function. This provides line-editing and history facilities.
562 Note that libreadline is GPL-licenced, so if you distribute a binary of
563 pcretest linked in this way, there may be licensing issues.
564
565 Setting this option causes the -lreadline option to be added to the
566 pcretest build. In many operating environments with a sytem-installed
567 libreadline this is sufficient. However, in some environments (e.g. if
568 an unmodified distribution version of readline is in use), some extra
569 configuration may be necessary. The INSTALL file for libreadline says
570 this:
571
572 "Readline uses the termcap functions, but does not link with the
573 termcap or curses library itself, allowing applications which link
574 with readline the to choose an appropriate library."
575
576 If your environment has not been set up so that an appropriate library
577 is automatically included, you may need to add something like
578
579 LIBS="-ncurses"
580
581 immediately before the configure command.
582
583
584 SEE ALSO
585
586 pcreapi(3), pcre_config(3).
587
588
589 AUTHOR
590
591 Philip Hazel
592 University Computing Service
593 Cambridge CB2 3QH, England.
594
595
596 REVISION
597
598 Last updated: 06 September 2009
599 Copyright (c) 1997-2009 University of Cambridge.
600 ------------------------------------------------------------------------------
601
602
603 PCREMATCHING(3) PCREMATCHING(3)
604
605
606 NAME
607 PCRE - Perl-compatible regular expressions
608
609
610 PCRE MATCHING ALGORITHMS
611
612 This document describes the two different algorithms that are available
613 in PCRE for matching a compiled regular expression against a given sub-
614 ject string. The "standard" algorithm is the one provided by the
615 pcre_exec() function. This works in the same was as Perl's matching
616 function, and provides a Perl-compatible matching operation.
617
618 An alternative algorithm is provided by the pcre_dfa_exec() function;
619 this operates in a different way, and is not Perl-compatible. It has
620 advantages and disadvantages compared with the standard algorithm, and
621 these are described below.
622
623 When there is only one possible way in which a given subject string can
624 match a pattern, the two algorithms give the same answer. A difference
625 arises, however, when there are multiple possibilities. For example, if
626 the pattern
627
628 ^<.*>
629
630 is matched against the string
631
632 <something> <something else> <something further>
633
634 there are three possible answers. The standard algorithm finds only one
635 of them, whereas the alternative algorithm finds all three.
636
637
638 REGULAR EXPRESSIONS AS TREES
639
640 The set of strings that are matched by a regular expression can be rep-
641 resented as a tree structure. An unlimited repetition in the pattern
642 makes the tree of infinite size, but it is still a tree. Matching the
643 pattern to a given subject string (from a given starting point) can be
644 thought of as a search of the tree. There are two ways to search a
645 tree: depth-first and breadth-first, and these correspond to the two
646 matching algorithms provided by PCRE.
647
648
649 THE STANDARD MATCHING ALGORITHM
650
651 In the terminology of Jeffrey Friedl's book "Mastering Regular Expres-
652 sions", the standard algorithm is an "NFA algorithm". It conducts a
653 depth-first search of the pattern tree. That is, it proceeds along a
654 single path through the tree, checking that the subject matches what is
655 required. When there is a mismatch, the algorithm tries any alterna-
656 tives at the current point, and if they all fail, it backs up to the
657 previous branch point in the tree, and tries the next alternative
658 branch at that level. This often involves backing up (moving to the
659 left) in the subject string as well. The order in which repetition
660 branches are tried is controlled by the greedy or ungreedy nature of
661 the quantifier.
662
663 If a leaf node is reached, a matching string has been found, and at
664 that point the algorithm stops. Thus, if there is more than one possi-
665 ble match, this algorithm returns the first one that it finds. Whether
666 this is the shortest, the longest, or some intermediate length depends
667 on the way the greedy and ungreedy repetition quantifiers are specified
668 in the pattern.
669
670 Because it ends up with a single path through the tree, it is rela-
671 tively straightforward for this algorithm to keep track of the sub-
672 strings that are matched by portions of the pattern in parentheses.
673 This provides support for capturing parentheses and back references.
674
675
676 THE ALTERNATIVE MATCHING ALGORITHM
677
678 This algorithm conducts a breadth-first search of the tree. Starting
679 from the first matching point in the subject, it scans the subject
680 string from left to right, once, character by character, and as it does
681 this, it remembers all the paths through the tree that represent valid
682 matches. In Friedl's terminology, this is a kind of "DFA algorithm",
683 though it is not implemented as a traditional finite state machine (it
684 keeps multiple states active simultaneously).
685
686 The scan continues until either the end of the subject is reached, or
687 there are no more unterminated paths. At this point, terminated paths
688 represent the different matching possibilities (if there are none, the
689 match has failed). Thus, if there is more than one possible match,
690 this algorithm finds all of them, and in particular, it finds the long-
691 est. In PCRE, there is an option to stop the algorithm after the first
692 match (which is necessarily the shortest) has been found.
693
694 Note that all the matches that are found start at the same point in the
695 subject. If the pattern
696
697 cat(er(pillar)?)
698
699 is matched against the string "the caterpillar catchment", the result
700 will be the three strings "cat", "cater", and "caterpillar" that start
701 at the fourth character of the subject. The algorithm does not automat-
702 ically move on to find matches that start at later positions.
703
704 Although the general principle of this matching algorithm is that it
705 scans the subject string only once, without backtracking, there is one
706 exception: when a lookbehind assertion is encountered, the preceding
707 characters have to be re-inspected.
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.
771
772
773 DISADVANTAGES OF THE ALTERNATIVE ALGORITHM
774
775 The alternative algorithm suffers from a number of disadvantages:
776
777 1. It is substantially slower than the standard algorithm. This is
778 partly because it has to search for all possible matches, but is also
779 because it is less susceptible to optimization.
780
781 2. Capturing parentheses and back references are not supported.
782
783 3. Although atomic groups are supported, their use does not provide the
784 performance advantage that it does for the standard algorithm.
785
786
787 AUTHOR
788
789 Philip Hazel
790 University Computing Service
791 Cambridge CB2 3QH, England.
792
793
794 REVISION
795
796 Last updated: 05 September 2009
797 Copyright (c) 1997-2009 University of Cambridge.
798 ------------------------------------------------------------------------------
799
800
801 PCREAPI(3) PCREAPI(3)
802
803
804 NAME
805 PCRE - Perl-compatible regular expressions
806
807
808 PCRE NATIVE API
809
810 #include <pcre.h>
811
812 pcre *pcre_compile(const char *pattern, int options,
813 const char **errptr, int *erroffset,
814 const unsigned char *tableptr);
815
816 pcre *pcre_compile2(const char *pattern, int options,
817 int *errorcodeptr,
818 const char **errptr, int *erroffset,
819 const unsigned char *tableptr);
820
821 pcre_extra *pcre_study(const pcre *code, int options,
822 const char **errptr);
823
824 int pcre_exec(const pcre *code, const pcre_extra *extra,
825 const char *subject, int length, int startoffset,
826 int options, int *ovector, int ovecsize);
827
828 int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
829 const char *subject, int length, int startoffset,
830 int options, int *ovector, int ovecsize,
831 int *workspace, int wscount);
832
833 int pcre_copy_named_substring(const pcre *code,
834 const char *subject, int *ovector,
835 int stringcount, const char *stringname,
836 char *buffer, int buffersize);
837
838 int pcre_copy_substring(const char *subject, int *ovector,
839 int stringcount, int stringnumber, char *buffer,
840 int buffersize);
841
842 int pcre_get_named_substring(const pcre *code,
843 const char *subject, int *ovector,
844 int stringcount, const char *stringname,
845 const char **stringptr);
846
847 int pcre_get_stringnumber(const pcre *code,
848 const char *name);
849
850 int pcre_get_stringtable_entries(const pcre *code,
851 const char *name, char **first, char **last);
852
853 int pcre_get_substring(const char *subject, int *ovector,
854 int stringcount, int stringnumber,
855 const char **stringptr);
856
857 int pcre_get_substring_list(const char *subject,
858 int *ovector, int stringcount, const char ***listptr);
859
860 void pcre_free_substring(const char *stringptr);
861
862 void pcre_free_substring_list(const char **stringptr);
863
864 const unsigned char *pcre_maketables(void);
865
866 int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
867 int what, void *where);
868
869 int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
870
871 int pcre_refcount(pcre *code, int adjust);
872
873 int pcre_config(int what, void *where);
874
875 char *pcre_version(void);
876
877 void *(*pcre_malloc)(size_t);
878
879 void (*pcre_free)(void *);
880
881 void *(*pcre_stack_malloc)(size_t);
882
883 void (*pcre_stack_free)(void *);
884
885 int (*pcre_callout)(pcre_callout_block *);
886
887
888 PCRE API OVERVIEW
889
890 PCRE has its own native API, which is described in this document. There
891 are also some wrapper functions that correspond to the POSIX regular
892 expression API. These are described in the pcreposix documentation.
893 Both of these APIs define a set of C function calls. A C++ wrapper is
894 distributed with PCRE. It is documented in the pcrecpp page.
895
896 The native API C function prototypes are defined in the header file
897 pcre.h, and on Unix systems the library itself is called libpcre. It
898 can normally be accessed by adding -lpcre to the command for linking an
899 application that uses PCRE. The header file defines the macros
900 PCRE_MAJOR and PCRE_MINOR to contain the major and minor release num-
901 bers for the library. Applications can use these to include support
902 for different releases of PCRE.
903
904 The functions pcre_compile(), pcre_compile2(), pcre_study(), and
905 pcre_exec() are used for compiling and matching regular expressions in
906 a Perl-compatible manner. A sample program that demonstrates the sim-
907 plest way of using them is provided in the file called pcredemo.c in
908 the PCRE source distribution. A listing of this program is given in the
909 pcredemo documentation, and the pcresample documentation describes how
910 to compile and run it.
911
912 A second matching function, pcre_dfa_exec(), which is not Perl-compati-
913 ble, is also provided. This uses a different algorithm for the match-
914 ing. The alternative algorithm finds all possible matches (at a given
915 point in the subject), and scans the subject just once (unless there
916 are lookbehind assertions). However, this algorithm does not return
917 captured substrings. A description of the two matching algorithms and
918 their advantages and disadvantages is given in the pcrematching docu-
919 mentation.
920
921 In addition to the main compiling and matching functions, there are
922 convenience functions for extracting captured substrings from a subject
923 string that is matched by pcre_exec(). They are:
924
925 pcre_copy_substring()
926 pcre_copy_named_substring()
927 pcre_get_substring()
928 pcre_get_named_substring()
929 pcre_get_substring_list()
930 pcre_get_stringnumber()
931 pcre_get_stringtable_entries()
932
933 pcre_free_substring() and pcre_free_substring_list() are also provided,
934 to free the memory used for extracted strings.
935
936 The function pcre_maketables() is used to build a set of character
937 tables in the current locale for passing to pcre_compile(),
938 pcre_exec(), or pcre_dfa_exec(). This is an optional facility that is
939 provided for specialist use. Most commonly, no special tables are
940 passed, in which case internal tables that are generated when PCRE is
941 built are used.
942
943 The function pcre_fullinfo() is used to find out information about a
944 compiled pattern; pcre_info() is an obsolete version that returns only
945 some of the available information, but is retained for backwards com-
946 patibility. The function pcre_version() returns a pointer to a string
947 containing the version of PCRE and its date of release.
948
949 The function pcre_refcount() maintains a reference count in a data
950 block containing a compiled pattern. This is provided for the benefit
951 of object-oriented applications.
952
953 The global variables pcre_malloc and pcre_free initially contain the
954 entry points of the standard malloc() and free() functions, respec-
955 tively. PCRE calls the memory management functions via these variables,
956 so a calling program can replace them if it wishes to intercept the
957 calls. This should be done before calling any PCRE functions.
958
959 The global variables pcre_stack_malloc and pcre_stack_free are also
960 indirections to memory management functions. These special functions
961 are used only when PCRE is compiled to use the heap for remembering
962 data, instead of recursive function calls, when running the pcre_exec()
963 function. See the pcrebuild documentation for details of how to do
964 this. It is a non-standard way of building PCRE, for use in environ-
965 ments that have limited stacks. Because of the greater use of memory
966 management, it runs more slowly. Separate functions are provided so
967 that special-purpose external code can be used for this case. When
968 used, these functions are always called in a stack-like manner (last
969 obtained, first freed), and always for memory blocks of the same size.
970 There is a discussion about PCRE's stack usage in the pcrestack docu-
971 mentation.
972
973 The global variable pcre_callout initially contains NULL. It can be set
974 by the caller to a "callout" function, which PCRE will then call at
975 specified points during a matching operation. Details are given in the
976 pcrecallout documentation.
977
978
979 NEWLINES
980
981 PCRE supports five different conventions for indicating line breaks in
982 strings: a single CR (carriage return) character, a single LF (line-
983 feed) character, the two-character sequence CRLF, any of the three pre-
984 ceding, or any Unicode newline sequence. The Unicode newline sequences
985 are the three just mentioned, plus the single characters VT (vertical
986 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
987 separator, U+2028), and PS (paragraph separator, U+2029).
988
989 Each of the first three conventions is used by at least one operating
990 system as its standard newline sequence. When PCRE is built, a default
991 can be specified. The default default is LF, which is the Unix stan-
992 dard. When PCRE is run, the default can be overridden, either when a
993 pattern is compiled, or when it is matched.
994
995 At compile time, the newline convention can be specified by the options
996 argument of pcre_compile(), or it can be specified by special text at
997 the start of the pattern itself; this overrides any other settings. See
998 the pcrepattern page for details of the special character sequences.
999
1000 In the PCRE documentation the word "newline" is used to mean "the char-
1001 acter or pair of characters that indicate a line break". The choice of
1002 newline convention affects the handling of the dot, circumflex, and
1003 dollar metacharacters, the handling of #-comments in /x mode, and, when
1004 CRLF is a recognized line ending sequence, the match position advance-
1005 ment for a non-anchored pattern. There is more detail about this in the
1006 section on pcre_exec() options below.
1007
1008 The choice of newline convention does not affect the interpretation of
1009 the \n or \r escape sequences, nor does it affect what \R matches,
1010 which is controlled in a similar way, but by separate options.
1011
1012
1013 MULTITHREADING
1014
1015 The PCRE functions can be used in multi-threading applications, with
1016 the proviso that the memory management functions pointed to by
1017 pcre_malloc, pcre_free, pcre_stack_malloc, and pcre_stack_free, and the
1018 callout function pointed to by pcre_callout, are shared by all threads.
1019
1020 The compiled form of a regular expression is not altered during match-
1021 ing, so the same compiled pattern can safely be used by several threads
1022 at once.
1023
1024
1025 SAVING PRECOMPILED PATTERNS FOR LATER USE
1026
1027 The compiled form of a regular expression can be saved and re-used at a
1028 later time, possibly by a different program, and even on a host other
1029 than the one on which it was compiled. Details are given in the
1030 pcreprecompile documentation. However, compiling a regular expression
1031 with one version of PCRE for use with a different version is not guar-
1032 anteed to work and may cause crashes.
1033
1034
1035 CHECKING BUILD-TIME OPTIONS
1036
1037 int pcre_config(int what, void *where);
1038
1039 The function pcre_config() makes it possible for a PCRE client to dis-
1040 cover which optional features have been compiled into the PCRE library.
1041 The pcrebuild documentation has more details about these optional fea-
1042 tures.
1043
1044 The first argument for pcre_config() is an integer, specifying which
1045 information is required; the second argument is a pointer to a variable
1046 into which the information is placed. The following information is
1047 available:
1048
1049 PCRE_CONFIG_UTF8
1050
1051 The output is an integer that is set to one if UTF-8 support is avail-
1052 able; otherwise it is set to zero.
1053
1054 PCRE_CONFIG_UNICODE_PROPERTIES
1055
1056 The output is an integer that is set to one if support for Unicode
1057 character properties is available; otherwise it is set to zero.
1058
1059 PCRE_CONFIG_NEWLINE
1060
1061 The output is an integer whose value specifies the default character
1062 sequence that is recognized as meaning "newline". The four values that
1063 are supported are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF,
1064 and -1 for ANY. Though they are derived from ASCII, the same values
1065 are returned in EBCDIC environments. The default should normally corre-
1066 spond to the standard sequence for your operating system.
1067
1068 PCRE_CONFIG_BSR
1069
1070 The output is an integer whose value indicates what character sequences
1071 the \R escape sequence matches by default. A value of 0 means that \R
1072 matches any Unicode line ending sequence; a value of 1 means that \R
1073 matches only CR, LF, or CRLF. The default can be overridden when a pat-
1074 tern is compiled or matched.
1075
1076 PCRE_CONFIG_LINK_SIZE
1077
1078 The output is an integer that contains the number of bytes used for
1079 internal linkage in compiled regular expressions. The value is 2, 3, or
1080 4. Larger values allow larger regular expressions to be compiled, at
1081 the expense of slower matching. The default value of 2 is sufficient
1082 for all but the most massive patterns, since it allows the compiled
1083 pattern to be up to 64K in size.
1084
1085 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1086
1087 The output is an integer that contains the threshold above which the
1088 POSIX interface uses malloc() for output vectors. Further details are
1089 given in the pcreposix documentation.
1090
1091 PCRE_CONFIG_MATCH_LIMIT
1092
1093 The output is a long integer that gives the default limit for the num-
1094 ber of internal matching function calls in a pcre_exec() execution.
1095 Further details are given with pcre_exec() below.
1096
1097 PCRE_CONFIG_MATCH_LIMIT_RECURSION
1098
1099 The output is a long integer that gives the default limit for the depth
1100 of recursion when calling the internal matching function in a
1101 pcre_exec() execution. Further details are given with pcre_exec()
1102 below.
1103
1104 PCRE_CONFIG_STACKRECURSE
1105
1106 The output is an integer that is set to one if internal recursion when
1107 running pcre_exec() is implemented by recursive function calls that use
1108 the stack to remember their state. This is the usual way that PCRE is
1109 compiled. The output is zero if PCRE was compiled to use blocks of data
1110 on the heap instead of recursive function calls. In this case,
1111 pcre_stack_malloc and pcre_stack_free are called to manage memory
1112 blocks on the heap, thus avoiding the use of the stack.
1113
1114
1115 COMPILING A PATTERN
1116
1117 pcre *pcre_compile(const char *pattern, int options,
1118 const char **errptr, int *erroffset,
1119 const unsigned char *tableptr);
1120
1121 pcre *pcre_compile2(const char *pattern, int options,
1122 int *errorcodeptr,
1123 const char **errptr, int *erroffset,
1124 const unsigned char *tableptr);
1125
1126 Either of the functions pcre_compile() or pcre_compile2() can be called
1127 to compile a pattern into an internal form. The only difference between
1128 the two interfaces is that pcre_compile2() has an additional argument,
1129 errorcodeptr, via which a numerical error code can be returned.
1130
1131 The pattern is a C string terminated by a binary zero, and is passed in
1132 the pattern argument. A pointer to a single block of memory that is
1133 obtained via pcre_malloc is returned. This contains the compiled code
1134 and related data. The pcre type is defined for the returned block; this
1135 is a typedef for a structure whose contents are not externally defined.
1136 It is up to the caller to free the memory (via pcre_free) when it is no
1137 longer required.
1138
1139 Although the compiled code of a PCRE regex is relocatable, that is, it
1140 does not depend on memory location, the complete pcre data block is not
1141 fully relocatable, because it may contain a copy of the tableptr argu-
1142 ment, which is an address (see below).
1143
1144 The options argument contains various bit settings that affect the com-
1145 pilation. It should be zero if no options are required. The available
1146 options are described below. Some of them (in particular, those that
1147 are compatible with Perl, but also some others) can also be set and
1148 unset from within the pattern (see the detailed description in the
1149 pcrepattern documentation). For those options that can be different in
1150 different parts of the pattern, the contents of the options argument
1151 specifies their initial settings at the start of compilation and execu-
1152 tion. The PCRE_ANCHORED and PCRE_NEWLINE_xxx options can be set at the
1153 time of matching as well as at compile time.
1154
1155 If errptr is NULL, pcre_compile() returns NULL immediately. Otherwise,
1156 if compilation of a pattern fails, pcre_compile() returns NULL, and
1157 sets the variable pointed to by errptr to point to a textual error mes-
1158 sage. This is a static string that is part of the library. You must not
1159 try to free it. The byte offset from the start of the pattern to the
1160 character that was being processes when the error was discovered is
1161 placed in the variable pointed to by erroffset, which must not be NULL.
1162 If it is, an immediate error is given. Some errors are not detected
1163 until checks are carried out when the whole pattern has been scanned;
1164 in this case the offset is set to the end of the pattern.
1165
1166 If pcre_compile2() is used instead of pcre_compile(), and the error-
1167 codeptr argument is not NULL, a non-zero error code number is returned
1168 via this argument in the event of an error. This is in addition to the
1169 textual error message. Error codes and messages are listed below.
1170
1171 If the final argument, tableptr, is NULL, PCRE uses a default set of
1172 character tables that are built when PCRE is compiled, using the
1173 default C locale. Otherwise, tableptr must be an address that is the
1174 result of a call to pcre_maketables(). This value is stored with the
1175 compiled pattern, and used again by pcre_exec(), unless another table
1176 pointer is passed to it. For more discussion, see the section on locale
1177 support below.
1178
1179 This code fragment shows a typical straightforward call to pcre_com-
1180 pile():
1181
1182 pcre *re;
1183 const char *error;
1184 int erroffset;
1185 re = pcre_compile(
1186 "^A.*Z", /* the pattern */
1187 0, /* default options */
1188 &error, /* for error message */
1189 &erroffset, /* for error offset */
1190 NULL); /* use default character tables */
1191
1192 The following names for option bits are defined in the pcre.h header
1193 file:
1194
1195 PCRE_ANCHORED
1196
1197 If this bit is set, the pattern is forced to be "anchored", that is, it
1198 is constrained to match only at the first matching point in the string
1199 that is being searched (the "subject string"). This effect can also be
1200 achieved by appropriate constructs in the pattern itself, which is the
1201 only way to do it in Perl.
1202
1203 PCRE_AUTO_CALLOUT
1204
1205 If this bit is set, pcre_compile() automatically inserts callout items,
1206 all with number 255, before each pattern item. For discussion of the
1207 callout facility, see the pcrecallout documentation.
1208
1209 PCRE_BSR_ANYCRLF
1210 PCRE_BSR_UNICODE
1211
1212 These options (which are mutually exclusive) control what the \R escape
1213 sequence matches. The choice is either to match only CR, LF, or CRLF,
1214 or to match any Unicode newline sequence. The default is specified when
1215 PCRE is built. It can be overridden from within the pattern, or by set-
1216 ting an option when a compiled pattern is matched.
1217
1218 PCRE_CASELESS
1219
1220 If this bit is set, letters in the pattern match both upper and lower
1221 case letters. It is equivalent to Perl's /i option, and it can be
1222 changed within a pattern by a (?i) option setting. In UTF-8 mode, PCRE
1223 always understands the concept of case for characters whose values are
1224 less than 128, so caseless matching is always possible. For characters
1225 with higher values, the concept of case is supported if PCRE is com-
1226 piled with Unicode property support, but not otherwise. If you want to
1227 use caseless matching for characters 128 and above, you must ensure
1228 that PCRE is compiled with Unicode property support as well as with
1229 UTF-8 support.
1230
1231 PCRE_DOLLAR_ENDONLY
1232
1233 If this bit is set, a dollar metacharacter in the pattern matches only
1234 at the end of the subject string. Without this option, a dollar also
1235 matches immediately before a newline at the end of the string (but not
1236 before any other newlines). The PCRE_DOLLAR_ENDONLY option is ignored
1237 if PCRE_MULTILINE is set. There is no equivalent to this option in
1238 Perl, and no way to set it within a pattern.
1239
1240 PCRE_DOTALL
1241
1242 If this bit is set, a dot metacharater in the pattern matches all char-
1243 acters, including those that indicate newline. Without it, a dot does
1244 not match when the current position is at a newline. This option is
1245 equivalent to Perl's /s option, and it can be changed within a pattern
1246 by a (?s) option setting. A negative class such as [^a] always matches
1247 newline characters, independent of the setting of this option.
1248
1249 PCRE_DUPNAMES
1250
1251 If this bit is set, names used to identify capturing subpatterns need
1252 not be unique. This can be helpful for certain types of pattern when it
1253 is known that only one instance of the named subpattern can ever be
1254 matched. There are more details of named subpatterns below; see also
1255 the pcrepattern documentation.
1256
1257 PCRE_EXTENDED
1258
1259 If this bit is set, whitespace data characters in the pattern are
1260 totally ignored except when escaped or inside a character class. White-
1261 space does not include the VT character (code 11). In addition, charac-
1262 ters between an unescaped # outside a character class and the next new-
1263 line, inclusive, are also ignored. This is equivalent to Perl's /x
1264 option, and it can be changed within a pattern by a (?x) option set-
1265 ting.
1266
1267 This option makes it possible to include comments inside complicated
1268 patterns. Note, however, that this applies only to data characters.
1269 Whitespace characters may never appear within special character
1270 sequences in a pattern, for example within the sequence (?( which
1271 introduces a conditional subpattern.
1272
1273 PCRE_EXTRA
1274
1275 This option was invented in order to turn on additional functionality
1276 of PCRE that is incompatible with Perl, but it is currently of very
1277 little use. When set, any backslash in a pattern that is followed by a
1278 letter that has no special meaning causes an error, thus reserving
1279 these combinations for future expansion. By default, as in Perl, a
1280 backslash followed by a letter with no special meaning is treated as a
1281 literal. (Perl can, however, be persuaded to give a warning for this.)
1282 There are at present no other features controlled by this option. It
1283 can also be set by a (?X) option setting within a pattern.
1284
1285 PCRE_FIRSTLINE
1286
1287 If this option is set, an unanchored pattern is required to match
1288 before or at the first newline in the subject string, though the
1289 matched text may continue over the newline.
1290
1291 PCRE_JAVASCRIPT_COMPAT
1292
1293 If this option is set, PCRE's behaviour is changed in some ways so that
1294 it is compatible with JavaScript rather than Perl. The changes are as
1295 follows:
1296
1297 (1) A lone closing square bracket in a pattern causes a compile-time
1298 error, because this is illegal in JavaScript (by default it is treated
1299 as a data character). Thus, the pattern AB]CD becomes illegal when this
1300 option is set.
1301
1302 (2) At run time, a back reference to an unset subpattern group matches
1303 an empty string (by default this causes the current matching alterna-
1304 tive to fail). A pattern such as (\1)(a) succeeds when this option is
1305 set (assuming it can find an "a" in the subject), whereas it fails by
1306 default, for Perl compatibility.
1307
1308 PCRE_MULTILINE
1309
1310 By default, PCRE treats the subject string as consisting of a single
1311 line of characters (even if it actually contains newlines). The "start
1312 of line" metacharacter (^) matches only at the start of the string,
1313 while the "end of line" metacharacter ($) matches only at the end of
1314 the string, or before a terminating newline (unless PCRE_DOLLAR_ENDONLY
1315 is set). This is the same as Perl.
1316
1317 When PCRE_MULTILINE it is set, the "start of line" and "end of line"
1318 constructs match immediately following or immediately before internal
1319 newlines in the subject string, respectively, as well as at the very
1320 start and end. This is equivalent to Perl's /m option, and it can be
1321 changed within a pattern by a (?m) option setting. If there are no new-
1322 lines in a subject string, or no occurrences of ^ or $ in a pattern,
1323 setting PCRE_MULTILINE has no effect.
1324
1325 PCRE_NEWLINE_CR
1326 PCRE_NEWLINE_LF
1327 PCRE_NEWLINE_CRLF
1328 PCRE_NEWLINE_ANYCRLF
1329 PCRE_NEWLINE_ANY
1330
1331 These options override the default newline definition that was chosen
1332 when PCRE was built. Setting the first or the second specifies that a
1333 newline is indicated by a single character (CR or LF, respectively).
1334 Setting PCRE_NEWLINE_CRLF specifies that a newline is indicated by the
1335 two-character CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies
1336 that any of the three preceding sequences should be recognized. Setting
1337 PCRE_NEWLINE_ANY specifies that any Unicode newline sequence should be
1338 recognized. The Unicode newline sequences are the three just mentioned,
1339 plus the single characters VT (vertical tab, U+000B), FF (formfeed,
1340 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
1341 (paragraph separator, U+2029). The last two are recognized only in
1342 UTF-8 mode.
1343
1344 The newline setting in the options word uses three bits that are
1345 treated as a number, giving eight possibilities. Currently only six are
1346 used (default plus the five values above). This means that if you set
1347 more than one newline option, the combination may or may not be sensi-
1348 ble. For example, PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to
1349 PCRE_NEWLINE_CRLF, but other combinations may yield unused numbers and
1350 cause an error.
1351
1352 The only time that a line break is specially recognized when compiling
1353 a pattern is if PCRE_EXTENDED is set, and an unescaped # outside a
1354 character class is encountered. This indicates a comment that lasts
1355 until after the next line break sequence. In other circumstances, line
1356 break sequences are treated as literal data, except that in
1357 PCRE_EXTENDED mode, both CR and LF are treated as whitespace characters
1358 and are therefore ignored.
1359
1360 The newline option that is set at compile time becomes the default that
1361 is used for pcre_exec() and pcre_dfa_exec(), but it can be overridden.
1362
1363 PCRE_NO_AUTO_CAPTURE
1364
1365 If this option is set, it disables the use of numbered capturing paren-
1366 theses in the pattern. Any opening parenthesis that is not followed by
1367 ? behaves as if it were followed by ?: but named parentheses can still
1368 be used for capturing (and they acquire numbers in the usual way).
1369 There is no equivalent of this option in Perl.
1370
1371 PCRE_UNGREEDY
1372
1373 This option inverts the "greediness" of the quantifiers so that they
1374 are not greedy by default, but become greedy if followed by "?". It is
1375 not compatible with Perl. It can also be set by a (?U) option setting
1376 within the pattern.
1377
1378 PCRE_UTF8
1379
1380 This option causes PCRE to regard both the pattern and the subject as
1381 strings of UTF-8 characters instead of single-byte character strings.
1382 However, it is available only when PCRE is built to include UTF-8 sup-
1383 port. If not, the use of this option provokes an error. Details of how
1384 this option changes the behaviour of PCRE are given in the section on
1385 UTF-8 support in the main pcre page.
1386
1387 PCRE_NO_UTF8_CHECK
1388
1389 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
1390 automatically checked. There is a discussion about the validity of
1391 UTF-8 strings in the main pcre page. If an invalid UTF-8 sequence of
1392 bytes is found, pcre_compile() returns an error. If you already know
1393 that your pattern is valid, and you want to skip this check for perfor-
1394 mance reasons, you can set the PCRE_NO_UTF8_CHECK option. When it is
1395 set, the effect of passing an invalid UTF-8 string as a pattern is
1396 undefined. It may cause your program to crash. Note that this option
1397 can also be passed to pcre_exec() and pcre_dfa_exec(), to suppress the
1398 UTF-8 validity checking of subject strings.
1399
1400
1401 COMPILATION ERROR CODES
1402
1403 The following table lists the error codes than may be returned by
1404 pcre_compile2(), along with the error messages that may be returned by
1405 both compiling functions. As PCRE has developed, some error codes have
1406 fallen out of use. To avoid confusion, they have not been re-used.
1407
1408 0 no error
1409 1 \ at end of pattern
1410 2 \c at end of pattern
1411 3 unrecognized character follows \
1412 4 numbers out of order in {} quantifier
1413 5 number too big in {} quantifier
1414 6 missing terminating ] for character class
1415 7 invalid escape sequence in character class
1416 8 range out of order in character class
1417 9 nothing to repeat
1418 10 [this code is not in use]
1419 11 internal error: unexpected repeat
1420 12 unrecognized character after (? or (?-
1421 13 POSIX named classes are supported only within a class
1422 14 missing )
1423 15 reference to non-existent subpattern
1424 16 erroffset passed as NULL
1425 17 unknown option bit(s) set
1426 18 missing ) after comment
1427 19 [this code is not in use]
1428 20 regular expression is too large
1429 21 failed to get memory
1430 22 unmatched parentheses
1431 23 internal error: code overflow
1432 24 unrecognized character after (?<
1433 25 lookbehind assertion is not fixed length
1434 26 malformed number or name after (?(
1435 27 conditional group contains more than two branches
1436 28 assertion expected after (?(
1437 29 (?R or (?[+-]digits must be followed by )
1438 30 unknown POSIX class name
1439 31 POSIX collating elements are not supported
1440 32 this version of PCRE is not compiled with PCRE_UTF8 support
1441 33 [this code is not in use]
1442 34 character value in \x{...} sequence is too large
1443 35 invalid condition (?(0)
1444 36 \C not allowed in lookbehind assertion
1445 37 PCRE does not support \L, \l, \N, \U, or \u
1446 38 number after (?C is > 255
1447 39 closing ) for (?C expected
1448 40 recursive call could loop indefinitely
1449 41 unrecognized character after (?P
1450 42 syntax error in subpattern name (missing terminator)
1451 43 two named subpatterns have the same name
1452 44 invalid UTF-8 string
1453 45 support for \P, \p, and \X has not been compiled
1454 46 malformed \P or \p sequence
1455 47 unknown property name after \P or \p
1456 48 subpattern name is too long (maximum 32 characters)
1457 49 too many named subpatterns (maximum 10000)
1458 50 [this code is not in use]
1459 51 octal value is greater than \377 (not in UTF-8 mode)
1460 52 internal error: overran compiling workspace
1461 53 internal error: previously-checked referenced subpattern not
1462 found
1463 54 DEFINE group contains more than one branch
1464 55 repeating a DEFINE group is not allowed
1465 56 inconsistent NEWLINE options
1466 57 \g is not followed by a braced, angle-bracketed, or quoted
1467 name/number or by a plain number
1468 58 a numbered reference must not be zero
1469 59 (*VERB) with an argument is not supported
1470 60 (*VERB) not recognized
1471 61 number is too big
1472 62 subpattern name expected
1473 63 digit expected after (?+
1474 64 ] is an invalid data character in JavaScript compatibility mode
1475
1476 The numbers 32 and 10000 in errors 48 and 49 are defaults; different
1477 values may be used if the limits were changed when PCRE was built.
1478
1479
1480 STUDYING A PATTERN
1481
1482 pcre_extra *pcre_study(const pcre *code, int options
1483 const char **errptr);
1484
1485 If a compiled pattern is going to be used several times, it is worth
1486 spending more time analyzing it in order to speed up the time taken for
1487 matching. The function pcre_study() takes a pointer to a compiled pat-
1488 tern as its first argument. If studying the pattern produces additional
1489 information that will help speed up matching, pcre_study() returns a
1490 pointer to a pcre_extra block, in which the study_data field points to
1491 the results of the study.
1492
1493 The returned value from pcre_study() can be passed directly to
1494 pcre_exec(). However, a pcre_extra block also contains other fields
1495 that can be set by the caller before the block is passed; these are
1496 described below in the section on matching a pattern.
1497
1498 If studying the pattern does not produce any additional information
1499 pcre_study() returns NULL. In that circumstance, if the calling program
1500 wants to pass any of the other fields to pcre_exec(), it must set up
1501 its own pcre_extra block.
1502
1503 The second argument of pcre_study() contains option bits. At present,
1504 no options are defined, and this argument should always be zero.
1505
1506 The third argument for pcre_study() is a pointer for an error message.
1507 If studying succeeds (even if no data is returned), the variable it
1508 points to is set to NULL. Otherwise it is set to point to a textual
1509 error message. This is a static string that is part of the library. You
1510 must not try to free it. You should test the error pointer for NULL
1511 after calling pcre_study(), to be sure that it has run successfully.
1512
1513 This is a typical call to pcre_study():
1514
1515 pcre_extra *pe;
1516 pe = pcre_study(
1517 re, /* result of pcre_compile() */
1518 0, /* no options exist */
1519 &error); /* set to NULL or points to a message */
1520
1521 At present, studying a pattern is useful only for non-anchored patterns
1522 that do not have a single fixed starting character. A bitmap of possi-
1523 ble starting bytes is created.
1524
1525
1526 LOCALE SUPPORT
1527
1528 PCRE handles caseless matching, and determines whether characters are
1529 letters, digits, or whatever, by reference to a set of tables, indexed
1530 by character value. When running in UTF-8 mode, this applies only to
1531 characters with codes less than 128. Higher-valued codes never match
1532 escapes such as \w or \d, but can be tested with \p if PCRE is built
1533 with Unicode character property support. The use of locales with Uni-
1534 code is discouraged. If you are handling characters with codes greater
1535 than 128, you should either use UTF-8 and Unicode, or use locales, but
1536 not try to mix the two.
1537
1538 PCRE contains an internal set of tables that are used when the final
1539 argument of pcre_compile() is NULL. These are sufficient for many
1540 applications. Normally, the internal tables recognize only ASCII char-
1541 acters. However, when PCRE is built, it is possible to cause the inter-
1542 nal tables to be rebuilt in the default "C" locale of the local system,
1543 which may cause them to be different.
1544
1545 The internal tables can always be overridden by tables supplied by the
1546 application that calls PCRE. These may be created in a different locale
1547 from the default. As more and more applications change to using Uni-
1548 code, the need for this locale support is expected to die away.
1549
1550 External tables are built by calling the pcre_maketables() function,
1551 which has no arguments, in the relevant locale. The result can then be
1552 passed to pcre_compile() or pcre_exec() as often as necessary. For
1553 example, to build and use tables that are appropriate for the French
1554 locale (where accented characters with values greater than 128 are
1555 treated as letters), the following code could be used:
1556
1557 setlocale(LC_CTYPE, "fr_FR");
1558 tables = pcre_maketables();
1559 re = pcre_compile(..., tables);
1560
1561 The locale name "fr_FR" is used on Linux and other Unix-like systems;
1562 if you are using Windows, the name for the French locale is "french".
1563
1564 When pcre_maketables() runs, the tables are built in memory that is
1565 obtained via pcre_malloc. It is the caller's responsibility to ensure
1566 that the memory containing the tables remains available for as long as
1567 it is needed.
1568
1569 The pointer that is passed to pcre_compile() is saved with the compiled
1570 pattern, and the same tables are used via this pointer by pcre_study()
1571 and normally also by pcre_exec(). Thus, by default, for any single pat-
1572 tern, compilation, studying and matching all happen in the same locale,
1573 but different patterns can be compiled in different locales.
1574
1575 It is possible to pass a table pointer or NULL (indicating the use of
1576 the internal tables) to pcre_exec(). Although not intended for this
1577 purpose, this facility could be used to match a pattern in a different
1578 locale from the one in which it was compiled. Passing table pointers at
1579 run time is discussed below in the section on matching a pattern.
1580
1581
1582 INFORMATION ABOUT A PATTERN
1583
1584 int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
1585 int what, void *where);
1586
1587 The pcre_fullinfo() function returns information about a compiled pat-
1588 tern. It replaces the obsolete pcre_info() function, which is neverthe-
1589 less retained for backwards compability (and is documented below).
1590
1591 The first argument for pcre_fullinfo() is a pointer to the compiled
1592 pattern. The second argument is the result of pcre_study(), or NULL if
1593 the pattern was not studied. The third argument specifies which piece
1594 of information is required, and the fourth argument is a pointer to a
1595 variable to receive the data. The yield of the function is zero for
1596 success, or one of the following negative numbers:
1597
1598 PCRE_ERROR_NULL the argument code was NULL
1599 the argument where was NULL
1600 PCRE_ERROR_BADMAGIC the "magic number" was not found
1601 PCRE_ERROR_BADOPTION the value of what was invalid
1602
1603 The "magic number" is placed at the start of each compiled pattern as
1604 an simple check against passing an arbitrary memory pointer. Here is a
1605 typical call of pcre_fullinfo(), to obtain the length of the compiled
1606 pattern:
1607
1608 int rc;
1609 size_t length;
1610 rc = pcre_fullinfo(
1611 re, /* result of pcre_compile() */
1612 pe, /* result of pcre_study(), or NULL */
1613 PCRE_INFO_SIZE, /* what is required */
1614 &length); /* where to put the data */
1615
1616 The possible values for the third argument are defined in pcre.h, and
1617 are as follows:
1618
1619 PCRE_INFO_BACKREFMAX
1620
1621 Return the number of the highest back reference in the pattern. The
1622 fourth argument should point to an int variable. Zero is returned if
1623 there are no back references.
1624
1625 PCRE_INFO_CAPTURECOUNT
1626
1627 Return the number of capturing subpatterns in the pattern. The fourth
1628 argument should point to an int variable.
1629
1630 PCRE_INFO_DEFAULT_TABLES
1631
1632 Return a pointer to the internal default character tables within PCRE.
1633 The fourth argument should point to an unsigned char * variable. This
1634 information call is provided for internal use by the pcre_study() func-
1635 tion. External callers can cause PCRE to use its internal tables by
1636 passing a NULL table pointer.
1637
1638 PCRE_INFO_FIRSTBYTE
1639
1640 Return information about the first byte of any matched string, for a
1641 non-anchored pattern. The fourth argument should point to an int vari-
1642 able. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name
1643 is still recognized for backwards compatibility.)
1644
1645 If there is a fixed first byte, for example, from a pattern such as
1646 (cat|cow|coyote), its value is returned. Otherwise, if either
1647
1648 (a) the pattern was compiled with the PCRE_MULTILINE option, and every
1649 branch starts with "^", or
1650
1651 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not
1652 set (if it were set, the pattern would be anchored),
1653
1654 -1 is returned, indicating that the pattern matches only at the start
1655 of a subject string or after any newline within the string. Otherwise
1656 -2 is returned. For anchored patterns, -2 is returned.
1657
1658 PCRE_INFO_FIRSTTABLE
1659
1660 If the pattern was studied, and this resulted in the construction of a
1661 256-bit table indicating a fixed set of bytes for the first byte in any
1662 matching string, a pointer to the table is returned. Otherwise NULL is
1663 returned. The fourth argument should point to an unsigned char * vari-
1664 able.
1665
1666 PCRE_INFO_HASCRORLF
1667
1668 Return 1 if the pattern contains any explicit matches for CR or LF
1669 characters, otherwise 0. The fourth argument should point to an int
1670 variable. An explicit match is either a literal CR or LF character, or
1671 \r or \n.
1672
1673 PCRE_INFO_JCHANGED
1674
1675 Return 1 if the (?J) or (?-J) option setting is used in the pattern,
1676 otherwise 0. The fourth argument should point to an int variable. (?J)
1677 and (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1678
1679 PCRE_INFO_LASTLITERAL
1680
1681 Return the value of the rightmost literal byte that must exist in any
1682 matched string, other than at its start, if such a byte has been
1683 recorded. The fourth argument should point to an int variable. If there
1684 is no such byte, -1 is returned. For anchored patterns, a last literal
1685 byte is recorded only if it follows something of variable length. For
1686 example, for the pattern /^a\d+z\d+/ the returned value is "z", but for
1687 /^a\dz\d/ the returned value is -1.
1688
1689 PCRE_INFO_NAMECOUNT
1690 PCRE_INFO_NAMEENTRYSIZE
1691 PCRE_INFO_NAMETABLE
1692
1693 PCRE supports the use of named as well as numbered capturing parenthe-
1694 ses. The names are just an additional way of identifying the parenthe-
1695 ses, which still acquire numbers. Several convenience functions such as
1696 pcre_get_named_substring() are provided for extracting captured sub-
1697 strings by name. It is also possible to extract the data directly, by
1698 first converting the name to a number in order to access the correct
1699 pointers in the output vector (described with pcre_exec() below). To do
1700 the conversion, you need to use the name-to-number map, which is
1701 described by these three values.
1702
1703 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT
1704 gives the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size
1705 of each entry; both of these return an int value. The entry size
1706 depends on the length of the longest name. PCRE_INFO_NAMETABLE returns
1707 a pointer to the first entry of the table (a pointer to char). The
1708 first two bytes of each entry are the number of the capturing parenthe-
1709 sis, most significant byte first. The rest of the entry is the corre-
1710 sponding name, zero terminated. The names are in alphabetical order.
1711 When PCRE_DUPNAMES is set, duplicate names are in order of their paren-
1712 theses numbers. For example, consider the following pattern (assume
1713 PCRE_EXTENDED is set, so white space - including newlines - is
1714 ignored):
1715
1716 (?<date> (?<year>(\d\d)?\d\d) -
1717 (?<month>\d\d) - (?<day>\d\d) )
1718
1719 There are four named subpatterns, so the table has four entries, and
1720 each entry in the table is eight bytes long. The table is as follows,
1721 with non-printing bytes shows in hexadecimal, and undefined bytes shown
1722 as ??:
1723
1724 00 01 d a t e 00 ??
1725 00 05 d a y 00 ?? ??
1726 00 04 m o n t h 00
1727 00 02 y e a r 00 ??
1728
1729 When writing code to extract data from named subpatterns using the
1730 name-to-number map, remember that the length of the entries is likely
1731 to be different for each compiled pattern.
1732
1733 PCRE_INFO_OKPARTIAL
1734
1735 Return 1 if the pattern can be used for partial matching with
1736 pcre_exec(), otherwise 0. The fourth argument should point to an int
1737 variable. From release 8.00, this always returns 1, because the
1738 restrictions that previously applied to partial matching have been
1739 lifted. The pcrepartial documentation gives details of partial match-
1740 ing.
1741
1742 PCRE_INFO_OPTIONS
1743
1744 Return a copy of the options with which the pattern was compiled. The
1745 fourth argument should point to an unsigned long int variable. These
1746 option bits are those specified in the call to pcre_compile(), modified
1747 by any top-level option settings at the start of the pattern itself. In
1748 other words, they are the options that will be in force when matching
1749 starts. For example, if the pattern /(?im)abc(?-i)d/ is compiled with
1750 the PCRE_EXTENDED option, the result is PCRE_CASELESS, PCRE_MULTILINE,
1751 and PCRE_EXTENDED.
1752
1753 A pattern is automatically anchored by PCRE if all of its top-level
1754 alternatives begin with one of the following:
1755
1756 ^ unless PCRE_MULTILINE is set
1757 \A always
1758 \G always
1759 .* if PCRE_DOTALL is set and there are no back
1760 references to the subpattern in which .* appears
1761
1762 For such patterns, the PCRE_ANCHORED bit is set in the options returned
1763 by pcre_fullinfo().
1764
1765 PCRE_INFO_SIZE
1766
1767 Return the size of the compiled pattern, that is, the value that was
1768 passed as the argument to pcre_malloc() when PCRE was getting memory in
1769 which to place the compiled data. The fourth argument should point to a
1770 size_t variable.
1771
1772 PCRE_INFO_STUDYSIZE
1773
1774 Return the size of the data block pointed to by the study_data field in
1775 a pcre_extra block. That is, it is the value that was passed to
1776 pcre_malloc() when PCRE was getting memory into which to place the data
1777 created by pcre_study(). The fourth argument should point to a size_t
1778 variable.
1779
1780
1781 OBSOLETE INFO FUNCTION
1782
1783 int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
1784
1785 The pcre_info() function is now obsolete because its interface is too
1786 restrictive to return all the available data about a compiled pattern.
1787 New programs should use pcre_fullinfo() instead. The yield of
1788 pcre_info() is the number of capturing subpatterns, or one of the fol-
1789 lowing negative numbers:
1790
1791 PCRE_ERROR_NULL the argument code was NULL
1792 PCRE_ERROR_BADMAGIC the "magic number" was not found
1793
1794 If the optptr argument is not NULL, a copy of the options with which
1795 the pattern was compiled is placed in the integer it points to (see
1796 PCRE_INFO_OPTIONS above).
1797
1798 If the pattern is not anchored and the firstcharptr argument is not
1799 NULL, it is used to pass back information about the first character of
1800 any matched string (see PCRE_INFO_FIRSTBYTE above).
1801
1802
1803 REFERENCE COUNTS
1804
1805 int pcre_refcount(pcre *code, int adjust);
1806
1807 The pcre_refcount() function is used to maintain a reference count in
1808 the data block that contains a compiled pattern. It is provided for the
1809 benefit of applications that operate in an object-oriented manner,
1810 where different parts of the application may be using the same compiled
1811 pattern, but you want to free the block when they are all done.
1812
1813 When a pattern is compiled, the reference count field is initialized to
1814 zero. It is changed only by calling this function, whose action is to
1815 add the adjust value (which may be positive or negative) to it. The
1816 yield of the function is the new value. However, the value of the count
1817 is constrained to lie between 0 and 65535, inclusive. If the new value
1818 is outside these limits, it is forced to the appropriate limit value.
1819
1820 Except when it is zero, the reference count is not correctly preserved
1821 if a pattern is compiled on one host and then transferred to a host
1822 whose byte-order is different. (This seems a highly unlikely scenario.)
1823
1824
1825 MATCHING A PATTERN: THE TRADITIONAL FUNCTION
1826
1827 int pcre_exec(const pcre *code, const pcre_extra *extra,
1828 const char *subject, int length, int startoffset,
1829 int options, int *ovector, int ovecsize);
1830
1831 The function pcre_exec() is called to match a subject string against a
1832 compiled pattern, which is passed in the code argument. If the pattern
1833 has been studied, the result of the study should be passed in the extra
1834 argument. This function is the main matching facility of the library,
1835 and it operates in a Perl-like manner. For specialist use there is also
1836 an alternative matching function, which is described below in the sec-
1837 tion about the pcre_dfa_exec() function.
1838
1839 In most applications, the pattern will have been compiled (and option-
1840 ally studied) in the same process that calls pcre_exec(). However, it
1841 is possible to save compiled patterns and study data, and then use them
1842 later in different processes, possibly even on different hosts. For a
1843 discussion about this, see the pcreprecompile documentation.
1844
1845 Here is an example of a simple call to pcre_exec():
1846
1847 int rc;
1848 int ovector[30];
1849 rc = pcre_exec(
1850 re, /* result of pcre_compile() */
1851 NULL, /* we didn't study the pattern */
1852 "some string", /* the subject string */
1853 11, /* the length of the subject string */
1854 0, /* start at offset 0 in the subject */
1855 0, /* default options */
1856 ovector, /* vector of integers for substring information */
1857 30); /* number of elements (NOT size in bytes) */
1858
1859 Extra data for pcre_exec()
1860
1861 If the extra argument is not NULL, it must point to a pcre_extra data
1862 block. The pcre_study() function returns such a block (when it doesn't
1863 return NULL), but you can also create one for yourself, and pass addi-
1864 tional information in it. The pcre_extra block contains the following
1865 fields (not necessarily in this order):
1866
1867 unsigned long int flags;
1868 void *study_data;
1869 unsigned long int match_limit;
1870 unsigned long int match_limit_recursion;
1871 void *callout_data;
1872 const unsigned char *tables;
1873
1874 The flags field is a bitmap that specifies which of the other fields
1875 are set. The flag bits are:
1876
1877 PCRE_EXTRA_STUDY_DATA
1878 PCRE_EXTRA_MATCH_LIMIT
1879 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1880 PCRE_EXTRA_CALLOUT_DATA
1881 PCRE_EXTRA_TABLES
1882
1883 Other flag bits should be set to zero. The study_data field is set in
1884 the pcre_extra block that is returned by pcre_study(), together with
1885 the appropriate flag bit. You should not set this yourself, but you may
1886 add to the block by setting the other fields and their corresponding
1887 flag bits.
1888
1889 The match_limit field provides a means of preventing PCRE from using up
1890 a vast amount of resources when running patterns that are not going to
1891 match, but which have a very large number of possibilities in their
1892 search trees. The classic example is the use of nested unlimited
1893 repeats.
1894
1895 Internally, PCRE uses a function called match() which it calls repeat-
1896 edly (sometimes recursively). The limit set by match_limit is imposed
1897 on the number of times this function is called during a match, which
1898 has the effect of limiting the amount of backtracking that can take
1899 place. For patterns that are not anchored, the count restarts from zero
1900 for each position in the subject string.
1901
1902 The default value for the limit can be set when PCRE is built; the
1903 default default is 10 million, which handles all but the most extreme
1904 cases. You can override the default by suppling pcre_exec() with a
1905 pcre_extra block in which match_limit is set, and
1906 PCRE_EXTRA_MATCH_LIMIT is set in the flags field. If the limit is
1907 exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1908
1909 The match_limit_recursion field is similar to match_limit, but instead
1910 of limiting the total number of times that match() is called, it limits
1911 the depth of recursion. The recursion depth is a smaller number than
1912 the total number of calls, because not all calls to match() are recur-
1913 sive. This limit is of use only if it is set smaller than match_limit.
1914
1915 Limiting the recursion depth limits the amount of stack that can be
1916 used, or, when PCRE has been compiled to use memory on the heap instead
1917 of the stack, the amount of heap memory that can be used.
1918
1919 The default value for match_limit_recursion can be set when PCRE is
1920 built; the default default is the same value as the default for
1921 match_limit. You can override the default by suppling pcre_exec() with
1922 a pcre_extra block in which match_limit_recursion is set, and
1923 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the flags field. If the
1924 limit is exceeded, pcre_exec() returns PCRE_ERROR_RECURSIONLIMIT.
1925
1926 The callout_data field is used in conjunction with the "callout" fea-
1927 ture, and is described in the pcrecallout documentation.
1928
1929 The tables field is used to pass a character tables pointer to
1930 pcre_exec(); this overrides the value that is stored with the compiled
1931 pattern. A non-NULL value is stored with the compiled pattern only if
1932 custom tables were supplied to pcre_compile() via its tableptr argu-
1933 ment. If NULL is passed to pcre_exec() using this mechanism, it forces
1934 PCRE's internal tables to be used. This facility is helpful when re-
1935 using patterns that have been saved after compiling with an external
1936 set of tables, because the external tables might be at a different
1937 address when pcre_exec() is called. See the pcreprecompile documenta-
1938 tion for a discussion of saving compiled patterns for later use.
1939
1940 Option bits for pcre_exec()
1941
1942 The unused bits of the options argument for pcre_exec() must be zero.
1943 The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_xxx,
1944 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1945 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
1946 PCRE_PARTIAL_HARD.
1947
1948 PCRE_ANCHORED
1949
1950 The PCRE_ANCHORED option limits pcre_exec() to matching at the first
1951 matching position. If a pattern was compiled with PCRE_ANCHORED, or
1952 turned out to be anchored by virtue of its contents, it cannot be made
1953 unachored at matching time.
1954
1955 PCRE_BSR_ANYCRLF
1956 PCRE_BSR_UNICODE
1957
1958 These options (which are mutually exclusive) control what the \R escape
1959 sequence matches. The choice is either to match only CR, LF, or CRLF,
1960 or to match any Unicode newline sequence. These options override the
1961 choice that was made or defaulted when the pattern was compiled.
1962
1963 PCRE_NEWLINE_CR
1964 PCRE_NEWLINE_LF
1965 PCRE_NEWLINE_CRLF
1966 PCRE_NEWLINE_ANYCRLF
1967 PCRE_NEWLINE_ANY
1968
1969 These options override the newline definition that was chosen or
1970 defaulted when the pattern was compiled. For details, see the descrip-
1971 tion of pcre_compile() above. During matching, the newline choice
1972 affects the behaviour of the dot, circumflex, and dollar metacharac-
1973 ters. It may also alter the way the match position is advanced after a
1974 match failure for an unanchored pattern.
1975
1976 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is
1977 set, and a match attempt for an unanchored pattern fails when the cur-
1978 rent position is at a CRLF sequence, and the pattern contains no
1979 explicit matches for CR or LF characters, the match position is
1980 advanced by two characters instead of one, in other words, to after the
1981 CRLF.
1982
1983 The above rule is a compromise that makes the most common cases work as
1984 expected. For example, if the pattern is .+A (and the PCRE_DOTALL
1985 option is not set), it does not match the string "\r\nA" because, after
1986 failing at the start, it skips both the CR and the LF before retrying.
1987 However, the pattern [\r\n]A does match that string, because it con-
1988 tains an explicit CR or LF reference, and so advances only by one char-
1989 acter after the first failure.
1990
1991 An explicit match for CR of LF is either a literal appearance of one of
1992 those characters, or one of the \r or \n escape sequences. Implicit
1993 matches such as [^X] do not count, nor does \s (which includes CR and
1994 LF in the characters that it matches).
1995
1996 Notwithstanding the above, anomalous effects may still occur when CRLF
1997 is a valid newline sequence and explicit \r or \n escapes appear in the
1998 pattern.
1999
2000 PCRE_NOTBOL
2001
2002 This option specifies that first character of the subject string is not
2003 the beginning of a line, so the circumflex metacharacter should not
2004 match before it. Setting this without PCRE_MULTILINE (at compile time)
2005 causes circumflex never to match. This option affects only the behav-
2006 iour of the circumflex metacharacter. It does not affect \A.
2007
2008 PCRE_NOTEOL
2009
2010 This option specifies that the end of the subject string is not the end
2011 of a line, so the dollar metacharacter should not match it nor (except
2012 in multiline mode) a newline immediately before it. Setting this with-
2013 out PCRE_MULTILINE (at compile time) causes dollar never to match. This
2014 option affects only the behaviour of the dollar metacharacter. It does
2015 not affect \Z or \z.
2016
2017 PCRE_NOTEMPTY
2018
2019 An empty string is not considered to be a valid match if this option is
2020 set. If there are alternatives in the pattern, they are tried. If all
2021 the alternatives match the empty string, the entire match fails. For
2022 example, if the pattern
2023
2024 a?b?
2025
2026 is applied to a string not beginning with "a" or "b", it matches an
2027 empty string at the start of the subject. With PCRE_NOTEMPTY set, this
2028 match is not valid, so PCRE searches further into the string for occur-
2029 rences of "a" or "b".
2030
2031 PCRE_NOTEMPTY_ATSTART
2032
2033 This is like PCRE_NOTEMPTY, except that an empty string match that is
2034 not at the start of the subject is permitted. If the pattern is
2035 anchored, such a match can occur only if the pattern contains \K.
2036
2037 Perl has no direct equivalent of PCRE_NOTEMPTY or
2038 PCRE_NOTEMPTY_ATSTART, but it does make a special case of a pattern
2039 match of the empty string within its split() function, and when using
2040 the /g modifier. It is possible to emulate Perl's behaviour after
2041 matching a null string by first trying the match again at the same off-
2042 set with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then if that
2043 fails, by advancing the starting offset (see below) and trying an ordi-
2044 nary match again. There is some code that demonstrates how to do this
2045 in the pcredemo sample program.
2046
2047 PCRE_NO_START_OPTIMIZE
2048
2049 There are a number of optimizations that pcre_exec() uses at the start
2050 of a match, in order to speed up the process. For example, if it is
2051 known that a match must start with a specific character, it searches
2052 the subject for that character, and fails immediately if it cannot find
2053 it, without actually running the main matching function. When callouts
2054 are in use, these optimizations can cause them to be skipped. This
2055 option disables the "start-up" optimizations, causing performance to
2056 suffer, but ensuring that the callouts do occur.
2057
2058 PCRE_NO_UTF8_CHECK
2059
2060 When PCRE_UTF8 is set at compile time, the validity of the subject as a
2061 UTF-8 string is automatically checked when pcre_exec() is subsequently
2062 called. The value of startoffset is also checked to ensure that it
2063 points to the start of a UTF-8 character. There is a discussion about
2064 the validity of UTF-8 strings in the section on UTF-8 support in the
2065 main pcre page. If an invalid UTF-8 sequence of bytes is found,
2066 pcre_exec() returns the error PCRE_ERROR_BADUTF8. If startoffset con-
2067 tains an invalid value, PCRE_ERROR_BADUTF8_OFFSET is returned.
2068
2069 If you already know that your subject is valid, and you want to skip
2070 these checks for performance reasons, you can set the
2071 PCRE_NO_UTF8_CHECK option when calling pcre_exec(). You might want to
2072 do this for the second and subsequent calls to pcre_exec() if you are
2073 making repeated calls to find all the matches in a single subject
2074 string. However, you should be sure that the value of startoffset
2075 points to the start of a UTF-8 character. When PCRE_NO_UTF8_CHECK is
2076 set, the effect of passing an invalid UTF-8 string as a subject, or a
2077 value of startoffset that does not point to the start of a UTF-8 char-
2078 acter, is undefined. Your program may crash.
2079
2080 PCRE_PARTIAL_HARD
2081 PCRE_PARTIAL_SOFT
2082
2083 These options turn on the partial matching feature. For backwards com-
2084 patibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial
2085 match occurs if the end of the subject string is reached successfully,
2086 but there are not enough subject characters to complete the match. If
2087 this happens when PCRE_PARTIAL_HARD is set, pcre_exec() immediately
2088 returns PCRE_ERROR_PARTIAL. Otherwise, if PCRE_PARTIAL_SOFT is set,
2089 matching continues by testing any other alternatives. Only if they all
2090 fail is PCRE_ERROR_PARTIAL returned (instead of PCRE_ERROR_NOMATCH).
2091 The portion of the string that was inspected when the partial match was
2092 found is set as the first matching string. There is a more detailed
2093 discussion in the pcrepartial documentation.
2094
2095 The string to be matched by pcre_exec()
2096
2097 The subject string is passed to pcre_exec() as a pointer in subject, a
2098 length (in bytes) in length, and a starting byte offset in startoffset.
2099 In UTF-8 mode, the byte offset must point to the start of a UTF-8 char-
2100 acter. Unlike the pattern string, the subject may contain binary zero
2101 bytes. When the starting offset is zero, the search for a match starts
2102 at the beginning of the subject, and this is by far the most common
2103 case.
2104
2105 A non-zero starting offset is useful when searching for another match
2106 in the same subject by calling pcre_exec() again after a previous suc-
2107 cess. Setting startoffset differs from just passing over a shortened
2108 string and setting PCRE_NOTBOL in the case of a pattern that begins
2109 with any kind of lookbehind. For example, consider the pattern
2110
2111 \Biss\B
2112
2113 which finds occurrences of "iss" in the middle of words. (\B matches
2114 only if the current position in the subject is not a word boundary.)
2115 When applied to the string "Mississipi" the first call to pcre_exec()
2116 finds the first occurrence. If pcre_exec() is called again with just
2117 the remainder of the subject, namely "issipi", it does not match,
2118 because \B is always false at the start of the subject, which is deemed
2119 to be a word boundary. However, if pcre_exec() is passed the entire
2120 string again, but with startoffset set to 4, it finds the second occur-
2121 rence of "iss" because it is able to look behind the starting point to
2122 discover that it is preceded by a letter.
2123
2124 If a non-zero starting offset is passed when the pattern is anchored,
2125 one attempt to match at the given offset is made. This can only succeed
2126 if the pattern does not require the match to be at the start of the
2127 subject.
2128
2129 How pcre_exec() returns captured substrings
2130
2131 In general, a pattern matches a certain portion of the subject, and in
2132 addition, further substrings from the subject may be picked out by
2133 parts of the pattern. Following the usage in Jeffrey Friedl's book,
2134 this is called "capturing" in what follows, and the phrase "capturing
2135 subpattern" is used for a fragment of a pattern that picks out a sub-
2136 string. PCRE supports several other kinds of parenthesized subpattern
2137 that do not cause substrings to be captured.
2138
2139 Captured substrings are returned to the caller via a vector of integers
2140 whose address is passed in ovector. The number of elements in the vec-
2141 tor is passed in ovecsize, which must be a non-negative number. Note:
2142 this argument is NOT the size of ovector in bytes.
2143
2144 The first two-thirds of the vector is used to pass back captured sub-
2145 strings, each substring using a pair of integers. The remaining third
2146 of the vector is used as workspace by pcre_exec() while matching cap-
2147 turing subpatterns, and is not available for passing back information.
2148 The number passed in ovecsize should always be a multiple of three. If
2149 it is not, it is rounded down.
2150
2151 When a match is successful, information about captured substrings is
2152 returned in pairs of integers, starting at the beginning of ovector,
2153 and continuing up to two-thirds of its length at the most. The first
2154 element of each pair is set to the byte offset of the first character
2155 in a substring, and the second is set to the byte offset of the first
2156 character after the end of a substring. Note: these values are always
2157 byte offsets, even in UTF-8 mode. They are not character counts.
2158
2159 The first pair of integers, ovector[0] and ovector[1], identify the
2160 portion of the subject string matched by the entire pattern. The next
2161 pair is used for the first capturing subpattern, and so on. The value
2162 returned by pcre_exec() is one more than the highest numbered pair that
2163 has been set. For example, if two substrings have been captured, the
2164 returned value is 3. If there are no capturing subpatterns, the return
2165 value from a successful match is 1, indicating that just the first pair
2166 of offsets has been set.
2167
2168 If a capturing subpattern is matched repeatedly, it is the last portion
2169 of the string that it matched that is returned.
2170
2171 If the vector is too small to hold all the captured substring offsets,
2172 it is used as far as possible (up to two-thirds of its length), and the
2173 function returns a value of zero. If the substring offsets are not of
2174 interest, pcre_exec() may be called with ovector passed as NULL and
2175 ovecsize as zero. However, if the pattern contains back references and
2176 the ovector is not big enough to remember the related substrings, PCRE
2177 has to get additional memory for use during matching. Thus it is usu-
2178 ally advisable to supply an ovector.
2179
2180 The pcre_info() function can be used to find out how many capturing
2181 subpatterns there are in a compiled pattern. The smallest size for
2182 ovector that will allow for n captured substrings, in addition to the
2183 offsets of the substring matched by the whole pattern, is (n+1)*3.
2184
2185 It is possible for capturing subpattern number n+1 to match some part
2186 of the subject when subpattern n has not been used at all. For example,
2187 if the string "abc" is matched against the pattern (a|(z))(bc) the
2188 return from the function is 4, and subpatterns 1 and 3 are matched, but
2189 2 is not. When this happens, both values in the offset pairs corre-
2190 sponding to unused subpatterns are set to -1.
2191
2192 Offset values that correspond to unused subpatterns at the end of the
2193 expression are also set to -1. For example, if the string "abc" is
2194 matched against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not
2195 matched. The return from the function is 2, because the highest used
2196 capturing subpattern number is 1. However, you can refer to the offsets
2197 for the second and third capturing subpatterns if you wish (assuming
2198 the vector is large enough, of course).
2199
2200 Some convenience functions are provided for extracting the captured
2201 substrings as separate strings. These are described below.
2202
2203 Error return values from pcre_exec()
2204
2205 If pcre_exec() fails, it returns a negative number. The following are
2206 defined in the header file:
2207
2208 PCRE_ERROR_NOMATCH (-1)
2209
2210 The subject string did not match the pattern.
2211
2212 PCRE_ERROR_NULL (-2)
2213
2214 Either code or subject was passed as NULL, or ovector was NULL and
2215 ovecsize was not zero.
2216
2217 PCRE_ERROR_BADOPTION (-3)
2218
2219 An unrecognized bit was set in the options argument.
2220
2221 PCRE_ERROR_BADMAGIC (-4)
2222
2223 PCRE stores a 4-byte "magic number" at the start of the compiled code,
2224 to catch the case when it is passed a junk pointer and to detect when a
2225 pattern that was compiled in an environment of one endianness is run in
2226 an environment with the other endianness. This is the error that PCRE
2227 gives when the magic number is not present.
2228
2229 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2230
2231 While running the pattern match, an unknown item was encountered in the
2232 compiled pattern. This error could be caused by a bug in PCRE or by
2233 overwriting of the compiled pattern.
2234
2235 PCRE_ERROR_NOMEMORY (-6)
2236
2237 If a pattern contains back references, but the ovector that is passed
2238 to pcre_exec() is not big enough to remember the referenced substrings,
2239 PCRE gets a block of memory at the start of matching to use for this
2240 purpose. If the call via pcre_malloc() fails, this error is given. The
2241 memory is automatically freed at the end of matching.
2242
2243 PCRE_ERROR_NOSUBSTRING (-7)
2244
2245 This error is used by the pcre_copy_substring(), pcre_get_substring(),
2246 and pcre_get_substring_list() functions (see below). It is never
2247 returned by pcre_exec().
2248
2249 PCRE_ERROR_MATCHLIMIT (-8)
2250
2251 The backtracking limit, as specified by the match_limit field in a
2252 pcre_extra structure (or defaulted) was reached. See the description
2253 above.
2254
2255 PCRE_ERROR_CALLOUT (-9)
2256
2257 This error is never generated by pcre_exec() itself. It is provided for
2258 use by callout functions that want to yield a distinctive error code.
2259 See the pcrecallout documentation for details.
2260
2261 PCRE_ERROR_BADUTF8 (-10)
2262
2263 A string that contains an invalid UTF-8 byte sequence was passed as a
2264 subject.
2265
2266 PCRE_ERROR_BADUTF8_OFFSET (-11)
2267
2268 The UTF-8 byte sequence that was passed as a subject was valid, but the
2269 value of startoffset did not point to the beginning of a UTF-8 charac-
2270 ter.
2271
2272 PCRE_ERROR_PARTIAL (-12)
2273
2274 The subject string did not match, but it did match partially. See the
2275 pcrepartial documentation for details of partial matching.
2276
2277 PCRE_ERROR_BADPARTIAL (-13)
2278
2279 This code is no longer in use. It was formerly returned when the
2280 PCRE_PARTIAL option was used with a compiled pattern containing items
2281 that were not supported for partial matching. From release 8.00
2282 onwards, there are no restrictions on partial matching.
2283
2284 PCRE_ERROR_INTERNAL (-14)
2285
2286 An unexpected internal error has occurred. This error could be caused
2287 by a bug in PCRE or by overwriting of the compiled pattern.
2288
2289 PCRE_ERROR_BADCOUNT (-15)
2290
2291 This error is given if the value of the ovecsize argument is negative.
2292
2293 PCRE_ERROR_RECURSIONLIMIT (-21)
2294
2295 The internal recursion limit, as specified by the match_limit_recursion
2296 field in a pcre_extra structure (or defaulted) was reached. See the
2297 description above.
2298
2299 PCRE_ERROR_BADNEWLINE (-23)
2300
2301 An invalid combination of PCRE_NEWLINE_xxx options was given.
2302
2303 Error numbers -16 to -20 and -22 are not used by pcre_exec().
2304
2305
2306 EXTRACTING CAPTURED SUBSTRINGS BY NUMBER
2307
2308 int pcre_copy_substring(const char *subject, int *ovector,
2309 int stringcount, int stringnumber, char *buffer,
2310 int buffersize);
2311
2312 int pcre_get_substring(const char *subject, int *ovector,
2313 int stringcount, int stringnumber,
2314 const char **stringptr);
2315
2316 int pcre_get_substring_list(const char *subject,
2317 int *ovector, int stringcount, const char ***listptr);
2318
2319 Captured substrings can be accessed directly by using the offsets
2320 returned by pcre_exec() in ovector. For convenience, the functions
2321 pcre_copy_substring(), pcre_get_substring(), and pcre_get_sub-
2322 string_list() are provided for extracting captured substrings as new,
2323 separate, zero-terminated strings. These functions identify substrings
2324 by number. The next section describes functions for extracting named
2325 substrings.
2326
2327 A substring that contains a binary zero is correctly extracted and has
2328 a further zero added on the end, but the result is not, of course, a C
2329 string. However, you can process such a string by referring to the
2330 length that is returned by pcre_copy_substring() and pcre_get_sub-
2331 string(). Unfortunately, the interface to pcre_get_substring_list() is
2332 not adequate for handling strings containing binary zeros, because the
2333 end of the final string is not independently indicated.
2334
2335 The first three arguments are the same for all three of these func-
2336 tions: subject is the subject string that has just been successfully
2337 matched, ovector is a pointer to the vector of integer offsets that was
2338 passed to pcre_exec(), and stringcount is the number of substrings that
2339 were captured by the match, including the substring that matched the
2340 entire regular expression. This is the value returned by pcre_exec() if
2341 it is greater than zero. If pcre_exec() returned zero, indicating that
2342 it ran out of space in ovector, the value passed as stringcount should
2343 be the number of elements in the vector divided by three.
2344
2345 The functions pcre_copy_substring() and pcre_get_substring() extract a
2346 single substring, whose number is given as stringnumber. A value of
2347 zero extracts the substring that matched the entire pattern, whereas
2348 higher values extract the captured substrings. For pcre_copy_sub-
2349 string(), the string is placed in buffer, whose length is given by
2350 buffersize, while for pcre_get_substring() a new block of memory is
2351 obtained via pcre_malloc, and its address is returned via stringptr.
2352 The yield of the function is the length of the string, not including
2353 the terminating zero, or one of these error codes:
2354
2355 PCRE_ERROR_NOMEMORY (-6)
2356
2357 The buffer was too small for pcre_copy_substring(), or the attempt to
2358 get memory failed for pcre_get_substring().
2359
2360 PCRE_ERROR_NOSUBSTRING (-7)
2361
2362 There is no substring whose number is stringnumber.
2363
2364 The pcre_get_substring_list() function extracts all available sub-
2365 strings and builds a list of pointers to them. All this is done in a
2366 single block of memory that is obtained via pcre_malloc. The address of
2367 the memory block is returned via listptr, which is also the start of
2368 the list of string pointers. The end of the list is marked by a NULL
2369 pointer. The yield of the function is zero if all went well, or the
2370 error code
2371
2372 PCRE_ERROR_NOMEMORY (-6)
2373
2374 if the attempt to get the memory block failed.
2375
2376 When any of these functions encounter a substring that is unset, which
2377 can happen when capturing subpattern number n+1 matches some part of
2378 the subject, but subpattern n has not been used at all, they return an
2379 empty string. This can be distinguished from a genuine zero-length sub-
2380 string by inspecting the appropriate offset in ovector, which is nega-
2381 tive for unset substrings.
2382
2383 The two convenience functions pcre_free_substring() and pcre_free_sub-
2384 string_list() can be used to free the memory returned by a previous
2385 call of pcre_get_substring() or pcre_get_substring_list(), respec-
2386 tively. They do nothing more than call the function pointed to by
2387 pcre_free, which of course could be called directly from a C program.
2388 However, PCRE is used in some situations where it is linked via a spe-
2389 cial interface to another programming language that cannot use
2390 pcre_free directly; it is for these cases that the functions are pro-
2391 vided.
2392
2393
2394 EXTRACTING CAPTURED SUBSTRINGS BY NAME
2395
2396 int pcre_get_stringnumber(const pcre *code,
2397 const char *name);
2398
2399 int pcre_copy_named_substring(const pcre *code,
2400 const char *subject, int *ovector,
2401 int stringcount, const char *stringname,
2402 char *buffer, int buffersize);
2403
2404 int pcre_get_named_substring(const pcre *code,
2405 const char *subject, int *ovector,
2406 int stringcount, const char *stringname,
2407 const char **stringptr);
2408
2409 To extract a substring by name, you first have to find associated num-
2410 ber. For example, for this pattern
2411
2412 (a+)b(?<xxx>\d+)...
2413
2414 the number of the subpattern called "xxx" is 2. If the name is known to
2415 be unique (PCRE_DUPNAMES was not set), you can find the number from the
2416 name by calling pcre_get_stringnumber(). The first argument is the com-
2417 piled pattern, and the second is the name. The yield of the function is
2418 the subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no
2419 subpattern of that name.
2420
2421 Given the number, you can extract the substring directly, or use one of
2422 the functions described in the previous section. For convenience, there
2423 are also two functions that do the whole job.
2424
2425 Most of the arguments of pcre_copy_named_substring() and
2426 pcre_get_named_substring() are the same as those for the similarly
2427 named functions that extract by number. As these are described in the
2428 previous section, they are not re-described here. There are just two
2429 differences:
2430
2431 First, instead of a substring number, a substring name is given. Sec-
2432 ond, there is an extra argument, given at the start, which is a pointer
2433 to the compiled pattern. This is needed in order to gain access to the
2434 name-to-number translation table.
2435
2436 These functions call pcre_get_stringnumber(), and if it succeeds, they
2437 then call pcre_copy_substring() or pcre_get_substring(), as appropri-
2438 ate. NOTE: If PCRE_DUPNAMES is set and there are duplicate names, the
2439 behaviour may not be what you want (see the next section).
2440
2441 Warning: If the pattern uses the "(?|" feature to set up multiple sub-
2442 patterns with the same number, you cannot use names to distinguish
2443 them, because names are not included in the compiled code. The matching
2444 process uses only numbers.
2445
2446
2447 DUPLICATE SUBPATTERN NAMES
2448
2449 int pcre_get_stringtable_entries(const pcre *code,
2450 const char *name, char **first, char **last);
2451
2452 When a pattern is compiled with the PCRE_DUPNAMES option, names for
2453 subpatterns are not required to be unique. Normally, patterns with
2454 duplicate names are such that in any one match, only one of the named
2455 subpatterns participates. An example is shown in the pcrepattern docu-
2456 mentation.
2457
2458 When duplicates are present, pcre_copy_named_substring() and
2459 pcre_get_named_substring() return the first substring corresponding to
2460 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING
2461 (-7) is returned; no data is returned. The pcre_get_stringnumber()
2462 function returns one of the numbers that are associated with the name,
2463 but it is not defined which it is.
2464
2465 If you want to get full details of all captured substrings for a given
2466 name, you must use the pcre_get_stringtable_entries() function. The
2467 first argument is the compiled pattern, and the second is the name. The
2468 third and fourth are pointers to variables which are updated by the
2469 function. After it has run, they point to the first and last entries in
2470 the name-to-number table for the given name. The function itself
2471 returns the length of each entry, or PCRE_ERROR_NOSUBSTRING (-7) if
2472 there are none. The format of the table is described above in the sec-
2473 tion entitled Information about a pattern. Given all the relevant
2474 entries for the name, you can extract each of their numbers, and hence
2475 the captured data, if any.
2476
2477
2478 FINDING ALL POSSIBLE MATCHES
2479
2480 The traditional matching function uses a similar algorithm to Perl,
2481 which stops when it finds the first match, starting at a given point in
2482 the subject. If you want to find all possible matches, or the longest
2483 possible match, consider using the alternative matching function (see
2484 below) instead. If you cannot use the alternative function, but still
2485 need to find all possible matches, you can kludge it up by making use
2486 of the callout facility, which is described in the pcrecallout documen-
2487 tation.
2488
2489 What you have to do is to insert a callout right at the end of the pat-
2490 tern. When your callout function is called, extract and save the cur-
2491 rent matched substring. Then return 1, which forces pcre_exec() to
2492 backtrack and try other alternatives. Ultimately, when it runs out of
2493 matches, pcre_exec() will yield PCRE_ERROR_NOMATCH.
2494
2495
2496 MATCHING A PATTERN: THE ALTERNATIVE FUNCTION
2497
2498 int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
2499 const char *subject, int length, int startoffset,
2500 int options, int *ovector, int ovecsize,
2501 int *workspace, int wscount);
2502
2503 The function pcre_dfa_exec() is called to match a subject string
2504 against a compiled pattern, using a matching algorithm that scans the
2505 subject string just once, and does not backtrack. This has different
2506 characteristics to the normal algorithm, and is not compatible with
2507 Perl. Some of the features of PCRE patterns are not supported. Never-
2508 theless, there are times when this kind of matching can be useful. For
2509 a discussion of the two matching algorithms, and a list of features
2510 that pcre_dfa_exec() does not support, see the pcrematching documenta-
2511 tion.
2512
2513 The arguments for the pcre_dfa_exec() function are the same as for
2514 pcre_exec(), plus two extras. The ovector argument is used in a differ-
2515 ent way, and this is described below. The other common arguments are
2516 used in the same way as for pcre_exec(), so their description is not
2517 repeated here.
2518
2519 The two additional arguments provide workspace for the function. The
2520 workspace vector should contain at least 20 elements. It is used for
2521 keeping track of multiple paths through the pattern tree. More
2522 workspace will be needed for patterns and subjects where there are a
2523 lot of potential matches.
2524
2525 Here is an example of a simple call to pcre_dfa_exec():
2526
2527 int rc;
2528 int ovector[10];
2529 int wspace[20];
2530 rc = pcre_dfa_exec(
2531 re, /* result of pcre_compile() */
2532 NULL, /* we didn't study the pattern */
2533 "some string", /* the subject string */
2534 11, /* the length of the subject string */
2535 0, /* start at offset 0 in the subject */
2536 0, /* default options */
2537 ovector, /* vector of integers for substring information */
2538 10, /* number of elements (NOT size in bytes) */
2539 wspace, /* working space vector */
2540 20); /* number of elements (NOT size in bytes) */
2541
2542 Option bits for pcre_dfa_exec()
2543
2544 The unused bits of the options argument for pcre_dfa_exec() must be
2545 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEW-
2546 LINE_xxx, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
2547 PCRE_NOTEMPTY_ATSTART, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, PCRE_PAR-
2548 TIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART. All but the last
2549 four of these are exactly the same as for pcre_exec(), so their
2550 description is not repeated here.
2551
2552 PCRE_PARTIAL_HARD
2553 PCRE_PARTIAL_SOFT
2554
2555 These have the same general effect as they do for pcre_exec(), but the
2556 details are slightly different. When PCRE_PARTIAL_HARD is set for
2557 pcre_dfa_exec(), it returns PCRE_ERROR_PARTIAL if the end of the sub-
2558 ject is reached and there is still at least one matching possibility
2559 that requires additional characters. This happens even if some complete
2560 matches have also been found. When PCRE_PARTIAL_SOFT is set, the return
2561 code PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if the end
2562 of the subject is reached, there have been no complete matches, but
2563 there is still at least one matching possibility. The portion of the
2564 string that was inspected when the longest partial match was found is
2565 set as the first matching string in both cases.
2566
2567 PCRE_DFA_SHORTEST
2568
2569 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to
2570 stop as soon as it has found one match. Because of the way the alterna-
2571 tive algorithm works, this is necessarily the shortest possible match
2572 at the first possible matching point in the subject string.
2573
2574 PCRE_DFA_RESTART
2575
2576 When pcre_dfa_exec() returns a partial match, it is possible to call it
2577 again, with additional subject characters, and have it continue with
2578 the same match. The PCRE_DFA_RESTART option requests this action; when
2579 it is set, the workspace and wscount options must reference the same
2580 vector as before because data about the match so far is left in them
2581 after a partial match. There is more discussion of this facility in the
2582 pcrepartial documentation.
2583
2584 Successful returns from pcre_dfa_exec()
2585
2586 When pcre_dfa_exec() succeeds, it may have matched more than one sub-
2587 string in the subject. Note, however, that all the matches from one run
2588 of the function start at the same point in the subject. The shorter
2589 matches are all initial substrings of the longer matches. For example,
2590 if the pattern
2591
2592 <.*>
2593
2594 is matched against the string
2595
2596 This is <something> <something else> <something further> no more
2597
2598 the three matched strings are
2599
2600 <something>
2601 <something> <something else>
2602 <something> <something else> <something further>
2603
2604 On success, the yield of the function is a number greater than zero,
2605 which is the number of matched substrings. The substrings themselves
2606 are returned in ovector. Each string uses two elements; the first is
2607 the offset to the start, and the second is the offset to the end. In
2608 fact, all the strings have the same start offset. (Space could have
2609 been saved by giving this only once, but it was decided to retain some
2610 compatibility with the way pcre_exec() returns data, even though the
2611 meaning of the strings is different.)
2612
2613 The strings are returned in reverse order of length; that is, the long-
2614 est matching string is given first. If there were too many matches to
2615 fit into ovector, the yield of the function is zero, and the vector is
2616 filled with the longest matches.
2617
2618 Error returns from pcre_dfa_exec()
2619
2620 The pcre_dfa_exec() function returns a negative number when it fails.
2621 Many of the errors are the same as for pcre_exec(), and these are
2622 described above. There are in addition the following errors that are
2623 specific to pcre_dfa_exec():
2624
2625 PCRE_ERROR_DFA_UITEM (-16)
2626
2627 This return is given if pcre_dfa_exec() encounters an item in the pat-
2628 tern that it does not support, for instance, the use of \C or a back
2629 reference.
2630
2631 PCRE_ERROR_DFA_UCOND (-17)
2632
2633 This return is given if pcre_dfa_exec() encounters a condition item
2634 that uses a back reference for the condition, or a test for recursion
2635 in a specific group. These are not supported.
2636
2637 PCRE_ERROR_DFA_UMLIMIT (-18)
2638
2639 This return is given if pcre_dfa_exec() is called with an extra block
2640 that contains a setting of the match_limit field. This is not supported
2641 (it is meaningless).
2642
2643 PCRE_ERROR_DFA_WSSIZE (-19)
2644
2645 This return is given if pcre_dfa_exec() runs out of space in the
2646 workspace vector.
2647
2648 PCRE_ERROR_DFA_RECURSE (-20)
2649
2650 When a recursive subpattern is processed, the matching function calls
2651 itself recursively, using private vectors for ovector and workspace.
2652 This error is given if the output vector is not large enough. This
2653 should be extremely rare, as a vector of size 1000 is used.
2654
2655
2656 SEE ALSO
2657
2658 pcrebuild(3), pcrecallout(3), pcrecpp(3)(3), pcrematching(3), pcrepar-
2659 tial(3), pcreposix(3), pcreprecompile(3), pcresample(3), pcrestack(3).
2660
2661
2662 AUTHOR
2663
2664 Philip Hazel
2665 University Computing Service
2666 Cambridge CB2 3QH, England.
2667
2668
2669 REVISION
2670
2671 Last updated: 22 September 2009
2672 Copyright (c) 1997-2009 University of Cambridge.
2673 ------------------------------------------------------------------------------
2674
2675
2676 PCRECALLOUT(3) PCRECALLOUT(3)
2677
2678
2679 NAME
2680 PCRE - Perl-compatible regular expressions
2681
2682
2683 PCRE CALLOUTS
2684
2685 int (*pcre_callout)(pcre_callout_block *);
2686
2687 PCRE provides a feature called "callout", which is a means of temporar-
2688 ily passing control to the caller of PCRE in the middle of pattern
2689 matching. The caller of PCRE provides an external function by putting
2690 its entry point in the global variable pcre_callout. By default, this
2691 variable contains NULL, which disables all calling out.
2692
2693 Within a regular expression, (?C) indicates the points at which the
2694 external function is to be called. Different callout points can be
2695 identified by putting a number less than 256 after the letter C. The
2696 default value is zero. For example, this pattern has two callout
2697 points:
2698
2699 (?C1)abc(?C2)def
2700
2701 If the PCRE_AUTO_CALLOUT option bit is set when pcre_compile() is
2702 called, PCRE automatically inserts callouts, all with number 255,
2703 before each item in the pattern. For example, if PCRE_AUTO_CALLOUT is
2704 used with the pattern
2705
2706 A(\d{2}|--)
2707
2708 it is processed as if it were
2709
2710 (?C255)A(?C255)((?C255)\d{2}(?C255)|(?C255)-(?C255)-(?C255))(?C255)
2711
2712 Notice that there is a callout before and after each parenthesis and
2713 alternation bar. Automatic callouts can be used for tracking the
2714 progress of pattern matching. The pcretest command has an option that
2715 sets automatic callouts; when it is used, the output indicates how the
2716 pattern is matched. This is useful information when you are trying to
2717 optimize the performance of a particular pattern.
2718
2719
2720 MISSING CALLOUTS
2721
2722 You should be aware that, because of optimizations in the way PCRE
2723 matches patterns by default, callouts sometimes do not happen. For
2724 example, if the pattern is
2725
2726 ab(?C4)cd
2727
2728 PCRE knows that any matching string must contain the letter "d". If the
2729 subject string is "abyz", the lack of "d" means that matching doesn't
2730 ever start, and the callout is never reached. However, with "abyd",
2731 though the result is still no match, the callout is obeyed.
2732
2733 You can disable these optimizations by passing the PCRE_NO_START_OPTI-
2734 MIZE option to pcre_exec() or pcre_dfa_exec(). This slows down the
2735 matching process, but does ensure that callouts such as the example
2736 above are obeyed.
2737
2738
2739 THE CALLOUT INTERFACE
2740
2741 During matching, when PCRE reaches a callout point, the external func-
2742 tion defined by pcre_callout is called (if it is set). This applies to
2743 both the pcre_exec() and the pcre_dfa_exec() matching functions. The
2744 only argument to the callout function is a pointer to a pcre_callout
2745 block. This structure contains the following fields:
2746
2747 int version;
2748 int callout_number;
2749 int *offset_vector;
2750 const char *subject;
2751 int subject_length;
2752 int start_match;
2753 int current_position;
2754 int capture_top;
2755 int capture_last;
2756 void *callout_data;
2757 int pattern_position;
2758 int next_item_length;
2759
2760 The version field is an integer containing the version number of the
2761 block format. The initial version was 0; the current version is 1. The
2762 version number will change again in future if additional fields are
2763 added, but the intention is never to remove any of the existing fields.
2764
2765 The callout_number field contains the number of the callout, as com-
2766 piled into the pattern (that is, the number after ?C for manual call-
2767 outs, and 255 for automatically generated callouts).
2768
2769 The offset_vector field is a pointer to the vector of offsets that was
2770 passed by the caller to pcre_exec() or pcre_dfa_exec(). When
2771 pcre_exec() is used, the contents can be inspected in order to extract
2772 substrings that have been matched so far, in the same way as for
2773 extracting substrings after a match has completed. For pcre_dfa_exec()
2774 this field is not useful.
2775
2776 The subject and subject_length fields contain copies of the values that
2777 were passed to pcre_exec().
2778
2779 The start_match field normally contains the offset within the subject
2780 at which the current match attempt started. However, if the escape
2781 sequence \K has been encountered, this value is changed to reflect the
2782 modified starting point. If the pattern is not anchored, the callout
2783 function may be called several times from the same point in the pattern
2784 for different starting points in the subject.
2785
2786 The current_position field contains the offset within the subject of
2787 the current match pointer.
2788
2789 When the pcre_exec() function is used, the capture_top field contains
2790 one more than the number of the highest numbered captured substring so
2791 far. If no substrings have been captured, the value of capture_top is
2792 one. This is always the case when pcre_dfa_exec() is used, because it
2793 does not support captured substrings.
2794
2795 The capture_last field contains the number of the most recently cap-
2796 tured substring. If no substrings have been captured, its value is -1.
2797 This is always the case when pcre_dfa_exec() is used.
2798
2799 The callout_data field contains a value that is passed to pcre_exec()
2800 or pcre_dfa_exec() specifically so that it can be passed back in call-
2801 outs. It is passed in the pcre_callout field of the pcre_extra data
2802 structure. If no such data was passed, the value of callout_data in a
2803 pcre_callout block is NULL. There is a description of the pcre_extra
2804 structure in the pcreapi documentation.
2805
2806 The pattern_position field is present from version 1 of the pcre_call-
2807 out structure. It contains the offset to the next item to be matched in
2808 the pattern string.
2809
2810 The next_item_length field is present from version 1 of the pcre_call-
2811 out structure. It contains the length of the next item to be matched in
2812 the pattern string. When the callout immediately precedes an alterna-
2813 tion bar, a closing parenthesis, or the end of the pattern, the length
2814 is zero. When the callout precedes an opening parenthesis, the length
2815 is that of the entire subpattern.
2816
2817 The pattern_position and next_item_length fields are intended to help
2818 in distinguishing between different automatic callouts, which all have
2819 the same callout number. However, they are set for all callouts.
2820
2821
2822 RETURN VALUES
2823
2824 The external callout function returns an integer to PCRE. If the value
2825 is zero, matching proceeds as normal. If the value is greater than
2826 zero, matching fails at the current point, but the testing of other
2827 matching possibilities goes ahead, just as if a lookahead assertion had
2828 failed. If the value is less than zero, the match is abandoned, and
2829 pcre_exec() (or pcre_dfa_exec()) returns the negative value.
2830
2831 Negative values should normally be chosen from the set of
2832 PCRE_ERROR_xxx values. In particular, PCRE_ERROR_NOMATCH forces a stan-
2833 dard "no match" failure. The error number PCRE_ERROR_CALLOUT is
2834 reserved for use by callout functions; it will never be used by PCRE
2835 itself.
2836
2837
2838 AUTHOR
2839
2840 Philip Hazel
2841 University Computing Service
2842 Cambridge CB2 3QH, England.
2843
2844
2845 REVISION
2846
2847 Last updated: 15 March 2009
2848 Copyright (c) 1997-2009 University of Cambridge.
2849 ------------------------------------------------------------------------------
2850
2851
2852 PCRECOMPAT(3) PCRECOMPAT(3)
2853
2854
2855 NAME
2856 PCRE - Perl-compatible regular expressions
2857
2858
2859 DIFFERENCES BETWEEN PCRE AND PERL
2860
2861 This document describes the differences in the ways that PCRE and Perl
2862 handle regular expressions. The differences described here are mainly
2863 with respect to Perl 5.8, though PCRE versions 7.0 and later contain
2864 some features that are in Perl 5.10.
2865
2866 1. PCRE has only a subset of Perl's UTF-8 and Unicode support. Details
2867 of what it does have are given in the section on UTF-8 support in the
2868 main pcre page.
2869
2870 2. PCRE does not allow repeat quantifiers on lookahead assertions. Perl
2871 permits them, but they do not mean what you might think. For example,
2872 (?!a){3} does not assert that the next three characters are not "a". It
2873 just asserts that the next character is not "a" three times.
2874
2875 3. Capturing subpatterns that occur inside negative lookahead asser-
2876 tions are counted, but their entries in the offsets vector are never
2877 set. Perl sets its numerical variables from any such patterns that are
2878 matched before the assertion fails to match something (thereby succeed-
2879 ing), but only if the negative lookahead assertion contains just one
2880 branch.
2881
2882 4. Though binary zero characters are supported in the subject string,
2883 they are not allowed in a pattern string because it is passed as a nor-
2884 mal C string, terminated by zero. The escape sequence \0 can be used in
2885 the pattern to represent a binary zero.
2886
2887 5. The following Perl escape sequences are not supported: \l, \u, \L,
2888 \U, and \N. In fact these are implemented by Perl's general string-han-
2889 dling and are not part of its pattern matching engine. If any of these
2890 are encountered by PCRE, an error is generated.
2891
2892 6. The Perl escape sequences \p, \P, and \X are supported only if PCRE
2893 is built with Unicode character property support. The properties that
2894 can be tested with \p and \P are limited to the general category prop-
2895 erties such as Lu and Nd, script names such as Greek or Han, and the
2896 derived properties Any and L&. PCRE does support the Cs (surrogate)
2897 property, which Perl does not; the Perl documentation says "Because
2898 Perl hides the need for the user to understand the internal representa-
2899 tion of Unicode characters, there is no need to implement the somewhat
2900 messy concept of surrogates."
2901
2902 7. PCRE does support the \Q...\E escape for quoting substrings. Charac-
2903 ters in between are treated as literals. This is slightly different
2904 from Perl in that $ and @ are also handled as literals inside the
2905 quotes. In Perl, they cause variable interpolation (but of course PCRE
2906 does not have variables). Note the following examples:
2907
2908 Pattern PCRE matches Perl matches
2909
2910 \Qabc$xyz\E abc$xyz abc followed by the
2911 contents of $xyz
2912 \Qabc\$xyz\E abc\$xyz abc\$xyz
2913 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
2914
2915 The \Q...\E sequence is recognized both inside and outside character
2916 classes.
2917
2918 8. Fairly obviously, PCRE does not support the (?{code}) and (??{code})
2919 constructions. However, there is support for recursive patterns. This
2920 is not available in Perl 5.8, but it is in Perl 5.10. Also, the PCRE
2921 "callout" feature allows an external function to be called during pat-
2922 tern matching. See the pcrecallout documentation for details.
2923
2924 9. Subpatterns that are called recursively or as "subroutines" are
2925 always treated as atomic groups in PCRE. This is like Python, but
2926 unlike Perl. There is a discussion of an example that explains this in
2927 more detail in the section on recursion differences from Perl in the
2928 pcrecompat page.
2929
2930 10. There are some differences that are concerned with the settings of
2931 captured strings when part of a pattern is repeated. For example,
2932 matching "aba" against the pattern /^(a(b)?)+$/ in Perl leaves $2
2933 unset, but in PCRE it is set to "b".
2934
2935 11. PCRE does support Perl 5.10's backtracking verbs (*ACCEPT),
2936 (*FAIL), (*F), (*COMMIT), (*PRUNE), (*SKIP), and (*THEN), but only in
2937 the forms without an argument. PCRE does not support (*MARK).
2938
2939 12. PCRE provides some extensions to the Perl regular expression facil-
2940 ities. Perl 5.10 will include new features that are not in earlier
2941 versions, some of which (such as named parentheses) have been in PCRE
2942 for some time. This list is with respect to Perl 5.10:
2943
2944 (a) Although lookbehind assertions must match fixed length strings,
2945 each alternative branch of a lookbehind assertion can match a different
2946 length of string. Perl requires them all to have the same length.
2947
2948 (b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $
2949 meta-character matches only at the very end of the string.
2950
2951 (c) If PCRE_EXTRA is set, a backslash followed by a letter with no spe-
2952 cial meaning is faulted. Otherwise, like Perl, the backslash is quietly
2953 ignored. (Perl can be made to issue a warning.)
2954
2955 (d) If PCRE_UNGREEDY is set, the greediness of the repetition quanti-
2956 fiers is inverted, that is, by default they are not greedy, but if fol-
2957 lowed by a question mark they are.
2958
2959 (e) PCRE_ANCHORED can be used at matching time to force a pattern to be
2960 tried only at the first matching position in the subject string.
2961
2962 (f) The PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2963 and PCRE_NO_AUTO_CAPTURE options for pcre_exec() have no Perl equiva-
2964 lents.
2965
2966 (g) The \R escape sequence can be restricted to match only CR, LF, or
2967 CRLF by the PCRE_BSR_ANYCRLF option.
2968
2969 (h) The callout facility is PCRE-specific.
2970
2971 (i) The partial matching facility is PCRE-specific.
2972
2973 (j) Patterns compiled by PCRE can be saved and re-used at a later time,
2974 even on different hosts that have the other endianness.
2975
2976 (k) The alternative matching function (pcre_dfa_exec()) matches in a
2977 different way and is not Perl-compatible.
2978
2979 (l) PCRE recognizes some special sequences such as (*CR) at the start
2980 of a pattern that set overall options that cannot be changed within the
2981 pattern.
2982
2983
2984 AUTHOR
2985
2986 Philip Hazel
2987 University Computing Service
2988 Cambridge CB2 3QH, England.
2989
2990
2991 REVISION
2992
2993 Last updated: 18 September 2009
2994 Copyright (c) 1997-2009 University of Cambridge.
2995 ------------------------------------------------------------------------------
2996
2997
2998 PCREPATTERN(3) PCREPATTERN(3)
2999
3000
3001 NAME
3002 PCRE - Perl-compatible regular expressions
3003
3004
3005 PCRE REGULAR EXPRESSION DETAILS
3006
3007 The syntax and semantics of the regular expressions that are supported
3008 by PCRE are described in detail below. There is a quick-reference syn-
3009 tax summary in the pcresyntax page. PCRE tries to match Perl syntax and
3010 semantics as closely as it can. PCRE also supports some alternative
3011 regular expression syntax (which does not conflict with the Perl syn-
3012 tax) in order to provide some compatibility with regular expressions in
3013 Python, .NET, and Oniguruma.
3014
3015 Perl's regular expressions are described in its own documentation, and
3016 regular expressions in general are covered in a number of books, some
3017 of which have copious examples. Jeffrey Friedl's "Mastering Regular
3018 Expressions", published by O'Reilly, covers regular expressions in
3019 great detail. This description of PCRE's regular expressions is
3020 intended as reference material.
3021
3022 The original operation of PCRE was on strings of one-byte characters.
3023 However, there is now also support for UTF-8 character strings. To use
3024 this, you must build PCRE to include UTF-8 support, and then call
3025 pcre_compile() with the PCRE_UTF8 option. There is also a special
3026 sequence that can be given at the start of a pattern:
3027
3028 (*UTF8)
3029
3030 Starting a pattern with this sequence is equivalent to setting the
3031 PCRE_UTF8 option. This feature is not Perl-compatible. How setting
3032 UTF-8 mode affects pattern matching is mentioned in several places
3033 below. There is also a summary of UTF-8 features in the section on
3034 UTF-8 support in the main pcre page.
3035
3036 The remainder of this document discusses the patterns that are sup-
3037 ported by PCRE when its main matching function, pcre_exec(), is used.
3038 From release 6.0, PCRE offers a second matching function,
3039 pcre_dfa_exec(), which matches using a different algorithm that is not
3040 Perl-compatible. Some of the features discussed below are not available
3041 when pcre_dfa_exec() is used. The advantages and disadvantages of the
3042 alternative function, and how it differs from the normal function, are
3043 discussed in the pcrematching page.
3044
3045
3046 NEWLINE CONVENTIONS
3047
3048 PCRE supports five different conventions for indicating line breaks in
3049 strings: a single CR (carriage return) character, a single LF (line-
3050 feed) character, the two-character sequence CRLF, any of the three pre-
3051 ceding, or any Unicode newline sequence. The pcreapi page has further
3052 discussion about newlines, and shows how to set the newline convention
3053 in the options arguments for the compiling and matching functions.
3054
3055 It is also possible to specify a newline convention by starting a pat-
3056 tern string with one of the following five sequences:
3057
3058 (*CR) carriage return
3059 (*LF) linefeed
3060 (*CRLF) carriage return, followed by linefeed
3061 (*ANYCRLF) any of the three above
3062 (*ANY) all Unicode newline sequences
3063
3064 These override the default and the options given to pcre_compile(). For
3065 example, on a Unix system where LF is the default newline sequence, the
3066 pattern
3067
3068 (*CR)a.b
3069
3070 changes the convention to CR. That pattern matches "a\nb" because LF is
3071 no longer a newline. Note that these special settings, which are not
3072 Perl-compatible, are recognized only at the very start of a pattern,
3073 and that they must be in upper case. If more than one of them is
3074 present, the last one is used.
3075
3076 The newline convention does not affect what the \R escape sequence
3077 matches. By default, this is any Unicode newline sequence, for Perl
3078 compatibility. However, this can be changed; see the description of \R
3079 in the section entitled "Newline sequences" below. A change of \R set-
3080 ting can be combined with a change of newline convention.
3081
3082
3083 CHARACTERS AND METACHARACTERS
3084
3085 A regular expression is a pattern that is matched against a subject
3086 string from left to right. Most characters stand for themselves in a
3087 pattern, and match the corresponding characters in the subject. As a
3088 trivial example, the pattern
3089
3090 The quick brown fox
3091
3092 matches a portion of a subject string that is identical to itself. When
3093 caseless matching is specified (the PCRE_CASELESS option), letters are
3094 matched independently of case. In UTF-8 mode, PCRE always understands
3095 the concept of case for characters whose values are less than 128, so
3096 caseless matching is always possible. For characters with higher val-
3097 ues, the concept of case is supported if PCRE is compiled with Unicode
3098 property support, but not otherwise. If you want to use caseless
3099 matching for characters 128 and above, you must ensure that PCRE is
3100 compiled with Unicode property support as well as with UTF-8 support.
3101
3102 The power of regular expressions comes from the ability to include
3103 alternatives and repetitions in the pattern. These are encoded in the
3104 pattern by the use of metacharacters, which do not stand for themselves
3105 but instead are interpreted in some special way.
3106
3107 There are two different sets of metacharacters: those that are recog-
3108 nized anywhere in the pattern except within square brackets, and those
3109 that are recognized within square brackets. Outside square brackets,
3110 the metacharacters are as follows:
3111
3112 \ general escape character with several uses
3113 ^ assert start of string (or line, in multiline mode)
3114 $ assert end of string (or line, in multiline mode)
3115 . match any character except newline (by default)
3116 [ start character class definition
3117 | start of alternative branch
3118 ( start subpattern
3119 ) end subpattern
3120 ? extends the meaning of (
3121 also 0 or 1 quantifier
3122 also quantifier minimizer
3123 * 0 or more quantifier
3124 + 1 or more quantifier
3125 also "possessive quantifier"
3126 { start min/max quantifier
3127
3128 Part of a pattern that is in square brackets is called a "character
3129 class". In a character class the only metacharacters are:
3130
3131 \ general escape character
3132 ^ negate the class, but only if the first character
3133 - indicates character range
3134 [ POSIX character class (only if followed by POSIX
3135 syntax)
3136 ] terminates the character class
3137
3138 The following sections describe the use of each of the metacharacters.
3139
3140
3141 BACKSLASH
3142
3143 The backslash character has several uses. Firstly, if it is followed by
3144 a non-alphanumeric character, it takes away any special meaning that
3145 character may have. This use of backslash as an escape character
3146 applies both inside and outside character classes.
3147
3148 For example, if you want to match a * character, you write \* in the
3149 pattern. This escaping action applies whether or not the following
3150 character would otherwise be interpreted as a metacharacter, so it is
3151 always safe to precede a non-alphanumeric with backslash to specify
3152 that it stands for itself. In particular, if you want to match a back-
3153 slash, you write \\.
3154
3155 If a pattern is compiled with the PCRE_EXTENDED option, whitespace in
3156 the pattern (other than in a character class) and characters between a
3157 # outside a character class and the next newline are ignored. An escap-
3158 ing backslash can be used to include a whitespace or # character as
3159 part of the pattern.
3160
3161 If you want to remove the special meaning from a sequence of charac-
3162 ters, you can do so by putting them between \Q and \E. This is differ-
3163 ent from Perl in that $ and @ are handled as literals in \Q...\E
3164 sequences in PCRE, whereas in Perl, $ and @ cause variable interpola-
3165 tion. Note the following examples:
3166
3167 Pattern PCRE matches Perl matches
3168
3169 \Qabc$xyz\E abc$xyz abc followed by the
3170 contents of $xyz
3171 \Qabc\$xyz\E abc\$xyz abc\$xyz
3172 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3173
3174 The \Q...\E sequence is recognized both inside and outside character
3175 classes.
3176
3177 Non-printing characters
3178
3179 A second use of backslash provides a way of encoding non-printing char-
3180 acters in patterns in a visible manner. There is no restriction on the
3181 appearance of non-printing characters, apart from the binary zero that
3182 terminates a pattern, but when a pattern is being prepared by text
3183 editing, it is usually easier to use one of the following escape
3184 sequences than the binary character it represents:
3185
3186 \a alarm, that is, the BEL character (hex 07)
3187 \cx "control-x", where x is any character
3188 \e escape (hex 1B)
3189 \f formfeed (hex 0C)
3190 \n linefeed (hex 0A)
3191 \r carriage return (hex 0D)
3192 \t tab (hex 09)
3193 \ddd character with octal code ddd, or backreference
3194 \xhh character with hex code hh
3195 \x{hhh..} character with hex code hhh..
3196
3197 The precise effect of \cx is as follows: if x is a lower case letter,
3198 it is converted to upper case. Then bit 6 of the character (hex 40) is
3199 inverted. Thus \cz becomes hex 1A, but \c{ becomes hex 3B, while \c;
3200 becomes hex 7B.
3201
3202 After \x, from zero to two hexadecimal digits are read (letters can be
3203 in upper or lower case). Any number of hexadecimal digits may appear
3204 between \x{ and }, but the value of the character code must be less
3205 than 256 in non-UTF-8 mode, and less than 2**31 in UTF-8 mode. That is,
3206 the maximum value in hexadecimal is 7FFFFFFF. Note that this is bigger
3207 than the largest Unicode code point, which is 10FFFF.
3208
3209 If characters other than hexadecimal digits appear between \x{ and },
3210 or if there is no terminating }, this form of escape is not recognized.
3211 Instead, the initial \x will be interpreted as a basic hexadecimal
3212 escape, with no following digits, giving a character whose value is
3213 zero.
3214
3215 Characters whose value is less than 256 can be defined by either of the
3216 two syntaxes for \x. There is no difference in the way they are han-
3217 dled. For example, \xdc is exactly the same as \x{dc}.
3218
3219 After \0 up to two further octal digits are read. If there are fewer
3220 than two digits, just those that are present are used. Thus the
3221 sequence \0\x\07 specifies two binary zeros followed by a BEL character
3222 (code value 7). Make sure you supply two digits after the initial zero
3223 if the pattern character that follows is itself an octal digit.
3224
3225 The handling of a backslash followed by a digit other than 0 is compli-
3226 cated. Outside a character class, PCRE reads it and any following dig-
3227 its as a decimal number. If the number is less than 10, or if there
3228 have been at least that many previous capturing left parentheses in the
3229 expression, the entire sequence is taken as a back reference. A
3230 description of how this works is given later, following the discussion
3231 of parenthesized subpatterns.
3232
3233 Inside a character class, or if the decimal number is greater than 9
3234 and there have not been that many capturing subpatterns, PCRE re-reads
3235 up to three octal digits following the backslash, and uses them to gen-
3236 erate a data character. Any subsequent digits stand for themselves. In
3237 non-UTF-8 mode, the value of a character specified in octal must be
3238 less than \400. In UTF-8 mode, values up to \777 are permitted. For
3239 example:
3240
3241 \040 is another way of writing a space
3242 \40 is the same, provided there are fewer than 40
3243 previous capturing subpatterns
3244 \7 is always a back reference
3245 \11 might be a back reference, or another way of
3246 writing a tab
3247 \011 is always a tab
3248 \0113 is a tab followed by the character "3"
3249 \113 might be a back reference, otherwise the
3250 character with octal code 113
3251 \377 might be a back reference, otherwise
3252 the byte consisting entirely of 1 bits
3253 \81 is either a back reference, or a binary zero
3254 followed by the two characters "8" and "1"
3255
3256 Note that octal values of 100 or greater must not be introduced by a
3257 leading zero, because no more than three octal digits are ever read.
3258
3259 All the sequences that define a single character value can be used both
3260 inside and outside character classes. In addition, inside a character
3261 class, the sequence \b is interpreted as the backspace character (hex
3262 08), and the sequences \R and \X are interpreted as the characters "R"
3263 and "X", respectively. Outside a character class, these sequences have
3264 different meanings (see below).
3265
3266 Absolute and relative back references
3267
3268 The sequence \g followed by an unsigned or a negative number, option-
3269 ally enclosed in braces, is an absolute or relative back reference. A
3270 named back reference can be coded as \g{name}. Back references are dis-
3271 cussed later, following the discussion of parenthesized subpatterns.
3272
3273 Absolute and relative subroutine calls
3274
3275 For compatibility with Oniguruma, the non-Perl syntax \g followed by a
3276 name or a number enclosed either in angle brackets or single quotes, is
3277 an alternative syntax for referencing a subpattern as a "subroutine".
3278 Details are discussed later. Note that \g{...} (Perl syntax) and
3279 \g<...> (Oniguruma syntax) are not synonymous. The former is a back
3280 reference; the latter is a subroutine call.
3281
3282 Generic character types
3283
3284 Another use of backslash is for specifying generic character types. The
3285 following are always recognized:
3286
3287 \d any decimal digit
3288 \D any character that is not a decimal digit
3289 \h any horizontal whitespace character
3290 \H any character that is not a horizontal whitespace character
3291 \s any whitespace character
3292 \S any character that is not a whitespace character
3293 \v any vertical whitespace character
3294 \V any character that is not a vertical whitespace character
3295 \w any "word" character
3296 \W any "non-word" character
3297
3298 Each pair of escape sequences partitions the complete set of characters
3299 into two disjoint sets. Any given character matches one, and only one,
3300 of each pair.
3301
3302 These character type sequences can appear both inside and outside char-
3303 acter classes. They each match one character of the appropriate type.
3304 If the current matching point is at the end of the subject string, all
3305 of them fail, since there is no character to match.
3306
3307 For compatibility with Perl, \s does not match the VT character (code
3308 11). This makes it different from the the POSIX "space" class. The \s
3309 characters are HT (9), LF (10), FF (12), CR (13), and space (32). If
3310 "use locale;" is included in a Perl script, \s may match the VT charac-
3311 ter. In PCRE, it never does.
3312
3313 In UTF-8 mode, characters with values greater than 128 never match \d,
3314 \s, or \w, and always match \D, \S, and \W. This is true even when Uni-
3315 code character property support is available. These sequences retain
3316 their original meanings from before UTF-8 support was available, mainly
3317 for efficiency reasons. Note that this also affects \b, because it is
3318 defined in terms of \w and \W.
3319
3320 The sequences \h, \H, \v, and \V are Perl 5.10 features. In contrast to
3321 the other sequences, these do match certain high-valued codepoints in
3322 UTF-8 mode. The horizontal space characters are:
3323
3324 U+0009 Horizontal tab
3325 U+0020 Space
3326 U+00A0 Non-break space
3327 U+1680 Ogham space mark
3328 U+180E Mongolian vowel separator
3329 U+2000 En quad
3330 U+2001 Em quad
3331 U+2002 En space
3332 U+2003 Em space
3333 U+2004 Three-per-em space
3334 U+2005 Four-per-em space
3335 U+2006 Six-per-em space
3336 U+2007 Figure space
3337 U+2008 Punctuation space
3338 U+2009 Thin space
3339 U+200A Hair space
3340 U+202F Narrow no-break space
3341 U+205F Medium mathematical space
3342 U+3000 Ideographic space
3343
3344 The vertical space characters are:
3345
3346 U+000A Linefeed
3347 U+000B Vertical tab
3348 U+000C Formfeed
3349 U+000D Carriage return
3350 U+0085 Next line
3351 U+2028 Line separator
3352 U+2029 Paragraph separator
3353
3354 A "word" character is an underscore or any character less than 256 that
3355 is a letter or digit. The definition of letters and digits is con-
3356 trolled by PCRE's low-valued character tables, and may vary if locale-
3357 specific matching is taking place (see "Locale support" in the pcreapi
3358 page). For example, in a French locale such as "fr_FR" in Unix-like
3359 systems, or "french" in Windows, some character codes greater than 128
3360 are used for accented letters, and these are matched by \w. The use of
3361 locales with Unicode is discouraged.
3362
3363 Newline sequences
3364
3365 Outside a character class, by default, the escape sequence \R matches
3366 any Unicode newline sequence. This is a Perl 5.10 feature. In non-UTF-8
3367 mode \R is equivalent to the following:
3368
3369 (?>\r\n|\n|\x0b|\f|\r|\x85)
3370
3371 This is an example of an "atomic group", details of which are given
3372 below. This particular group matches either the two-character sequence
3373 CR followed by LF, or one of the single characters LF (linefeed,
3374 U+000A), VT (vertical tab, U+000B), FF (formfeed, U+000C), CR (carriage
3375 return, U+000D), or NEL (next line, U+0085). The two-character sequence
3376 is treated as a single unit that cannot be split.
3377
3378 In UTF-8 mode, two additional characters whose codepoints are greater
3379 than 255 are added: LS (line separator, U+2028) and PS (paragraph sepa-
3380 rator, U+2029). Unicode character property support is not needed for
3381 these characters to be recognized.
3382
3383 It is possible to restrict \R to match only CR, LF, or CRLF (instead of
3384 the complete set of Unicode line endings) by setting the option
3385 PCRE_BSR_ANYCRLF either at compile time or when the pattern is matched.
3386 (BSR is an abbrevation for "backslash R".) This can be made the default
3387 when PCRE is built; if this is the case, the other behaviour can be
3388 requested via the PCRE_BSR_UNICODE option. It is also possible to
3389 specify these settings by starting a pattern string with one of the
3390 following sequences:
3391
3392 (*BSR_ANYCRLF) CR, LF, or CRLF only
3393 (*BSR_UNICODE) any Unicode newline sequence
3394
3395 These override the default and the options given to pcre_compile(), but
3396 they can be overridden by options given to pcre_exec(). Note that these
3397 special settings, which are not Perl-compatible, are recognized only at
3398 the very start of a pattern, and that they must be in upper case. If
3399 more than one of them is present, the last one is used. They can be
3400 combined with a change of newline convention, for example, a pattern
3401 can start with:
3402
3403 (*ANY)(*BSR_ANYCRLF)
3404
3405 Inside a character class, \R matches the letter "R".
3406
3407 Unicode character properties
3408
3409 When PCRE is built with Unicode character property support, three addi-
3410 tional escape sequences that match characters with specific properties
3411 are available. When not in UTF-8 mode, these sequences are of course
3412 limited to testing characters whose codepoints are less than 256, but
3413 they do work in this mode. The extra escape sequences are:
3414
3415 \p{xx} a character with the xx property
3416 \P{xx} a character without the xx property
3417 \X an extended Unicode sequence
3418
3419 The property names represented by xx above are limited to the Unicode
3420 script names, the general category properties, and "Any", which matches
3421 any character (including newline). Other properties such as "InMusical-
3422 Symbols" are not currently supported by PCRE. Note that \P{Any} does
3423 not match any characters, so always causes a match failure.
3424
3425 Sets of Unicode characters are defined as belonging to certain scripts.
3426 A character from one of these sets can be matched using a script name.
3427 For example:
3428
3429 \p{Greek}
3430 \P{Han}
3431
3432 Those that are not part of an identified script are lumped together as
3433 "Common". The current list of scripts is:
3434
3435 Arabic, Armenian, Balinese, Bengali, Bopomofo, Braille, Buginese,
3436 Buhid, Canadian_Aboriginal, Cherokee, Common, Coptic, Cuneiform,
3437 Cypriot, Cyrillic, Deseret, Devanagari, Ethiopic, Georgian, Glagolitic,
3438 Gothic, Greek, Gujarati, Gurmukhi, Han, Hangul, Hanunoo, Hebrew, Hira-
3439 gana, Inherited, Kannada, Katakana, Kharoshthi, Khmer, Lao, Latin,
3440 Limbu, Linear_B, Malayalam, Mongolian, Myanmar, New_Tai_Lue, Nko,
3441 Ogham, Old_Italic, Old_Persian, Oriya, Osmanya, Phags_Pa, Phoenician,
3442 Runic, Shavian, Sinhala, Syloti_Nagri, Syriac, Tagalog, Tagbanwa,
3443 Tai_Le, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh, Ugaritic, Yi.
3444
3445 Each character has exactly one general category property, specified by
3446 a two-letter abbreviation. For compatibility with Perl, negation can be
3447 specified by including a circumflex between the opening brace and the
3448 property name. For example, \p{^Lu} is the same as \P{Lu}.
3449
3450 If only one letter is specified with \p or \P, it includes all the gen-
3451 eral category properties that start with that letter. In this case, in
3452 the absence of negation, the curly brackets in the escape sequence are
3453 optional; these two examples have the same effect:
3454
3455 \p{L}
3456 \pL
3457
3458 The following general category property codes are supported:
3459
3460 C Other
3461 Cc Control
3462 Cf Format
3463 Cn Unassigned
3464 Co Private use
3465 Cs Surrogate
3466
3467 L Letter
3468 Ll Lower case letter
3469 Lm Modifier letter
3470 Lo Other letter
3471 Lt Title case letter
3472 Lu Upper case letter
3473
3474 M Mark
3475 Mc Spacing mark
3476 Me Enclosing mark
3477 Mn Non-spacing mark
3478
3479 N Number
3480 Nd Decimal number
3481 Nl Letter number
3482 No Other number
3483
3484 P Punctuation
3485 Pc Connector punctuation
3486 Pd Dash punctuation
3487 Pe Close punctuation
3488 Pf Final punctuation
3489 Pi Initial punctuation
3490 Po Other punctuation
3491 Ps Open punctuation
3492
3493 S Symbol
3494 Sc Currency symbol
3495 Sk Modifier symbol
3496 Sm Mathematical symbol
3497 So Other symbol
3498
3499 Z Separator
3500 Zl Line separator
3501 Zp Paragraph separator
3502 Zs Space separator
3503
3504 The special property L& is also supported: it matches a character that
3505 has the Lu, Ll, or Lt property, in other words, a letter that is not
3506 classified as a modifier or "other".
3507
3508 The Cs (Surrogate) property applies only to characters in the range
3509 U+D800 to U+DFFF. Such characters are not valid in UTF-8 strings (see
3510 RFC 3629) and so cannot be tested by PCRE, unless UTF-8 validity check-
3511 ing has been turned off (see the discussion of PCRE_NO_UTF8_CHECK in
3512 the pcreapi page). Perl does not support the Cs property.
3513
3514 The long synonyms for property names that Perl supports (such as
3515 \p{Letter}) are not supported by PCRE, nor is it permitted to prefix
3516 any of these properties with "Is".
3517
3518 No character that is in the Unicode table has the Cn (unassigned) prop-
3519 erty. Instead, this property is assumed for any code point that is not
3520 in the Unicode table.
3521
3522 Specifying caseless matching does not affect these escape sequences.
3523 For example, \p{Lu} always matches only upper case letters.
3524
3525 The \X escape matches any number of Unicode characters that form an
3526 extended Unicode sequence. \X is equivalent to
3527
3528 (?>\PM\pM*)
3529
3530 That is, it matches a character without the "mark" property, followed
3531 by zero or more characters with the "mark" property, and treats the
3532 sequence as an atomic group (see below). Characters with the "mark"
3533 property are typically accents that affect the preceding character.
3534 None of them have codepoints less than 256, so in non-UTF-8 mode \X
3535 matches any one character.
3536
3537 Matching characters by Unicode property is not fast, because PCRE has
3538 to search a structure that contains data for over fifteen thousand
3539 characters. That is why the traditional escape sequences such as \d and
3540 \w do not use Unicode properties in PCRE.
3541
3542 Resetting the match start
3543
3544 The escape sequence \K, which is a Perl 5.10 feature, causes any previ-
3545 ously matched characters not to be included in the final matched
3546 sequence. For example, the pattern:
3547
3548 foo\Kbar
3549
3550 matches "foobar", but reports that it has matched "bar". This feature
3551 is similar to a lookbehind assertion (described below). However, in
3552 this case, the part of the subject before the real match does not have
3553 to be of fixed length, as lookbehind assertions do. The use of \K does
3554 not interfere with the setting of captured substrings. For example,
3555 when the pattern
3556
3557 (foo)\Kbar
3558
3559 matches "foobar", the first substring is still set to "foo".
3560
3561 Simple assertions
3562
3563 The final use of backslash is for certain simple assertions. An asser-
3564 tion specifies a condition that has to be met at a particular point in
3565 a match, without consuming any characters from the subject string. The
3566 use of subpatterns for more complicated assertions is described below.
3567 The backslashed assertions are:
3568
3569 \b matches at a word boundary
3570 \B matches when not at a word boundary
3571 \A matches at the start of the subject
3572 \Z matches at the end of the subject
3573 also matches before a newline at the end of the subject
3574 \z matches only at the end of the subject
3575 \G matches at the first matching position in the subject
3576
3577 These assertions may not appear in character classes (but note that \b
3578 has a different meaning, namely the backspace character, inside a char-
3579 acter class).
3580
3581 A word boundary is a position in the subject string where the current
3582 character and the previous character do not both match \w or \W (i.e.
3583 one matches \w and the other matches \W), or the start or end of the
3584 string if the first or last character matches \w, respectively.
3585
3586 The \A, \Z, and \z assertions differ from the traditional circumflex
3587 and dollar (described in the next section) in that they only ever match
3588 at the very start and end of the subject string, whatever options are
3589 set. Thus, they are independent of multiline mode. These three asser-
3590 tions are not affected by the PCRE_NOTBOL or PCRE_NOTEOL options, which
3591 affect only the behaviour of the circumflex and dollar metacharacters.
3592 However, if the startoffset argument of pcre_exec() is non-zero, indi-
3593 cating that matching is to start at a point other than the beginning of
3594 the subject, \A can never match. The difference between \Z and \z is
3595 that \Z matches before a newline at the end of the string as well as at
3596 the very end, whereas \z matches only at the end.
3597
3598 The \G assertion is true only when the current matching position is at
3599 the start point of the match, as specified by the startoffset argument
3600 of pcre_exec(). It differs from \A when the value of startoffset is
3601 non-zero. By calling pcre_exec() multiple times with appropriate argu-
3602 ments, you can mimic Perl's /g option, and it is in this kind of imple-
3603 mentation where \G can be useful.
3604
3605 Note, however, that PCRE's interpretation of \G, as the start of the
3606 current match, is subtly different from Perl's, which defines it as the
3607 end of the previous match. In Perl, these can be different when the
3608 previously matched string was empty. Because PCRE does just one match
3609 at a time, it cannot reproduce this behaviour.
3610
3611 If all the alternatives of a pattern begin with \G, the expression is
3612 anchored to the starting match position, and the "anchored" flag is set
3613 in the compiled regular expression.
3614
3615
3616 CIRCUMFLEX AND DOLLAR
3617
3618 Outside a character class, in the default matching mode, the circumflex
3619 character is an assertion that is true only if the current matching
3620 point is at the start of the subject string. If the startoffset argu-
3621 ment of pcre_exec() is non-zero, circumflex can never match if the
3622 PCRE_MULTILINE option is unset. Inside a character class, circumflex
3623 has an entirely different meaning (see below).
3624
3625 Circumflex need not be the first character of the pattern if a number
3626 of alternatives are involved, but it should be the first thing in each
3627 alternative in which it appears if the pattern is ever to match that
3628 branch. If all possible alternatives start with a circumflex, that is,
3629 if the pattern is constrained to match only at the start of the sub-
3630 ject, it is said to be an "anchored" pattern. (There are also other
3631 constructs that can cause a pattern to be anchored.)
3632
3633 A dollar character is an assertion that is true only if the current
3634 matching point is at the end of the subject string, or immediately
3635 before a newline at the end of the string (by default). Dollar need not
3636 be the last character of the pattern if a number of alternatives are
3637 involved, but it should be the last item in any branch in which it
3638 appears. Dollar has no special meaning in a character class.
3639
3640 The meaning of dollar can be changed so that it matches only at the
3641 very end of the string, by setting the PCRE_DOLLAR_ENDONLY option at
3642 compile time. This does not affect the \Z assertion.
3643
3644 The meanings of the circumflex and dollar characters are changed if the
3645 PCRE_MULTILINE option is set. When this is the case, a circumflex
3646 matches immediately after internal newlines as well as at the start of
3647 the subject string. It does not match after a newline that ends the
3648 string. A dollar matches before any newlines in the string, as well as
3649 at the very end, when PCRE_MULTILINE is set. When newline is specified
3650 as the two-character sequence CRLF, isolated CR and LF characters do
3651 not indicate newlines.
3652
3653 For example, the pattern /^abc$/ matches the subject string "def\nabc"
3654 (where \n represents a newline) in multiline mode, but not otherwise.
3655 Consequently, patterns that are anchored in single line mode because
3656 all branches start with ^ are not anchored in multiline mode, and a
3657 match for circumflex is possible when the startoffset argument of
3658 pcre_exec() is non-zero. The PCRE_DOLLAR_ENDONLY option is ignored if
3659 PCRE_MULTILINE is set.
3660
3661 Note that the sequences \A, \Z, and \z can be used to match the start
3662 and end of the subject in both modes, and if all branches of a pattern
3663 start with \A it is always anchored, whether or not PCRE_MULTILINE is
3664 set.
3665
3666
3667 FULL STOP (PERIOD, DOT)
3668
3669 Outside a character class, a dot in the pattern matches any one charac-
3670 ter in the subject string except (by default) a character that signi-
3671 fies the end of a line. In UTF-8 mode, the matched character may be
3672 more than one byte long.
3673
3674 When a line ending is defined as a single character, dot never matches
3675 that character; when the two-character sequence CRLF is used, dot does
3676 not match CR if it is immediately followed by LF, but otherwise it
3677 matches all characters (including isolated CRs and LFs). When any Uni-
3678 code line endings are being recognized, dot does not match CR or LF or
3679 any of the other line ending characters.
3680
3681 The behaviour of dot with regard to newlines can be changed. If the
3682 PCRE_DOTALL option is set, a dot matches any one character, without
3683 exception. If the two-character sequence CRLF is present in the subject
3684 string, it takes two dots to match it.
3685
3686 The handling of dot is entirely independent of the handling of circum-
3687 flex and dollar, the only relationship being that they both involve
3688 newlines. Dot has no special meaning in a character class.
3689
3690
3691 MATCHING A SINGLE BYTE
3692
3693 Outside a character class, the escape sequence \C matches any one byte,
3694 both in and out of UTF-8 mode. Unlike a dot, it always matches any
3695 line-ending characters. The feature is provided in Perl in order to
3696 match individual bytes in UTF-8 mode. Because it breaks up UTF-8 char-
3697 acters into individual bytes, what remains in the string may be a mal-
3698 formed UTF-8 string. For this reason, the \C escape sequence is best
3699 avoided.
3700
3701 PCRE does not allow \C to appear in lookbehind assertions (described
3702 below), because in UTF-8 mode this would make it impossible to calcu-
3703 late the length of the lookbehind.
3704
3705
3706 SQUARE BRACKETS AND CHARACTER CLASSES
3707
3708 An opening square bracket introduces a character class, terminated by a
3709 closing square bracket. A closing square bracket on its own is not spe-
3710 cial. If a closing square bracket is required as a member of the class,
3711 it should be the first data character in the class (after an initial
3712 circumflex, if present) or escaped with a backslash.
3713
3714 A character class matches a single character in the subject. In UTF-8
3715 mode, the character may occupy more than one byte. A matched character
3716 must be in the set of characters defined by the class, unless the first
3717 character in the class definition is a circumflex, in which case the
3718 subject character must not be in the set defined by the class. If a
3719 circumflex is actually required as a member of the class, ensure it is
3720 not the first character, or escape it with a backslash.
3721
3722 For example, the character class [aeiou] matches any lower case vowel,
3723 while [^aeiou] matches any character that is not a lower case vowel.
3724 Note that a circumflex is just a convenient notation for specifying the
3725 characters that are in the class by enumerating those that are not. A
3726 class that starts with a circumflex is not an assertion: it still con-
3727 sumes a character from the subject string, and therefore it fails if
3728 the current pointer is at the end of the string.
3729
3730 In UTF-8 mode, characters with values greater than 255 can be included
3731 in a class as a literal string of bytes, or by using the \x{ escaping
3732 mechanism.
3733
3734 When caseless matching is set, any letters in a class represent both
3735 their upper case and lower case versions, so for example, a caseless
3736 [aeiou] matches "A" as well as "a", and a caseless [^aeiou] does not
3737 match "A", whereas a caseful version would. In UTF-8 mode, PCRE always
3738 understands the concept of case for characters whose values are less
3739 than 128, so caseless matching is always possible. For characters with
3740 higher values, the concept of case is supported if PCRE is compiled
3741 with Unicode property support, but not otherwise. If you want to use
3742 caseless matching for characters 128 and above, you must ensure that
3743 PCRE is compiled with Unicode property support as well as with UTF-8
3744 support.
3745
3746 Characters that might indicate line breaks are never treated in any
3747 special way when matching character classes, whatever line-ending
3748 sequence is in use, and whatever setting of the PCRE_DOTALL and
3749 PCRE_MULTILINE options is used. A class such as [^a] always matches one
3750 of these characters.
3751
3752 The minus (hyphen) character can be used to specify a range of charac-
3753 ters in a character class. For example, [d-m] matches any letter
3754 between d and m, inclusive. If a minus character is required in a
3755 class, it must be escaped with a backslash or appear in a position
3756 where it cannot be interpreted as indicating a range, typically as the
3757 first or last character in the class.
3758
3759 It is not possible to have the literal character "]" as the end charac-
3760 ter of a range. A pattern such as [W-]46] is interpreted as a class of
3761 two characters ("W" and "-") followed by a literal string "46]", so it
3762 would match "W46]" or "-46]". However, if the "]" is escaped with a
3763 backslash it is interpreted as the end of range, so [W-\]46] is inter-
3764 preted as a class containing a range followed by two other characters.
3765 The octal or hexadecimal representation of "]" can also be used to end
3766 a range.
3767
3768 Ranges operate in the collating sequence of character values. They can
3769 also be used for characters specified numerically, for example
3770 [\000-\037]. In UTF-8 mode, ranges can include characters whose values
3771 are greater than 255, for example [\x{100}-\x{2ff}].
3772
3773 If a range that includes letters is used when caseless matching is set,
3774 it matches the letters in either case. For example, [W-c] is equivalent
3775 to [][\\^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if
3776 character tables for a French locale are in use, [\xc8-\xcb] matches
3777 accented E characters in both cases. In UTF-8 mode, PCRE supports the
3778 concept of case for characters with values greater than 128 only when
3779 it is compiled with Unicode property support.
3780
3781 The character types \d, \D, \p, \P, \s, \S, \w, and \W may also appear
3782 in a character class, and add the characters that they match to the
3783 class. For example, [\dABCDEF] matches any hexadecimal digit. A circum-
3784 flex can conveniently be used with the upper case character types to
3785 specify a more restricted set of characters than the matching lower
3786 case type. For example, the class [^\W_] matches any letter or digit,
3787 but not underscore.
3788
3789 The only metacharacters that are recognized in character classes are
3790 backslash, hyphen (only where it can be interpreted as specifying a
3791 range), circumflex (only at the start), opening square bracket (only
3792 when it can be interpreted as introducing a POSIX class name - see the
3793 next section), and the terminating closing square bracket. However,
3794 escaping other non-alphanumeric characters does no harm.
3795
3796
3797 POSIX CHARACTER CLASSES
3798
3799 Perl supports the POSIX notation for character classes. This uses names
3800 enclosed by [: and :] within the enclosing square brackets. PCRE also
3801 supports this notation. For example,
3802
3803 [01[:alpha:]%]
3804
3805 matches "0", "1", any alphabetic character, or "%". The supported class
3806 names are
3807
3808 alnum letters and digits
3809 alpha letters
3810 ascii character codes 0 - 127
3811 blank space or tab only
3812 cntrl control characters
3813 digit decimal digits (same as \d)
3814 graph printing characters, excluding space
3815 lower lower case letters
3816 print printing characters, including space
3817 punct printing characters, excluding letters and digits
3818 space white space (not quite the same as \s)
3819 upper upper case letters
3820 word "word" characters (same as \w)
3821 xdigit hexadecimal digits
3822
3823 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13),
3824 and space (32). Notice that this list includes the VT character (code
3825 11). This makes "space" different to \s, which does not include VT (for
3826 Perl compatibility).
3827
3828 The name "word" is a Perl extension, and "blank" is a GNU extension
3829 from Perl 5.8. Another Perl extension is negation, which is indicated
3830 by a ^ character after the colon. For example,
3831
3832 [12[:^digit:]]
3833
3834 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the
3835 POSIX syntax [.ch.] and [=ch=] where "ch" is a "collating element", but
3836 these are not supported, and an error is given if they are encountered.
3837
3838 In UTF-8 mode, characters with values greater than 128 do not match any
3839 of the POSIX character classes.
3840
3841
3842 VERTICAL BAR
3843
3844 Vertical bar characters are used to separate alternative patterns. For
3845 example, the pattern
3846
3847 gilbert|sullivan
3848
3849 matches either "gilbert" or "sullivan". Any number of alternatives may
3850 appear, and an empty alternative is permitted (matching the empty
3851 string). The matching process tries each alternative in turn, from left
3852 to right, and the first one that succeeds is used. If the alternatives
3853 are within a subpattern (defined below), "succeeds" means matching the
3854 rest of the main pattern as well as the alternative in the subpattern.
3855
3856
3857 INTERNAL OPTION SETTING
3858
3859 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
3860 PCRE_EXTENDED options (which are Perl-compatible) can be changed from
3861 within the pattern by a sequence of Perl option letters enclosed
3862 between "(?" and ")". The option letters are
3863
3864 i for PCRE_CASELESS
3865 m for PCRE_MULTILINE
3866 s for PCRE_DOTALL
3867 x for PCRE_EXTENDED
3868
3869 For example, (?im) sets caseless, multiline matching. It is also possi-
3870 ble to unset these options by preceding the letter with a hyphen, and a
3871 combined setting and unsetting such as (?im-sx), which sets PCRE_CASE-
3872 LESS and PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED,
3873 is also permitted. If a letter appears both before and after the
3874 hyphen, the option is unset.
3875
3876 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA
3877 can be changed in the same way as the Perl-compatible options by using
3878 the characters J, U and X respectively.
3879
3880 When one of these option changes occurs at top level (that is, not
3881 inside subpattern parentheses), the change applies to the remainder of
3882 the pattern that follows. If the change is placed right at the start of
3883 a pattern, PCRE extracts it into the global options (and it will there-
3884 fore show up in data extracted by the pcre_fullinfo() function).
3885
3886 An option change within a subpattern (see below for a description of
3887 subpatterns) affects only that part of the current pattern that follows
3888 it, so
3889
3890 (a(?i)b)c
3891
3892 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not
3893 used). By this means, options can be made to have different settings
3894 in different parts of the pattern. Any changes made in one alternative
3895 do carry on into subsequent branches within the same subpattern. For
3896 example,
3897
3898 (a(?i)b|c)
3899
3900 matches "ab", "aB", "c", and "C", even though when matching "C" the
3901 first branch is abandoned before the option setting. This is because
3902 the effects of option settings happen at compile time. There would be
3903 some very weird behaviour otherwise.
3904
3905 Note: There are other PCRE-specific options that can be set by the
3906 application when the compile or match functions are called. In some
3907 cases the pattern can contain special leading sequences such as (*CRLF)
3908 to override what the application has set or what has been defaulted.
3909 Details are given in the section entitled "Newline sequences" above.
3910 There is also the (*UTF8) leading sequence that can be used to set
3911 UTF-8 mode; this is equivalent to setting the PCRE_UTF8 option.
3912
3913
3914 SUBPATTERNS
3915
3916 Subpatterns are delimited by parentheses (round brackets), which can be
3917 nested. Turning part of a pattern into a subpattern does two things:
3918
3919 1. It localizes a set of alternatives. For example, the pattern
3920
3921 cat(aract|erpillar|)
3922
3923 matches one of the words "cat", "cataract", or "caterpillar". Without
3924 the parentheses, it would match "cataract", "erpillar" or an empty
3925 string.
3926
3927 2. It sets up the subpattern as a capturing subpattern. This means
3928 that, when the whole pattern matches, that portion of the subject
3929 string that matched the subpattern is passed back to the caller via the
3930 ovector argument of pcre_exec(). Opening parentheses are counted from
3931 left to right (starting from 1) to obtain numbers for the capturing
3932 subpatterns.
3933
3934 For example, if the string "the red king" is matched against the pat-
3935 tern
3936
3937 the ((red|white) (king|queen))
3938
3939 the captured substrings are "red king", "red", and "king", and are num-
3940 bered 1, 2, and 3, respectively.
3941
3942 The fact that plain parentheses fulfil two functions is not always
3943 helpful. There are often times when a grouping subpattern is required
3944 without a capturing requirement. If an opening parenthesis is followed
3945 by a question mark and a colon, the subpattern does not do any captur-
3946 ing, and is not counted when computing the number of any subsequent
3947 capturing subpatterns. For example, if the string "the white queen" is
3948 matched against the pattern
3949
3950 the ((?:red|white) (king|queen))
3951
3952 the captured substrings are "white queen" and "queen", and are numbered
3953 1 and 2. The maximum number of capturing subpatterns is 65535.
3954
3955 As a convenient shorthand, if any option settings are required at the
3956 start of a non-capturing subpattern, the option letters may appear
3957 between the "?" and the ":". Thus the two patterns
3958
3959 (?i:saturday|sunday)
3960 (?:(?i)saturday|sunday)
3961
3962 match exactly the same set of strings. Because alternative branches are
3963 tried from left to right, and options are not reset until the end of
3964 the subpattern is reached, an option setting in one branch does affect
3965 subsequent branches, so the above patterns match "SUNDAY" as well as
3966 "Saturday".
3967
3968
3969 DUPLICATE SUBPATTERN NUMBERS
3970
3971 Perl 5.10 introduced a feature whereby each alternative in a subpattern
3972 uses the same numbers for its capturing parentheses. Such a subpattern
3973 starts with (?| and is itself a non-capturing subpattern. For example,
3974 consider this pattern:
3975
3976 (?|(Sat)ur|(Sun))day
3977
3978 Because the two alternatives are inside a (?| group, both sets of cap-
3979 turing parentheses are numbered one. Thus, when the pattern matches,
3980 you can look at captured substring number one, whichever alternative
3981 matched. This construct is useful when you want to capture part, but
3982 not all, of one of a number of alternatives. Inside a (?| group, paren-
3983 theses are numbered as usual, but the number is reset at the start of
3984 each branch. The numbers of any capturing buffers that follow the sub-
3985 pattern start after the highest number used in any branch. The follow-
3986 ing example is taken from the Perl documentation. The numbers under-
3987 neath show in which buffer the captured content will be stored.
3988
3989 # before ---------------branch-reset----------- after
3990 / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
3991 # 1 2 2 3 2 3 4
3992
3993 A backreference or a recursive call to a numbered subpattern always
3994 refers to the first one in the pattern with the given number.
3995
3996 An alternative approach to using this "branch reset" feature is to use
3997 duplicate named subpatterns, as described in the next section.
3998
3999
4000 NAMED SUBPATTERNS
4001
4002 Identifying capturing parentheses by number is simple, but it can be
4003 very hard to keep track of the numbers in complicated regular expres-
4004 sions. Furthermore, if an expression is modified, the numbers may
4005 change. To help with this difficulty, PCRE supports the naming of sub-
4006 patterns. This feature was not added to Perl until release 5.10. Python
4007 had the feature earlier, and PCRE introduced it at release 4.0, using
4008 the Python syntax. PCRE now supports both the Perl and the Python syn-
4009 tax.
4010
4011 In PCRE, a subpattern can be named in one of three ways: (?<name>...)
4012 or (?'name'...) as in Perl, or (?P<name>...) as in Python. References
4013 to capturing parentheses from other parts of the pattern, such as back-
4014 references, recursion, and conditions, can be made by name as well as
4015 by number.
4016
4017 Names consist of up to 32 alphanumeric characters and underscores.
4018 Named capturing parentheses are still allocated numbers as well as
4019 names, exactly as if the names were not present. The PCRE API provides
4020 function calls for extracting the name-to-number translation table from
4021 a compiled pattern. There is also a convenience function for extracting
4022 a captured substring by name.
4023
4024 By default, a name must be unique within a pattern, but it is possible
4025 to relax this constraint by setting the PCRE_DUPNAMES option at compile
4026 time. This can be useful for patterns where only one instance of the
4027 named parentheses can match. Suppose you want to match the name of a
4028 weekday, either as a 3-letter abbreviation or as the full name, and in
4029 both cases you want to extract the abbreviation. This pattern (ignoring
4030 the line breaks) does the job:
4031
4032 (?<DN>Mon|Fri|Sun)(?:day)?|
4033 (?<DN>Tue)(?:sday)?|
4034 (?<DN>Wed)(?:nesday)?|
4035 (?<DN>Thu)(?:rsday)?|
4036 (?<DN>Sat)(?:urday)?
4037
4038 There are five capturing substrings, but only one is ever set after a
4039 match. (An alternative way of solving this problem is to use a "branch
4040 reset" subpattern, as described in the previous section.)
4041
4042 The convenience function for extracting the data by name returns the
4043 substring for the first (and in this example, the only) subpattern of
4044 that name that matched. This saves searching to find which numbered
4045 subpattern it was. If you make a reference to a non-unique named sub-
4046 pattern from elsewhere in the pattern, the one that corresponds to the
4047 lowest number is used. For further details of the interfaces for han-
4048 dling named subpatterns, see the pcreapi documentation.
4049
4050 Warning: You cannot use different names to distinguish between two sub-
4051 patterns with the same number (see the previous section) because PCRE
4052 uses only the numbers when matching.
4053
4054
4055 REPETITION
4056
4057 Repetition is specified by quantifiers, which can follow any of the
4058 following items:
4059
4060 a literal data character
4061 the dot metacharacter
4062 the \C escape sequence
4063 the \X escape sequence (in UTF-8 mode with Unicode properties)
4064 the \R escape sequence
4065 an escape such as \d that matches a single character
4066 a character class
4067 a back reference (see next section)
4068 a parenthesized subpattern (unless it is an assertion)
4069
4070 The general repetition quantifier specifies a minimum and maximum num-
4071 ber of permitted matches, by giving the two numbers in curly brackets
4072 (braces), separated by a comma. The numbers must be less than 65536,
4073 and the first must be less than or equal to the second. For example:
4074
4075 z{2,4}
4076
4077 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a
4078 special character. If the second number is omitted, but the comma is
4079 present, there is no upper limit; if the second number and the comma
4080 are both omitted, the quantifier specifies an exact number of required
4081 matches. Thus
4082
4083 [aeiou]{3,}
4084
4085 matches at least 3 successive vowels, but may match many more, while
4086
4087 \d{8}
4088
4089 matches exactly 8 digits. An opening curly bracket that appears in a
4090 position where a quantifier is not allowed, or one that does not match
4091 the syntax of a quantifier, is taken as a literal character. For exam-
4092 ple, {,6} is not a quantifier, but a literal string of four characters.
4093
4094 In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to
4095 individual bytes. Thus, for example, \x{100}{2} matches two UTF-8 char-
4096 acters, each of which is represented by a two-byte sequence. Similarly,
4097 when Unicode property support is available, \X{3} matches three Unicode
4098 extended sequences, each of which may be several bytes long (and they
4099 may be of different lengths).
4100
4101 The quantifier {0} is permitted, causing the expression to behave as if
4102 the previous item and the quantifier were not present. This may be use-
4103 ful for subpatterns that are referenced as subroutines from elsewhere
4104 in the pattern. Items other than subpatterns that have a {0} quantifier
4105 are omitted from the compiled pattern.
4106
4107 For convenience, the three most common quantifiers have single-charac-
4108 ter abbreviations:
4109
4110 * is equivalent to {0,}
4111 + is equivalent to {1,}
4112 ? is equivalent to {0,1}
4113
4114 It is possible to construct infinite loops by following a subpattern
4115 that can match no characters with a quantifier that has no upper limit,
4116 for example:
4117
4118 (a?)*
4119
4120 Earlier versions of Perl and PCRE used to give an error at compile time
4121 for such patterns. However, because there are cases where this can be
4122 useful, such patterns are now accepted, but if any repetition of the
4123 subpattern does in fact match no characters, the loop is forcibly bro-
4124 ken.
4125
4126 By default, the quantifiers are "greedy", that is, they match as much
4127 as possible (up to the maximum number of permitted times), without
4128 causing the rest of the pattern to fail. The classic example of where
4129 this gives problems is in trying to match comments in C programs. These
4130 appear between /* and */ and within the comment, individual * and /
4131 characters may appear. An attempt to match C comments by applying the
4132 pattern
4133
4134 /\*.*\*/
4135
4136 to the string
4137
4138 /* first comment */ not comment /* second comment */
4139
4140 fails, because it matches the entire string owing to the greediness of
4141 the .* item.
4142
4143 However, if a quantifier is followed by a question mark, it ceases to
4144 be greedy, and instead matches the minimum number of times possible, so
4145 the pattern
4146
4147 /\*.*?\*/
4148
4149 does the right thing with the C comments. The meaning of the various
4150 quantifiers is not otherwise changed, just the preferred number of
4151 matches. Do not confuse this use of question mark with its use as a
4152 quantifier in its own right. Because it has two uses, it can sometimes
4153 appear doubled, as in
4154
4155 \d??\d
4156
4157 which matches one digit by preference, but can match two if that is the
4158 only way the rest of the pattern matches.
4159
4160 If the PCRE_UNGREEDY option is set (an option that is not available in
4161 Perl), the quantifiers are not greedy by default, but individual ones
4162 can be made greedy by following them with a question mark. In other
4163 words, it inverts the default behaviour.
4164
4165 When a parenthesized subpattern is quantified with a minimum repeat
4166 count that is greater than 1 or with a limited maximum, more memory is
4167 required for the compiled pattern, in proportion to the size of the
4168 minimum or maximum.
4169
4170 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equiv-
4171 alent to Perl's /s) is set, thus allowing the dot to match newlines,
4172 the pattern is implicitly anchored, because whatever follows will be
4173 tried against every character position in the subject string, so there
4174 is no point in retrying the overall match at any position after the
4175 first. PCRE normally treats such a pattern as though it were preceded
4176 by \A.
4177
4178 In cases where it is known that the subject string contains no new-
4179 lines, it is worth setting PCRE_DOTALL in order to obtain this opti-
4180 mization, or alternatively using ^ to indicate anchoring explicitly.
4181
4182 However, there is one situation where the optimization cannot be used.
4183 When .* is inside capturing parentheses that are the subject of a
4184 backreference elsewhere in the pattern, a match at the start may fail
4185 where a later one succeeds. Consider, for example:
4186
4187 (.*)abc\1
4188
4189 If the subject is "xyz123abc123" the match point is the fourth charac-
4190 ter. For this reason, such a pattern is not implicitly anchored.
4191
4192 When a capturing subpattern is repeated, the value captured is the sub-
4193 string that matched the final iteration. For example, after
4194
4195 (tweedle[dume]{3}\s*)+
4196
4197 has matched "tweedledum tweedledee" the value of the captured substring
4198 is "tweedledee". However, if there are nested capturing subpatterns,
4199 the corresponding captured values may have been set in previous itera-
4200 tions. For example, after
4201
4202 /(a|(b))+/
4203
4204 matches "aba" the value of the second captured substring is "b".
4205
4206
4207 ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS
4208
4209 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
4210 repetition, failure of what follows normally causes the repeated item
4211 to be re-evaluated to see if a different number of repeats allows the
4212 rest of the pattern to match. Sometimes it is useful to prevent this,
4213 either to change the nature of the match, or to cause it fail earlier
4214 than it otherwise might, when the author of the pattern knows there is
4215 no point in carrying on.
4216
4217 Consider, for example, the pattern \d+foo when applied to the subject
4218 line
4219
4220 123456bar
4221
4222 After matching all 6 digits and then failing to match "foo", the normal
4223 action of the matcher is to try again with only 5 digits matching the
4224 \d+ item, and then with 4, and so on, before ultimately failing.
4225 "Atomic grouping" (a term taken from Jeffrey Friedl's book) provides
4226 the means for specifying that once a subpattern has matched, it is not
4227 to be re-evaluated in this way.
4228
4229 If we use atomic grouping for the previous example, the matcher gives
4230 up immediately on failing to match "foo" the first time. The notation
4231 is a kind of special parenthesis, starting with (?> as in this example:
4232
4233 (?>\d+)foo
4234
4235 This kind of parenthesis "locks up" the part of the pattern it con-
4236 tains once it has matched, and a failure further into the pattern is
4237 prevented from backtracking into it. Backtracking past it to previous
4238 items, however, works as normal.
4239
4240 An alternative description is that a subpattern of this type matches
4241 the string of characters that an identical standalone pattern would
4242 match, if anchored at the current point in the subject string.
4243
4244 Atomic grouping subpatterns are not capturing subpatterns. Simple cases
4245 such as the above example can be thought of as a maximizing repeat that
4246 must swallow everything it can. So, while both \d+ and \d+? are pre-
4247 pared to adjust the number of digits they match in order to make the
4248 rest of the pattern match, (?>\d+) can only match an entire sequence of
4249 digits.
4250
4251 Atomic groups in general can of course contain arbitrarily complicated
4252 subpatterns, and can be nested. However, when the subpattern for an
4253 atomic group is just a single repeated item, as in the example above, a
4254 simpler notation, called a "possessive quantifier" can be used. This
4255 consists of an additional + character following a quantifier. Using
4256 this notation, the previous example can be rewritten as
4257
4258 \d++foo
4259
4260 Note that a possessive quantifier can be used with an entire group, for
4261 example:
4262
4263 (abc|xyz){2,3}+
4264
4265 Possessive quantifiers are always greedy; the setting of the
4266 PCRE_UNGREEDY option is ignored. They are a convenient notation for the
4267 simpler forms of atomic group. However, there is no difference in the
4268 meaning of a possessive quantifier and the equivalent atomic group,
4269 though there may be a performance difference; possessive quantifiers
4270 should be slightly faster.
4271
4272 The possessive quantifier syntax is an extension to the Perl 5.8 syn-
4273 tax. Jeffrey Friedl originated the idea (and the name) in the first
4274 edition of his book. Mike McCloskey liked it, so implemented it when he
4275 built Sun's Java package, and PCRE copied it from there. It ultimately
4276 found its way into Perl at release 5.10.
4277
4278 PCRE has an optimization that automatically "possessifies" certain sim-
4279 ple pattern constructs. For example, the sequence A+B is treated as
4280 A++B because there is no point in backtracking into a sequence of A's
4281 when B must follow.
4282
4283 When a pattern contains an unlimited repeat inside a subpattern that
4284 can itself be repeated an unlimited number of times, the use of an
4285 atomic group is the only way to avoid some failing matches taking a
4286 very long time indeed. The pattern
4287
4288 (\D+|<\d+>)*[!?]
4289
4290 matches an unlimited number of substrings that either consist of non-
4291 digits, or digits enclosed in <>, followed by either ! or ?. When it
4292 matches, it runs quickly. However, if it is applied to
4293
4294 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
4295
4296 it takes a long time before reporting failure. This is because the
4297 string can be divided between the internal \D+ repeat and the external
4298 * repeat in a large number of ways, and all have to be tried. (The
4299 example uses [!?] rather than a single character at the end, because
4300 both PCRE and Perl have an optimization that allows for fast failure
4301 when a single character is used. They remember the last single charac-
4302 ter that is required for a match, and fail early if it is not present
4303 in the string.) If the pattern is changed so that it uses an atomic
4304 group, like this:
4305
4306 ((?>\D+)|<\d+>)*[!?]
4307
4308 sequences of non-digits cannot be broken, and failure happens quickly.
4309
4310
4311 BACK REFERENCES
4312
4313 Outside a character class, a backslash followed by a digit greater than
4314 0 (and possibly further digits) is a back reference to a capturing sub-
4315 pattern earlier (that is, to its left) in the pattern, provided there
4316 have been that many previous capturing left parentheses.
4317
4318 However, if the decimal number following the backslash is less than 10,
4319 it is always taken as a back reference, and causes an error only if
4320 there are not that many capturing left parentheses in the entire pat-
4321 tern. In other words, the parentheses that are referenced need not be
4322 to the left of the reference for numbers less than 10. A "forward back
4323 reference" of this type can make sense when a repetition is involved
4324 and the subpattern to the right has participated in an earlier itera-
4325 tion.
4326
4327 It is not possible to have a numerical "forward back reference" to a
4328 subpattern whose number is 10 or more using this syntax because a
4329 sequence such as \50 is interpreted as a character defined in octal.
4330 See the subsection entitled "Non-printing characters" above for further
4331 details of the handling of digits following a backslash. There is no
4332 such problem when named parentheses are used. A back reference to any
4333 subpattern is possible using named parentheses (see below).
4334
4335 Another way of avoiding the ambiguity inherent in the use of digits
4336 following a backslash is to use the \g escape sequence, which is a fea-
4337 ture introduced in Perl 5.10. This escape must be followed by an
4338 unsigned number or a negative number, optionally enclosed in braces.
4339 These examples are all identical:
4340
4341 (ring), \1
4342 (ring), \g1
4343 (ring), \g{1}
4344
4345 An unsigned number specifies an absolute reference without the ambigu-
4346 ity that is present in the older syntax. It is also useful when literal
4347 digits follow the reference. A negative number is a relative reference.
4348 Consider this example:
4349
4350 (abc(def)ghi)\g{-1}
4351
4352 The sequence \g{-1} is a reference to the most recently started captur-
4353 ing subpattern before \g, that is, is it equivalent to \2. Similarly,
4354 \g{-2} would be equivalent to \1. The use of relative references can be
4355 helpful in long patterns, and also in patterns that are created by
4356 joining together fragments that contain references within themselves.
4357
4358 A back reference matches whatever actually matched the capturing sub-
4359 pattern in the current subject string, rather than anything matching
4360 the subpattern itself (see "Subpatterns as subroutines" below for a way
4361 of doing that). So the pattern
4362
4363 (sens|respons)e and \1ibility
4364
4365 matches "sense and sensibility" and "response and responsibility", but
4366 not "sense and responsibility". If caseful matching is in force at the
4367 time of the back reference, the case of letters is relevant. For exam-
4368 ple,
4369
4370 ((?i)rah)\s+\1
4371
4372 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the
4373 original capturing subpattern is matched caselessly.
4374
4375 There are several different ways of writing back references to named
4376 subpatterns. The .NET syntax \k{name} and the Perl syntax \k<name> or
4377 \k'name' are supported, as is the Python syntax (?P=name). Perl 5.10's
4378 unified back reference syntax, in which \g can be used for both numeric
4379 and named references, is also supported. We could rewrite the above
4380 example in any of the following ways:
4381
4382 (?<p1>(?i)rah)\s+\k<p1>
4383 (?'p1'(?i)rah)\s+\k{p1}
4384 (?P<p1>(?i)rah)\s+(?P=p1)
4385 (?<p1>(?i)rah)\s+\g{p1}
4386
4387 A subpattern that is referenced by name may appear in the pattern
4388 before or after the reference.
4389
4390 There may be more than one back reference to the same subpattern. If a
4391 subpattern has not actually been used in a particular match, any back
4392 references to it always fail. For example, the pattern
4393
4394 (a|(bc))\2
4395
4396 always fails if it starts to match "a" rather than "bc". Because there
4397 may be many capturing parentheses in a pattern, all digits following
4398 the backslash are taken as part of a potential back reference number.
4399 If the pattern continues with a digit character, some delimiter must be
4400 used to terminate the back reference. If the PCRE_EXTENDED option is
4401 set, this can be whitespace. Otherwise an empty comment (see "Com-
4402 ments" below) can be used.
4403
4404 A back reference that occurs inside the parentheses to which it refers
4405 fails when the subpattern is first used, so, for example, (a\1) never
4406 matches. However, such references can be useful inside repeated sub-
4407 patterns. For example, the pattern
4408
4409 (a|b\1)+
4410
4411 matches any number of "a"s and also "aba", "ababbaa" etc. At each iter-
4412 ation of the subpattern, the back reference matches the character
4413 string corresponding to the previous iteration. In order for this to
4414 work, the pattern must be such that the first iteration does not need
4415 to match the back reference. This can be done using alternation, as in
4416 the example above, or by a quantifier with a minimum of zero.
4417
4418
4419 ASSERTIONS
4420
4421 An assertion is a test on the characters following or preceding the
4422 current matching point that does not actually consume any characters.
4423 The simple assertions coded as \b, \B, \A, \G, \Z, \z, ^ and $ are
4424 described above.
4425
4426 More complicated assertions are coded as subpatterns. There are two
4427 kinds: those that look ahead of the current position in the subject
4428 string, and those that look behind it. An assertion subpattern is
4429 matched in the normal way, except that it does not cause the current
4430 matching position to be changed.
4431
4432 Assertion subpatterns are not capturing subpatterns, and may not be
4433 repeated, because it makes no sense to assert the same thing several
4434 times. If any kind of assertion contains capturing subpatterns within
4435 it, these are counted for the purposes of numbering the capturing sub-
4436 patterns in the whole pattern. However, substring capturing is carried
4437 out only for positive assertions, because it does not make sense for
4438 negative assertions.
4439
4440 Lookahead assertions
4441
4442 Lookahead assertions start with (?= for positive assertions and (?! for
4443 negative assertions. For example,
4444
4445 \w+(?=;)
4446
4447 matches a word followed by a semicolon, but does not include the semi-
4448 colon in the match, and
4449
4450 foo(?!bar)
4451
4452 matches any occurrence of "foo" that is not followed by "bar". Note
4453 that the apparently similar pattern
4454
4455 (?!foo)bar
4456
4457 does not find an occurrence of "bar" that is preceded by something
4458 other than "foo"; it finds any occurrence of "bar" whatsoever, because
4459 the assertion (?!foo) is always true when the next three characters are
4460 "bar". A lookbehind assertion is needed to achieve the other effect.
4461
4462 If you want to force a matching failure at some point in a pattern, the
4463 most convenient way to do it is with (?!) because an empty string
4464 always matches, so an assertion that requires there not to be an empty
4465 string must always fail.
4466
4467 Lookbehind assertions
4468
4469 Lookbehind assertions start with (?<= for positive assertions and (?<!
4470 for negative assertions. For example,
4471
4472 (?<!foo)bar
4473
4474 does find an occurrence of "bar" that is not preceded by "foo". The
4475 contents of a lookbehind assertion are restricted such that all the
4476 strings it matches must have a fixed length. However, if there are sev-
4477 eral top-level alternatives, they do not all have to have the same
4478 fixed length. Thus
4479
4480 (?<=bullock|donkey)
4481
4482 is permitted, but
4483
4484 (?<!dogs?|cats?)
4485
4486 causes an error at compile time. Branches that match different length
4487 strings are permitted only at the top level of a lookbehind assertion.
4488 This is an extension compared with Perl (5.8 and 5.10), which requires
4489 all branches to match the same length of string. An assertion such as
4490
4491 (?<=ab(c|de))
4492
4493 is not permitted, because its single top-level branch can match two
4494 different lengths, but it is acceptable to PCRE if rewritten to use two
4495 top-level branches:
4496
4497 (?<=abc|abde)
4498
4499 In some cases, the Perl 5.10 escape sequence \K (see above) can be used
4500 instead of a lookbehind assertion to get round the fixed-length
4501 restriction.
4502
4503 The implementation of lookbehind assertions is, for each alternative,
4504 to temporarily move the current position back by the fixed length and
4505 then try to match. If there are insufficient characters before the cur-
4506 rent position, the assertion fails.
4507
4508 PCRE does not allow the \C escape (which matches a single byte in UTF-8
4509 mode) to appear in lookbehind assertions, because it makes it impossi-
4510 ble to calculate the length of the lookbehind. The \X and \R escapes,
4511 which can match different numbers of bytes, are also not permitted.
4512
4513 "Subroutine" calls (see below) such as (?2) or (?&X) are permitted in
4514 lookbehinds, as long as the subpattern matches a fixed-length string.
4515 Recursion, however, is not supported.
4516
4517 Possessive quantifiers can be used in conjunction with lookbehind
4518 assertions to specify efficient matching at the end of the subject
4519 string. Consider a simple pattern such as
4520
4521 abcd$
4522
4523 when applied to a long string that does not match. Because matching
4524 proceeds from left to right, PCRE will look for each "a" in the subject
4525 and then see if what follows matches the rest of the pattern. If the
4526 pattern is specified as
4527
4528 ^.*abcd$
4529
4530 the initial .* matches the entire string at first, but when this fails
4531 (because there is no following "a"), it backtracks to match all but the
4532 last character, then all but the last two characters, and so on. Once
4533 again the search for "a" covers the entire string, from right to left,
4534 so we are no better off. However, if the pattern is written as
4535
4536 ^.*+(?<=abcd)
4537
4538 there can be no backtracking for the .*+ item; it can match only the
4539 entire string. The subsequent lookbehind assertion does a single test
4540 on the last four characters. If it fails, the match fails immediately.
4541 For long strings, this approach makes a significant difference to the
4542 processing time.
4543
4544 Using multiple assertions
4545
4546 Several assertions (of any sort) may occur in succession. For example,
4547
4548 (?<=\d{3})(?<!999)foo
4549
4550 matches "foo" preceded by three digits that are not "999". Notice that
4551 each of the assertions is applied independently at the same point in
4552 the subject string. First there is a check that the previous three
4553 characters are all digits, and then there is a check that the same
4554 three characters are not "999". This pattern does not match "foo" pre-
4555 ceded by six characters, the first of which are digits and the last
4556 three of which are not "999". For example, it doesn't match "123abc-
4557 foo". A pattern to do that is
4558
4559 (?<=\d{3}...)(?<!999)foo
4560
4561 This time the first assertion looks at the preceding six characters,
4562 checking that the first three are digits, and then the second assertion
4563 checks that the preceding three characters are not "999".
4564
4565 Assertions can be nested in any combination. For example,
4566
4567 (?<=(?<!foo)bar)baz
4568
4569 matches an occurrence of "baz" that is preceded by "bar" which in turn
4570 is not preceded by "foo", while
4571
4572 (?<=\d{3}(?!999)...)foo
4573
4574 is another pattern that matches "foo" preceded by three digits and any
4575 three characters that are not "999".
4576
4577
4578 CONDITIONAL SUBPATTERNS
4579
4580 It is possible to cause the matching process to obey a subpattern con-
4581 ditionally or to choose between two alternative subpatterns, depending
4582 on the result of an assertion, or whether a previous capturing subpat-
4583 tern matched or not. The two possible forms of conditional subpattern
4584 are
4585
4586 (?(condition)yes-pattern)
4587 (?(condition)yes-pattern|no-pattern)
4588
4589 If the condition is satisfied, the yes-pattern is used; otherwise the
4590 no-pattern (if present) is used. If there are more than two alterna-
4591 tives in the subpattern, a compile-time error occurs.
4592
4593 There are four kinds of condition: references to subpatterns, refer-
4594 ences to recursion, a pseudo-condition called DEFINE, and assertions.
4595
4596 Checking for a used subpattern by number
4597
4598 If the text between the parentheses consists of a sequence of digits,
4599 the condition is true if the capturing subpattern of that number has
4600 previously matched. An alternative notation is to precede the digits
4601 with a plus or minus sign. In this case, the subpattern number is rela-
4602 tive rather than absolute. The most recently opened parentheses can be
4603 referenced by (?(-1), the next most recent by (?(-2), and so on. In
4604 looping constructs it can also make sense to refer to subsequent groups
4605 with constructs such as (?(+2).
4606
4607 Consider the following pattern, which contains non-significant white
4608 space to make it more readable (assume the PCRE_EXTENDED option) and to
4609 divide it into three parts for ease of discussion:
4610
4611 ( \( )? [^()]+ (?(1) \) )
4612
4613 The first part matches an optional opening parenthesis, and if that
4614 character is present, sets it as the first captured substring. The sec-
4615 ond part matches one or more characters that are not parentheses. The
4616 third part is a conditional subpattern that tests whether the first set
4617 of parentheses matched or not. If they did, that is, if subject started
4618 with an opening parenthesis, the condition is true, and so the yes-pat-
4619 tern is executed and a closing parenthesis is required. Otherwise,
4620 since no-pattern is not present, the subpattern matches nothing. In
4621 other words, this pattern matches a sequence of non-parentheses,
4622 optionally enclosed in parentheses.
4623
4624 If you were embedding this pattern in a larger one, you could use a
4625 relative reference:
4626
4627 ...other stuff... ( \( )? [^()]+ (?(-1) \) ) ...
4628
4629 This makes the fragment independent of the parentheses in the larger
4630 pattern.
4631
4632 Checking for a used subpattern by name
4633
4634 Perl uses the syntax (?(<name>)...) or (?('name')...) to test for a
4635 used subpattern by name. For compatibility with earlier versions of
4636 PCRE, which had this facility before Perl, the syntax (?(name)...) is
4637 also recognized. However, there is a possible ambiguity with this syn-
4638 tax, because subpattern names may consist entirely of digits. PCRE
4639 looks first for a named subpattern; if it cannot find one and the name
4640 consists entirely of digits, PCRE looks for a subpattern of that num-
4641 ber, which must be greater than zero. Using subpattern names that con-
4642 sist entirely of digits is not recommended.
4643
4644 Rewriting the above example to use a named subpattern gives this:
4645
4646 (?<OPEN> \( )? [^()]+ (?(<OPEN>) \) )
4647
4648
4649 Checking for pattern recursion
4650
4651 If the condition is the string (R), and there is no subpattern with the
4652 name R, the condition is true if a recursive call to the whole pattern
4653 or any subpattern has been made. If digits or a name preceded by amper-
4654 sand follow the letter R, for example:
4655
4656 (?(R3)...) or (?(R&name)...)
4657
4658 the condition is true if the most recent recursion is into the subpat-
4659 tern whose number or name is given. This condition does not check the
4660 entire recursion stack.
4661
4662 At "top level", all these recursion test conditions are false. Recur-
4663 sive patterns are described below.
4664
4665 Defining subpatterns for use by reference only
4666
4667 If the condition is the string (DEFINE), and there is no subpattern
4668 with the name DEFINE, the condition is always false. In this case,
4669 there may be only one alternative in the subpattern. It is always
4670 skipped if control reaches this point in the pattern; the idea of
4671 DEFINE is that it can be used to define "subroutines" that can be ref-
4672 erenced from elsewhere. (The use of "subroutines" is described below.)
4673 For example, a pattern to match an IPv4 address could be written like
4674 this (ignore whitespace and line breaks):
4675
4676 (?(DEFINE) (?<byte> 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
4677 \b (?&byte) (\.(?&byte)){3} \b
4678
4679 The first part of the pattern is a DEFINE group inside which a another
4680 group named "byte" is defined. This matches an individual component of
4681 an IPv4 address (a number less than 256). When matching takes place,
4682 this part of the pattern is skipped because DEFINE acts like a false
4683 condition.
4684
4685 The rest of the pattern uses references to the named group to match the
4686 four dot-separated components of an IPv4 address, insisting on a word
4687 boundary at each end.
4688
4689 Assertion conditions
4690
4691 If the condition is not in any of the above formats, it must be an
4692 assertion. This may be a positive or negative lookahead or lookbehind
4693 assertion. Consider this pattern, again containing non-significant
4694 white space, and with the two alternatives on the second line:
4695
4696 (?(?=[^a-z]*[a-z])
4697 \d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} )
4698
4699 The condition is a positive lookahead assertion that matches an
4700 optional sequence of non-letters followed by a letter. In other words,
4701 it tests for the presence of at least one letter in the subject. If a
4702 letter is found, the subject is matched against the first alternative;
4703 otherwise it is matched against the second. This pattern matches
4704 strings in one of the two forms dd-aaa-dd or dd-dd-dd, where aaa are
4705 letters and dd are digits.
4706
4707
4708 COMMENTS
4709
4710 The sequence (?# marks the start of a comment that continues up to the
4711 next closing parenthesis. Nested parentheses are not permitted. The
4712 characters that make up a comment play no part in the pattern matching
4713 at all.
4714
4715 If the PCRE_EXTENDED option is set, an unescaped # character outside a
4716 character class introduces a comment that continues to immediately
4717 after the next newline in the pattern.
4718
4719
4720 RECURSIVE PATTERNS
4721
4722 Consider the problem of matching a string in parentheses, allowing for
4723 unlimited nested parentheses. Without the use of recursion, the best
4724 that can be done is to use a pattern that matches up to some fixed
4725 depth of nesting. It is not possible to handle an arbitrary nesting
4726 depth.
4727
4728 For some time, Perl has provided a facility that allows regular expres-
4729 sions to recurse (amongst other things). It does this by interpolating
4730 Perl code in the expression at run time, and the code can refer to the
4731 expression itself. A Perl pattern using code interpolation to solve the
4732 parentheses problem can be created like this:
4733
4734 $re = qr{\( (?: (?>[^()]+) | (?p{$re}) )* \)}x;
4735
4736 The (?p{...}) item interpolates Perl code at run time, and in this case
4737 refers recursively to the pattern in which it appears.
4738
4739 Obviously, PCRE cannot support the interpolation of Perl code. Instead,
4740 it supports special syntax for recursion of the entire pattern, and
4741 also for individual subpattern recursion. After its introduction in
4742 PCRE and Python, this kind of recursion was subsequently introduced
4743 into Perl at release 5.10.
4744
4745 A special item that consists of (? followed by a number greater than
4746 zero and a closing parenthesis is a recursive call of the subpattern of
4747 the given number, provided that it occurs inside that subpattern. (If
4748 not, it is a "subroutine" call, which is described in the next sec-
4749 tion.) The special item (?R) or (?0) is a recursive call of the entire
4750 regular expression.
4751
4752 This PCRE pattern solves the nested parentheses problem (assume the
4753 PCRE_EXTENDED option is set so that white space is ignored):
4754
4755 \( ( (?>[^()]+) | (?R) )* \)
4756
4757 First it matches an opening parenthesis. Then it matches any number of
4758 substrings which can either be a sequence of non-parentheses, or a
4759 recursive match of the pattern itself (that is, a correctly parenthe-
4760 sized substring). Finally there is a closing parenthesis.
4761
4762 If this were part of a larger pattern, you would not want to recurse
4763 the entire pattern, so instead you could use this:
4764
4765 ( \( ( (?>[^()]+) | (?1) )* \) )
4766
4767 We have put the pattern into parentheses, and caused the recursion to
4768 refer to them instead of the whole pattern.
4769
4770 In a larger pattern, keeping track of parenthesis numbers can be
4771 tricky. This is made easier by the use of relative references. (A Perl
4772 5.10 feature.) Instead of (?1) in the pattern above you can write
4773 (?-2) to refer to the second most recently opened parentheses preceding
4774 the recursion. In other words, a negative number counts capturing
4775 parentheses leftwards from the point at which it is encountered.
4776
4777 It is also possible to refer to subsequently opened parentheses, by
4778 writing references such as (?+2). However, these cannot be recursive
4779 because the reference is not inside the parentheses that are refer-
4780 enced. They are always "subroutine" calls, as described in the next
4781 section.
4782
4783 An alternative approach is to use named parentheses instead. The Perl
4784 syntax for this is (?&name); PCRE's earlier syntax (?P>name) is also
4785 supported. We could rewrite the above example as follows:
4786
4787 (?<pn> \( ( (?>[^()]+) | (?&pn) )* \) )
4788
4789 If there is more than one subpattern with the same name, the earliest
4790 one is used.
4791
4792 This particular example pattern that we have been looking at contains
4793 nested unlimited repeats, and so the use of atomic grouping for match-
4794 ing strings of non-parentheses is important when applying the pattern
4795 to strings that do not match. For example, when this pattern is applied
4796 to
4797
4798 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
4799
4800 it yields "no match" quickly. However, if atomic grouping is not used,
4801 the match runs for a very long time indeed because there are so many
4802 different ways the + and * repeats can carve up the subject, and all
4803 have to be tested before failure can be reported.
4804
4805 At the end of a match, the values set for any capturing subpatterns are
4806 those from the outermost level of the recursion at which the subpattern
4807 value is set. If you want to obtain intermediate values, a callout
4808 function can be used (see below and the pcrecallout documentation). If
4809 the pattern above is matched against
4810
4811 (ab(cd)ef)
4812
4813 the value for the capturing parentheses is "ef", which is the last
4814 value taken on at the top level. If additional parentheses are added,
4815 giving
4816
4817 \( ( ( (?>[^()]+) | (?R) )* ) \)
4818 ^ ^
4819 ^ ^
4820
4821 the string they capture is "ab(cd)ef", the contents of the top level
4822 parentheses. If there are more than 15 capturing parentheses in a pat-
4823 tern, PCRE has to obtain extra memory to store data during a recursion,
4824 which it does by using pcre_malloc, freeing it via pcre_free after-
4825 wards. If no memory can be obtained, the match fails with the
4826 PCRE_ERROR_NOMEMORY error.
4827
4828 Do not confuse the (?R) item with the condition (R), which tests for
4829 recursion. Consider this pattern, which matches text in angle brack-
4830 ets, allowing for arbitrary nesting. Only digits are allowed in nested
4831 brackets (that is, when recursing), whereas any characters are permit-
4832 ted at the outer level.
4833
4834 < (?: (?(R) \d++ | [^<>]*+) | (?R)) * >
4835
4836 In this pattern, (?(R) is the start of a conditional subpattern, with
4837 two different alternatives for the recursive and non-recursive cases.
4838 The (?R) item is the actual recursive call.
4839
4840 Recursion difference from Perl
4841
4842 In PCRE (like Python, but unlike Perl), a recursive subpattern call is
4843 always treated as an atomic group. That is, once it has matched some of
4844 the subject string, it is never re-entered, even if it contains untried
4845 alternatives and there is a subsequent matching failure. This can be
4846 illustrated by the following pattern, which purports to match a palin-
4847 dromic string that contains an odd number of characters (for example,
4848 "a", "aba", "abcba", "abcdcba"):
4849
4850 ^(.|(.)(?1)\2)$
4851
4852 The idea is that it either matches a single character, or two identical
4853 characters surrounding a sub-palindrome. In Perl, this pattern works;
4854 in PCRE it does not if the pattern is longer than three characters.
4855 Consider the subject string "abcba":
4856
4857 At the top level, the first character is matched, but as it is not at
4858 the end of the string, the first alternative fails; the second alterna-
4859 tive is taken and the recursion kicks in. The recursive call to subpat-
4860 tern 1 successfully matches the next character ("b"). (Note that the
4861 beginning and end of line tests are not part of the recursion).
4862
4863 Back at the top level, the next character ("c") is compared with what
4864 subpattern 2 matched, which was "a". This fails. Because the recursion
4865 is treated as an atomic group, there are now no backtracking points,
4866 and so the entire match fails. (Perl is able, at this point, to re-
4867 enter the recursion and try the second alternative.) However, if the
4868 pattern is written with the alternatives in the other order, things are
4869 different:
4870
4871 ^((.)(?1)\2|.)$
4872
4873 This time, the recursing alternative is tried first, and continues to
4874 recurse until it runs out of characters, at which point the recursion
4875 fails. But this time we do have another alternative to try at the
4876 higher level. That is the big difference: in the previous case the
4877 remaining alternative is at a deeper recursion level, which PCRE cannot
4878 use.
4879
4880 To change the pattern so that matches all palindromic strings, not just
4881 those with an odd number of characters, it is tempting to change the
4882 pattern to this:
4883
4884 ^((.)(?1)\2|.?)$
4885
4886 Again, this works in Perl, but not in PCRE, and for the same reason.
4887 When a deeper recursion has matched a single character, it cannot be
4888 entered again in order to match an empty string. The solution is to
4889 separate the two cases, and write out the odd and even cases as alter-
4890 natives at the higher level:
4891
4892 ^(?:((.)(?1)\2|)|((.)(?3)\4|.))
4893
4894 If you want to match typical palindromic phrases, the pattern has to
4895 ignore all non-word characters, which can be done like this:
4896
4897 ^\W*+(?:((.)\W*+(?1)\W*+\2|)|((.)\W*+(?3)\W*+4|\W*+.\W*+))\W*+$
4898
4899 If run with the PCRE_CASELESS option, this pattern matches phrases such
4900 as "A man, a plan, a canal: Panama!" and it works well in both PCRE and
4901 Perl. Note the use of the possessive quantifier *+ to avoid backtrack-
4902 ing into sequences of non-word characters. Without this, PCRE takes a
4903 great deal longer (ten times or more) to match typical phrases, and
4904 Perl takes so long that you think it has gone into a loop.
4905
4906
4907 SUBPATTERNS AS SUBROUTINES
4908
4909 If the syntax for a recursive subpattern reference (either by number or
4910 by name) is used outside the parentheses to which it refers, it oper-
4911 ates like a subroutine in a programming language. The "called" subpat-
4912 tern may be defined before or after the reference. A numbered reference
4913 can be absolute or relative, as in these examples:
4914
4915 (...(absolute)...)...(?2)...
4916 (...(relative)...)...(?-1)...
4917 (...(?+1)...(relative)...
4918
4919 An earlier example pointed out that the pattern
4920
4921 (sens|respons)e and \1ibility
4922
4923 matches "sense and sensibility" and "response and responsibility", but
4924 not "sense and responsibility". If instead the pattern
4925
4926 (sens|respons)e and (?1)ibility
4927
4928 is used, it does match "sense and responsibility" as well as the other
4929 two strings. Another example is given in the discussion of DEFINE
4930 above.
4931
4932 Like recursive subpatterns, a "subroutine" call is always treated as an
4933 atomic group. That is, once it has matched some of the subject string,
4934 it is never re-entered, even if it contains untried alternatives and
4935 there is a subsequent matching failure.
4936
4937 When a subpattern is used as a subroutine, processing options such as
4938 case-independence are fixed when the subpattern is defined. They cannot
4939 be changed for different calls. For example, consider this pattern:
4940
4941 (abc)(?i:(?-1))
4942
4943 It matches "abcabc". It does not match "abcABC" because the change of
4944 processing option does not affect the called subpattern.
4945
4946
4947 ONIGURUMA SUBROUTINE SYNTAX
4948
4949 For compatibility with Oniguruma, the non-Perl syntax \g followed by a
4950 name or a number enclosed either in angle brackets or single quotes, is
4951 an alternative syntax for referencing a subpattern as a subroutine,
4952 possibly recursively. Here are two of the examples used above, rewrit-
4953 ten using this syntax:
4954
4955 (?<pn> \( ( (?>[^()]+) | \g<pn> )* \) )
4956 (sens|respons)e and \g'1'ibility
4957
4958 PCRE supports an extension to Oniguruma: if a number is preceded by a
4959 plus or a minus sign it is taken as a relative reference. For example:
4960
4961 (abc)(?i:\g<-1>)
4962
4963 Note that \g{...} (Perl syntax) and \g<...> (Oniguruma syntax) are not
4964 synonymous. The former is a back reference; the latter is a subroutine
4965 call.
4966
4967
4968 CALLOUTS
4969
4970 Perl has a feature whereby using the sequence (?{...}) causes arbitrary
4971 Perl code to be obeyed in the middle of matching a regular expression.
4972 This makes it possible, amongst other things, to extract different sub-
4973 strings that match the same pair of parentheses when there is a repeti-
4974 tion.
4975
4976 PCRE provides a similar feature, but of course it cannot obey arbitrary
4977 Perl code. The feature is called "callout". The caller of PCRE provides
4978 an external function by putting its entry point in the global variable
4979 pcre_callout. By default, this variable contains NULL, which disables
4980 all calling out.
4981
4982 Within a regular expression, (?C) indicates the points at which the
4983 external function is to be called. If you want to identify different
4984 callout points, you can put a number less than 256 after the letter C.
4985 The default value is zero. For example, this pattern has two callout
4986 points:
4987
4988 (?C1)abc(?C2)def
4989
4990 If the PCRE_AUTO_CALLOUT flag is passed to pcre_compile(), callouts are
4991 automatically installed before each item in the pattern. They are all
4992 numbered 255.
4993
4994 During matching, when PCRE reaches a callout point (and pcre_callout is
4995 set), the external function is called. It is provided with the number
4996 of the callout, the position in the pattern, and, optionally, one item
4997 of data originally supplied by the caller of pcre_exec(). The callout
4998 function may cause matching to proceed, to backtrack, or to fail alto-
4999 gether. A complete description of the interface to the callout function
5000 is given in the pcrecallout documentation.
5001
5002
5003 BACKTRACKING CONTROL
5004
5005 Perl 5.10 introduced a number of "Special Backtracking Control Verbs",
5006 which are described in the Perl documentation as "experimental and sub-
5007 ject to change or removal in a future version of Perl". It goes on to
5008 say: "Their usage in production code should be noted to avoid problems
5009 during upgrades." The same remarks apply to the PCRE features described
5010 in this section.
5011
5012 Since these verbs are specifically related to backtracking, most of
5013 them can be used only when the pattern is to be matched using
5014 pcre_exec(), which uses a backtracking algorithm. With the exception of
5015 (*FAIL), which behaves like a failing negative assertion, they cause an
5016 error if encountered by pcre_dfa_exec().
5017
5018 If any of these verbs are used in an assertion subpattern, their effect
5019 is confined to that subpattern; it does not extend to the surrounding
5020 pattern. Note that assertion subpatterns are processed as anchored at
5021 the point where they are tested.
5022
5023 The new verbs make use of what was previously invalid syntax: an open-
5024 ing parenthesis followed by an asterisk. In Perl, they are generally of
5025 the form (*VERB:ARG) but PCRE does not support the use of arguments, so
5026 its general form is just (*VERB). Any number of these verbs may occur
5027 in a pattern. There are two kinds:
5028
5029 Verbs that act immediately
5030
5031 The following verbs act as soon as they are encountered:
5032
5033 (*ACCEPT)
5034
5035 This verb causes the match to end successfully, skipping the remainder
5036 of the pattern. When inside a recursion, only the innermost pattern is
5037 ended immediately. If the (*ACCEPT) is inside capturing parentheses,
5038 the data so far is captured. (This feature was added to PCRE at release
5039 8.00.) For example:
5040
5041 A((?:A|B(*ACCEPT)|C)D)
5042
5043 This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is cap-
5044 tured by the outer parentheses.
5045
5046 (*FAIL) or (*F)
5047
5048 This verb causes the match to fail, forcing backtracking to occur. It
5049 is equivalent to (?!) but easier to read. The Perl documentation notes
5050 that it is probably useful only when combined with (?{}) or (??{}).
5051 Those are, of course, Perl features that are not present in PCRE. The
5052 nearest equivalent is the callout feature, as for example in this pat-
5053 tern:
5054
5055 a+(?C)(*FAIL)
5056
5057 A match with the string "aaaa" always fails, but the callout is taken
5058 before each backtrack happens (in this example, 10 times).
5059
5060 Verbs that act after backtracking
5061
5062 The following verbs do nothing when they are encountered. Matching con-
5063 tinues with what follows, but if there is no subsequent match, a fail-
5064 ure is forced. The verbs differ in exactly what kind of failure
5065 occurs.
5066
5067 (*COMMIT)
5068
5069 This verb causes the whole match to fail outright if the rest of the
5070 pattern does not match. Even if the pattern is unanchored, no further
5071 attempts to find a match by advancing the start point take place. Once
5072 (*COMMIT) has been passed, pcre_exec() is committed to finding a match
5073 at the current starting point, or not at all. For example:
5074
5075 a+(*COMMIT)b
5076
5077 This matches "xxaab" but not "aacaab". It can be thought of as a kind
5078 of dynamic anchor, or "I've started, so I must finish."
5079
5080 (*PRUNE)
5081
5082 This verb causes the match to fail at the current position if the rest
5083 of the pattern does not match. If the pattern is unanchored, the normal
5084 "bumpalong" advance to the next starting character then happens. Back-
5085 tracking can occur as usual to the left of (*PRUNE), or when matching
5086 to the right of (*PRUNE), but if there is no match to the right, back-
5087 tracking cannot cross (*PRUNE). In simple cases, the use of (*PRUNE)
5088 is just an alternative to an atomic group or possessive quantifier, but
5089 there are some uses of (*PRUNE) that cannot be expressed in any other
5090 way.
5091
5092 (*SKIP)
5093
5094 This verb is like (*PRUNE), except that if the pattern is unanchored,
5095 the "bumpalong" advance is not to the next character, but to the posi-
5096 tion in the subject where (*SKIP) was encountered. (*SKIP) signifies
5097 that whatever text was matched leading up to it cannot be part of a
5098 successful match. Consider:
5099
5100 a+(*SKIP)b
5101
5102 If the subject is "aaaac...", after the first match attempt fails
5103 (starting at the first character in the string), the starting point
5104 skips on to start the next attempt at "c". Note that a possessive quan-
5105 tifer does not have the same effect in this example; although it would
5106 suppress backtracking during the first match attempt, the second
5107 attempt would start at the second character instead of skipping on to
5108 "c".
5109
5110 (*THEN)
5111
5112 This verb causes a skip to the next alternation if the rest of the pat-
5113 tern does not match. That is, it cancels pending backtracking, but only
5114 within the current alternation. Its name comes from the observation
5115 that it can be used for a pattern-based if-then-else block:
5116
5117 ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
5118
5119 If the COND1 pattern matches, FOO is tried (and possibly further items
5120 after the end of the group if FOO succeeds); on failure the matcher
5121 skips to the second alternative and tries COND2, without backtracking
5122 into COND1. If (*THEN) is used outside of any alternation, it acts
5123 exactly like (*PRUNE).
5124
5125
5126 SEE ALSO
5127
5128 pcreapi(3), pcrecallout(3), pcrematching(3), pcre(3).
5129
5130
5131 AUTHOR
5132
5133 Philip Hazel
5134 University Computing Service
5135 Cambridge CB2 3QH, England.
5136
5137
5138 REVISION
5139
5140 Last updated: 22 September 2009
5141 Copyright (c) 1997-2009 University of Cambridge.
5142 ------------------------------------------------------------------------------
5143
5144
5145 PCRESYNTAX(3) PCRESYNTAX(3)
5146
5147
5148 NAME
5149 PCRE - Perl-compatible regular expressions
5150
5151
5152 PCRE REGULAR EXPRESSION SYNTAX SUMMARY
5153
5154 The full syntax and semantics of the regular expressions that are sup-
5155 ported by PCRE are described in the pcrepattern documentation. This
5156 document contains just a quick-reference summary of the syntax.
5157
5158
5159 QUOTING
5160
5161 \x where x is non-alphanumeric is a literal x
5162 \Q...\E treat enclosed characters as literal
5163
5164
5165 CHARACTERS
5166
5167 \a alarm, that is, the BEL character (hex 07)
5168 \cx "control-x", where x is any character
5169 \e escape (hex 1B)
5170 \f formfeed (hex 0C)
5171 \n newline (hex 0A)
5172 \r carriage return (hex 0D)
5173 \t tab (hex 09)
5174 \ddd character with octal code ddd, or backreference
5175 \xhh character with hex code hh
5176 \x{hhh..} character with hex code hhh..
5177
5178
5179 CHARACTER TYPES
5180
5181 . any character except newline;
5182 in dotall mode, any character whatsoever
5183 \C one byte, even in UTF-8 mode (best avoided)
5184 \d a decimal digit
5185 \D a character that is not a decimal digit
5186 \h a horizontal whitespace character
5187 \H a character that is not a horizontal whitespace character
5188 \p{xx} a character with the xx property
5189 \P{xx} a character without the xx property
5190 \R a newline sequence
5191 \s a whitespace character
5192 \S a character that is not a whitespace character
5193 \v a vertical whitespace character
5194 \V a character that is not a vertical whitespace character
5195 \w a "word" character
5196 \W a "non-word" character
5197 \X an extended Unicode sequence
5198
5199 In PCRE, \d, \D, \s, \S, \w, and \W recognize only ASCII characters.
5200
5201
5202 GENERAL CATEGORY PROPERTY CODES FOR \p and \P
5203
5204 C Other
5205 Cc Control
5206 Cf Format
5207 Cn Unassigned
5208 Co Private use
5209 Cs Surrogate
5210
5211 L Letter
5212 Ll Lower case letter
5213 Lm Modifier letter
5214 Lo Other letter
5215 Lt Title case letter
5216 Lu Upper case letter
5217 L& Ll, Lu, or Lt
5218
5219 M Mark
5220 Mc Spacing mark
5221 Me Enclosing mark
5222 Mn Non-spacing mark
5223
5224 N Number
5225 Nd Decimal number
5226 Nl Letter number
5227 No Other number
5228
5229 P Punctuation
5230 Pc Connector punctuation
5231 Pd Dash punctuation
5232 Pe Close punctuation
5233 Pf Final punctuation
5234 Pi Initial punctuation
5235 Po Other punctuation
5236 Ps Open punctuation
5237
5238 S Symbol
5239 Sc Currency symbol
5240 Sk Modifier symbol
5241 Sm Mathematical symbol
5242 So Other symbol
5243
5244 Z Separator
5245 Zl Line separator
5246 Zp Paragraph separator
5247 Zs Space separator
5248
5249
5250 SCRIPT NAMES FOR \p AND \P
5251
5252 Arabic, Armenian, Balinese, Bengali, Bopomofo, Braille, Buginese,
5253 Buhid, Canadian_Aboriginal, Carian, Cham, Cherokee, Common, Coptic, Cu-
5254 neiform, Cypriot, Cyrillic, Deseret, Devanagari, Ethiopic, Georgian,
5255 Glagolitic, Gothic, Greek, Gujarati, Gurmukhi, Han, Hangul, Hanunoo,
5256 Hebrew, Hiragana, Inherited, Kannada, Katakana, Kayah_Li, Kharoshthi,
5257 Khmer, Lao, Latin, Lepcha, Limbu, Linear_B, Lycian, Lydian, Malayalam,
5258 Mongolian, Myanmar, New_Tai_Lue, Nko, Ogham, Old_Italic, Old_Persian,
5259 Ol_Chiki, Oriya, Osmanya, Phags_Pa, Phoenician, Rejang, Runic, Saurash-
5260 tra, Shavian, Sinhala, Sudanese, Syloti_Nagri, Syriac, Tagalog, Tag-
5261 banwa, Tai_Le, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh,
5262 Ugaritic, Vai, Yi.
5263
5264
5265 CHARACTER CLASSES
5266
5267 [...] positive character class
5268 [^...] negative character class
5269 [x-y] range (can be used for hex characters)
5270 [[:xxx:]] positive POSIX named set
5271 [[:^xxx:]] negative POSIX named set
5272
5273 alnum alphanumeric
5274 alpha alphabetic
5275 ascii 0-127
5276 blank space or tab
5277 cntrl control character
5278 digit decimal digit
5279 graph printing, excluding space
5280 lower lower case letter
5281 print printing, including space
5282 punct printing, excluding alphanumeric
5283 space whitespace
5284 upper upper case letter
5285 word same as \w
5286 xdigit hexadecimal digit
5287
5288 In PCRE, POSIX character set names recognize only ASCII characters. You
5289 can use \Q...\E inside a character class.
5290
5291
5292 QUANTIFIERS
5293
5294 ? 0 or 1, greedy
5295 ?+ 0 or 1, possessive
5296 ?? 0 or 1, lazy
5297 * 0 or more, greedy
5298 *+ 0 or more, possessive
5299 *? 0 or more, lazy
5300 + 1 or more, greedy
5301 ++ 1 or more, possessive
5302 +? 1 or more, lazy
5303 {n} exactly n
5304 {n,m} at least n, no more than m, greedy
5305 {n,m}+ at least n, no more than m, possessive
5306 {n,m}? at least n, no more than m, lazy
5307 {n,} n or more, greedy
5308 {n,}+ n or more, possessive
5309 {n,}? n or more, lazy
5310
5311
5312 ANCHORS AND SIMPLE ASSERTIONS
5313
5314 \b word boundary (only ASCII letters recognized)
5315 \B not a word boundary
5316 ^ start of subject
5317 also after internal newline in multiline mode
5318 \A start of subject
5319 $ end of subject
5320 also before newline at end of subject
5321 also before internal newline in multiline mode
5322 \Z end of subject
5323 also before newline at end of subject
5324 \z end of subject
5325 \G first matching position in subject
5326
5327
5328 MATCH POINT RESET
5329
5330 \K reset start of match
5331
5332
5333 ALTERNATION
5334
5335 expr|expr|expr...
5336
5337
5338 CAPTURING
5339
5340 (...) capturing group
5341 (?<name>...) named capturing group (Perl)
5342 (?'name'...) named capturing group (Perl)
5343 (?P<name>...) named capturing group (Python)
5344 (?:...) non-capturing group
5345 (?|...) non-capturing group; reset group numbers for
5346 capturing groups in each alternative
5347
5348
5349 ATOMIC GROUPS
5350
5351 (?>...) atomic, non-capturing group
5352
5353
5354 COMMENT
5355
5356 (?#....) comment (not nestable)
5357
5358
5359 OPTION SETTING
5360
5361 (?i) caseless
5362 (?J) allow duplicate names
5363 (?m) multiline
5364 (?s) single line (dotall)
5365 (?U) default ungreedy (lazy)
5366 (?x) extended (ignore white space)
5367 (?-...) unset option(s)
5368
5369 The following is recognized only at the start of a pattern or after one
5370 of the newline-setting options with similar syntax:
5371
5372 (*UTF8) set UTF-8 mode
5373
5374
5375 LOOKAHEAD AND LOOKBEHIND ASSERTIONS
5376
5377 (?=...) positive look ahead
5378 (?!...) negative look ahead
5379 (?<=...) positive look behind
5380 (?<!...) negative look behind
5381
5382 Each top-level branch of a look behind must be of a fixed length.
5383
5384
5385 BACKREFERENCES
5386
5387 \n reference by number (can be ambiguous)
5388 \gn reference by number
5389 \g{n} reference by number
5390 \g{-n} relative reference by number
5391 \k<name> reference by name (Perl)
5392 \k'name' reference by name (Perl)
5393 \g{name} reference by name (Perl)
5394 \k{name} reference by name (.NET)
5395 (?P=name) reference by name (Python)
5396
5397
5398 SUBROUTINE REFERENCES (POSSIBLY RECURSIVE)
5399
5400 (?R) recurse whole pattern
5401 (?n) call subpattern by absolute number
5402 (?+n) call subpattern by relative number
5403 (?-n) call subpattern by relative number
5404 (?&name) call subpattern by name (Perl)
5405 (?P>name) call subpattern by name (Python)
5406 \g<name> call subpattern by name (Oniguruma)
5407 \g'name' call subpattern by name (Oniguruma)
5408 \g<n> call subpattern by absolute number (Oniguruma)
5409 \g'n' call subpattern by absolute number (Oniguruma)
5410 \g<+n> call subpattern by relative number (PCRE extension)
5411 \g'+n' call subpattern by relative number (PCRE extension)
5412 \g<-n> call subpattern by relative number (PCRE extension)
5413 \g'-n' call subpattern by relative number (PCRE extension)
5414
5415
5416 CONDITIONAL PATTERNS
5417
5418 (?(condition)yes-pattern)
5419 (?(condition)yes-pattern|no-pattern)
5420
5421 (?(n)... absolute reference condition
5422 (?(+n)... relative reference condition
5423 (?(-n)... relative reference condition
5424 (?(<name>)... named reference condition (Perl)
5425 (?('name')... named reference condition (Perl)
5426 (?(name)... named reference condition (PCRE)
5427 (?(R)... overall recursion condition
5428 (?(Rn)... specific group recursion condition
5429 (?(R&name)... specific recursion condition
5430 (?(DEFINE)... define subpattern for reference
5431 (?(assert)... assertion condition
5432
5433
5434 BACKTRACKING CONTROL
5435
5436 The following act immediately they are reached:
5437
5438 (*ACCEPT) force successful match
5439 (*FAIL) force backtrack; synonym (*F)
5440
5441 The following act only when a subsequent match failure causes a back-
5442 track to reach them. They all force a match failure, but they differ in
5443 what happens afterwards. Those that advance the start-of-match point do
5444 so only if the pattern is not anchored.
5445
5446 (*COMMIT) overall failure, no advance of starting point
5447 (*PRUNE) advance to next starting character
5448 (*SKIP) advance start to current matching position
5449 (*THEN) local failure, backtrack to next alternation
5450
5451
5452 NEWLINE CONVENTIONS
5453
5454 These are recognized only at the very start of the pattern or after a
5455 (*BSR_...) or (*UTF8) option.
5456
5457 (*CR) carriage return only
5458 (*LF) linefeed only
5459 (*CRLF) carriage return followed by linefeed
5460 (*ANYCRLF) all three of the above
5461 (*ANY) any Unicode newline sequence
5462
5463
5464 WHAT \R MATCHES
5465
5466 These are recognized only at the very start of the pattern or after a
5467 (*...) option that sets the newline convention or UTF-8 mode.
5468
5469 (*BSR_ANYCRLF) CR, LF, or CRLF
5470 (*BSR_UNICODE) any Unicode newline sequence
5471
5472
5473 CALLOUTS
5474
5475 (?C) callout
5476 (?Cn) callout with data n
5477
5478
5479 SEE ALSO
5480
5481 pcrepattern(3), pcreapi(3), pcrecallout(3), pcrematching(3), pcre(3).
5482
5483
5484 AUTHOR
5485
5486 Philip Hazel
5487 University Computing Service
5488 Cambridge CB2 3QH, England.
5489
5490
5491 REVISION
5492
5493 Last updated: 11 April 2009
5494 Copyright (c) 1997-2009 University of Cambridge.
5495 ------------------------------------------------------------------------------
5496
5497
5498 PCREPARTIAL(3) PCREPARTIAL(3)
5499
5500
5501 NAME
5502 PCRE - Perl-compatible regular expressions
5503
5504
5505 PARTIAL MATCHING IN PCRE
5506
5507 In normal use of PCRE, if the subject string that is passed to
5508 pcre_exec() or pcre_dfa_exec() matches as far as it goes, but is too
5509 short to match the entire pattern, PCRE_ERROR_NOMATCH is returned.
5510 There are circumstances where it might be helpful to distinguish this
5511 case from other cases in which there is no match.
5512
5513 Consider, for example, an application where a human is required to type
5514 in data for a field with specific formatting requirements. An example
5515 might be a date in the form ddmmmyy, defined by this pattern:
5516
5517 ^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$
5518
5519 If the application sees the user's keystrokes one by one, and can check
5520 that what has been typed so far is potentially valid, it is able to
5521 raise an error as soon as a mistake is made, by beeping and not
5522 reflecting the character that has been typed, for example. This immedi-
5523 ate feedback is likely to be a better user interface than a check that
5524 is delayed until the entire string has been entered. Partial matching
5525 can also sometimes be useful when the subject string is very long and
5526 is not all available at once.
5527
5528 PCRE supports partial matching by means of the PCRE_PARTIAL_SOFT and
5529 PCRE_PARTIAL_HARD options, which can be set when calling pcre_exec() or
5530 pcre_dfa_exec(). For backwards compatibility, PCRE_PARTIAL is a synonym
5531 for PCRE_PARTIAL_SOFT. The essential difference between the two options
5532 is whether or not a partial match is preferred to an alternative com-
5533 plete match, though the details differ between the two matching func-
5534 tions. If both options are set, PCRE_PARTIAL_HARD takes precedence.
5535
5536 Setting a partial matching option disables one of PCRE's optimizations.
5537 PCRE remembers the last literal byte in a pattern, and abandons match-
5538 ing immediately if such a byte is not present in the subject string.
5539 This optimization cannot be used for a subject string that might match
5540 only partially.
5541
5542
5543 PARTIAL MATCHING USING pcre_exec()
5544
5545 A partial match occurs during a call to pcre_exec() whenever the end of
5546 the subject string is reached successfully, but matching cannot con-
5547 tinue because more characters are needed. However, at least one charac-
5548 ter must have been matched. (In other words, a partial match can never
5549 be an empty string.)
5550
5551 If PCRE_PARTIAL_SOFT is set, the partial match is remembered, but
5552 matching continues as normal, and other alternatives in the pattern are
5553 tried. If no complete match can be found, pcre_exec() returns
5554 PCRE_ERROR_PARTIAL instead of PCRE_ERROR_NOMATCH. If there are at least
5555 two slots in the offsets vector, the first of them is set to the offset
5556 of the earliest character that was inspected when the partial match was
5557 found. For convenience, the second offset points to the end of the
5558 string so that a substring can easily be extracted.
5559
5560 For the majority of patterns, the first offset identifies the start of
5561 the partially matched string. However, for patterns that contain look-
5562 behind assertions, or \K, or begin with \b or \B, earlier characters
5563 have been inspected while carrying out the match. For example:
5564
5565 /(?<=abc)123/
5566
5567 This pattern matches "123", but only if it is preceded by "abc". If the
5568 subject string is "xyzabc12", the offsets after a partial match are for
5569 the substring "abc12", because all these characters are needed if
5570 another match is tried with extra characters added.
5571
5572 If there is more than one partial match, the first one that was found
5573 provides the data that is returned. Consider this pattern:
5574
5575 /123\w+X|dogY/
5576
5577 If this is matched against the subject string "abc123dog", both alter-
5578 natives fail to match, but the end of the subject is reached during
5579 matching, so PCRE_ERROR_PARTIAL is returned instead of
5580 PCRE_ERROR_NOMATCH. The offsets are set to 3 and 9, identifying
5581 "123dog" as the first partial match that was found. (In this example,
5582 there are two partial matches, because "dog" on its own partially
5583 matches the second alternative.)
5584
5585 If PCRE_PARTIAL_HARD is set for pcre_exec(), it returns PCRE_ERROR_PAR-
5586 TIAL as soon as a partial match is found, without continuing to search
5587 for possible complete matches. The difference between the two options
5588 can be illustrated by a pattern such as:
5589
5590 /dog(sbody)?/
5591
5592 This matches either "dog" or "dogsbody", greedily (that is, it prefers
5593 the longer string if possible). If it is matched against the string
5594 "dog" with PCRE_PARTIAL_SOFT, it yields a complete match for "dog".
5595 However, if PCRE_PARTIAL_HARD is set, the result is PCRE_ERROR_PARTIAL.
5596 On the other hand, if the pattern is made ungreedy the result is dif-
5597 ferent:
5598
5599 /dog(sbody)??/
5600
5601 In this case the result is always a complete match because pcre_exec()
5602 finds that first, and it never continues after finding a match. It
5603 might be easier to follow this explanation by thinking of the two pat-
5604 terns like this:
5605
5606 /dog(sbody)?/ is the same as /dogsbody|dog/
5607 /dog(sbody)??/ is the same as /dog|dogsbody/
5608
5609 The second pattern will never match "dogsbody" when pcre_exec() is
5610 used, because it will always find the shorter match first.
5611
5612
5613 PARTIAL MATCHING USING pcre_dfa_exec()
5614
5615 The pcre_dfa_exec() function moves along the subject string character
5616 by character, without backtracking, searching for all possible matches
5617 simultaneously. If the end of the subject is reached before the end of
5618 the pattern, there is the possibility of a partial match, again pro-
5619 vided that at least one character has matched.
5620
5621 When PCRE_PARTIAL_SOFT is set, PCRE_ERROR_PARTIAL is returned only if
5622 there have been no complete matches. Otherwise, the complete matches
5623 are returned. However, if PCRE_PARTIAL_HARD is set, a partial match
5624 takes precedence over any complete matches. The portion of the string
5625 that was inspected when the longest partial match was found is set as
5626 the first matching string, provided there are at least two slots in the
5627 offsets vector.
5628
5629 Because pcre_dfa_exec() always searches for all possible matches, and
5630 there is no difference between greedy and ungreedy repetition, its be-
5631 haviour is different from pcre_exec when PCRE_PARTIAL_HARD is set. Con-
5632 sider the string "dog" matched against the ungreedy pattern shown
5633 above:
5634
5635 /dog(sbody)??/
5636
5637 Whereas pcre_exec() stops as soon as it finds the complete match for
5638 "dog", pcre_dfa_exec() also finds the partial match for "dogsbody", and
5639 so returns that when PCRE_PARTIAL_HARD is set.
5640
5641
5642 PARTIAL MATCHING AND WORD BOUNDARIES
5643
5644 If a pattern ends with one of sequences \w or \W, which test for word
5645 boundaries, partial matching with PCRE_PARTIAL_SOFT can give counter-
5646 intuitive results. Consider this pattern:
5647
5648 /\bcat\b/
5649
5650 This matches "cat", provided there is a word boundary at either end. If
5651 the subject string is "the cat", the comparison of the final "t" with a
5652 following character cannot take place, so a partial match is found.
5653 However, pcre_exec() carries on with normal matching, which matches \b
5654 at the end of the subject when the last character is a letter, thus
5655 finding a complete match. The result, therefore, is not PCRE_ERROR_PAR-
5656 TIAL. The same thing happens with pcre_dfa_exec(), because it also
5657 finds the complete match.
5658
5659 Using PCRE_PARTIAL_HARD in this case does yield PCRE_ERROR_PARTIAL,
5660 because then the partial match takes precedence.
5661
5662
5663 FORMERLY RESTRICTED PATTERNS
5664
5665 For releases of PCRE prior to 8.00, because of the way certain internal
5666 optimizations were implemented in the pcre_exec() function, the
5667 PCRE_PARTIAL option (predecessor of PCRE_PARTIAL_SOFT) could not be
5668 used with all patterns. From release 8.00 onwards, the restrictions no
5669 longer apply, and partial matching with pcre_exec() can be requested
5670 for any pattern.
5671
5672 Items that were formerly restricted were repeated single characters and
5673 repeated metasequences. If PCRE_PARTIAL was set for a pattern that did
5674 not conform to the restrictions, pcre_exec() returned the error code
5675 PCRE_ERROR_BADPARTIAL (-13). This error code is no longer in use. The
5676 PCRE_INFO_OKPARTIAL call to pcre_fullinfo() to find out if a compiled
5677 pattern can be used for partial matching now always returns 1.
5678
5679
5680 EXAMPLE OF PARTIAL MATCHING USING PCRETEST
5681
5682 If the escape sequence \P is present in a pcretest data line, the
5683 PCRE_PARTIAL_SOFT option is used for the match. Here is a run of
5684 pcretest that uses the date example quoted above:
5685
5686 re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/
5687 data> 25jun04\P
5688 0: 25jun04
5689 1: jun
5690 data> 25dec3\P
5691 Partial match: 23dec3
5692 data> 3ju\P
5693 Partial match: 3ju
5694 data> 3juj\P
5695 No match
5696 data> j\P
5697 No match
5698
5699 The first data string is matched completely, so pcretest shows the
5700 matched substrings. The remaining four strings do not match the com-
5701 plete pattern, but the first two are partial matches. Similar output is
5702 obtained when pcre_dfa_exec() is used.
5703
5704 If the escape sequence \P is present more than once in a pcretest data
5705 line, the PCRE_PARTIAL_HARD option is set for the match.
5706
5707
5708 MULTI-SEGMENT MATCHING WITH pcre_dfa_exec()
5709
5710 When a partial match has been found using pcre_dfa_exec(), it is possi-
5711 ble to continue the match by providing additional subject data and
5712 calling pcre_dfa_exec() again with the same compiled regular expres-
5713 sion, this time setting the PCRE_DFA_RESTART option. You must pass the
5714 same working space as before, because this is where details of the pre-
5715 vious partial match are stored. Here is an example using pcretest,
5716 using the \R escape sequence to set the PCRE_DFA_RESTART option (\D
5717 specifies the use of pcre_dfa_exec()):
5718
5719 re> /^\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d$/
5720 data> 23ja\P\D
5721 Partial match: 23ja
5722 data> n05\R\D
5723 0: n05
5724
5725 The first call has "23ja" as the subject, and requests partial match-
5726 ing; the second call has "n05" as the subject for the continued
5727 (restarted) match. Notice that when the match is complete, only the
5728 last part is shown; PCRE does not retain the previously partially-
5729 matched string. It is up to the calling program to do that if it needs
5730 to.
5731
5732 You can set the PCRE_PARTIAL_SOFT or PCRE_PARTIAL_HARD options with
5733 PCRE_DFA_RESTART to continue partial matching over multiple segments.
5734 This facility can be used to pass very long subject strings to
5735 pcre_dfa_exec().
5736
5737
5738 MULTI-SEGMENT MATCHING WITH pcre_exec()
5739
5740 From release 8.00, pcre_exec() can also be used to do multi-segment
5741 matching. Unlike pcre_dfa_exec(), it is not possible to restart the
5742 previous match with a new segment of data. Instead, new data must be
5743 added to the previous subject string, and the entire match re-run,
5744 starting from the point where the partial match occurred. Earlier data
5745 can be discarded. Consider an unanchored pattern that matches dates:
5746
5747 re> /\d?\d(jan|feb|mar|apr|may|jun|jul|aug|sep|oct|nov|dec)\d\d/
5748 data> The date is 23ja\P
5749 Partial match: 23ja
5750
5751 The this stage, an application could discard the text preceding "23ja",
5752 add on text from the next segment, and call pcre_exec() again. Unlike
5753 pcre_dfa_exec(), the entire matching string must always be available,
5754 and the complete matching process occurs for each call, so more memory
5755 and more processing time is needed.
5756
5757 Note: If the pattern contains lookbehind assertions, or \K, or starts
5758 with \b or \B, the string that is returned for a partial match will
5759 include characters that precede the partially matched string itself,
5760 because these must be retained when adding on more characters for a
5761 subsequent matching attempt.
5762
5763
5764 ISSUES WITH MULTI-SEGMENT MATCHING
5765
5766 Certain types of pattern may give problems with multi-segment matching,
5767 whichever matching function is used.
5768
5769 1. If the pattern contains tests for the beginning or end of a line,
5770 you need to pass the PCRE_NOTBOL or PCRE_NOTEOL options, as appropri-
5771 ate, when the subject string for any call does not contain the begin-
5772 ning or end of a line.
5773
5774 2. Lookbehind assertions at the start of a pattern are catered for in
5775 the offsets that are returned for a partial match. However, in theory,
5776 a lookbehind assertion later in the pattern could require even earlier
5777 characters to be inspected, and it might not have been reached when a
5778 partial match occurs. This is probably an extremely unlikely case; you
5779 could guard against it to a certain extent by always including extra
5780 characters at the start.
5781
5782 3. Matching a subject string that is split into multiple segments may
5783 not always produce exactly the same result as matching over one single
5784 long string, especially when PCRE_PARTIAL_SOFT is used. The section
5785 "Partial Matching and Word Boundaries" above describes an issue that
5786 arises if the pattern ends with \b or \B. Another kind of difference
5787 may occur when there are multiple matching possibilities, because a
5788 partial match result is given only when there are no completed matches.
5789 This means that as soon as the shortest match has been found, continua-
5790 tion to a new subject segment is no longer possible. Consider again
5791 this pcretest example:
5792
5793 re> /dog(sbody)?/
5794 data> dogsb\P
5795 0: dog
5796 data> do\P\D
5797 Partial match: do
5798 data> gsb\R\P\D
5799 0: g
5800 data> dogsbody\D
5801 0: dogsbody
5802 1: dog
5803
5804 The first data line passes the string "dogsb" to pcre_exec(), setting
5805 the PCRE_PARTIAL_SOFT option. Although the string is a partial match
5806 for "dogsbody", the result is not PCRE_ERROR_PARTIAL, because the
5807 shorter string "dog" is a complete match. Similarly, when the subject
5808 is presented to pcre_dfa_exec() in several parts ("do" and "gsb" being
5809 the first two) the match stops when "dog" has been found, and it is not
5810 possible to continue. On the other hand, if "dogsbody" is presented as
5811 a single string, pcre_dfa_exec() finds both matches.
5812
5813 Because of these problems, it is probably best to use PCRE_PARTIAL_HARD
5814 when matching multi-segment data. The example above then behaves dif-
5815 ferently:
5816
5817 re> /dog(sbody)?/
5818 data> dogsb\P\P
5819 Partial match: dogsb
5820 data> do\P\D
5821 Partial match: do
5822 data> gsb\R\P\P\D
5823 Partial match: gsb
5824
5825
5826 4. Patterns that contain alternatives at the top level which do not all
5827 start with the same pattern item may not work as expected when
5828 pcre_dfa_exec() is used. For example, consider this pattern:
5829
5830 1234|3789
5831
5832 If the first part of the subject is "ABC123", a partial match of the
5833 first alternative is found at offset 3. There is no partial match for
5834 the second alternative, because such a match does not start at the same
5835 point in the subject string. Attempting to continue with the string
5836 "7890" does not yield a match because only those alternatives that
5837 match at one point in the subject are remembered. The problem arises
5838 because the start of the second alternative matches within the first
5839 alternative. There is no problem with anchored patterns or patterns
5840 such as:
5841
5842 1234|ABCD
5843
5844 where no string can be a partial match for both alternatives. This is
5845 not a problem if pcre_exec() is used, because the entire match has to
5846 be rerun each time:
5847
5848 re> /1234|3789/
5849 data> ABC123\P
5850 Partial match: 123
5851 data> 1237890
5852 0: 3789
5853
5854
5855 AUTHOR
5856
5857 Philip Hazel
5858 University Computing Service
5859 Cambridge CB2 3QH, England.
5860
5861
5862 REVISION
5863
5864 Last updated: 05 September 2009
5865 Copyright (c) 1997-2009 University of Cambridge.
5866 ------------------------------------------------------------------------------
5867
5868
5869 PCREPRECOMPILE(3) PCREPRECOMPILE(3)
5870
5871
5872 NAME
5873 PCRE - Perl-compatible regular expressions
5874
5875
5876 SAVING AND RE-USING PRECOMPILED PCRE PATTERNS
5877
5878 If you are running an application that uses a large number of regular
5879 expression patterns, it may be useful to store them in a precompiled
5880 form instead of having to compile them every time the application is
5881 run. If you are not using any private character tables (see the
5882 pcre_maketables() documentation), this is relatively straightforward.
5883 If you are using private tables, it is a little bit more complicated.
5884
5885 If you save compiled patterns to a file, you can copy them to a differ-
5886 ent host and run them there. This works even if the new host has the
5887 opposite endianness to the one on which the patterns were compiled.
5888 There may be a small performance penalty, but it should be insignifi-
5889 cant. However, compiling regular expressions with one version of PCRE
5890 for use with a different version is not guaranteed to work and may
5891 cause crashes.
5892
5893
5894 SAVING A COMPILED PATTERN
5895 The value returned by pcre_compile() points to a single block of memory
5896 that holds the compiled pattern and associated data. You can find the
5897 length of this block in bytes by calling pcre_fullinfo() with an argu-
5898 ment of PCRE_INFO_SIZE. You can then save the data in any appropriate
5899 manner. Here is sample code that compiles a pattern and writes it to a
5900 file. It assumes that the variable fd refers to a file that is open for
5901 output:
5902
5903 int erroroffset, rc, size;
5904 char *error;
5905 pcre *re;
5906
5907 re = pcre_compile("my pattern", 0, &error, &erroroffset, NULL);
5908 if (re == NULL) { ... handle errors ... }
5909 rc = pcre_fullinfo(re, NULL, PCRE_INFO_SIZE, &size);
5910 if (rc < 0) { ... handle errors ... }
5911 rc = fwrite(re, 1, size, fd);
5912 if (rc != size) { ... handle errors ... }
5913
5914 In this example, the bytes that comprise the compiled pattern are
5915 copied exactly. Note that this is binary data that may contain any of
5916 the 256 possible byte values. On systems that make a distinction
5917 between binary and non-binary data, be sure that the file is opened for
5918 binary output.
5919
5920 If you want to write more than one pattern to a file, you will have to
5921 devise a way of separating them. For binary data, preceding each pat-
5922 tern with its length is probably the most straightforward approach.
5923 Another possibility is to write out the data in hexadecimal instead of
5924 binary, one pattern to a line.
5925
5926 Saving compiled patterns in a file is only one possible way of storing
5927 them for later use. They could equally well be saved in a database, or
5928 in the memory of some daemon process that passes them via sockets to
5929 the processes that want them.
5930
5931 If the pattern has been studied, it is also possible to save the study
5932 data in a similar way to the compiled pattern itself. When studying
5933 generates additional information, pcre_study() returns a pointer to a
5934 pcre_extra data block. Its format is defined in the section on matching
5935 a pattern in the pcreapi documentation. The study_data field points to
5936 the binary study data, and this is what you must save (not the
5937 pcre_extra block itself). The length of the study data can be obtained
5938 by calling pcre_fullinfo() with an argument of PCRE_INFO_STUDYSIZE.
5939 Remember to check that pcre_study() did return a non-NULL value before
5940 trying to save the study data.
5941
5942
5943 RE-USING A PRECOMPILED PATTERN
5944
5945 Re-using a precompiled pattern is straightforward. Having reloaded it
5946 into main memory, you pass its pointer to pcre_exec() or
5947 pcre_dfa_exec() in the usual way. This should work even on another
5948 host, and even if that host has the opposite endianness to the one
5949 where the pattern was compiled.
5950
5951 However, if you passed a pointer to custom character tables when the
5952 pattern was compiled (the tableptr argument of pcre_compile()), you
5953 must now pass a similar pointer to pcre_exec() or pcre_dfa_exec(),
5954 because the value saved with the compiled pattern will obviously be
5955 nonsense. A field in a pcre_extra() block is used to pass this data, as
5956 described in the section on matching a pattern in the pcreapi documen-
5957 tation.
5958
5959 If you did not provide custom character tables when the pattern was
5960 compiled, the pointer in the compiled pattern is NULL, which causes
5961 pcre_exec() to use PCRE's internal tables. Thus, you do not need to
5962 take any special action at run time in this case.
5963
5964 If you saved study data with the compiled pattern, you need to create
5965 your own pcre_extra data block and set the study_data field to point to
5966 the reloaded study data. You must also set the PCRE_EXTRA_STUDY_DATA
5967 bit in the flags field to indicate that study data is present. Then
5968 pass the pcre_extra block to pcre_exec() or pcre_dfa_exec() in the
5969 usual way.
5970
5971
5972 COMPATIBILITY WITH DIFFERENT PCRE RELEASES
5973
5974 In general, it is safest to recompile all saved patterns when you
5975 update to a new PCRE release, though not all updates actually require
5976 this. Recompiling is definitely needed for release 7.2.
5977
5978
5979 AUTHOR
5980
5981 Philip Hazel
5982 University Computing Service
5983 Cambridge CB2 3QH, England.
5984
5985
5986 REVISION
5987
5988 Last updated: 13 June 2007
5989 Copyright (c) 1997-2007 University of Cambridge.
5990 ------------------------------------------------------------------------------
5991
5992
5993 PCREPERFORM(3) PCREPERFORM(3)
5994
5995
5996 NAME
5997 PCRE - Perl-compatible regular expressions
5998
5999
6000 PCRE PERFORMANCE
6001
6002 Two aspects of performance are discussed below: memory usage and pro-
6003 cessing time. The way you express your pattern as a regular expression
6004 can affect both of them.
6005
6006
6007 MEMORY USAGE
6008
6009 Patterns are compiled by PCRE into a reasonably efficient byte code, so
6010 that most simple patterns do not use much memory. However, there is one
6011 case where memory usage can be unexpectedly large. When a parenthesized
6012 subpattern has a quantifier with a minimum greater than 1 and/or a lim-
6013 ited maximum, the whole subpattern is repeated in the compiled code.
6014 For example, the pattern
6015
6016 (abc|def){2,4}
6017
6018 is compiled as if it were
6019
6020 (abc|def)(abc|def)((abc|def)(abc|def)?)?
6021
6022 (Technical aside: It is done this way so that backtrack points within
6023 each of the repetitions can be independently maintained.)
6024
6025 For regular expressions whose quantifiers use only small numbers, this
6026 is not usually a problem. However, if the numbers are large, and par-
6027 ticularly if such repetitions are nested, the memory usage can become
6028 an embarrassment. For example, the very simple pattern
6029
6030 ((ab){1,1000}c){1,3}
6031
6032 uses 51K bytes when compiled. When PCRE is compiled with its default
6033 internal pointer size of two bytes, the size limit on a compiled pat-
6034 tern is 64K, and this is reached with the above pattern if the outer
6035 repetition is increased from 3 to 4. PCRE can be compiled to use larger
6036 internal pointers and thus handle larger compiled patterns, but it is
6037 better to try to rewrite your pattern to use less memory if you can.
6038
6039 One way of reducing the memory usage for such patterns is to make use
6040 of PCRE's "subroutine" facility. Re-writing the above pattern as
6041
6042 ((ab)(?2){0,999}c)(?1){0,2}
6043
6044 reduces the memory requirements to 18K, and indeed it remains under 20K
6045 even with the outer repetition increased to 100. However, this pattern
6046 is not exactly equivalent, because the "subroutine" calls are treated
6047 as atomic groups into which there can be no backtracking if there is a
6048 subsequent matching failure. Therefore, PCRE cannot do this kind of
6049 rewriting automatically. Furthermore, there is a noticeable loss of
6050 speed when executing the modified pattern. Nevertheless, if the atomic
6051 grouping is not a problem and the loss of speed is acceptable, this
6052 kind of rewriting will allow you to process patterns that PCRE cannot
6053 otherwise handle.
6054
6055
6056 PROCESSING TIME
6057
6058 Certain items in regular expression patterns are processed more effi-
6059 ciently than others. It is more efficient to use a character class like
6060 [aeiou] than a set of single-character alternatives such as
6061 (a|e|i|o|u). In general, the simplest construction that provides the
6062 required behaviour is usually the most efficient. Jeffrey Friedl's book
6063 contains a lot of useful general discussion about optimizing regular
6064 expressions for efficient performance. This document contains a few
6065 observations about PCRE.
6066
6067 Using Unicode character properties (the \p, \P, and \X escapes) is
6068 slow, because PCRE has to scan a structure that contains data for over
6069 fifteen thousand characters whenever it needs a character's property.
6070 If you can find an alternative pattern that does not use character
6071 properties, it will probably be faster.
6072
6073 When a pattern begins with .* not in parentheses, or in parentheses
6074 that are not the subject of a backreference, and the PCRE_DOTALL option
6075 is set, the pattern is implicitly anchored by PCRE, since it can match
6076 only at the start of a subject string. However, if PCRE_DOTALL is not
6077 set, PCRE cannot make this optimization, because the . metacharacter
6078 does not then match a newline, and if the subject string contains new-