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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 572 - (show annotations) (download)
Wed Nov 17 17:55:57 2010 UTC (3 years, 5 months ago) by ph10
File MIME type: text/plain
File size: 365440 byte(s)
Documentation updates and tidies.

1 -----------------------------------------------------------------------------
2 This file contains a concatenation of the PCRE man pages, converted to plain
3 text format for ease of searching with a text editor, or for use on systems
4 that do not have a man page processor. The small individual files that give
5 synopses of each function in the library have not been included. Neither has
6 the pcredemo program. There are separate text files for the pcregrep and
7 pcretest commands.
8 -----------------------------------------------------------------------------
9
10
11 PCRE(3) PCRE(3)
12
13
14 NAME
15 PCRE - Perl-compatible regular expressions
16
17
18 INTRODUCTION
19
20 The PCRE library is a set of functions that implement regular expres-
21 sion pattern matching using the same syntax and semantics as Perl, with
22 just a few differences. Some features that appeared in Python and PCRE
23 before they appeared in Perl are also available using the Python syn-
24 tax, there is some support for one or two .NET and Oniguruma syntax
25 items, and there is an option for requesting some minor changes that
26 give better JavaScript compatibility.
27
28 The current implementation of PCRE corresponds approximately with Perl
29 5.12, including support for UTF-8 encoded strings and Unicode general
30 category properties. However, UTF-8 and Unicode support has to be
31 explicitly enabled; it is not the default. The Unicode tables corre-
32 spond to Unicode release 5.2.0.
33
34 In addition to the Perl-compatible matching function, PCRE contains an
35 alternative function that matches the same compiled patterns in a dif-
36 ferent way. In certain circumstances, the alternative function has some
37 advantages. For a discussion of the two matching algorithms, see the
38 pcrematching page.
39
40 PCRE is written in C and released as a C library. A number of people
41 have written wrappers and interfaces of various kinds. In particular,
42 Google Inc. have provided a comprehensive C++ wrapper. This is now
43 included as part of the PCRE distribution. The pcrecpp page has details
44 of this interface. Other people's contributions can be found in the
45 Contrib directory at the primary FTP site, which is:
46
47 ftp://ftp.csx.cam.ac.uk/pub/software/programming/pcre
48
49 Details of exactly which Perl regular expression features are and are
50 not supported by PCRE are given in separate documents. See the pcrepat-
51 tern and pcrecompat pages. There is a syntax summary in the pcresyntax
52 page.
53
54 Some features of PCRE can be included, excluded, or changed when the
55 library is built. The pcre_config() function makes it possible for a
56 client to discover which features are available. The features them-
57 selves are described in the pcrebuild page. Documentation about build-
58 ing PCRE for various operating systems can be found in the README and
59 NON-UNIX-USE files in the source distribution.
60
61 The library contains a number of undocumented internal functions and
62 data tables that are used by more than one of the exported external
63 functions, but which are not intended for use by external callers.
64 Their names all begin with "_pcre_", which hopefully will not provoke
65 any name clashes. In some environments, it is possible to control which
66 external symbols are exported when a shared library is built, and in
67 these cases the undocumented symbols are not exported.
68
69
70 USER DOCUMENTATION
71
72 The user documentation for PCRE comprises a number of different sec-
73 tions. In the "man" format, each of these is a separate "man page". In
74 the HTML format, each is a separate page, linked from the index page.
75 In the plain text format, all the sections, except the pcredemo sec-
76 tion, are concatenated, for ease of searching. The sections are as fol-
77 lows:
78
79 pcre this document
80 pcre-config show PCRE installation configuration information
81 pcreapi details of PCRE's native C API
82 pcrebuild options for building PCRE
83 pcrecallout details of the callout feature
84 pcrecompat discussion of Perl compatibility
85 pcrecpp details of the C++ wrapper
86 pcredemo a demonstration C program that uses PCRE
87 pcregrep description of the pcregrep command
88 pcrematching discussion of the two matching algorithms
89 pcrepartial details of the partial matching facility
90 pcrepattern syntax and semantics of supported
91 regular expressions
92 pcreperform discussion of performance issues
93 pcreposix the POSIX-compatible C API
94 pcreprecompile details of saving and re-using precompiled patterns
95 pcresample discussion of the pcredemo program
96 pcrestack discussion of stack usage
97 pcresyntax quick syntax reference
98 pcretest description of the pcretest testing command
99
100 In addition, in the "man" and HTML formats, there is a short page for
101 each C library function, listing its arguments and results.
102
103
104 LIMITATIONS
105
106 There are some size limitations in PCRE but it is hoped that they will
107 never in practice be relevant.
108
109 The maximum length of a compiled pattern is 65539 (sic) bytes if PCRE
110 is compiled with the default internal linkage size of 2. If you want to
111 process regular expressions that are truly enormous, you can compile
112 PCRE with an internal linkage size of 3 or 4 (see the README file in
113 the source distribution and the pcrebuild documentation for details).
114 In these cases the limit is substantially larger. However, the speed
115 of execution is slower.
116
117 All values in repeating quantifiers must be less than 65536.
118
119 There is no limit to the number of parenthesized subpatterns, but there
120 can be no more than 65535 capturing subpatterns.
121
122 The maximum length of name for a named subpattern is 32 characters, and
123 the maximum number of named subpatterns is 10000.
124
125 The maximum length of a subject string is the largest positive number
126 that an integer variable can hold. However, when using the traditional
127 matching function, PCRE uses recursion to handle subpatterns and indef-
128 inite repetition. This means that the available stack space may limit
129 the size of a subject string that can be processed by certain patterns.
130 For a discussion of stack issues, see the pcrestack documentation.
131
132
133 UTF-8 AND UNICODE PROPERTY SUPPORT
134
135 From release 3.3, PCRE has had some support for character strings
136 encoded in the UTF-8 format. For release 4.0 this was greatly extended
137 to cover most common requirements, and in release 5.0 additional sup-
138 port for Unicode general category properties was added.
139
140 In order process UTF-8 strings, you must build PCRE to include UTF-8
141 support in the code, and, in addition, you must call pcre_compile()
142 with the PCRE_UTF8 option flag, or the pattern must start with the
143 sequence (*UTF8). When either of these is the case, both the pattern
144 and any subject strings that are matched against it are treated as
145 UTF-8 strings instead of strings of 1-byte characters.
146
147 If you compile PCRE with UTF-8 support, but do not use it at run time,
148 the library will be a bit bigger, but the additional run time overhead
149 is limited to testing the PCRE_UTF8 flag occasionally, so should not be
150 very big.
151
152 If PCRE is built with Unicode character property support (which implies
153 UTF-8 support), the escape sequences \p{..}, \P{..}, and \X are sup-
154 ported. The available properties that can be tested are limited to the
155 general category properties such as Lu for an upper case letter or Nd
156 for a decimal number, the Unicode script names such as Arabic or Han,
157 and the derived properties Any and L&. A full list is given in the
158 pcrepattern documentation. Only the short names for properties are sup-
159 ported. For example, \p{L} matches a letter. Its Perl synonym, \p{Let-
160 ter}, is not supported. Furthermore, in Perl, many properties may
161 optionally be prefixed by "Is", for compatibility with Perl 5.6. PCRE
162 does not support this.
163
164 Validity of UTF-8 strings
165
166 When you set the PCRE_UTF8 flag, the strings passed as patterns and
167 subjects are (by default) checked for validity on entry to the relevant
168 functions. From release 7.3 of PCRE, the check is according the rules
169 of RFC 3629, which are themselves derived from the Unicode specifica-
170 tion. Earlier releases of PCRE followed the rules of RFC 2279, which
171 allows the full range of 31-bit values (0 to 0x7FFFFFFF). The current
172 check allows only values in the range U+0 to U+10FFFF, excluding U+D800
173 to U+DFFF.
174
175 The excluded code points are the "Low Surrogate Area" of Unicode, of
176 which the Unicode Standard says this: "The Low Surrogate Area does not
177 contain any character assignments, consequently no character code
178 charts or namelists are provided for this area. Surrogates are reserved
179 for use with UTF-16 and then must be used in pairs." The code points
180 that are encoded by UTF-16 pairs are available as independent code
181 points in the UTF-8 encoding. (In other words, the whole surrogate
182 thing is a fudge for UTF-16 which unfortunately messes up UTF-8.)
183
184 If an invalid UTF-8 string is passed to PCRE, an error return
185 (PCRE_ERROR_BADUTF8) is given. In some situations, you may already know
186 that your strings are valid, and therefore want to skip these checks in
187 order to improve performance. If you set the PCRE_NO_UTF8_CHECK flag at
188 compile time or at run time, PCRE assumes that the pattern or subject
189 it is given (respectively) contains only valid UTF-8 codes. In this
190 case, it does not diagnose an invalid UTF-8 string.
191
192 If you pass an invalid UTF-8 string when PCRE_NO_UTF8_CHECK is set,
193 what happens depends on why the string is invalid. If the string con-
194 forms to the "old" definition of UTF-8 (RFC 2279), it is processed as a
195 string of characters in the range 0 to 0x7FFFFFFF. In other words,
196 apart from the initial validity test, PCRE (when in UTF-8 mode) handles
197 strings according to the more liberal rules of RFC 2279. However, if
198 the string does not even conform to RFC 2279, the result is undefined.
199 Your program may crash.
200
201 If you want to process strings of values in the full range 0 to
202 0x7FFFFFFF, encoded in a UTF-8-like manner as per the old RFC, you can
203 set PCRE_NO_UTF8_CHECK to bypass the more restrictive test. However, in
204 this situation, you will have to apply your own validity check.
205
206 General comments about UTF-8 mode
207
208 1. An unbraced hexadecimal escape sequence (such as \xb3) matches a
209 two-byte UTF-8 character if the value is greater than 127.
210
211 2. Octal numbers up to \777 are recognized, and match two-byte UTF-8
212 characters for values greater than \177.
213
214 3. Repeat quantifiers apply to complete UTF-8 characters, not to indi-
215 vidual bytes, for example: \x{100}{3}.
216
217 4. The dot metacharacter matches one UTF-8 character instead of a sin-
218 gle byte.
219
220 5. The escape sequence \C can be used to match a single byte in UTF-8
221 mode, but its use can lead to some strange effects. This facility is
222 not available in the alternative matching function, pcre_dfa_exec().
223
224 6. The character escapes \b, \B, \d, \D, \s, \S, \w, and \W correctly
225 test characters of any code value, but, by default, the characters that
226 PCRE recognizes as digits, spaces, or word characters remain the same
227 set as before, all with values less than 256. This remains true even
228 when PCRE is built to include Unicode property support, because to do
229 otherwise would slow down PCRE in many common cases. Note in particular
230 that this applies to \b and \B, because they are defined in terms of \w
231 and \W. If you really want to test for a wider sense of, say, "digit",
232 you can use explicit Unicode property tests such as \p{Nd}. Alterna-
233 tively, if you set the PCRE_UCP option, the way that the character
234 escapes work is changed so that Unicode properties are used to deter-
235 mine which characters match. There are more details in the section on
236 generic character types in the pcrepattern documentation.
237
238 7. Similarly, characters that match the POSIX named character classes
239 are all low-valued characters, unless the PCRE_UCP option is set.
240
241 8. However, the horizontal and vertical whitespace matching escapes
242 (\h, \H, \v, and \V) do match all the appropriate Unicode characters,
243 whether or not PCRE_UCP is set.
244
245 9. Case-insensitive matching applies only to characters whose values
246 are less than 128, unless PCRE is built with Unicode property support.
247 Even when Unicode property support is available, PCRE still uses its
248 own character tables when checking the case of low-valued characters,
249 so as not to degrade performance. The Unicode property information is
250 used only for characters with higher values. Furthermore, PCRE supports
251 case-insensitive matching only when there is a one-to-one mapping
252 between a letter's cases. There are a small number of many-to-one map-
253 pings in Unicode; these are not supported by PCRE.
254
255
256 AUTHOR
257
258 Philip Hazel
259 University Computing Service
260 Cambridge CB2 3QH, England.
261
262 Putting an actual email address here seems to have been a spam magnet,
263 so I've taken it away. If you want to email me, use my two initials,
264 followed by the two digits 10, at the domain cam.ac.uk.
265
266
267 REVISION
268
269 Last updated: 13 November 2010
270 Copyright (c) 1997-2010 University of Cambridge.
271 ------------------------------------------------------------------------------
272
273
274 PCREBUILD(3) PCREBUILD(3)
275
276
277 NAME
278 PCRE - Perl-compatible regular expressions
279
280
281 PCRE BUILD-TIME OPTIONS
282
283 This document describes the optional features of PCRE that can be
284 selected when the library is compiled. It assumes use of the configure
285 script, where the optional features are selected or deselected by pro-
286 viding options to configure before running the make command. However,
287 the same options can be selected in both Unix-like and non-Unix-like
288 environments using the GUI facility of cmake-gui if you are using CMake
289 instead of configure to build PCRE.
290
291 There is a lot more information about building PCRE in non-Unix-like
292 environments in the file called NON_UNIX_USE, which is part of the PCRE
293 distribution. You should consult this file as well as the README file
294 if you are building in a non-Unix-like environment.
295
296 The complete list of options for configure (which includes the standard
297 ones such as the selection of the installation directory) can be
298 obtained by running
299
300 ./configure --help
301
302 The following sections include descriptions of options whose names
303 begin with --enable or --disable. These settings specify changes to the
304 defaults for the configure command. Because of the way that configure
305 works, --enable and --disable always come in pairs, so the complemen-
306 tary option always exists as well, but as it specifies the default, it
307 is not described.
308
309
310 C++ SUPPORT
311
312 By default, the configure script will search for a C++ compiler and C++
313 header files. If it finds them, it automatically builds the C++ wrapper
314 library for PCRE. You can disable this by adding
315
316 --disable-cpp
317
318 to the configure command.
319
320
321 UTF-8 SUPPORT
322
323 To build PCRE with support for UTF-8 Unicode character strings, add
324
325 --enable-utf8
326
327 to the configure command. Of itself, this does not make PCRE treat
328 strings as UTF-8. As well as compiling PCRE with this option, you also
329 have have to set the PCRE_UTF8 option when you call the pcre_compile()
330 or pcre_compile2() functions.
331
332 If you set --enable-utf8 when compiling in an EBCDIC environment, PCRE
333 expects its input to be either ASCII or UTF-8 (depending on the runtime
334 option). It is not possible to support both EBCDIC and UTF-8 codes in
335 the same version of the library. Consequently, --enable-utf8 and
336 --enable-ebcdic are mutually exclusive.
337
338
339 UNICODE CHARACTER PROPERTY SUPPORT
340
341 UTF-8 support allows PCRE to process character values greater than 255
342 in the strings that it handles. On its own, however, it does not pro-
343 vide any facilities for accessing the properties of such characters. If
344 you want to be able to use the pattern escapes \P, \p, and \X, which
345 refer to Unicode character properties, you must add
346
347 --enable-unicode-properties
348
349 to the configure command. This implies UTF-8 support, even if you have
350 not explicitly requested it.
351
352 Including Unicode property support adds around 30K of tables to the
353 PCRE library. Only the general category properties such as Lu and Nd
354 are supported. Details are given in the pcrepattern documentation.
355
356
357 CODE VALUE OF NEWLINE
358
359 By default, PCRE interprets the linefeed (LF) character as indicating
360 the end of a line. This is the normal newline character on Unix-like
361 systems. You can compile PCRE to use carriage return (CR) instead, by
362 adding
363
364 --enable-newline-is-cr
365
366 to the configure command. There is also a --enable-newline-is-lf
367 option, which explicitly specifies linefeed as the newline character.
368
369 Alternatively, you can specify that line endings are to be indicated by
370 the two character sequence CRLF. If you want this, add
371
372 --enable-newline-is-crlf
373
374 to the configure command. There is a fourth option, specified by
375
376 --enable-newline-is-anycrlf
377
378 which causes PCRE to recognize any of the three sequences CR, LF, or
379 CRLF as indicating a line ending. Finally, a fifth option, specified by
380
381 --enable-newline-is-any
382
383 causes PCRE to recognize any Unicode newline sequence.
384
385 Whatever line ending convention is selected when PCRE is built can be
386 overridden when the library functions are called. At build time it is
387 conventional to use the standard for your operating system.
388
389
390 WHAT \R MATCHES
391
392 By default, the sequence \R in a pattern matches any Unicode newline
393 sequence, whatever has been selected as the line ending sequence. If
394 you specify
395
396 --enable-bsr-anycrlf
397
398 the default is changed so that \R matches only CR, LF, or CRLF. What-
399 ever is selected when PCRE is built can be overridden when the library
400 functions are called.
401
402
403 BUILDING SHARED AND STATIC LIBRARIES
404
405 The PCRE building process uses libtool to build both shared and static
406 Unix libraries by default. You can suppress one of these by adding one
407 of
408
409 --disable-shared
410 --disable-static
411
412 to the configure command, as required.
413
414
415 POSIX MALLOC USAGE
416
417 When PCRE is called through the POSIX interface (see the pcreposix doc-
418 umentation), additional working storage is required for holding the
419 pointers to capturing substrings, because PCRE requires three integers
420 per substring, whereas the POSIX interface provides only two. If the
421 number of expected substrings is small, the wrapper function uses space
422 on the stack, because this is faster than using malloc() for each call.
423 The default threshold above which the stack is no longer used is 10; it
424 can be changed by adding a setting such as
425
426 --with-posix-malloc-threshold=20
427
428 to the configure command.
429
430
431 HANDLING VERY LARGE PATTERNS
432
433 Within a compiled pattern, offset values are used to point from one
434 part to another (for example, from an opening parenthesis to an alter-
435 nation metacharacter). By default, two-byte values are used for these
436 offsets, leading to a maximum size for a compiled pattern of around
437 64K. This is sufficient to handle all but the most gigantic patterns.
438 Nevertheless, some people do want to process truyl enormous patterns,
439 so it is possible to compile PCRE to use three-byte or four-byte off-
440 sets by adding a setting such as
441
442 --with-link-size=3
443
444 to the configure command. The value given must be 2, 3, or 4. Using
445 longer offsets slows down the operation of PCRE because it has to load
446 additional bytes when handling them.
447
448
449 AVOIDING EXCESSIVE STACK USAGE
450
451 When matching with the pcre_exec() function, PCRE implements backtrack-
452 ing by making recursive calls to an internal function called match().
453 In environments where the size of the stack is limited, this can se-
454 verely limit PCRE's operation. (The Unix environment does not usually
455 suffer from this problem, but it may sometimes be necessary to increase
456 the maximum stack size. There is a discussion in the pcrestack docu-
457 mentation.) An alternative approach to recursion that uses memory from
458 the heap to remember data, instead of using recursive function calls,
459 has been implemented to work round the problem of limited stack size.
460 If you want to build a version of PCRE that works this way, add
461
462 --disable-stack-for-recursion
463
464 to the configure command. With this configuration, PCRE will use the
465 pcre_stack_malloc and pcre_stack_free variables to call memory manage-
466 ment functions. By default these point to malloc() and free(), but you
467 can replace the pointers so that your own functions are used instead.
468
469 Separate functions are provided rather than using pcre_malloc and
470 pcre_free because the usage is very predictable: the block sizes
471 requested are always the same, and the blocks are always freed in
472 reverse order. A calling program might be able to implement optimized
473 functions that perform better than malloc() and free(). PCRE runs
474 noticeably more slowly when built in this way. This option affects only
475 the pcre_exec() function; it is not relevant for pcre_dfa_exec().
476
477
478 LIMITING PCRE RESOURCE USAGE
479
480 Internally, PCRE has a function called match(), which it calls repeat-
481 edly (sometimes recursively) when matching a pattern with the
482 pcre_exec() function. By controlling the maximum number of times this
483 function may be called during a single matching operation, a limit can
484 be placed on the resources used by a single call to pcre_exec(). The
485 limit can be changed at run time, as described in the pcreapi documen-
486 tation. The default is 10 million, but this can be changed by adding a
487 setting such as
488
489 --with-match-limit=500000
490
491 to the configure command. This setting has no effect on the
492 pcre_dfa_exec() matching function.
493
494 In some environments it is desirable to limit the depth of recursive
495 calls of match() more strictly than the total number of calls, in order
496 to restrict the maximum amount of stack (or heap, if --disable-stack-
497 for-recursion is specified) that is used. A second limit controls this;
498 it defaults to the value that is set for --with-match-limit, which
499 imposes no additional constraints. However, you can set a lower limit
500 by adding, for example,
501
502 --with-match-limit-recursion=10000
503
504 to the configure command. This value can also be overridden at run
505 time.
506
507
508 CREATING CHARACTER TABLES AT BUILD TIME
509
510 PCRE uses fixed tables for processing characters whose code values are
511 less than 256. By default, PCRE is built with a set of tables that are
512 distributed in the file pcre_chartables.c.dist. These tables are for
513 ASCII codes only. If you add
514
515 --enable-rebuild-chartables
516
517 to the configure command, the distributed tables are no longer used.
518 Instead, a program called dftables is compiled and run. This outputs
519 the source for new set of tables, created in the default locale of your
520 C runtime system. (This method of replacing the tables does not work if
521 you are cross compiling, because dftables is run on the local host. If
522 you need to create alternative tables when cross compiling, you will
523 have to do so "by hand".)
524
525
526 USING EBCDIC CODE
527
528 PCRE assumes by default that it will run in an environment where the
529 character code is ASCII (or Unicode, which is a superset of ASCII).
530 This is the case for most computer operating systems. PCRE can, how-
531 ever, be compiled to run in an EBCDIC environment by adding
532
533 --enable-ebcdic
534
535 to the configure command. This setting implies --enable-rebuild-charta-
536 bles. You should only use it if you know that you are in an EBCDIC
537 environment (for example, an IBM mainframe operating system). The
538 --enable-ebcdic option is incompatible with --enable-utf8.
539
540
541 PCREGREP OPTIONS FOR COMPRESSED FILE SUPPORT
542
543 By default, pcregrep reads all files as plain text. You can build it so
544 that it recognizes files whose names end in .gz or .bz2, and reads them
545 with libz or libbz2, respectively, by adding one or both of
546
547 --enable-pcregrep-libz
548 --enable-pcregrep-libbz2
549
550 to the configure command. These options naturally require that the rel-
551 evant libraries are installed on your system. Configuration will fail
552 if they are not.
553
554
555 PCRETEST OPTION FOR LIBREADLINE SUPPORT
556
557 If you add
558
559 --enable-pcretest-libreadline
560
561 to the configure command, pcretest is linked with the libreadline
562 library, and when its input is from a terminal, it reads it using the
563 readline() function. This provides line-editing and history facilities.
564 Note that libreadline is GPL-licensed, so if you distribute a binary of
565 pcretest linked in this way, there may be licensing issues.
566
567 Setting this option causes the -lreadline option to be added to the
568 pcretest build. In many operating environments with a sytem-installed
569 libreadline this is sufficient. However, in some environments (e.g. if
570 an unmodified distribution version of readline is in use), some extra
571 configuration may be necessary. The INSTALL file for libreadline says
572 this:
573
574 "Readline uses the termcap functions, but does not link with the
575 termcap or curses library itself, allowing applications which link
576 with readline the to choose an appropriate library."
577
578 If your environment has not been set up so that an appropriate library
579 is automatically included, you may need to add something like
580
581 LIBS="-ncurses"
582
583 immediately before the configure command.
584
585
586 SEE ALSO
587
588 pcreapi(3), pcre_config(3).
589
590
591 AUTHOR
592
593 Philip Hazel
594 University Computing Service
595 Cambridge CB2 3QH, England.
596
597
598 REVISION
599
600 Last updated: 29 September 2009
601 Copyright (c) 1997-2009 University of Cambridge.
602 ------------------------------------------------------------------------------
603
604
605 PCREMATCHING(3) PCREMATCHING(3)
606
607
608 NAME
609 PCRE - Perl-compatible regular expressions
610
611
612 PCRE MATCHING ALGORITHMS
613
614 This document describes the two different algorithms that are available
615 in PCRE for matching a compiled regular expression against a given sub-
616 ject string. The "standard" algorithm is the one provided by the
617 pcre_exec() function. This works in the same was as Perl's matching
618 function, and provides a Perl-compatible matching operation.
619
620 An alternative algorithm is provided by the pcre_dfa_exec() function;
621 this operates in a different way, and is not Perl-compatible. It has
622 advantages and disadvantages compared with the standard algorithm, and
623 these are described below.
624
625 When there is only one possible way in which a given subject string can
626 match a pattern, the two algorithms give the same answer. A difference
627 arises, however, when there are multiple possibilities. For example, if
628 the pattern
629
630 ^<.*>
631
632 is matched against the string
633
634 <something> <something else> <something further>
635
636 there are three possible answers. The standard algorithm finds only one
637 of them, whereas the alternative algorithm finds all three.
638
639
640 REGULAR EXPRESSIONS AS TREES
641
642 The set of strings that are matched by a regular expression can be rep-
643 resented as a tree structure. An unlimited repetition in the pattern
644 makes the tree of infinite size, but it is still a tree. Matching the
645 pattern to a given subject string (from a given starting point) can be
646 thought of as a search of the tree. There are two ways to search a
647 tree: depth-first and breadth-first, and these correspond to the two
648 matching algorithms provided by PCRE.
649
650
651 THE STANDARD MATCHING ALGORITHM
652
653 In the terminology of Jeffrey Friedl's book "Mastering Regular Expres-
654 sions", the standard algorithm is an "NFA algorithm". It conducts a
655 depth-first search of the pattern tree. That is, it proceeds along a
656 single path through the tree, checking that the subject matches what is
657 required. When there is a mismatch, the algorithm tries any alterna-
658 tives at the current point, and if they all fail, it backs up to the
659 previous branch point in the tree, and tries the next alternative
660 branch at that level. This often involves backing up (moving to the
661 left) in the subject string as well. The order in which repetition
662 branches are tried is controlled by the greedy or ungreedy nature of
663 the quantifier.
664
665 If a leaf node is reached, a matching string has been found, and at
666 that point the algorithm stops. Thus, if there is more than one possi-
667 ble match, this algorithm returns the first one that it finds. Whether
668 this is the shortest, the longest, or some intermediate length depends
669 on the way the greedy and ungreedy repetition quantifiers are specified
670 in the pattern.
671
672 Because it ends up with a single path through the tree, it is rela-
673 tively straightforward for this algorithm to keep track of the sub-
674 strings that are matched by portions of the pattern in parentheses.
675 This provides support for capturing parentheses and back references.
676
677
678 THE ALTERNATIVE MATCHING ALGORITHM
679
680 This algorithm conducts a breadth-first search of the tree. Starting
681 from the first matching point in the subject, it scans the subject
682 string from left to right, once, character by character, and as it does
683 this, it remembers all the paths through the tree that represent valid
684 matches. In Friedl's terminology, this is a kind of "DFA algorithm",
685 though it is not implemented as a traditional finite state machine (it
686 keeps multiple states active simultaneously).
687
688 Although the general principle of this matching algorithm is that it
689 scans the subject string only once, without backtracking, there is one
690 exception: when a lookaround assertion is encountered, the characters
691 following or preceding the current point have to be independently
692 inspected.
693
694 The scan continues until either the end of the subject is reached, or
695 there are no more unterminated paths. At this point, terminated paths
696 represent the different matching possibilities (if there are none, the
697 match has failed). Thus, if there is more than one possible match,
698 this algorithm finds all of them, and in particular, it finds the long-
699 est. The matches are returned in decreasing order of length. There is
700 an option to stop the algorithm after the first match (which is neces-
701 sarily the shortest) is found.
702
703 Note that all the matches that are found start at the same point in the
704 subject. If the pattern
705
706 cat(er(pillar)?)?
707
708 is matched against the string "the caterpillar catchment", the result
709 will be the three strings "caterpillar", "cater", and "cat" that start
710 at the fifth character of the subject. The algorithm does not automati-
711 cally move on to find matches that start at later positions.
712
713 There are a number of features of PCRE regular expressions that are not
714 supported by the alternative matching algorithm. They are as follows:
715
716 1. Because the algorithm finds all possible matches, the greedy or
717 ungreedy nature of repetition quantifiers is not relevant. Greedy and
718 ungreedy quantifiers are treated in exactly the same way. However, pos-
719 sessive quantifiers can make a difference when what follows could also
720 match what is quantified, for example in a pattern like this:
721
722 ^a++\w!
723
724 This pattern matches "aaab!" but not "aaa!", which would be matched by
725 a non-possessive quantifier. Similarly, if an atomic group is present,
726 it is matched as if it were a standalone pattern at the current point,
727 and the longest match is then "locked in" for the rest of the overall
728 pattern.
729
730 2. When dealing with multiple paths through the tree simultaneously, it
731 is not straightforward to keep track of captured substrings for the
732 different matching possibilities, and PCRE's implementation of this
733 algorithm does not attempt to do this. This means that no captured sub-
734 strings are available.
735
736 3. Because no substrings are captured, back references within the pat-
737 tern are not supported, and cause errors if encountered.
738
739 4. For the same reason, conditional expressions that use a backrefer-
740 ence as the condition or test for a specific group recursion are not
741 supported.
742
743 5. Because many paths through the tree may be active, the \K escape
744 sequence, which resets the start of the match when encountered (but may
745 be on some paths and not on others), is not supported. It causes an
746 error if encountered.
747
748 6. Callouts are supported, but the value of the capture_top field is
749 always 1, and the value of the capture_last field is always -1.
750
751 7. The \C escape sequence, which (in the standard algorithm) matches a
752 single byte, even in UTF-8 mode, is not supported because the alterna-
753 tive algorithm moves through the subject string one character at a
754 time, for all active paths through the tree.
755
756 8. Except for (*FAIL), the backtracking control verbs such as (*PRUNE)
757 are not supported. (*FAIL) is supported, and behaves like a failing
758 negative assertion.
759
760
761 ADVANTAGES OF THE ALTERNATIVE ALGORITHM
762
763 Using the alternative matching algorithm provides the following advan-
764 tages:
765
766 1. All possible matches (at a single point in the subject) are automat-
767 ically found, and in particular, the longest match is found. To find
768 more than one match using the standard algorithm, you have to do kludgy
769 things with callouts.
770
771 2. Because the alternative algorithm scans the subject string just
772 once, and never needs to backtrack, it is possible to pass very long
773 subject strings to the matching function in several pieces, checking
774 for partial matching each time. Although it is possible to do multi-
775 segment matching using the standard algorithm (pcre_exec()), by retain-
776 ing partially matched substrings, it is more complicated. The pcrepar-
777 tial documentation gives details of partial matching and discusses
778 multi-segment matching.
779
780
781 DISADVANTAGES OF THE ALTERNATIVE ALGORITHM
782
783 The alternative algorithm suffers from a number of disadvantages:
784
785 1. It is substantially slower than the standard algorithm. This is
786 partly because it has to search for all possible matches, but is also
787 because it is less susceptible to optimization.
788
789 2. Capturing parentheses and back references are not supported.
790
791 3. Although atomic groups are supported, their use does not provide the
792 performance advantage that it does for the standard algorithm.
793
794
795 AUTHOR
796
797 Philip Hazel
798 University Computing Service
799 Cambridge CB2 3QH, England.
800
801
802 REVISION
803
804 Last updated: 17 November 2010
805 Copyright (c) 1997-2010 University of Cambridge.
806 ------------------------------------------------------------------------------
807
808
809 PCREAPI(3) PCREAPI(3)
810
811
812 NAME
813 PCRE - Perl-compatible regular expressions
814
815
816 PCRE NATIVE API
817
818 #include <pcre.h>
819
820 pcre *pcre_compile(const char *pattern, int options,
821 const char **errptr, int *erroffset,
822 const unsigned char *tableptr);
823
824 pcre *pcre_compile2(const char *pattern, int options,
825 int *errorcodeptr,
826 const char **errptr, int *erroffset,
827 const unsigned char *tableptr);
828
829 pcre_extra *pcre_study(const pcre *code, int options,
830 const char **errptr);
831
832 int pcre_exec(const pcre *code, const pcre_extra *extra,
833 const char *subject, int length, int startoffset,
834 int options, int *ovector, int ovecsize);
835
836 int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
837 const char *subject, int length, int startoffset,
838 int options, int *ovector, int ovecsize,
839 int *workspace, int wscount);
840
841 int pcre_copy_named_substring(const pcre *code,
842 const char *subject, int *ovector,
843 int stringcount, const char *stringname,
844 char *buffer, int buffersize);
845
846 int pcre_copy_substring(const char *subject, int *ovector,
847 int stringcount, int stringnumber, char *buffer,
848 int buffersize);
849
850 int pcre_get_named_substring(const pcre *code,
851 const char *subject, int *ovector,
852 int stringcount, const char *stringname,
853 const char **stringptr);
854
855 int pcre_get_stringnumber(const pcre *code,
856 const char *name);
857
858 int pcre_get_stringtable_entries(const pcre *code,
859 const char *name, char **first, char **last);
860
861 int pcre_get_substring(const char *subject, int *ovector,
862 int stringcount, int stringnumber,
863 const char **stringptr);
864
865 int pcre_get_substring_list(const char *subject,
866 int *ovector, int stringcount, const char ***listptr);
867
868 void pcre_free_substring(const char *stringptr);
869
870 void pcre_free_substring_list(const char **stringptr);
871
872 const unsigned char *pcre_maketables(void);
873
874 int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
875 int what, void *where);
876
877 int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
878
879 int pcre_refcount(pcre *code, int adjust);
880
881 int pcre_config(int what, void *where);
882
883 char *pcre_version(void);
884
885 void *(*pcre_malloc)(size_t);
886
887 void (*pcre_free)(void *);
888
889 void *(*pcre_stack_malloc)(size_t);
890
891 void (*pcre_stack_free)(void *);
892
893 int (*pcre_callout)(pcre_callout_block *);
894
895
896 PCRE API OVERVIEW
897
898 PCRE has its own native API, which is described in this document. There
899 are also some wrapper functions that correspond to the POSIX regular
900 expression API. These are described in the pcreposix documentation.
901 Both of these APIs define a set of C function calls. A C++ wrapper is
902 distributed with PCRE. It is documented in the pcrecpp page.
903
904 The native API C function prototypes are defined in the header file
905 pcre.h, and on Unix systems the library itself is called libpcre. It
906 can normally be accessed by adding -lpcre to the command for linking an
907 application that uses PCRE. The header file defines the macros
908 PCRE_MAJOR and PCRE_MINOR to contain the major and minor release num-
909 bers for the library. Applications can use these to include support
910 for different releases of PCRE.
911
912 In a Windows environment, if you want to statically link an application
913 program against a non-dll pcre.a file, you must define PCRE_STATIC
914 before including pcre.h or pcrecpp.h, because otherwise the pcre_mal-
915 loc() and pcre_free() exported functions will be declared
916 __declspec(dllimport), with unwanted results.
917
918 The functions pcre_compile(), pcre_compile2(), pcre_study(), and
919 pcre_exec() are used for compiling and matching regular expressions in
920 a Perl-compatible manner. A sample program that demonstrates the sim-
921 plest way of using them is provided in the file called pcredemo.c in
922 the PCRE source distribution. A listing of this program is given in the
923 pcredemo documentation, and the pcresample documentation describes how
924 to compile and run it.
925
926 A second matching function, pcre_dfa_exec(), which is not Perl-compati-
927 ble, is also provided. This uses a different algorithm for the match-
928 ing. The alternative algorithm finds all possible matches (at a given
929 point in the subject), and scans the subject just once (unless there
930 are lookbehind assertions). However, this algorithm does not return
931 captured substrings. A description of the two matching algorithms and
932 their advantages and disadvantages is given in the pcrematching docu-
933 mentation.
934
935 In addition to the main compiling and matching functions, there are
936 convenience functions for extracting captured substrings from a subject
937 string that is matched by pcre_exec(). They are:
938
939 pcre_copy_substring()
940 pcre_copy_named_substring()
941 pcre_get_substring()
942 pcre_get_named_substring()
943 pcre_get_substring_list()
944 pcre_get_stringnumber()
945 pcre_get_stringtable_entries()
946
947 pcre_free_substring() and pcre_free_substring_list() are also provided,
948 to free the memory used for extracted strings.
949
950 The function pcre_maketables() is used to build a set of character
951 tables in the current locale for passing to pcre_compile(),
952 pcre_exec(), or pcre_dfa_exec(). This is an optional facility that is
953 provided for specialist use. Most commonly, no special tables are
954 passed, in which case internal tables that are generated when PCRE is
955 built are used.
956
957 The function pcre_fullinfo() is used to find out information about a
958 compiled pattern; pcre_info() is an obsolete version that returns only
959 some of the available information, but is retained for backwards com-
960 patibility. The function pcre_version() returns a pointer to a string
961 containing the version of PCRE and its date of release.
962
963 The function pcre_refcount() maintains a reference count in a data
964 block containing a compiled pattern. This is provided for the benefit
965 of object-oriented applications.
966
967 The global variables pcre_malloc and pcre_free initially contain the
968 entry points of the standard malloc() and free() functions, respec-
969 tively. PCRE calls the memory management functions via these variables,
970 so a calling program can replace them if it wishes to intercept the
971 calls. This should be done before calling any PCRE functions.
972
973 The global variables pcre_stack_malloc and pcre_stack_free are also
974 indirections to memory management functions. These special functions
975 are used only when PCRE is compiled to use the heap for remembering
976 data, instead of recursive function calls, when running the pcre_exec()
977 function. See the pcrebuild documentation for details of how to do
978 this. It is a non-standard way of building PCRE, for use in environ-
979 ments that have limited stacks. Because of the greater use of memory
980 management, it runs more slowly. Separate functions are provided so
981 that special-purpose external code can be used for this case. When
982 used, these functions are always called in a stack-like manner (last
983 obtained, first freed), and always for memory blocks of the same size.
984 There is a discussion about PCRE's stack usage in the pcrestack docu-
985 mentation.
986
987 The global variable pcre_callout initially contains NULL. It can be set
988 by the caller to a "callout" function, which PCRE will then call at
989 specified points during a matching operation. Details are given in the
990 pcrecallout documentation.
991
992
993 NEWLINES
994
995 PCRE supports five different conventions for indicating line breaks in
996 strings: a single CR (carriage return) character, a single LF (line-
997 feed) character, the two-character sequence CRLF, any of the three pre-
998 ceding, or any Unicode newline sequence. The Unicode newline sequences
999 are the three just mentioned, plus the single characters VT (vertical
1000 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
1001 separator, U+2028), and PS (paragraph separator, U+2029).
1002
1003 Each of the first three conventions is used by at least one operating
1004 system as its standard newline sequence. When PCRE is built, a default
1005 can be specified. The default default is LF, which is the Unix stan-
1006 dard. When PCRE is run, the default can be overridden, either when a
1007 pattern is compiled, or when it is matched.
1008
1009 At compile time, the newline convention can be specified by the options
1010 argument of pcre_compile(), or it can be specified by special text at
1011 the start of the pattern itself; this overrides any other settings. See
1012 the pcrepattern page for details of the special character sequences.
1013
1014 In the PCRE documentation the word "newline" is used to mean "the char-
1015 acter or pair of characters that indicate a line break". The choice of
1016 newline convention affects the handling of the dot, circumflex, and
1017 dollar metacharacters, the handling of #-comments in /x mode, and, when
1018 CRLF is a recognized line ending sequence, the match position advance-
1019 ment for a non-anchored pattern. There is more detail about this in the
1020 section on pcre_exec() options below.
1021
1022 The choice of newline convention does not affect the interpretation of
1023 the \n or \r escape sequences, nor does it affect what \R matches,
1024 which is controlled in a similar way, but by separate options.
1025
1026
1027 MULTITHREADING
1028
1029 The PCRE functions can be used in multi-threading applications, with
1030 the proviso that the memory management functions pointed to by
1031 pcre_malloc, pcre_free, pcre_stack_malloc, and pcre_stack_free, and the
1032 callout function pointed to by pcre_callout, are shared by all threads.
1033
1034 The compiled form of a regular expression is not altered during match-
1035 ing, so the same compiled pattern can safely be used by several threads
1036 at once.
1037
1038
1039 SAVING PRECOMPILED PATTERNS FOR LATER USE
1040
1041 The compiled form of a regular expression can be saved and re-used at a
1042 later time, possibly by a different program, and even on a host other
1043 than the one on which it was compiled. Details are given in the
1044 pcreprecompile documentation. However, compiling a regular expression
1045 with one version of PCRE for use with a different version is not guar-
1046 anteed to work and may cause crashes.
1047
1048
1049 CHECKING BUILD-TIME OPTIONS
1050
1051 int pcre_config(int what, void *where);
1052
1053 The function pcre_config() makes it possible for a PCRE client to dis-
1054 cover which optional features have been compiled into the PCRE library.
1055 The pcrebuild documentation has more details about these optional fea-
1056 tures.
1057
1058 The first argument for pcre_config() is an integer, specifying which
1059 information is required; the second argument is a pointer to a variable
1060 into which the information is placed. The following information is
1061 available:
1062
1063 PCRE_CONFIG_UTF8
1064
1065 The output is an integer that is set to one if UTF-8 support is avail-
1066 able; otherwise it is set to zero.
1067
1068 PCRE_CONFIG_UNICODE_PROPERTIES
1069
1070 The output is an integer that is set to one if support for Unicode
1071 character properties is available; otherwise it is set to zero.
1072
1073 PCRE_CONFIG_NEWLINE
1074
1075 The output is an integer whose value specifies the default character
1076 sequence that is recognized as meaning "newline". The four values that
1077 are supported are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF,
1078 and -1 for ANY. Though they are derived from ASCII, the same values
1079 are returned in EBCDIC environments. The default should normally corre-
1080 spond to the standard sequence for your operating system.
1081
1082 PCRE_CONFIG_BSR
1083
1084 The output is an integer whose value indicates what character sequences
1085 the \R escape sequence matches by default. A value of 0 means that \R
1086 matches any Unicode line ending sequence; a value of 1 means that \R
1087 matches only CR, LF, or CRLF. The default can be overridden when a pat-
1088 tern is compiled or matched.
1089
1090 PCRE_CONFIG_LINK_SIZE
1091
1092 The output is an integer that contains the number of bytes used for
1093 internal linkage in compiled regular expressions. The value is 2, 3, or
1094 4. Larger values allow larger regular expressions to be compiled, at
1095 the expense of slower matching. The default value of 2 is sufficient
1096 for all but the most massive patterns, since it allows the compiled
1097 pattern to be up to 64K in size.
1098
1099 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
1100
1101 The output is an integer that contains the threshold above which the
1102 POSIX interface uses malloc() for output vectors. Further details are
1103 given in the pcreposix documentation.
1104
1105 PCRE_CONFIG_MATCH_LIMIT
1106
1107 The output is a long integer that gives the default limit for the num-
1108 ber of internal matching function calls in a pcre_exec() execution.
1109 Further details are given with pcre_exec() below.
1110
1111 PCRE_CONFIG_MATCH_LIMIT_RECURSION
1112
1113 The output is a long integer that gives the default limit for the depth
1114 of recursion when calling the internal matching function in a
1115 pcre_exec() execution. Further details are given with pcre_exec()
1116 below.
1117
1118 PCRE_CONFIG_STACKRECURSE
1119
1120 The output is an integer that is set to one if internal recursion when
1121 running pcre_exec() is implemented by recursive function calls that use
1122 the stack to remember their state. This is the usual way that PCRE is
1123 compiled. The output is zero if PCRE was compiled to use blocks of data
1124 on the heap instead of recursive function calls. In this case,
1125 pcre_stack_malloc and pcre_stack_free are called to manage memory
1126 blocks on the heap, thus avoiding the use of the stack.
1127
1128
1129 COMPILING A PATTERN
1130
1131 pcre *pcre_compile(const char *pattern, int options,
1132 const char **errptr, int *erroffset,
1133 const unsigned char *tableptr);
1134
1135 pcre *pcre_compile2(const char *pattern, int options,
1136 int *errorcodeptr,
1137 const char **errptr, int *erroffset,
1138 const unsigned char *tableptr);
1139
1140 Either of the functions pcre_compile() or pcre_compile2() can be called
1141 to compile a pattern into an internal form. The only difference between
1142 the two interfaces is that pcre_compile2() has an additional argument,
1143 errorcodeptr, via which a numerical error code can be returned. To
1144 avoid too much repetition, we refer just to pcre_compile() below, but
1145 the information applies equally to pcre_compile2().
1146
1147 The pattern is a C string terminated by a binary zero, and is passed in
1148 the pattern argument. A pointer to a single block of memory that is
1149 obtained via pcre_malloc is returned. This contains the compiled code
1150 and related data. The pcre type is defined for the returned block; this
1151 is a typedef for a structure whose contents are not externally defined.
1152 It is up to the caller to free the memory (via pcre_free) when it is no
1153 longer required.
1154
1155 Although the compiled code of a PCRE regex is relocatable, that is, it
1156 does not depend on memory location, the complete pcre data block is not
1157 fully relocatable, because it may contain a copy of the tableptr argu-
1158 ment, which is an address (see below).
1159
1160 The options argument contains various bit settings that affect the com-
1161 pilation. It should be zero if no options are required. The available
1162 options are described below. Some of them (in particular, those that
1163 are compatible with Perl, but some others as well) can also be set and
1164 unset from within the pattern (see the detailed description in the
1165 pcrepattern documentation). For those options that can be different in
1166 different parts of the pattern, the contents of the options argument
1167 specifies their settings at the start of compilation and execution. The
1168 PCRE_ANCHORED, PCRE_BSR_xxx, and PCRE_NEWLINE_xxx options can be set at
1169 the time of matching as well as at compile time.
1170
1171 If errptr is NULL, pcre_compile() returns NULL immediately. Otherwise,
1172 if compilation of a pattern fails, pcre_compile() returns NULL, and
1173 sets the variable pointed to by errptr to point to a textual error mes-
1174 sage. This is a static string that is part of the library. You must not
1175 try to free it. The offset from the start of the pattern to the byte
1176 that was being processed when the error was discovered is placed in the
1177 variable pointed to by erroffset, which must not be NULL. If it is, an
1178 immediate error is given. Some errors are not detected until checks are
1179 carried out when the whole pattern has been scanned; in this case the
1180 offset is set to the end of the pattern.
1181
1182 Note that the offset is in bytes, not characters, even in UTF-8 mode.
1183 It may point into the middle of a UTF-8 character (for example, when
1184 PCRE_ERROR_BADUTF8 is returned for an invalid UTF-8 string).
1185
1186 If pcre_compile2() is used instead of pcre_compile(), and the error-
1187 codeptr argument is not NULL, a non-zero error code number is returned
1188 via this argument in the event of an error. This is in addition to the
1189 textual error message. Error codes and messages are listed below.
1190
1191 If the final argument, tableptr, is NULL, PCRE uses a default set of
1192 character tables that are built when PCRE is compiled, using the
1193 default C locale. Otherwise, tableptr must be an address that is the
1194 result of a call to pcre_maketables(). This value is stored with the
1195 compiled pattern, and used again by pcre_exec(), unless another table
1196 pointer is passed to it. For more discussion, see the section on locale
1197 support below.
1198
1199 This code fragment shows a typical straightforward call to pcre_com-
1200 pile():
1201
1202 pcre *re;
1203 const char *error;
1204 int erroffset;
1205 re = pcre_compile(
1206 "^A.*Z", /* the pattern */
1207 0, /* default options */
1208 &error, /* for error message */
1209 &erroffset, /* for error offset */
1210 NULL); /* use default character tables */
1211
1212 The following names for option bits are defined in the pcre.h header
1213 file:
1214
1215 PCRE_ANCHORED
1216
1217 If this bit is set, the pattern is forced to be "anchored", that is, it
1218 is constrained to match only at the first matching point in the string
1219 that is being searched (the "subject string"). This effect can also be
1220 achieved by appropriate constructs in the pattern itself, which is the
1221 only way to do it in Perl.
1222
1223 PCRE_AUTO_CALLOUT
1224
1225 If this bit is set, pcre_compile() automatically inserts callout items,
1226 all with number 255, before each pattern item. For discussion of the
1227 callout facility, see the pcrecallout documentation.
1228
1229 PCRE_BSR_ANYCRLF
1230 PCRE_BSR_UNICODE
1231
1232 These options (which are mutually exclusive) control what the \R escape
1233 sequence matches. The choice is either to match only CR, LF, or CRLF,
1234 or to match any Unicode newline sequence. The default is specified when
1235 PCRE is built. It can be overridden from within the pattern, or by set-
1236 ting an option when a compiled pattern is matched.
1237
1238 PCRE_CASELESS
1239
1240 If this bit is set, letters in the pattern match both upper and lower
1241 case letters. It is equivalent to Perl's /i option, and it can be
1242 changed within a pattern by a (?i) option setting. In UTF-8 mode, PCRE
1243 always understands the concept of case for characters whose values are
1244 less than 128, so caseless matching is always possible. For characters
1245 with higher values, the concept of case is supported if PCRE is com-
1246 piled with Unicode property support, but not otherwise. If you want to
1247 use caseless matching for characters 128 and above, you must ensure
1248 that PCRE is compiled with Unicode property support as well as with
1249 UTF-8 support.
1250
1251 PCRE_DOLLAR_ENDONLY
1252
1253 If this bit is set, a dollar metacharacter in the pattern matches only
1254 at the end of the subject string. Without this option, a dollar also
1255 matches immediately before a newline at the end of the string (but not
1256 before any other newlines). The PCRE_DOLLAR_ENDONLY option is ignored
1257 if PCRE_MULTILINE is set. There is no equivalent to this option in
1258 Perl, and no way to set it within a pattern.
1259
1260 PCRE_DOTALL
1261
1262 If this bit is set, a dot metacharacter in the pattern matches a char-
1263 acter of any value, including one that indicates a newline. However, it
1264 only ever matches one character, even if newlines are coded as CRLF.
1265 Without this option, a dot does not match when the current position is
1266 at a newline. This option is equivalent to Perl's /s option, and it can
1267 be changed within a pattern by a (?s) option setting. A negative class
1268 such as [^a] always matches newline characters, independent of the set-
1269 ting of this option.
1270
1271 PCRE_DUPNAMES
1272
1273 If this bit is set, names used to identify capturing subpatterns need
1274 not be unique. This can be helpful for certain types of pattern when it
1275 is known that only one instance of the named subpattern can ever be
1276 matched. There are more details of named subpatterns below; see also
1277 the pcrepattern documentation.
1278
1279 PCRE_EXTENDED
1280
1281 If this bit is set, whitespace data characters in the pattern are
1282 totally ignored except when escaped or inside a character class. White-
1283 space does not include the VT character (code 11). In addition, charac-
1284 ters between an unescaped # outside a character class and the next new-
1285 line, inclusive, are also ignored. This is equivalent to Perl's /x
1286 option, and it can be changed within a pattern by a (?x) option set-
1287 ting.
1288
1289 Which characters are interpreted as newlines is controlled by the
1290 options passed to pcre_compile() or by a special sequence at the start
1291 of the pattern, as described in the section entitled "Newline conven-
1292 tions" in the pcrepattern documentation. Note that the end of this type
1293 of comment is a literal newline sequence in the pattern; escape
1294 sequences that happen to represent a newline do not count.
1295
1296 This option makes it possible to include comments inside complicated
1297 patterns. Note, however, that this applies only to data characters.
1298 Whitespace characters may never appear within special character
1299 sequences in a pattern, for example within the sequence (?( that intro-
1300 duces a conditional subpattern.
1301
1302 PCRE_EXTRA
1303
1304 This option was invented in order to turn on additional functionality
1305 of PCRE that is incompatible with Perl, but it is currently of very
1306 little use. When set, any backslash in a pattern that is followed by a
1307 letter that has no special meaning causes an error, thus reserving
1308 these combinations for future expansion. By default, as in Perl, a
1309 backslash followed by a letter with no special meaning is treated as a
1310 literal. (Perl can, however, be persuaded to give an error for this, by
1311 running it with the -w option.) There are at present no other features
1312 controlled by this option. It can also be set by a (?X) option setting
1313 within a pattern.
1314
1315 PCRE_FIRSTLINE
1316
1317 If this option is set, an unanchored pattern is required to match
1318 before or at the first newline in the subject string, though the
1319 matched text may continue over the newline.
1320
1321 PCRE_JAVASCRIPT_COMPAT
1322
1323 If this option is set, PCRE's behaviour is changed in some ways so that
1324 it is compatible with JavaScript rather than Perl. The changes are as
1325 follows:
1326
1327 (1) A lone closing square bracket in a pattern causes a compile-time
1328 error, because this is illegal in JavaScript (by default it is treated
1329 as a data character). Thus, the pattern AB]CD becomes illegal when this
1330 option is set.
1331
1332 (2) At run time, a back reference to an unset subpattern group matches
1333 an empty string (by default this causes the current matching alterna-
1334 tive to fail). A pattern such as (\1)(a) succeeds when this option is
1335 set (assuming it can find an "a" in the subject), whereas it fails by
1336 default, for Perl compatibility.
1337
1338 PCRE_MULTILINE
1339
1340 By default, PCRE treats the subject string as consisting of a single
1341 line of characters (even if it actually contains newlines). The "start
1342 of line" metacharacter (^) matches only at the start of the string,
1343 while the "end of line" metacharacter ($) matches only at the end of
1344 the string, or before a terminating newline (unless PCRE_DOLLAR_ENDONLY
1345 is set). This is the same as Perl.
1346
1347 When PCRE_MULTILINE it is set, the "start of line" and "end of line"
1348 constructs match immediately following or immediately before internal
1349 newlines in the subject string, respectively, as well as at the very
1350 start and end. This is equivalent to Perl's /m option, and it can be
1351 changed within a pattern by a (?m) option setting. If there are no new-
1352 lines in a subject string, or no occurrences of ^ or $ in a pattern,
1353 setting PCRE_MULTILINE has no effect.
1354
1355 PCRE_NEWLINE_CR
1356 PCRE_NEWLINE_LF
1357 PCRE_NEWLINE_CRLF
1358 PCRE_NEWLINE_ANYCRLF
1359 PCRE_NEWLINE_ANY
1360
1361 These options override the default newline definition that was chosen
1362 when PCRE was built. Setting the first or the second specifies that a
1363 newline is indicated by a single character (CR or LF, respectively).
1364 Setting PCRE_NEWLINE_CRLF specifies that a newline is indicated by the
1365 two-character CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies
1366 that any of the three preceding sequences should be recognized. Setting
1367 PCRE_NEWLINE_ANY specifies that any Unicode newline sequence should be
1368 recognized. The Unicode newline sequences are the three just mentioned,
1369 plus the single characters VT (vertical tab, U+000B), FF (formfeed,
1370 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
1371 (paragraph separator, U+2029). The last two are recognized only in
1372 UTF-8 mode.
1373
1374 The newline setting in the options word uses three bits that are
1375 treated as a number, giving eight possibilities. Currently only six are
1376 used (default plus the five values above). This means that if you set
1377 more than one newline option, the combination may or may not be sensi-
1378 ble. For example, PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to
1379 PCRE_NEWLINE_CRLF, but other combinations may yield unused numbers and
1380 cause an error.
1381
1382 The only time that a line break in a pattern is specially recognized
1383 when compiling is when PCRE_EXTENDED is set. CR and LF are whitespace
1384 characters, and so are ignored in this mode. Also, an unescaped # out-
1385 side a character class indicates a comment that lasts until after the
1386 next line break sequence. In other circumstances, line break sequences
1387 in patterns are treated as literal data.
1388
1389 The newline option that is set at compile time becomes the default that
1390 is used for pcre_exec() and pcre_dfa_exec(), but it can be overridden.
1391
1392 PCRE_NO_AUTO_CAPTURE
1393
1394 If this option is set, it disables the use of numbered capturing paren-
1395 theses in the pattern. Any opening parenthesis that is not followed by
1396 ? behaves as if it were followed by ?: but named parentheses can still
1397 be used for capturing (and they acquire numbers in the usual way).
1398 There is no equivalent of this option in Perl.
1399
1400 PCRE_UCP
1401
1402 This option changes the way PCRE processes \B, \b, \D, \d, \S, \s, \W,
1403 \w, and some of the POSIX character classes. By default, only ASCII
1404 characters are recognized, but if PCRE_UCP is set, Unicode properties
1405 are used instead to classify characters. More details are given in the
1406 section on generic character types in the pcrepattern page. If you set
1407 PCRE_UCP, matching one of the items it affects takes much longer. The
1408 option is available only if PCRE has been compiled with Unicode prop-
1409 erty support.
1410
1411 PCRE_UNGREEDY
1412
1413 This option inverts the "greediness" of the quantifiers so that they
1414 are not greedy by default, but become greedy if followed by "?". It is
1415 not compatible with Perl. It can also be set by a (?U) option setting
1416 within the pattern.
1417
1418 PCRE_UTF8
1419
1420 This option causes PCRE to regard both the pattern and the subject as
1421 strings of UTF-8 characters instead of single-byte character strings.
1422 However, it is available only when PCRE is built to include UTF-8 sup-
1423 port. If not, the use of this option provokes an error. Details of how
1424 this option changes the behaviour of PCRE are given in the section on
1425 UTF-8 support in the main pcre page.
1426
1427 PCRE_NO_UTF8_CHECK
1428
1429 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
1430 automatically checked. There is a discussion about the validity of
1431 UTF-8 strings in the main pcre page. If an invalid UTF-8 sequence of
1432 bytes is found, pcre_compile() returns an error. If you already know
1433 that your pattern is valid, and you want to skip this check for perfor-
1434 mance reasons, you can set the PCRE_NO_UTF8_CHECK option. When it is
1435 set, the effect of passing an invalid UTF-8 string as a pattern is
1436 undefined. It may cause your program to crash. Note that this option
1437 can also be passed to pcre_exec() and pcre_dfa_exec(), to suppress the
1438 UTF-8 validity checking of subject strings.
1439
1440
1441 COMPILATION ERROR CODES
1442
1443 The following table lists the error codes than may be returned by
1444 pcre_compile2(), along with the error messages that may be returned by
1445 both compiling functions. As PCRE has developed, some error codes have
1446 fallen out of use. To avoid confusion, they have not been re-used.
1447
1448 0 no error
1449 1 \ at end of pattern
1450 2 \c at end of pattern
1451 3 unrecognized character follows \
1452 4 numbers out of order in {} quantifier
1453 5 number too big in {} quantifier
1454 6 missing terminating ] for character class
1455 7 invalid escape sequence in character class
1456 8 range out of order in character class
1457 9 nothing to repeat
1458 10 [this code is not in use]
1459 11 internal error: unexpected repeat
1460 12 unrecognized character after (? or (?-
1461 13 POSIX named classes are supported only within a class
1462 14 missing )
1463 15 reference to non-existent subpattern
1464 16 erroffset passed as NULL
1465 17 unknown option bit(s) set
1466 18 missing ) after comment
1467 19 [this code is not in use]
1468 20 regular expression is too large
1469 21 failed to get memory
1470 22 unmatched parentheses
1471 23 internal error: code overflow
1472 24 unrecognized character after (?<
1473 25 lookbehind assertion is not fixed length
1474 26 malformed number or name after (?(
1475 27 conditional group contains more than two branches
1476 28 assertion expected after (?(
1477 29 (?R or (?[+-]digits must be followed by )
1478 30 unknown POSIX class name
1479 31 POSIX collating elements are not supported
1480 32 this version of PCRE is not compiled with PCRE_UTF8 support
1481 33 [this code is not in use]
1482 34 character value in \x{...} sequence is too large
1483 35 invalid condition (?(0)
1484 36 \C not allowed in lookbehind assertion
1485 37 PCRE does not support \L, \l, \N, \U, or \u
1486 38 number after (?C is > 255
1487 39 closing ) for (?C expected
1488 40 recursive call could loop indefinitely
1489 41 unrecognized character after (?P
1490 42 syntax error in subpattern name (missing terminator)
1491 43 two named subpatterns have the same name
1492 44 invalid UTF-8 string
1493 45 support for \P, \p, and \X has not been compiled
1494 46 malformed \P or \p sequence
1495 47 unknown property name after \P or \p
1496 48 subpattern name is too long (maximum 32 characters)
1497 49 too many named subpatterns (maximum 10000)
1498 50 [this code is not in use]
1499 51 octal value is greater than \377 (not in UTF-8 mode)
1500 52 internal error: overran compiling workspace
1501 53 internal error: previously-checked referenced subpattern
1502 not found
1503 54 DEFINE group contains more than one branch
1504 55 repeating a DEFINE group is not allowed
1505 56 inconsistent NEWLINE options
1506 57 \g is not followed by a braced, angle-bracketed, or quoted
1507 name/number or by a plain number
1508 58 a numbered reference must not be zero
1509 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
1510 60 (*VERB) not recognized
1511 61 number is too big
1512 62 subpattern name expected
1513 63 digit expected after (?+
1514 64 ] is an invalid data character in JavaScript compatibility mode
1515 65 different names for subpatterns of the same number are
1516 not allowed
1517 66 (*MARK) must have an argument
1518 67 this version of PCRE is not compiled with PCRE_UCP support
1519
1520 The numbers 32 and 10000 in errors 48 and 49 are defaults; different
1521 values may be used if the limits were changed when PCRE was built.
1522
1523
1524 STUDYING A PATTERN
1525
1526 pcre_extra *pcre_study(const pcre *code, int options
1527 const char **errptr);
1528
1529 If a compiled pattern is going to be used several times, it is worth
1530 spending more time analyzing it in order to speed up the time taken for
1531 matching. The function pcre_study() takes a pointer to a compiled pat-
1532 tern as its first argument. If studying the pattern produces additional
1533 information that will help speed up matching, pcre_study() returns a
1534 pointer to a pcre_extra block, in which the study_data field points to
1535 the results of the study.
1536
1537 The returned value from pcre_study() can be passed directly to
1538 pcre_exec() or pcre_dfa_exec(). However, a pcre_extra block also con-
1539 tains other fields that can be set by the caller before the block is
1540 passed; these are described below in the section on matching a pattern.
1541
1542 If studying the pattern does not produce any useful information,
1543 pcre_study() returns NULL. In that circumstance, if the calling program
1544 wants to pass any of the other fields to pcre_exec() or
1545 pcre_dfa_exec(), it must set up its own pcre_extra block.
1546
1547 The second argument of pcre_study() contains option bits. At present,
1548 no options are defined, and this argument should always be zero.
1549
1550 The third argument for pcre_study() is a pointer for an error message.
1551 If studying succeeds (even if no data is returned), the variable it
1552 points to is set to NULL. Otherwise it is set to point to a textual
1553 error message. This is a static string that is part of the library. You
1554 must not try to free it. You should test the error pointer for NULL
1555 after calling pcre_study(), to be sure that it has run successfully.
1556
1557 This is a typical call to pcre_study():
1558
1559 pcre_extra *pe;
1560 pe = pcre_study(
1561 re, /* result of pcre_compile() */
1562 0, /* no options exist */
1563 &error); /* set to NULL or points to a message */
1564
1565 Studying a pattern does two things: first, a lower bound for the length
1566 of subject string that is needed to match the pattern is computed. This
1567 does not mean that there are any strings of that length that match, but
1568 it does guarantee that no shorter strings match. The value is used by
1569 pcre_exec() and pcre_dfa_exec() to avoid wasting time by trying to
1570 match strings that are shorter than the lower bound. You can find out
1571 the value in a calling program via the pcre_fullinfo() function.
1572
1573 Studying a pattern is also useful for non-anchored patterns that do not
1574 have a single fixed starting character. A bitmap of possible starting
1575 bytes is created. This speeds up finding a position in the subject at
1576 which to start matching.
1577
1578 The two optimizations just described can be disabled by setting the
1579 PCRE_NO_START_OPTIMIZE option when calling pcre_exec() or
1580 pcre_dfa_exec(). You might want to do this if your pattern contains
1581 callouts or (*MARK), and you want to make use of these facilities in
1582 cases where matching fails. See the discussion of PCRE_NO_START_OPTI-
1583 MIZE below.
1584
1585
1586 LOCALE SUPPORT
1587
1588 PCRE handles caseless matching, and determines whether characters are
1589 letters, digits, or whatever, by reference to a set of tables, indexed
1590 by character value. When running in UTF-8 mode, this applies only to
1591 characters with codes less than 128. By default, higher-valued codes
1592 never match escapes such as \w or \d, but they can be tested with \p if
1593 PCRE is built with Unicode character property support. Alternatively,
1594 the PCRE_UCP option can be set at compile time; this causes \w and
1595 friends to use Unicode property support instead of built-in tables. The
1596 use of locales with Unicode is discouraged. If you are handling charac-
1597 ters with codes greater than 128, you should either use UTF-8 and Uni-
1598 code, or use locales, but not try to mix the two.
1599
1600 PCRE contains an internal set of tables that are used when the final
1601 argument of pcre_compile() is NULL. These are sufficient for many
1602 applications. Normally, the internal tables recognize only ASCII char-
1603 acters. However, when PCRE is built, it is possible to cause the inter-
1604 nal tables to be rebuilt in the default "C" locale of the local system,
1605 which may cause them to be different.
1606
1607 The internal tables can always be overridden by tables supplied by the
1608 application that calls PCRE. These may be created in a different locale
1609 from the default. As more and more applications change to using Uni-
1610 code, the need for this locale support is expected to die away.
1611
1612 External tables are built by calling the pcre_maketables() function,
1613 which has no arguments, in the relevant locale. The result can then be
1614 passed to pcre_compile() or pcre_exec() as often as necessary. For
1615 example, to build and use tables that are appropriate for the French
1616 locale (where accented characters with values greater than 128 are
1617 treated as letters), the following code could be used:
1618
1619 setlocale(LC_CTYPE, "fr_FR");
1620 tables = pcre_maketables();
1621 re = pcre_compile(..., tables);
1622
1623 The locale name "fr_FR" is used on Linux and other Unix-like systems;
1624 if you are using Windows, the name for the French locale is "french".
1625
1626 When pcre_maketables() runs, the tables are built in memory that is
1627 obtained via pcre_malloc. It is the caller's responsibility to ensure
1628 that the memory containing the tables remains available for as long as
1629 it is needed.
1630
1631 The pointer that is passed to pcre_compile() is saved with the compiled
1632 pattern, and the same tables are used via this pointer by pcre_study()
1633 and normally also by pcre_exec(). Thus, by default, for any single pat-
1634 tern, compilation, studying and matching all happen in the same locale,
1635 but different patterns can be compiled in different locales.
1636
1637 It is possible to pass a table pointer or NULL (indicating the use of
1638 the internal tables) to pcre_exec(). Although not intended for this
1639 purpose, this facility could be used to match a pattern in a different
1640 locale from the one in which it was compiled. Passing table pointers at
1641 run time is discussed below in the section on matching a pattern.
1642
1643
1644 INFORMATION ABOUT A PATTERN
1645
1646 int pcre_fullinfo(const pcre *code, const pcre_extra *extra,
1647 int what, void *where);
1648
1649 The pcre_fullinfo() function returns information about a compiled pat-
1650 tern. It replaces the obsolete pcre_info() function, which is neverthe-
1651 less retained for backwards compability (and is documented below).
1652
1653 The first argument for pcre_fullinfo() is a pointer to the compiled
1654 pattern. The second argument is the result of pcre_study(), or NULL if
1655 the pattern was not studied. The third argument specifies which piece
1656 of information is required, and the fourth argument is a pointer to a
1657 variable to receive the data. The yield of the function is zero for
1658 success, or one of the following negative numbers:
1659
1660 PCRE_ERROR_NULL the argument code was NULL
1661 the argument where was NULL
1662 PCRE_ERROR_BADMAGIC the "magic number" was not found
1663 PCRE_ERROR_BADOPTION the value of what was invalid
1664
1665 The "magic number" is placed at the start of each compiled pattern as
1666 an simple check against passing an arbitrary memory pointer. Here is a
1667 typical call of pcre_fullinfo(), to obtain the length of the compiled
1668 pattern:
1669
1670 int rc;
1671 size_t length;
1672 rc = pcre_fullinfo(
1673 re, /* result of pcre_compile() */
1674 pe, /* result of pcre_study(), or NULL */
1675 PCRE_INFO_SIZE, /* what is required */
1676 &length); /* where to put the data */
1677
1678 The possible values for the third argument are defined in pcre.h, and
1679 are as follows:
1680
1681 PCRE_INFO_BACKREFMAX
1682
1683 Return the number of the highest back reference in the pattern. The
1684 fourth argument should point to an int variable. Zero is returned if
1685 there are no back references.
1686
1687 PCRE_INFO_CAPTURECOUNT
1688
1689 Return the number of capturing subpatterns in the pattern. The fourth
1690 argument should point to an int variable.
1691
1692 PCRE_INFO_DEFAULT_TABLES
1693
1694 Return a pointer to the internal default character tables within PCRE.
1695 The fourth argument should point to an unsigned char * variable. This
1696 information call is provided for internal use by the pcre_study() func-
1697 tion. External callers can cause PCRE to use its internal tables by
1698 passing a NULL table pointer.
1699
1700 PCRE_INFO_FIRSTBYTE
1701
1702 Return information about the first byte of any matched string, for a
1703 non-anchored pattern. The fourth argument should point to an int vari-
1704 able. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name
1705 is still recognized for backwards compatibility.)
1706
1707 If there is a fixed first byte, for example, from a pattern such as
1708 (cat|cow|coyote), its value is returned. Otherwise, if either
1709
1710 (a) the pattern was compiled with the PCRE_MULTILINE option, and every
1711 branch starts with "^", or
1712
1713 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not
1714 set (if it were set, the pattern would be anchored),
1715
1716 -1 is returned, indicating that the pattern matches only at the start
1717 of a subject string or after any newline within the string. Otherwise
1718 -2 is returned. For anchored patterns, -2 is returned.
1719
1720 PCRE_INFO_FIRSTTABLE
1721
1722 If the pattern was studied, and this resulted in the construction of a
1723 256-bit table indicating a fixed set of bytes for the first byte in any
1724 matching string, a pointer to the table is returned. Otherwise NULL is
1725 returned. The fourth argument should point to an unsigned char * vari-
1726 able.
1727
1728 PCRE_INFO_HASCRORLF
1729
1730 Return 1 if the pattern contains any explicit matches for CR or LF
1731 characters, otherwise 0. The fourth argument should point to an int
1732 variable. An explicit match is either a literal CR or LF character, or
1733 \r or \n.
1734
1735 PCRE_INFO_JCHANGED
1736
1737 Return 1 if the (?J) or (?-J) option setting is used in the pattern,
1738 otherwise 0. The fourth argument should point to an int variable. (?J)
1739 and (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1740
1741 PCRE_INFO_LASTLITERAL
1742
1743 Return the value of the rightmost literal byte that must exist in any
1744 matched string, other than at its start, if such a byte has been
1745 recorded. The fourth argument should point to an int variable. If there
1746 is no such byte, -1 is returned. For anchored patterns, a last literal
1747 byte is recorded only if it follows something of variable length. For
1748 example, for the pattern /^a\d+z\d+/ the returned value is "z", but for
1749 /^a\dz\d/ the returned value is -1.
1750
1751 PCRE_INFO_MINLENGTH
1752
1753 If the pattern was studied and a minimum length for matching subject
1754 strings was computed, its value is returned. Otherwise the returned
1755 value is -1. The value is a number of characters, not bytes (this may
1756 be relevant in UTF-8 mode). The fourth argument should point to an int
1757 variable. A non-negative value is a lower bound to the length of any
1758 matching string. There may not be any strings of that length that do
1759 actually match, but every string that does match is at least that long.
1760
1761 PCRE_INFO_NAMECOUNT
1762 PCRE_INFO_NAMEENTRYSIZE
1763 PCRE_INFO_NAMETABLE
1764
1765 PCRE supports the use of named as well as numbered capturing parenthe-
1766 ses. The names are just an additional way of identifying the parenthe-
1767 ses, which still acquire numbers. Several convenience functions such as
1768 pcre_get_named_substring() are provided for extracting captured sub-
1769 strings by name. It is also possible to extract the data directly, by
1770 first converting the name to a number in order to access the correct
1771 pointers in the output vector (described with pcre_exec() below). To do
1772 the conversion, you need to use the name-to-number map, which is
1773 described by these three values.
1774
1775 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT
1776 gives the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size
1777 of each entry; both of these return an int value. The entry size
1778 depends on the length of the longest name. PCRE_INFO_NAMETABLE returns
1779 a pointer to the first entry of the table (a pointer to char). The
1780 first two bytes of each entry are the number of the capturing parenthe-
1781 sis, most significant byte first. The rest of the entry is the corre-
1782 sponding name, zero terminated.
1783
1784 The names are in alphabetical order. Duplicate names may appear if (?|
1785 is used to create multiple groups with the same number, as described in
1786 the section on duplicate subpattern numbers in the pcrepattern page.
1787 Duplicate names for subpatterns with different numbers are permitted
1788 only if PCRE_DUPNAMES is set. In all cases of duplicate names, they
1789 appear in the table in the order in which they were found in the pat-
1790 tern. In the absence of (?| this is the order of increasing number;
1791 when (?| is used this is not necessarily the case because later subpat-
1792 terns may have lower numbers.
1793
1794 As a simple example of the name/number table, consider the following
1795 pattern (assume PCRE_EXTENDED is set, so white space - including new-
1796 lines - is ignored):
1797
1798 (?<date> (?<year>(\d\d)?\d\d) -
1799 (?<month>\d\d) - (?<day>\d\d) )
1800
1801 There are four named subpatterns, so the table has four entries, and
1802 each entry in the table is eight bytes long. The table is as follows,
1803 with non-printing bytes shows in hexadecimal, and undefined bytes shown
1804 as ??:
1805
1806 00 01 d a t e 00 ??
1807 00 05 d a y 00 ?? ??
1808 00 04 m o n t h 00
1809 00 02 y e a r 00 ??
1810
1811 When writing code to extract data from named subpatterns using the
1812 name-to-number map, remember that the length of the entries is likely
1813 to be different for each compiled pattern.
1814
1815 PCRE_INFO_OKPARTIAL
1816
1817 Return 1 if the pattern can be used for partial matching with
1818 pcre_exec(), otherwise 0. The fourth argument should point to an int
1819 variable. From release 8.00, this always returns 1, because the
1820 restrictions that previously applied to partial matching have been
1821 lifted. The pcrepartial documentation gives details of partial match-
1822 ing.
1823
1824 PCRE_INFO_OPTIONS
1825
1826 Return a copy of the options with which the pattern was compiled. The
1827 fourth argument should point to an unsigned long int variable. These
1828 option bits are those specified in the call to pcre_compile(), modified
1829 by any top-level option settings at the start of the pattern itself. In
1830 other words, they are the options that will be in force when matching
1831 starts. For example, if the pattern /(?im)abc(?-i)d/ is compiled with
1832 the PCRE_EXTENDED option, the result is PCRE_CASELESS, PCRE_MULTILINE,
1833 and PCRE_EXTENDED.
1834
1835 A pattern is automatically anchored by PCRE if all of its top-level
1836 alternatives begin with one of the following:
1837
1838 ^ unless PCRE_MULTILINE is set
1839 \A always
1840 \G always
1841 .* if PCRE_DOTALL is set and there are no back
1842 references to the subpattern in which .* appears
1843
1844 For such patterns, the PCRE_ANCHORED bit is set in the options returned
1845 by pcre_fullinfo().
1846
1847 PCRE_INFO_SIZE
1848
1849 Return the size of the compiled pattern, that is, the value that was
1850 passed as the argument to pcre_malloc() when PCRE was getting memory in
1851 which to place the compiled data. The fourth argument should point to a
1852 size_t variable.
1853
1854 PCRE_INFO_STUDYSIZE
1855
1856 Return the size of the data block pointed to by the study_data field in
1857 a pcre_extra block. That is, it is the value that was passed to
1858 pcre_malloc() when PCRE was getting memory into which to place the data
1859 created by pcre_study(). If pcre_extra is NULL, or there is no study
1860 data, zero is returned. The fourth argument should point to a size_t
1861 variable.
1862
1863
1864 OBSOLETE INFO FUNCTION
1865
1866 int pcre_info(const pcre *code, int *optptr, int *firstcharptr);
1867
1868 The pcre_info() function is now obsolete because its interface is too
1869 restrictive to return all the available data about a compiled pattern.
1870 New programs should use pcre_fullinfo() instead. The yield of
1871 pcre_info() is the number of capturing subpatterns, or one of the fol-
1872 lowing negative numbers:
1873
1874 PCRE_ERROR_NULL the argument code was NULL
1875 PCRE_ERROR_BADMAGIC the "magic number" was not found
1876
1877 If the optptr argument is not NULL, a copy of the options with which
1878 the pattern was compiled is placed in the integer it points to (see
1879 PCRE_INFO_OPTIONS above).
1880
1881 If the pattern is not anchored and the firstcharptr argument is not
1882 NULL, it is used to pass back information about the first character of
1883 any matched string (see PCRE_INFO_FIRSTBYTE above).
1884
1885
1886 REFERENCE COUNTS
1887
1888 int pcre_refcount(pcre *code, int adjust);
1889
1890 The pcre_refcount() function is used to maintain a reference count in
1891 the data block that contains a compiled pattern. It is provided for the
1892 benefit of applications that operate in an object-oriented manner,
1893 where different parts of the application may be using the same compiled
1894 pattern, but you want to free the block when they are all done.
1895
1896 When a pattern is compiled, the reference count field is initialized to
1897 zero. It is changed only by calling this function, whose action is to
1898 add the adjust value (which may be positive or negative) to it. The
1899 yield of the function is the new value. However, the value of the count
1900 is constrained to lie between 0 and 65535, inclusive. If the new value
1901 is outside these limits, it is forced to the appropriate limit value.
1902
1903 Except when it is zero, the reference count is not correctly preserved
1904 if a pattern is compiled on one host and then transferred to a host
1905 whose byte-order is different. (This seems a highly unlikely scenario.)
1906
1907
1908 MATCHING A PATTERN: THE TRADITIONAL FUNCTION
1909
1910 int pcre_exec(const pcre *code, const pcre_extra *extra,
1911 const char *subject, int length, int startoffset,
1912 int options, int *ovector, int ovecsize);
1913
1914 The function pcre_exec() is called to match a subject string against a
1915 compiled pattern, which is passed in the code argument. If the pattern
1916 was studied, the result of the study should be passed in the extra
1917 argument. This function is the main matching facility of the library,
1918 and it operates in a Perl-like manner. For specialist use there is also
1919 an alternative matching function, which is described below in the sec-
1920 tion about the pcre_dfa_exec() function.
1921
1922 In most applications, the pattern will have been compiled (and option-
1923 ally studied) in the same process that calls pcre_exec(). However, it
1924 is possible to save compiled patterns and study data, and then use them
1925 later in different processes, possibly even on different hosts. For a
1926 discussion about this, see the pcreprecompile documentation.
1927
1928 Here is an example of a simple call to pcre_exec():
1929
1930 int rc;
1931 int ovector[30];
1932 rc = pcre_exec(
1933 re, /* result of pcre_compile() */
1934 NULL, /* we didn't study the pattern */
1935 "some string", /* the subject string */
1936 11, /* the length of the subject string */
1937 0, /* start at offset 0 in the subject */
1938 0, /* default options */
1939 ovector, /* vector of integers for substring information */
1940 30); /* number of elements (NOT size in bytes) */
1941
1942 Extra data for pcre_exec()
1943
1944 If the extra argument is not NULL, it must point to a pcre_extra data
1945 block. The pcre_study() function returns such a block (when it doesn't
1946 return NULL), but you can also create one for yourself, and pass addi-
1947 tional information in it. The pcre_extra block contains the following
1948 fields (not necessarily in this order):
1949
1950 unsigned long int flags;
1951 void *study_data;
1952 unsigned long int match_limit;
1953 unsigned long int match_limit_recursion;
1954 void *callout_data;
1955 const unsigned char *tables;
1956 unsigned char **mark;
1957
1958 The flags field is a bitmap that specifies which of the other fields
1959 are set. The flag bits are:
1960
1961 PCRE_EXTRA_STUDY_DATA
1962 PCRE_EXTRA_MATCH_LIMIT
1963 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1964 PCRE_EXTRA_CALLOUT_DATA
1965 PCRE_EXTRA_TABLES
1966 PCRE_EXTRA_MARK
1967
1968 Other flag bits should be set to zero. The study_data field is set in
1969 the pcre_extra block that is returned by pcre_study(), together with
1970 the appropriate flag bit. You should not set this yourself, but you may
1971 add to the block by setting the other fields and their corresponding
1972 flag bits.
1973
1974 The match_limit field provides a means of preventing PCRE from using up
1975 a vast amount of resources when running patterns that are not going to
1976 match, but which have a very large number of possibilities in their
1977 search trees. The classic example is a pattern that uses nested unlim-
1978 ited repeats.
1979
1980 Internally, PCRE uses a function called match() which it calls repeat-
1981 edly (sometimes recursively). The limit set by match_limit is imposed
1982 on the number of times this function is called during a match, which
1983 has the effect of limiting the amount of backtracking that can take
1984 place. For patterns that are not anchored, the count restarts from zero
1985 for each position in the subject string.
1986
1987 The default value for the limit can be set when PCRE is built; the
1988 default default is 10 million, which handles all but the most extreme
1989 cases. You can override the default by suppling pcre_exec() with a
1990 pcre_extra block in which match_limit is set, and
1991 PCRE_EXTRA_MATCH_LIMIT is set in the flags field. If the limit is
1992 exceeded, pcre_exec() returns PCRE_ERROR_MATCHLIMIT.
1993
1994 The match_limit_recursion field is similar to match_limit, but instead
1995 of limiting the total number of times that match() is called, it limits
1996 the depth of recursion. The recursion depth is a smaller number than
1997 the total number of calls, because not all calls to match() are recur-
1998 sive. This limit is of use only if it is set smaller than match_limit.
1999
2000 Limiting the recursion depth limits the amount of stack that can be
2001 used, or, when PCRE has been compiled to use memory on the heap instead
2002 of the stack, the amount of heap memory that can be used.
2003
2004 The default value for match_limit_recursion can be set when PCRE is
2005 built; the default default is the same value as the default for
2006 match_limit. You can override the default by suppling pcre_exec() with
2007 a pcre_extra block in which match_limit_recursion is set, and
2008 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the flags field. If the
2009 limit is exceeded, pcre_exec() returns PCRE_ERROR_RECURSIONLIMIT.
2010
2011 The callout_data field is used in conjunction with the "callout" fea-
2012 ture, and is described in the pcrecallout documentation.
2013
2014 The tables field is used to pass a character tables pointer to
2015 pcre_exec(); this overrides the value that is stored with the compiled
2016 pattern. A non-NULL value is stored with the compiled pattern only if
2017 custom tables were supplied to pcre_compile() via its tableptr argu-
2018 ment. If NULL is passed to pcre_exec() using this mechanism, it forces
2019 PCRE's internal tables to be used. This facility is helpful when re-
2020 using patterns that have been saved after compiling with an external
2021 set of tables, because the external tables might be at a different
2022 address when pcre_exec() is called. See the pcreprecompile documenta-
2023 tion for a discussion of saving compiled patterns for later use.
2024
2025 If PCRE_EXTRA_MARK is set in the flags field, the mark field must be
2026 set to point to a char * variable. If the pattern contains any back-
2027 tracking control verbs such as (*MARK:NAME), and the execution ends up
2028 with a name to pass back, a pointer to the name string (zero termi-
2029 nated) is placed in the variable pointed to by the mark field. The
2030 names are within the compiled pattern; if you wish to retain such a
2031 name you must copy it before freeing the memory of a compiled pattern.
2032 If there is no name to pass back, the variable pointed to by the mark
2033 field set to NULL. For details of the backtracking control verbs, see
2034 the section entitled "Backtracking control" in the pcrepattern documen-
2035 tation.
2036
2037 Option bits for pcre_exec()
2038
2039 The unused bits of the options argument for pcre_exec() must be zero.
2040 The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_xxx,
2041 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2042 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_SOFT, and
2043 PCRE_PARTIAL_HARD.
2044
2045 PCRE_ANCHORED
2046
2047 The PCRE_ANCHORED option limits pcre_exec() to matching at the first
2048 matching position. If a pattern was compiled with PCRE_ANCHORED, or
2049 turned out to be anchored by virtue of its contents, it cannot be made
2050 unachored at matching time.
2051
2052 PCRE_BSR_ANYCRLF
2053 PCRE_BSR_UNICODE
2054
2055 These options (which are mutually exclusive) control what the \R escape
2056 sequence matches. The choice is either to match only CR, LF, or CRLF,
2057 or to match any Unicode newline sequence. These options override the
2058 choice that was made or defaulted when the pattern was compiled.
2059
2060 PCRE_NEWLINE_CR
2061 PCRE_NEWLINE_LF
2062 PCRE_NEWLINE_CRLF
2063 PCRE_NEWLINE_ANYCRLF
2064 PCRE_NEWLINE_ANY
2065
2066 These options override the newline definition that was chosen or
2067 defaulted when the pattern was compiled. For details, see the descrip-
2068 tion of pcre_compile() above. During matching, the newline choice
2069 affects the behaviour of the dot, circumflex, and dollar metacharac-
2070 ters. It may also alter the way the match position is advanced after a
2071 match failure for an unanchored pattern.
2072
2073 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is
2074 set, and a match attempt for an unanchored pattern fails when the cur-
2075 rent position is at a CRLF sequence, and the pattern contains no
2076 explicit matches for CR or LF characters, the match position is
2077 advanced by two characters instead of one, in other words, to after the
2078 CRLF.
2079
2080 The above rule is a compromise that makes the most common cases work as
2081 expected. For example, if the pattern is .+A (and the PCRE_DOTALL
2082 option is not set), it does not match the string "\r\nA" because, after
2083 failing at the start, it skips both the CR and the LF before retrying.
2084 However, the pattern [\r\n]A does match that string, because it con-
2085 tains an explicit CR or LF reference, and so advances only by one char-
2086 acter after the first failure.
2087
2088 An explicit match for CR of LF is either a literal appearance of one of
2089 those characters, or one of the \r or \n escape sequences. Implicit
2090 matches such as [^X] do not count, nor does \s (which includes CR and
2091 LF in the characters that it matches).
2092
2093 Notwithstanding the above, anomalous effects may still occur when CRLF
2094 is a valid newline sequence and explicit \r or \n escapes appear in the
2095 pattern.
2096
2097 PCRE_NOTBOL
2098
2099 This option specifies that first character of the subject string is not
2100 the beginning of a line, so the circumflex metacharacter should not
2101 match before it. Setting this without PCRE_MULTILINE (at compile time)
2102 causes circumflex never to match. This option affects only the behav-
2103 iour of the circumflex metacharacter. It does not affect \A.
2104
2105 PCRE_NOTEOL
2106
2107 This option specifies that the end of the subject string is not the end
2108 of a line, so the dollar metacharacter should not match it nor (except
2109 in multiline mode) a newline immediately before it. Setting this with-
2110 out PCRE_MULTILINE (at compile time) causes dollar never to match. This
2111 option affects only the behaviour of the dollar metacharacter. It does
2112 not affect \Z or \z.
2113
2114 PCRE_NOTEMPTY
2115
2116 An empty string is not considered to be a valid match if this option is
2117 set. If there are alternatives in the pattern, they are tried. If all
2118 the alternatives match the empty string, the entire match fails. For
2119 example, if the pattern
2120
2121 a?b?
2122
2123 is applied to a string not beginning with "a" or "b", it matches an
2124 empty string at the start of the subject. With PCRE_NOTEMPTY set, this
2125 match is not valid, so PCRE searches further into the string for occur-
2126 rences of "a" or "b".
2127
2128 PCRE_NOTEMPTY_ATSTART
2129
2130 This is like PCRE_NOTEMPTY, except that an empty string match that is
2131 not at the start of the subject is permitted. If the pattern is
2132 anchored, such a match can occur only if the pattern contains \K.
2133
2134 Perl has no direct equivalent of PCRE_NOTEMPTY or
2135 PCRE_NOTEMPTY_ATSTART, but it does make a special case of a pattern
2136 match of the empty string within its split() function, and when using
2137 the /g modifier. It is possible to emulate Perl's behaviour after
2138 matching a null string by first trying the match again at the same off-
2139 set with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then if that
2140 fails, by advancing the starting offset (see below) and trying an ordi-
2141 nary match again. There is some code that demonstrates how to do this
2142 in the pcredemo sample program. In the most general case, you have to
2143 check to see if the newline convention recognizes CRLF as a newline,
2144 and if so, and the current character is CR followed by LF, advance the
2145 starting offset by two characters instead of one.
2146
2147 PCRE_NO_START_OPTIMIZE
2148
2149 There are a number of optimizations that pcre_exec() uses at the start
2150 of a match, in order to speed up the process. For example, if it is
2151 known that an unanchored match must start with a specific character, it
2152 searches the subject for that character, and fails immediately if it
2153 cannot find it, without actually running the main matching function.
2154 This means that a special item such as (*COMMIT) at the start of a pat-
2155 tern is not considered until after a suitable starting point for the
2156 match has been found. When callouts or (*MARK) items are in use, these
2157 "start-up" optimizations can cause them to be skipped if the pattern is
2158 never actually used. The start-up optimizations are in effect a pre-
2159 scan of the subject that takes place before the pattern is run.
2160
2161 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations,
2162 possibly causing performance to suffer, but ensuring that in cases
2163 where the result is "no match", the callouts do occur, and that items
2164 such as (*COMMIT) and (*MARK) are considered at every possible starting
2165 position in the subject string. Setting PCRE_NO_START_OPTIMIZE can
2166 change the outcome of a matching operation. Consider the pattern
2167
2168 (*COMMIT)ABC
2169
2170 When this is compiled, PCRE records the fact that a match must start
2171 with the character "A". Suppose the subject string is "DEFABC". The
2172 start-up optimization scans along the subject, finds "A" and runs the
2173 first match attempt from there. The (*COMMIT) item means that the pat-
2174 tern must match the current starting position, which in this case, it
2175 does. However, if the same match is run with PCRE_NO_START_OPTIMIZE
2176 set, the initial scan along the subject string does not happen. The
2177 first match attempt is run starting from "D" and when this fails,
2178 (*COMMIT) prevents any further matches being tried, so the overall
2179 result is "no match". If the pattern is studied, more start-up opti-
2180 mizations may be used. For example, a minimum length for the subject
2181 may be recorded. Consider the pattern
2182
2183 (*MARK:A)(X|Y)
2184
2185 The minimum length for a match is one character. If the subject is
2186 "ABC", there will be attempts to match "ABC", "BC", "C", and then
2187 finally an empty string. If the pattern is studied, the final attempt
2188 does not take place, because PCRE knows that the subject is too short,
2189 and so the (*MARK) is never encountered. In this case, studying the
2190 pattern does not affect the overall match result, which is still "no
2191 match", but it does affect the auxiliary information that is returned.
2192
2193 PCRE_NO_UTF8_CHECK
2194
2195 When PCRE_UTF8 is set at compile time, the validity of the subject as a
2196 UTF-8 string is automatically checked when pcre_exec() is subsequently
2197 called. The value of startoffset is also checked to ensure that it
2198 points to the start of a UTF-8 character. There is a discussion about
2199 the validity of UTF-8 strings in the section on UTF-8 support in the
2200 main pcre page. If an invalid UTF-8 sequence of bytes is found,
2201 pcre_exec() returns the error PCRE_ERROR_BADUTF8 or, if PCRE_PAR-
2202 TIAL_HARD is set and the problem is a truncated UTF-8 character at the
2203 end of the subject, PCRE_ERROR_SHORTUTF8. If startoffset contains a
2204 value that does not point to the start of a UTF-8 character (or to the
2205 end of the subject), PCRE_ERROR_BADUTF8_OFFSET is returned.
2206
2207 If you already know that your subject is valid, and you want to skip
2208 these checks for performance reasons, you can set the
2209 PCRE_NO_UTF8_CHECK option when calling pcre_exec(). You might want to
2210 do this for the second and subsequent calls to pcre_exec() if you are
2211 making repeated calls to find all the matches in a single subject
2212 string. However, you should be sure that the value of startoffset
2213 points to the start of a UTF-8 character (or the end of the subject).
2214 When PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8
2215 string as a subject or an invalid value of startoffset is undefined.
2216 Your program may crash.
2217
2218 PCRE_PARTIAL_HARD
2219 PCRE_PARTIAL_SOFT
2220
2221 These options turn on the partial matching feature. For backwards com-
2222 patibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial
2223 match occurs if the end of the subject string is reached successfully,
2224 but there are not enough subject characters to complete the match. If
2225 this happens when PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set,
2226 matching continues by testing any remaining alternatives. Only if no
2227 complete match can be found is PCRE_ERROR_PARTIAL returned instead of
2228 PCRE_ERROR_NOMATCH. In other words, PCRE_PARTIAL_SOFT says that the
2229 caller is prepared to handle a partial match, but only if no complete
2230 match can be found.
2231
2232 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this
2233 case, if a partial match is found, pcre_exec() immediately returns
2234 PCRE_ERROR_PARTIAL, without considering any other alternatives. In
2235 other words, when PCRE_PARTIAL_HARD is set, a partial match is consid-
2236 ered to be more important that an alternative complete match.
2237
2238 In both cases, the portion of the string that was inspected when the
2239 partial match was found is set as the first matching string. There is a
2240 more detailed discussion of partial and multi-segment matching, with
2241 examples, in the pcrepartial documentation.
2242
2243 The string to be matched by pcre_exec()
2244
2245 The subject string is passed to pcre_exec() as a pointer in subject, a
2246 length (in bytes) in length, and a starting byte offset in startoffset.
2247 If this is negative or greater than the length of the subject,
2248 pcre_exec() returns PCRE_ERROR_BADOFFSET. When the starting offset is
2249 zero, the search for a match starts at the beginning of the subject,
2250 and this is by far the most common case. In UTF-8 mode, the byte offset
2251 must point to the start of a UTF-8 character (or the end of the sub-
2252 ject). Unlike the pattern string, the subject may contain binary zero
2253 bytes.
2254
2255 A non-zero starting offset is useful when searching for another match
2256 in the same subject by calling pcre_exec() again after a previous suc-
2257 cess. Setting startoffset differs from just passing over a shortened
2258 string and setting PCRE_NOTBOL in the case of a pattern that begins
2259 with any kind of lookbehind. For example, consider the pattern
2260
2261 \Biss\B
2262
2263 which finds occurrences of "iss" in the middle of words. (\B matches
2264 only if the current position in the subject is not a word boundary.)
2265 When applied to the string "Mississipi" the first call to pcre_exec()
2266 finds the first occurrence. If pcre_exec() is called again with just
2267 the remainder of the subject, namely "issipi", it does not match,
2268 because \B is always false at the start of the subject, which is deemed
2269 to be a word boundary. However, if pcre_exec() is passed the entire
2270 string again, but with startoffset set to 4, it finds the second occur-
2271 rence of "iss" because it is able to look behind the starting point to
2272 discover that it is preceded by a letter.
2273
2274 Finding all the matches in a subject is tricky when the pattern can
2275 match an empty string. It is possible to emulate Perl's /g behaviour by
2276 first trying the match again at the same offset, with the
2277 PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED options, and then if that
2278 fails, advancing the starting offset and trying an ordinary match
2279 again. There is some code that demonstrates how to do this in the pcre-
2280 demo sample program. In the most general case, you have to check to see
2281 if the newline convention recognizes CRLF as a newline, and if so, and
2282 the current character is CR followed by LF, advance the starting offset
2283 by two characters instead of one.
2284
2285 If a non-zero starting offset is passed when the pattern is anchored,
2286 one attempt to match at the given offset is made. This can only succeed
2287 if the pattern does not require the match to be at the start of the
2288 subject.
2289
2290 How pcre_exec() returns captured substrings
2291
2292 In general, a pattern matches a certain portion of the subject, and in
2293 addition, further substrings from the subject may be picked out by
2294 parts of the pattern. Following the usage in Jeffrey Friedl's book,
2295 this is called "capturing" in what follows, and the phrase "capturing
2296 subpattern" is used for a fragment of a pattern that picks out a sub-
2297 string. PCRE supports several other kinds of parenthesized subpattern
2298 that do not cause substrings to be captured.
2299
2300 Captured substrings are returned to the caller via a vector of integers
2301 whose address is passed in ovector. The number of elements in the vec-
2302 tor is passed in ovecsize, which must be a non-negative number. Note:
2303 this argument is NOT the size of ovector in bytes.
2304
2305 The first two-thirds of the vector is used to pass back captured sub-
2306 strings, each substring using a pair of integers. The remaining third
2307 of the vector is used as workspace by pcre_exec() while matching cap-
2308 turing subpatterns, and is not available for passing back information.
2309 The number passed in ovecsize should always be a multiple of three. If
2310 it is not, it is rounded down.
2311
2312 When a match is successful, information about captured substrings is
2313 returned in pairs of integers, starting at the beginning of ovector,
2314 and continuing up to two-thirds of its length at the most. The first
2315 element of each pair is set to the byte offset of the first character
2316 in a substring, and the second is set to the byte offset of the first
2317 character after the end of a substring. Note: these values are always
2318 byte offsets, even in UTF-8 mode. They are not character counts.
2319
2320 The first pair of integers, ovector[0] and ovector[1], identify the
2321 portion of the subject string matched by the entire pattern. The next
2322 pair is used for the first capturing subpattern, and so on. The value
2323 returned by pcre_exec() is one more than the highest numbered pair that
2324 has been set. For example, if two substrings have been captured, the
2325 returned value is 3. If there are no capturing subpatterns, the return
2326 value from a successful match is 1, indicating that just the first pair
2327 of offsets has been set.
2328
2329 If a capturing subpattern is matched repeatedly, it is the last portion
2330 of the string that it matched that is returned.
2331
2332 If the vector is too small to hold all the captured substring offsets,
2333 it is used as far as possible (up to two-thirds of its length), and the
2334 function returns a value of zero. If the substring offsets are not of
2335 interest, pcre_exec() may be called with ovector passed as NULL and
2336 ovecsize as zero. However, if the pattern contains back references and
2337 the ovector is not big enough to remember the related substrings, PCRE
2338 has to get additional memory for use during matching. Thus it is usu-
2339 ally advisable to supply an ovector.
2340
2341 The pcre_fullinfo() function can be used to find out how many capturing
2342 subpatterns there are in a compiled pattern. The smallest size for
2343 ovector that will allow for n captured substrings, in addition to the
2344 offsets of the substring matched by the whole pattern, is (n+1)*3.
2345
2346 It is possible for capturing subpattern number n+1 to match some part
2347 of the subject when subpattern n has not been used at all. For example,
2348 if the string "abc" is matched against the pattern (a|(z))(bc) the
2349 return from the function is 4, and subpatterns 1 and 3 are matched, but
2350 2 is not. When this happens, both values in the offset pairs corre-
2351 sponding to unused subpatterns are set to -1.
2352
2353 Offset values that correspond to unused subpatterns at the end of the
2354 expression are also set to -1. For example, if the string "abc" is
2355 matched against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not
2356 matched. The return from the function is 2, because the highest used
2357 capturing subpattern number is 1, and the offsets for for the second
2358 and third capturing subpatterns (assuming the vector is large enough,
2359 of course) are set to -1.
2360
2361 Note: Elements of ovector that do not correspond to capturing parenthe-
2362 ses in the pattern are never changed. That is, if a pattern contains n
2363 capturing parentheses, no more than ovector[0] to ovector[2n+1] are set
2364 by pcre_exec(). The other elements retain whatever values they previ-
2365 ously had.
2366
2367 Some convenience functions are provided for extracting the captured
2368 substrings as separate strings. These are described below.
2369
2370 Error return values from pcre_exec()
2371
2372 If pcre_exec() fails, it returns a negative number. The following are
2373 defined in the header file:
2374
2375 PCRE_ERROR_NOMATCH (-1)
2376
2377 The subject string did not match the pattern.
2378
2379 PCRE_ERROR_NULL (-2)
2380
2381 Either code or subject was passed as NULL, or ovector was NULL and
2382 ovecsize was not zero.
2383
2384 PCRE_ERROR_BADOPTION (-3)
2385
2386 An unrecognized bit was set in the options argument.
2387
2388 PCRE_ERROR_BADMAGIC (-4)
2389
2390 PCRE stores a 4-byte "magic number" at the start of the compiled code,
2391 to catch the case when it is passed a junk pointer and to detect when a
2392 pattern that was compiled in an environment of one endianness is run in
2393 an environment with the other endianness. This is the error that PCRE
2394 gives when the magic number is not present.
2395
2396 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2397
2398 While running the pattern match, an unknown item was encountered in the
2399 compiled pattern. This error could be caused by a bug in PCRE or by
2400 overwriting of the compiled pattern.
2401
2402 PCRE_ERROR_NOMEMORY (-6)
2403
2404 If a pattern contains back references, but the ovector that is passed
2405 to pcre_exec() is not big enough to remember the referenced substrings,
2406 PCRE gets a block of memory at the start of matching to use for this
2407 purpose. If the call via pcre_malloc() fails, this error is given. The
2408 memory is automatically freed at the end of matching.
2409
2410 This error is also given if pcre_stack_malloc() fails in pcre_exec().
2411 This can happen only when PCRE has been compiled with --disable-stack-
2412 for-recursion.
2413
2414 PCRE_ERROR_NOSUBSTRING (-7)
2415
2416 This error is used by the pcre_copy_substring(), pcre_get_substring(),
2417 and pcre_get_substring_list() functions (see below). It is never
2418 returned by pcre_exec().
2419
2420 PCRE_ERROR_MATCHLIMIT (-8)
2421
2422 The backtracking limit, as specified by the match_limit field in a
2423 pcre_extra structure (or defaulted) was reached. See the description
2424 above.
2425
2426 PCRE_ERROR_CALLOUT (-9)
2427
2428 This error is never generated by pcre_exec() itself. It is provided for
2429 use by callout functions that want to yield a distinctive error code.
2430 See the pcrecallout documentation for details.
2431
2432 PCRE_ERROR_BADUTF8 (-10)
2433
2434 A string that contains an invalid UTF-8 byte sequence was passed as a
2435 subject. However, if PCRE_PARTIAL_HARD is set and the problem is a
2436 truncated UTF-8 character at the end of the subject, PCRE_ERROR_SHORT-
2437 UTF8 is used instead.
2438
2439 PCRE_ERROR_BADUTF8_OFFSET (-11)
2440
2441 The UTF-8 byte sequence that was passed as a subject was valid, but the
2442 value of startoffset did not point to the beginning of a UTF-8 charac-
2443 ter or the end of the subject.
2444
2445 PCRE_ERROR_PARTIAL (-12)
2446
2447 The subject string did not match, but it did match partially. See the
2448 pcrepartial documentation for details of partial matching.
2449
2450 PCRE_ERROR_BADPARTIAL (-13)
2451
2452 This code is no longer in use. It was formerly returned when the
2453 PCRE_PARTIAL option was used with a compiled pattern containing items
2454 that were not supported for partial matching. From release 8.00
2455 onwards, there are no restrictions on partial matching.
2456
2457 PCRE_ERROR_INTERNAL (-14)
2458
2459 An unexpected internal error has occurred. This error could be caused
2460 by a bug in PCRE or by overwriting of the compiled pattern.
2461
2462 PCRE_ERROR_BADCOUNT (-15)
2463
2464 This error is given if the value of the ovecsize argument is negative.
2465
2466 PCRE_ERROR_RECURSIONLIMIT (-21)
2467
2468 The internal recursion limit, as specified by the match_limit_recursion
2469 field in a pcre_extra structure (or defaulted) was reached. See the
2470 description above.
2471
2472 PCRE_ERROR_BADNEWLINE (-23)
2473
2474 An invalid combination of PCRE_NEWLINE_xxx options was given.
2475
2476 PCRE_ERROR_BADOFFSET (-24)
2477
2478 The value of startoffset was negative or greater than the length of the
2479 subject, that is, the value in length.
2480
2481 PCRE_ERROR_SHORTUTF8 (-25)
2482
2483 The subject string ended with an incomplete (truncated) UTF-8 charac-
2484 ter, and the PCRE_PARTIAL_HARD option was set. Without this option,
2485 PCRE_ERROR_BADUTF8 is returned in this situation.
2486
2487 Error numbers -16 to -20 and -22 are not used by pcre_exec().
2488
2489
2490 EXTRACTING CAPTURED SUBSTRINGS BY NUMBER
2491
2492 int pcre_copy_substring(const char *subject, int *ovector,
2493 int stringcount, int stringnumber, char *buffer,
2494 int buffersize);
2495
2496 int pcre_get_substring(const char *subject, int *ovector,
2497 int stringcount, int stringnumber,
2498 const char **stringptr);
2499
2500 int pcre_get_substring_list(const char *subject,
2501 int *ovector, int stringcount, const char ***listptr);
2502
2503 Captured substrings can be accessed directly by using the offsets
2504 returned by pcre_exec() in ovector. For convenience, the functions
2505 pcre_copy_substring(), pcre_get_substring(), and pcre_get_sub-
2506 string_list() are provided for extracting captured substrings as new,
2507 separate, zero-terminated strings. These functions identify substrings
2508 by number. The next section describes functions for extracting named
2509 substrings.
2510
2511 A substring that contains a binary zero is correctly extracted and has
2512 a further zero added on the end, but the result is not, of course, a C
2513 string. However, you can process such a string by referring to the
2514 length that is returned by pcre_copy_substring() and pcre_get_sub-
2515 string(). Unfortunately, the interface to pcre_get_substring_list() is
2516 not adequate for handling strings containing binary zeros, because the
2517 end of the final string is not independently indicated.
2518
2519 The first three arguments are the same for all three of these func-
2520 tions: subject is the subject string that has just been successfully
2521 matched, ovector is a pointer to the vector of integer offsets that was
2522 passed to pcre_exec(), and stringcount is the number of substrings that
2523 were captured by the match, including the substring that matched the
2524 entire regular expression. This is the value returned by pcre_exec() if
2525 it is greater than zero. If pcre_exec() returned zero, indicating that
2526 it ran out of space in ovector, the value passed as stringcount should
2527 be the number of elements in the vector divided by three.
2528
2529 The functions pcre_copy_substring() and pcre_get_substring() extract a
2530 single substring, whose number is given as stringnumber. A value of
2531 zero extracts the substring that matched the entire pattern, whereas
2532 higher values extract the captured substrings. For pcre_copy_sub-
2533 string(), the string is placed in buffer, whose length is given by
2534 buffersize, while for pcre_get_substring() a new block of memory is
2535 obtained via pcre_malloc, and its address is returned via stringptr.
2536 The yield of the function is the length of the string, not including
2537 the terminating zero, or one of these error codes:
2538
2539 PCRE_ERROR_NOMEMORY (-6)
2540
2541 The buffer was too small for pcre_copy_substring(), or the attempt to
2542 get memory failed for pcre_get_substring().
2543
2544 PCRE_ERROR_NOSUBSTRING (-7)
2545
2546 There is no substring whose number is stringnumber.
2547
2548 The pcre_get_substring_list() function extracts all available sub-
2549 strings and builds a list of pointers to them. All this is done in a
2550 single block of memory that is obtained via pcre_malloc. The address of
2551 the memory block is returned via listptr, which is also the start of
2552 the list of string pointers. The end of the list is marked by a NULL
2553 pointer. The yield of the function is zero if all went well, or the
2554 error code
2555
2556 PCRE_ERROR_NOMEMORY (-6)
2557
2558 if the attempt to get the memory block failed.
2559
2560 When any of these functions encounter a substring that is unset, which
2561 can happen when capturing subpattern number n+1 matches some part of
2562 the subject, but subpattern n has not been used at all, they return an
2563 empty string. This can be distinguished from a genuine zero-length sub-
2564 string by inspecting the appropriate offset in ovector, which is nega-
2565 tive for unset substrings.
2566
2567 The two convenience functions pcre_free_substring() and pcre_free_sub-
2568 string_list() can be used to free the memory returned by a previous
2569 call of pcre_get_substring() or pcre_get_substring_list(), respec-
2570 tively. They do nothing more than call the function pointed to by
2571 pcre_free, which of course could be called directly from a C program.
2572 However, PCRE is used in some situations where it is linked via a spe-
2573 cial interface to another programming language that cannot use
2574 pcre_free directly; it is for these cases that the functions are pro-
2575 vided.
2576
2577
2578 EXTRACTING CAPTURED SUBSTRINGS BY NAME
2579
2580 int pcre_get_stringnumber(const pcre *code,
2581 const char *name);
2582
2583 int pcre_copy_named_substring(const pcre *code,
2584 const char *subject, int *ovector,
2585 int stringcount, const char *stringname,
2586 char *buffer, int buffersize);
2587
2588 int pcre_get_named_substring(const pcre *code,
2589 const char *subject, int *ovector,
2590 int stringcount, const char *stringname,
2591 const char **stringptr);
2592
2593 To extract a substring by name, you first have to find associated num-
2594 ber. For example, for this pattern
2595
2596 (a+)b(?<xxx>\d+)...
2597
2598 the number of the subpattern called "xxx" is 2. If the name is known to
2599 be unique (PCRE_DUPNAMES was not set), you can find the number from the
2600 name by calling pcre_get_stringnumber(). The first argument is the com-
2601 piled pattern, and the second is the name. The yield of the function is
2602 the subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no
2603 subpattern of that name.
2604
2605 Given the number, you can extract the substring directly, or use one of
2606 the functions described in the previous section. For convenience, there
2607 are also two functions that do the whole job.
2608
2609 Most of the arguments of pcre_copy_named_substring() and
2610 pcre_get_named_substring() are the same as those for the similarly
2611 named functions that extract by number. As these are described in the
2612 previous section, they are not re-described here. There are just two
2613 differences:
2614
2615 First, instead of a substring number, a substring name is given. Sec-
2616 ond, there is an extra argument, given at the start, which is a pointer
2617 to the compiled pattern. This is needed in order to gain access to the
2618 name-to-number translation table.
2619
2620 These functions call pcre_get_stringnumber(), and if it succeeds, they
2621 then call pcre_copy_substring() or pcre_get_substring(), as appropri-
2622 ate. NOTE: If PCRE_DUPNAMES is set and there are duplicate names, the
2623 behaviour may not be what you want (see the next section).
2624
2625 Warning: If the pattern uses the (?| feature to set up multiple subpat-
2626 terns with the same number, as described in the section on duplicate
2627 subpattern numbers in the pcrepattern page, you cannot use names to
2628 distinguish the different subpatterns, because names are not included
2629 in the compiled code. The matching process uses only numbers. For this
2630 reason, the use of different names for subpatterns of the same number
2631 causes an error at compile time.
2632
2633
2634 DUPLICATE SUBPATTERN NAMES
2635
2636 int pcre_get_stringtable_entries(const pcre *code,
2637 const char *name, char **first, char **last);
2638
2639 When a pattern is compiled with the PCRE_DUPNAMES option, names for
2640 subpatterns are not required to be unique. (Duplicate names are always
2641 allowed for subpatterns with the same number, created by using the (?|
2642 feature. Indeed, if such subpatterns are named, they are required to
2643 use the same names.)
2644
2645 Normally, patterns with duplicate names are such that in any one match,
2646 only one of the named subpatterns participates. An example is shown in
2647 the pcrepattern documentation.
2648
2649 When duplicates are present, pcre_copy_named_substring() and
2650 pcre_get_named_substring() return the first substring corresponding to
2651 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING
2652 (-7) is returned; no data is returned. The pcre_get_stringnumber()
2653 function returns one of the numbers that are associated with the name,
2654 but it is not defined which it is.
2655
2656 If you want to get full details of all captured substrings for a given
2657 name, you must use the pcre_get_stringtable_entries() function. The
2658 first argument is the compiled pattern, and the second is the name. The
2659 third and fourth are pointers to variables which are updated by the
2660 function. After it has run, they point to the first and last entries in
2661 the name-to-number table for the given name. The function itself
2662 returns the length of each entry, or PCRE_ERROR_NOSUBSTRING (-7) if
2663 there are none. The format of the table is described above in the sec-
2664 tion entitled Information about a pattern. Given all the relevant
2665 entries for the name, you can extract each of their numbers, and hence
2666 the captured data, if any.
2667
2668
2669 FINDING ALL POSSIBLE MATCHES
2670
2671 The traditional matching function uses a similar algorithm to Perl,
2672 which stops when it finds the first match, starting at a given point in
2673 the subject. If you want to find all possible matches, or the longest
2674 possible match, consider using the alternative matching function (see
2675 below) instead. If you cannot use the alternative function, but still
2676 need to find all possible matches, you can kludge it up by making use
2677 of the callout facility, which is described in the pcrecallout documen-
2678 tation.
2679
2680 What you have to do is to insert a callout right at the end of the pat-
2681 tern. When your callout function is called, extract and save the cur-
2682 rent matched substring. Then return 1, which forces pcre_exec() to
2683 backtrack and try other alternatives. Ultimately, when it runs out of
2684 matches, pcre_exec() will yield PCRE_ERROR_NOMATCH.
2685
2686
2687 MATCHING A PATTERN: THE ALTERNATIVE FUNCTION
2688
2689 int pcre_dfa_exec(const pcre *code, const pcre_extra *extra,
2690 const char *subject, int length, int startoffset,
2691 int options, int *ovector, int ovecsize,
2692 int *workspace, int wscount);
2693
2694 The function pcre_dfa_exec() is called to match a subject string
2695 against a compiled pattern, using a matching algorithm that scans the
2696 subject string just once, and does not backtrack. This has different
2697 characteristics to the normal algorithm, and is not compatible with
2698 Perl. Some of the features of PCRE patterns are not supported. Never-
2699 theless, there are times when this kind of matching can be useful. For
2700 a discussion of the two matching algorithms, and a list of features
2701 that pcre_dfa_exec() does not support, see the pcrematching documenta-
2702 tion.
2703
2704 The arguments for the pcre_dfa_exec() function are the same as for
2705 pcre_exec(), plus two extras. The ovector argument is used in a differ-
2706 ent way, and this is described below. The other common arguments are
2707 used in the same way as for pcre_exec(), so their description is not
2708 repeated here.
2709
2710 The two additional arguments provide workspace for the function. The
2711 workspace vector should contain at least 20 elements. It is used for
2712 keeping track of multiple paths through the pattern tree. More
2713 workspace will be needed for patterns and subjects where there are a
2714 lot of potential matches.
2715
2716 Here is an example of a simple call to pcre_dfa_exec():
2717
2718 int rc;
2719 int ovector[10];
2720 int wspace[20];
2721 rc = pcre_dfa_exec(
2722 re, /* result of pcre_compile() */
2723 NULL, /* we didn't study the pattern */
2724 "some string", /* the subject string */
2725 11, /* the length of the subject string */
2726 0, /* start at offset 0 in the subject */
2727 0, /* default options */
2728 ovector, /* vector of integers for substring information */
2729 10, /* number of elements (NOT size in bytes) */
2730 wspace, /* working space vector */
2731 20); /* number of elements (NOT size in bytes) */
2732
2733 Option bits for pcre_dfa_exec()
2734
2735 The unused bits of the options argument for pcre_dfa_exec() must be
2736 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEW-
2737 LINE_xxx, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
2738 PCRE_NOTEMPTY_ATSTART, PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF,
2739 PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE, PCRE_PARTIAL_HARD, PCRE_PAR-
2740 TIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART. All but the last
2741 four of these are exactly the same as for pcre_exec(), so their
2742 description is not repeated here.
2743
2744 PCRE_PARTIAL_HARD
2745 PCRE_PARTIAL_SOFT
2746
2747 These have the same general effect as they do for pcre_exec(), but the
2748 details are slightly different. When PCRE_PARTIAL_HARD is set for
2749 pcre_dfa_exec(), it returns PCRE_ERROR_PARTIAL if the end of the sub-
2750 ject is reached and there is still at least one matching possibility
2751 that requires additional characters. This happens even if some complete
2752 matches have also been found. When PCRE_PARTIAL_SOFT is set, the return
2753 code PCRE_ERROR_NOMATCH is converted into PCRE_ERROR_PARTIAL if the end
2754 of the subject is reached, there have been no complete matches, but
2755 there is still at least one matching possibility. The portion of the
2756 string that was inspected when the longest partial match was found is
2757 set as the first matching string in both cases. There is a more
2758 detailed discussion of partial and multi-segment matching, with exam-
2759 ples, in the pcrepartial documentation.
2760
2761 PCRE_DFA_SHORTEST
2762
2763 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to
2764 stop as soon as it has found one match. Because of the way the alterna-
2765 tive algorithm works, this is necessarily the shortest possible match
2766 at the first possible matching point in the subject string.
2767
2768 PCRE_DFA_RESTART
2769
2770 When pcre_dfa_exec() returns a partial match, it is possible to call it
2771 again, with additional subject characters, and have it continue with
2772 the same match. The PCRE_DFA_RESTART option requests this action; when
2773 it is set, the workspace and wscount options must reference the same
2774 vector as before because data about the match so far is left in them
2775 after a partial match. There is more discussion of this facility in the
2776 pcrepartial documentation.
2777
2778 Successful returns from pcre_dfa_exec()
2779
2780 When pcre_dfa_exec() succeeds, it may have matched more than one sub-
2781 string in the subject. Note, however, that all the matches from one run
2782 of the function start at the same point in the subject. The shorter
2783 matches are all initial substrings of the longer matches. For example,
2784 if the pattern
2785
2786 <.*>
2787
2788 is matched against the string
2789
2790 This is <something> <something else> <something further> no more
2791
2792 the three matched strings are
2793
2794 <something>
2795 <something> <something else>
2796 <something> <something else> <something further>
2797
2798 On success, the yield of the function is a number greater than zero,
2799 which is the number of matched substrings. The substrings themselves
2800 are returned in ovector. Each string uses two elements; the first is
2801 the offset to the start, and the second is the offset to the end. In
2802 fact, all the strings have the same start offset. (Space could have
2803 been saved by giving this only once, but it was decided to retain some
2804 compatibility with the way pcre_exec() returns data, even though the
2805 meaning of the strings is different.)
2806
2807 The strings are returned in reverse order of length; that is, the long-
2808 est matching string is given first. If there were too many matches to
2809 fit into ovector, the yield of the function is zero, and the vector is
2810 filled with the longest matches.
2811
2812 Error returns from pcre_dfa_exec()
2813
2814 The pcre_dfa_exec() function returns a negative number when it fails.
2815 Many of the errors are the same as for pcre_exec(), and these are
2816 described above. There are in addition the following errors that are
2817 specific to pcre_dfa_exec():
2818
2819 PCRE_ERROR_DFA_UITEM (-16)
2820
2821 This return is given if pcre_dfa_exec() encounters an item in the pat-
2822 tern that it does not support, for instance, the use of \C or a back
2823 reference.
2824
2825 PCRE_ERROR_DFA_UCOND (-17)
2826
2827 This return is given if pcre_dfa_exec() encounters a condition item
2828 that uses a back reference for the condition, or a test for recursion
2829 in a specific group. These are not supported.
2830
2831 PCRE_ERROR_DFA_UMLIMIT (-18)
2832
2833 This return is given if pcre_dfa_exec() is called with an extra block
2834 that contains a setting of the match_limit field. This is not supported
2835 (it is meaningless).
2836
2837 PCRE_ERROR_DFA_WSSIZE (-19)
2838
2839 This return is given if pcre_dfa_exec() runs out of space in the
2840 workspace vector.
2841
2842 PCRE_ERROR_DFA_RECURSE (-20)
2843
2844 When a recursive subpattern is processed, the matching function calls
2845 itself recursively, using private vectors for ovector and workspace.
2846 This error is given if the output vector is not large enough. This
2847 should be extremely rare, as a vector of size 1000 is used.
2848
2849
2850 SEE ALSO
2851
2852 pcrebuild(3), pcrecallout(3), pcrecpp(3)(3), pcrematching(3), pcrepar-
2853 tial(3), pcreposix(3), pcreprecompile(3), pcresample(3), pcrestack(3).
2854
2855
2856 AUTHOR
2857
2858 Philip Hazel
2859 University Computing Service
2860 Cambridge CB2 3QH, England.
2861
2862
2863 REVISION
2864
2865 Last updated: 13 November 2010
2866 Copyright (c) 1997-2010 University of Cambridge.
2867 ------------------------------------------------------------------------------
2868
2869
2870 PCRECALLOUT(3) PCRECALLOUT(3)
2871
2872
2873 NAME
2874 PCRE - Perl-compatible regular expressions
2875
2876
2877 PCRE CALLOUTS
2878
2879 int (*pcre_callout)(pcre_callout_block *);
2880
2881 PCRE provides a feature called "callout", which is a means of temporar-
2882 ily passing control to the caller of PCRE in the middle of pattern
2883 matching. The caller of PCRE provides an external function by putting
2884 its entry point in the global variable pcre_callout. By default, this
2885 variable contains NULL, which disables all calling out.
2886
2887 Within a regular expression, (?C) indicates the points at which the
2888 external function is to be called. Different callout points can be
2889 identified by putting a number less than 256 after the letter C. The
2890 default value is zero. For example, this pattern has two callout
2891 points:
2892
2893 (?C1)abc(?C2)def
2894
2895 If the PCRE_AUTO_CALLOUT option bit is set when pcre_compile() or
2896 pcre_compile2() is called, PCRE automatically inserts callouts, all
2897 with number 255, before each item in the pattern. For example, if
2898 PCRE_AUTO_CALLOUT is used with the pattern
2899
2900 A(\d{2}|--)
2901
2902 it is processed as if it were
2903
2904 (?C255)A(?C255)((?C255)\d{2}(?C255)|(?C255)-(?C255)-(?C255))(?C255)
2905
2906 Notice that there is a callout before and after each parenthesis and
2907 alternation bar. Automatic callouts can be used for tracking the
2908 progress of pattern matching. The pcretest command has an option that
2909 sets automatic callouts; when it is used, the output indicates how the
2910 pattern is matched. This is useful information when you are trying to
2911 optimize the performance of a particular pattern.
2912
2913
2914 MISSING CALLOUTS
2915
2916 You should be aware that, because of optimizations in the way PCRE
2917 matches patterns by default, callouts sometimes do not happen. For
2918 example, if the pattern is
2919
2920 ab(?C4)cd
2921
2922 PCRE knows that any matching string must contain the letter "d". If the
2923 subject string is "abyz", the lack of "d" means that matching doesn't
2924 ever start, and the callout is never reached. However, with "abyd",
2925 though the result is still no match, the callout is obeyed.
2926
2927 If the pattern is studied, PCRE knows the minimum length of a matching
2928 string, and will immediately give a "no match" return without actually
2929 running a match if the subject is not long enough, or, for unanchored
2930 patterns, if it has been scanned far enough.
2931
2932 You can disable these optimizations by passing the PCRE_NO_START_OPTI-
2933 MIZE option to pcre_exec() or pcre_dfa_exec(). This slows down the
2934 matching process, but does ensure that callouts such as the example
2935 above are obeyed.
2936
2937
2938 THE CALLOUT INTERFACE
2939
2940 During matching, when PCRE reaches a callout point, the external func-
2941 tion defined by pcre_callout is called (if it is set). This applies to
2942 both the pcre_exec() and the pcre_dfa_exec() matching functions. The
2943 only argument to the callout function is a pointer to a pcre_callout
2944 block. This structure contains the following fields:
2945
2946 int version;
2947 int callout_number;
2948 int *offset_vector;
2949 const char *subject;
2950 int subject_length;
2951 int start_match;
2952 int current_position;
2953 int capture_top;
2954 int capture_last;
2955 void *callout_data;
2956 int pattern_position;
2957 int next_item_length;
2958
2959 The version field is an integer containing the version number of the
2960 block format. The initial version was 0; the current version is 1. The
2961 version number will change again in future if additional fields are
2962 added, but the intention is never to remove any of the existing fields.
2963
2964 The callout_number field contains the number of the callout, as com-
2965 piled into the pattern (that is, the number after ?C for manual call-
2966 outs, and 255 for automatically generated callouts).
2967
2968 The offset_vector field is a pointer to the vector of offsets that was
2969 passed by the caller to pcre_exec() or pcre_dfa_exec(). When
2970 pcre_exec() is used, the contents can be inspected in order to extract
2971 substrings that have been matched so far, in the same way as for
2972 extracting substrings after a match has completed. For pcre_dfa_exec()
2973 this field is not useful.
2974
2975 The subject and subject_length fields contain copies of the values that
2976 were passed to pcre_exec().
2977
2978 The start_match field normally contains the offset within the subject
2979 at which the current match attempt started. However, if the escape
2980 sequence \K has been encountered, this value is changed to reflect the
2981 modified starting point. If the pattern is not anchored, the callout
2982 function may be called several times from the same point in the pattern
2983 for different starting points in the subject.
2984
2985 The current_position field contains the offset within the subject of
2986 the current match pointer.
2987
2988 When the pcre_exec() function is used, the capture_top field contains
2989 one more than the number of the highest numbered captured substring so
2990 far. If no substrings have been captured, the value of capture_top is
2991 one. This is always the case when pcre_dfa_exec() is used, because it
2992 does not support captured substrings.
2993
2994 The capture_last field contains the number of the most recently cap-
2995 tured substring. If no substrings have been captured, its value is -1.
2996 This is always the case when pcre_dfa_exec() is used.
2997
2998 The callout_data field contains a value that is passed to pcre_exec()
2999 or pcre_dfa_exec() specifically so that it can be passed back in call-
3000 outs. It is passed in the pcre_callout field of the pcre_extra data
3001 structure. If no such data was passed, the value of callout_data in a
3002 pcre_callout block is NULL. There is a description of the pcre_extra
3003 structure in the pcreapi documentation.
3004
3005 The pattern_position field is present from version 1 of the pcre_call-
3006 out structure. It contains the offset to the next item to be matched in
3007 the pattern string.
3008
3009 The next_item_length field is present from version 1 of the pcre_call-
3010 out structure. It contains the length of the next item to be matched in
3011 the pattern string. When the callout immediately precedes an alterna-
3012 tion bar, a closing parenthesis, or the end of the pattern, the length
3013 is zero. When the callout precedes an opening parenthesis, the length
3014 is that of the entire subpattern.
3015
3016 The pattern_position and next_item_length fields are intended to help
3017 in distinguishing between different automatic callouts, which all have
3018 the same callout number. However, they are set for all callouts.
3019
3020
3021 RETURN VALUES
3022
3023 The external callout function returns an integer to PCRE. If the value
3024 is zero, matching proceeds as normal. If the value is greater than
3025 zero, matching fails at the current point, but the testing of other
3026 matching possibilities goes ahead, just as if a lookahead assertion had
3027 failed. If the value is less than zero, the match is abandoned, and
3028 pcre_exec() or pcre_dfa_exec() returns the negative value.
3029
3030 Negative values should normally be chosen from the set of
3031 PCRE_ERROR_xxx values. In particular, PCRE_ERROR_NOMATCH forces a stan-
3032 dard "no match" failure. The error number PCRE_ERROR_CALLOUT is
3033 reserved for use by callout functions; it will never be used by PCRE
3034 itself.
3035
3036
3037 AUTHOR
3038
3039 Philip Hazel
3040 University Computing Service
3041 Cambridge CB2 3QH, England.
3042
3043
3044 REVISION
3045
3046 Last updated: 29 September 2009
3047 Copyright (c) 1997-2009 University of Cambridge.
3048 ------------------------------------------------------------------------------
3049
3050
3051 PCRECOMPAT(3) PCRECOMPAT(3)
3052
3053
3054 NAME
3055 PCRE - Perl-compatible regular expressions
3056
3057
3058 DIFFERENCES BETWEEN PCRE AND PERL
3059
3060 This document describes the differences in the ways that PCRE and Perl
3061 handle regular expressions. The differences described here are with
3062 respect to Perl versions 5.10 and above.
3063
3064 1. PCRE has only a subset of Perl's UTF-8 and Unicode support. Details
3065 of what it does have are given in the section on UTF-8 support in the
3066 main pcre page.
3067
3068 2. PCRE does not allow repeat quantifiers on lookahead assertions. Perl
3069 permits them, but they do not mean what you might think. For example,
3070 (?!a){3} does not assert that the next three characters are not "a". It
3071 just asserts that the next character is not "a" three times.
3072
3073 3. Capturing subpatterns that occur inside negative lookahead asser-
3074 tions are counted, but their entries in the offsets vector are never
3075 set. Perl sets its numerical variables from any such patterns that are
3076 matched before the assertion fails to match something (thereby succeed-
3077 ing), but only if the negative lookahead assertion contains just one
3078 branch.
3079
3080 4. Though binary zero characters are supported in the subject string,
3081 they are not allowed in a pattern string because it is passed as a nor-
3082 mal C string, terminated by zero. The escape sequence \0 can be used in
3083 the pattern to represent a binary zero.
3084
3085 5. The following Perl escape sequences are not supported: \l, \u, \L,
3086 \U, and \N. In fact these are implemented by Perl's general string-han-
3087 dling and are not part of its pattern matching engine. If any of these
3088 are encountered by PCRE, an error is generated.
3089
3090 6. The Perl escape sequences \p, \P, and \X are supported only if PCRE
3091 is built with Unicode character property support. The properties that
3092 can be tested with \p and \P are limited to the general category prop-
3093 erties such as Lu and Nd, script names such as Greek or Han, and the
3094 derived properties Any and L&. PCRE does support the Cs (surrogate)
3095 property, which Perl does not; the Perl documentation says "Because
3096 Perl hides the need for the user to understand the internal representa-
3097 tion of Unicode characters, there is no need to implement the somewhat
3098 messy concept of surrogates."
3099
3100 7. PCRE does support the \Q...\E escape for quoting substrings. Charac-
3101 ters in between are treated as literals. This is slightly different
3102 from Perl in that $ and @ are also handled as literals inside the
3103 quotes. In Perl, they cause variable interpolation (but of course PCRE
3104 does not have variables). Note the following examples:
3105
3106 Pattern PCRE matches Perl matches
3107
3108 \Qabc$xyz\E abc$xyz abc followed by the
3109 contents of $xyz
3110 \Qabc\$xyz\E abc\$xyz abc\$xyz
3111 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3112
3113 The \Q...\E sequence is recognized both inside and outside character
3114 classes.
3115
3116 8. Fairly obviously, PCRE does not support the (?{code}) and (??{code})
3117 constructions. However, there is support for recursive patterns. This
3118 is not available in Perl 5.8, but it is in Perl 5.10. Also, the PCRE
3119 "callout" feature allows an external function to be called during pat-
3120 tern matching. See the pcrecallout documentation for details.
3121
3122 9. Subpatterns that are called recursively or as "subroutines" are
3123 always treated as atomic groups in PCRE. This is like Python, but
3124 unlike Perl. There is a discussion of an example that explains this in
3125 more detail in the section on recursion differences from Perl in the
3126 pcrepattern page.
3127
3128 10. There are some differences that are concerned with the settings of
3129 captured strings when part of a pattern is repeated. For example,
3130 matching "aba" against the pattern /^(a(b)?)+$/ in Perl leaves $2
3131 unset, but in PCRE it is set to "b".
3132
3133 11. PCRE's handling of duplicate subpattern numbers and duplicate sub-
3134 pattern names is not as general as Perl's. This is a consequence of the
3135 fact the PCRE works internally just with numbers, using an external ta-
3136 ble to translate between numbers and names. In particular, a pattern
3137 such as (?|(?<a>A)|(?<b)B), where the two capturing parentheses have
3138 the same number but different names, is not supported, and causes an
3139 error at compile time. If it were allowed, it would not be possible to
3140 distinguish which parentheses matched, because both names map to cap-
3141 turing subpattern number 1. To avoid this confusing situation, an error
3142 is given at compile time.
3143
3144 12. Perl recognizes comments in some places that PCRE doesn't, for
3145 example, between the ( and ? at the start of a subpattern.
3146
3147 13. PCRE provides some extensions to the Perl regular expression facil-
3148 ities. Perl 5.10 includes new features that are not in earlier ver-
3149 sions of Perl, some of which (such as named parentheses) have been in
3150 PCRE for some time. This list is with respect to Perl 5.10:
3151
3152 (a) Although lookbehind assertions in PCRE must match fixed length
3153 strings, each alternative branch of a lookbehind assertion can match a
3154 different length of string. Perl requires them all to have the same
3155 length.
3156
3157 (b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $
3158 meta-character matches only at the very end of the string.
3159
3160 (c) If PCRE_EXTRA is set, a backslash followed by a letter with no spe-
3161 cial meaning is faulted. Otherwise, like Perl, the backslash is quietly
3162 ignored. (Perl can be made to issue a warning.)
3163
3164 (d) If PCRE_UNGREEDY is set, the greediness of the repetition quanti-
3165 fiers is inverted, that is, by default they are not greedy, but if fol-
3166 lowed by a question mark they are.
3167
3168 (e) PCRE_ANCHORED can be used at matching time to force a pattern to be
3169 tried only at the first matching position in the subject string.
3170
3171 (f) The PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
3172 and PCRE_NO_AUTO_CAPTURE options for pcre_exec() have no Perl equiva-
3173 lents.
3174
3175 (g) The \R escape sequence can be restricted to match only CR, LF, or
3176 CRLF by the PCRE_BSR_ANYCRLF option.
3177
3178 (h) The callout facility is PCRE-specific.
3179
3180 (i) The partial matching facility is PCRE-specific.
3181
3182 (j) Patterns compiled by PCRE can be saved and re-used at a later time,
3183 even on different hosts that have the other endianness.
3184
3185 (k) The alternative matching function (pcre_dfa_exec()) matches in a
3186 different way and is not Perl-compatible.
3187
3188 (l) PCRE recognizes some special sequences such as (*CR) at the start
3189 of a pattern that set overall options that cannot be changed within the
3190 pattern.
3191
3192
3193 AUTHOR
3194
3195 Philip Hazel
3196 University Computing Service
3197 Cambridge CB2 3QH, England.
3198
3199
3200 REVISION
3201
3202 Last updated: 31 October 2010
3203 Copyright (c) 1997-2010 University of Cambridge.
3204 ------------------------------------------------------------------------------
3205
3206
3207 PCREPATTERN(3) PCREPATTERN(3)
3208
3209
3210 NAME
3211 PCRE - Perl-compatible regular expressions
3212
3213
3214 PCRE REGULAR EXPRESSION DETAILS
3215
3216 The syntax and semantics of the regular expressions that are supported
3217 by PCRE are described in detail below. There is a quick-reference syn-
3218 tax summary in the pcresyntax page. PCRE tries to match Perl syntax and
3219 semantics as closely as it can. PCRE also supports some alternative
3220 regular expression syntax (which does not conflict with the Perl syn-
3221 tax) in order to provide some compatibility with regular expressions in
3222 Python, .NET, and Oniguruma.
3223
3224 Perl's regular expressions are described in its own documentation, and
3225 regular expressions in general are covered in a number of books, some
3226 of which have copious examples. Jeffrey Friedl's "Mastering Regular
3227 Expressions", published by O'Reilly, covers regular expressions in
3228 great detail. This description of PCRE's regular expressions is
3229 intended as reference material.
3230
3231 The original operation of PCRE was on strings of one-byte characters.
3232 However, there is now also support for UTF-8 character strings. To use
3233 this, PCRE must be built to include UTF-8 support, and you must call
3234 pcre_compile() or pcre_compile2() with the PCRE_UTF8 option. There is
3235 also a special sequence that can be given at the start of a pattern:
3236
3237 (*UTF8)
3238
3239 Starting a pattern with this sequence is equivalent to setting the
3240 PCRE_UTF8 option. This feature is not Perl-compatible. How setting
3241 UTF-8 mode affects pattern matching is mentioned in several places
3242 below. There is also a summary of UTF-8 features in the section on
3243 UTF-8 support in the main pcre page.
3244
3245 Another special sequence that may appear at the start of a pattern or
3246 in combination with (*UTF8) is:
3247
3248 (*UCP)
3249
3250 This has the same effect as setting the PCRE_UCP option: it causes
3251 sequences such as \d and \w to use Unicode properties to determine
3252 character types, instead of recognizing only characters with codes less
3253 than 128 via a lookup table.
3254
3255 The remainder of this document discusses the patterns that are sup-
3256 ported by PCRE when its main matching function, pcre_exec(), is used.
3257 From release 6.0, PCRE offers a second matching function,
3258 pcre_dfa_exec(), which matches using a different algorithm that is not
3259 Perl-compatible. Some of the features discussed below are not available
3260 when pcre_dfa_exec() is used. The advantages and disadvantages of the
3261 alternative function, and how it differs from the normal function, are
3262 discussed in the pcrematching page.
3263
3264
3265 NEWLINE CONVENTIONS
3266
3267 PCRE supports five different conventions for indicating line breaks in
3268 strings: a single CR (carriage return) character, a single LF (line-
3269 feed) character, the two-character sequence CRLF, any of the three pre-
3270 ceding, or any Unicode newline sequence. The pcreapi page has further
3271 discussion about newlines, and shows how to set the newline convention
3272 in the options arguments for the compiling and matching functions.
3273
3274 It is also possible to specify a newline convention by starting a pat-
3275 tern string with one of the following five sequences:
3276
3277 (*CR) carriage return
3278 (*LF) linefeed
3279 (*CRLF) carriage return, followed by linefeed
3280 (*ANYCRLF) any of the three above
3281 (*ANY) all Unicode newline sequences
3282
3283 These override the default and the options given to pcre_compile() or
3284 pcre_compile2(). For example, on a Unix system where LF is the default
3285 newline sequence, the pattern
3286
3287 (*CR)a.b
3288
3289 changes the convention to CR. That pattern matches "a\nb" because LF is
3290 no longer a newline. Note that these special settings, which are not
3291 Perl-compatible, are recognized only at the very start of a pattern,
3292 and that they must be in upper case. If more than one of them is
3293 present, the last one is used.
3294
3295 The newline convention affects the interpretation of the dot metachar-
3296 acter when PCRE_DOTALL is not set, and also the behaviour of \N. How-
3297 ever, it does not affect what the \R escape sequence matches. By
3298 default, this is any Unicode newline sequence, for Perl compatibility.
3299 However, this can be changed; see the description of \R in the section
3300 entitled "Newline sequences" below. A change of \R setting can be com-
3301 bined with a change of newline convention.
3302
3303
3304 CHARACTERS AND METACHARACTERS
3305
3306 A regular expression is a pattern that is matched against a subject
3307 string from left to right. Most characters stand for themselves in a
3308 pattern, and match the corresponding characters in the subject. As a
3309 trivial example, the pattern
3310
3311 The quick brown fox
3312
3313 matches a portion of a subject string that is identical to itself. When
3314 caseless matching is specified (the PCRE_CASELESS option), letters are
3315 matched independently of case. In UTF-8 mode, PCRE always understands
3316 the concept of case for characters whose values are less than 128, so
3317 caseless matching is always possible. For characters with higher val-
3318 ues, the concept of case is supported if PCRE is compiled with Unicode
3319 property support, but not otherwise. If you want to use caseless
3320 matching for characters 128 and above, you must ensure that PCRE is
3321 compiled with Unicode property support as well as with UTF-8 support.
3322
3323 The power of regular expressions comes from the ability to include
3324 alternatives and repetitions in the pattern. These are encoded in the
3325 pattern by the use of metacharacters, which do not stand for themselves
3326 but instead are interpreted in some special way.
3327
3328 There are two different sets of metacharacters: those that are recog-
3329 nized anywhere in the pattern except within square brackets, and those
3330 that are recognized within square brackets. Outside square brackets,
3331 the metacharacters are as follows:
3332
3333 \ general escape character with several uses
3334 ^ assert start of string (or line, in multiline mode)
3335 $ assert end of string (or line, in multiline mode)
3336 . match any character except newline (by default)
3337 [ start character class definition
3338 | start of alternative branch
3339 ( start subpattern
3340 ) end subpattern
3341 ? extends the meaning of (
3342 also 0 or 1 quantifier
3343 also quantifier minimizer
3344 * 0 or more quantifier
3345 + 1 or more quantifier
3346 also "possessive quantifier"
3347 { start min/max quantifier
3348
3349 Part of a pattern that is in square brackets is called a "character
3350 class". In a character class the only metacharacters are:
3351
3352 \ general escape character
3353 ^ negate the class, but only if the first character
3354 - indicates character range
3355 [ POSIX character class (only if followed by POSIX
3356 syntax)
3357 ] terminates the character class
3358
3359 The following sections describe the use of each of the metacharacters.
3360
3361
3362 BACKSLASH
3363
3364 The backslash character has several uses. Firstly, if it is followed by
3365 a non-alphanumeric character, it takes away any special meaning that
3366 character may have. This use of backslash as an escape character
3367 applies both inside and outside character classes.
3368
3369 For example, if you want to match a * character, you write \* in the
3370 pattern. This escaping action applies whether or not the following
3371 character would otherwise be interpreted as a metacharacter, so it is
3372 always safe to precede a non-alphanumeric with backslash to specify
3373 that it stands for itself. In particular, if you want to match a back-
3374 slash, you write \\.
3375
3376 If a pattern is compiled with the PCRE_EXTENDED option, whitespace in
3377 the pattern (other than in a character class) and characters between a
3378 # outside a character class and the next newline are ignored. An escap-
3379 ing backslash can be used to include a whitespace or # character as
3380 part of the pattern.
3381
3382 If you want to remove the special meaning from a sequence of charac-
3383 ters, you can do so by putting them between \Q and \E. This is differ-
3384 ent from Perl in that $ and @ are handled as literals in \Q...\E
3385 sequences in PCRE, whereas in Perl, $ and @ cause variable interpola-
3386 tion. Note the following examples:
3387
3388 Pattern PCRE matches Perl matches
3389
3390 \Qabc$xyz\E abc$xyz abc followed by the
3391 contents of $xyz
3392 \Qabc\$xyz\E abc\$xyz abc\$xyz
3393 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
3394
3395 The \Q...\E sequence is recognized both inside and outside character
3396 classes. An isolated \E that is not preceded by \Q is ignored.
3397
3398 Non-printing characters
3399
3400 A second use of backslash provides a way of encoding non-printing char-
3401 acters in patterns in a visible manner. There is no restriction on the
3402 appearance of non-printing characters, apart from the binary zero that
3403 terminates a pattern, but when a pattern is being prepared by text
3404 editing, it is often easier to use one of the following escape
3405 sequences than the binary character it represents:
3406
3407 \a alarm, that is, the BEL character (hex 07)
3408 \cx "control-x", where x is any character
3409 \e escape (hex 1B)
3410 \f formfeed (hex 0C)
3411 \n linefeed (hex 0A)
3412 \r carriage return (hex 0D)
3413 \t tab (hex 09)
3414 \ddd character with octal code ddd, or back reference
3415 \xhh character with hex code hh
3416 \x{hhh..} character with hex code hhh..
3417
3418 The precise effect of \cx is as follows: if x is a lower case letter,
3419 it is converted to upper case. Then bit 6 of the character (hex 40) is
3420 inverted. Thus \cz becomes hex 1A, but \c{ becomes hex 3B, while \c;
3421 becomes hex 7B.
3422
3423 After \x, from zero to two hexadecimal digits are read (letters can be
3424 in upper or lower case). Any number of hexadecimal digits may appear
3425 between \x{ and }, but the value of the character code must be less
3426 than 256 in non-UTF-8 mode, and less than 2**31 in UTF-8 mode. That is,
3427 the maximum value in hexadecimal is 7FFFFFFF. Note that this is bigger
3428 than the largest Unicode code point, which is 10FFFF.
3429
3430 If characters other than hexadecimal digits appear between \x{ and },
3431 or if there is no terminating }, this form of escape is not recognized.
3432 Instead, the initial \x will be interpreted as a basic hexadecimal
3433 escape, with no following digits, giving a character whose value is
3434 zero.
3435
3436 Characters whose value is less than 256 can be defined by either of the
3437 two syntaxes for \x. There is no difference in the way they are han-
3438 dled. For example, \xdc is exactly the same as \x{dc}.
3439
3440 After \0 up to two further octal digits are read. If there are fewer
3441 than two digits, just those that are present are used. Thus the
3442 sequence \0\x\07 specifies two binary zeros followed by a BEL character
3443 (code value 7). Make sure you supply two digits after the initial zero
3444 if the pattern character that follows is itself an octal digit.
3445
3446 The handling of a backslash followed by a digit other than 0 is compli-
3447 cated. Outside a character class, PCRE reads it and any following dig-
3448 its as a decimal number. If the number is less than 10, or if there
3449 have been at least that many previous capturing left parentheses in the
3450 expression, the entire sequence is taken as a back reference. A
3451 description of how this works is given later, following the discussion
3452 of parenthesized subpatterns.
3453
3454 Inside a character class, or if the decimal number is greater than 9
3455 and there have not been that many capturing subpatterns, PCRE re-reads
3456 up to three octal digits following the backslash, and uses them to gen-
3457 erate a data character. Any subsequent digits stand for themselves. In
3458 non-UTF-8 mode, the value of a character specified in octal must be
3459 less than \400. In UTF-8 mode, values up to \777 are permitted. For
3460 example:
3461
3462 \040 is another way of writing a space
3463 \40 is the same, provided there are fewer than 40
3464 previous capturing subpatterns
3465 \7 is always a back reference
3466 \11 might be a back reference, or another way of
3467 writing a tab
3468 \011 is always a tab
3469 \0113 is a tab followed by the character "3"
3470 \113 might be a back reference, otherwise the
3471 character with octal code 113
3472 \377 might be a back reference, otherwise
3473 the byte consisting entirely of 1 bits
3474 \81 is either a back reference, or a binary zero
3475 followed by the two characters "8" and "1"
3476
3477 Note that octal values of 100 or greater must not be introduced by a
3478 leading zero, because no more than three octal digits are ever read.
3479
3480 All the sequences that define a single character value can be used both
3481 inside and outside character classes. In addition, inside a character
3482 class, the sequence \b is interpreted as the backspace character (hex
3483 08). The sequences \B, \N, \R, and \X are not special inside a charac-
3484 ter class. Like any other unrecognized escape sequences, they are
3485 treated as the literal characters "B", "N", "R", and "X" by default,
3486 but cause an error if the PCRE_EXTRA option is set. Outside a character
3487 class, these sequences have different meanings.
3488
3489 Absolute and relative back references
3490
3491 The sequence \g followed by an unsigned or a negative number, option-
3492 ally enclosed in braces, is an absolute or relative back reference. A
3493 named back reference can be coded as \g{name}. Back references are dis-
3494 cussed later, following the discussion of parenthesized subpatterns.
3495
3496 Absolute and relative subroutine calls
3497
3498 For compatibility with Oniguruma, the non-Perl syntax \g followed by a
3499 name or a number enclosed either in angle brackets or single quotes, is
3500 an alternative syntax for referencing a subpattern as a "subroutine".
3501 Details are discussed later. Note that \g{...} (Perl syntax) and
3502 \g<...> (Oniguruma syntax) are not synonymous. The former is a back
3503 reference; the latter is a subroutine call.
3504
3505 Generic character types
3506
3507 Another use of backslash is for specifying generic character types:
3508
3509 \d any decimal digit
3510 \D any character that is not a decimal digit
3511 \h any horizontal whitespace character
3512 \H any character that is not a horizontal whitespace character
3513 \s any whitespace character
3514 \S any character that is not a whitespace character
3515 \v any vertical whitespace character
3516 \V any character that is not a vertical whitespace character
3517 \w any "word" character
3518 \W any "non-word" character
3519
3520 There is also the single sequence \N, which matches a non-newline char-
3521 acter. This is the same as the "." metacharacter when PCRE_DOTALL is
3522 not set.
3523
3524 Each pair of lower and upper case escape sequences partitions the com-
3525 plete set of characters into two disjoint sets. Any given character
3526 matches one, and only one, of each pair. The sequences can appear both
3527 inside and outside character classes. They each match one character of
3528 the appropriate type. If the current matching point is at the end of
3529 the subject string, all of them fail, because there is no character to
3530 match.
3531
3532 For compatibility with Perl, \s does not match the VT character (code
3533 11). This makes it different from the the POSIX "space" class. The \s
3534 characters are HT (9), LF (10), FF (12), CR (13), and space (32). If
3535 "use locale;" is included in a Perl script, \s may match the VT charac-
3536 ter. In PCRE, it never does.
3537
3538 A "word" character is an underscore or any character that is a letter
3539 or digit. By default, the definition of letters and digits is con-
3540 trolled by PCRE's low-valued character tables, and may vary if locale-
3541 specific matching is taking place (see "Locale support" in the pcreapi
3542 page). For example, in a French locale such as "fr_FR" in Unix-like
3543 systems, or "french" in Windows, some character codes greater than 128
3544 are used for accented letters, and these are then matched by \w. The
3545 use of locales with Unicode is discouraged.
3546
3547 By default, in UTF-8 mode, characters with values greater than 128
3548 never match \d, \s, or \w, and always match \D, \S, and \W. These
3549 sequences retain their original meanings from before UTF-8 support was
3550 available, mainly for efficiency reasons. However, if PCRE is compiled
3551 with Unicode property support, and the PCRE_UCP option is set, the be-
3552 haviour is changed so that Unicode properties are used to determine
3553 character types, as follows:
3554
3555 \d any character that \p{Nd} matches (decimal digit)
3556 \s any character that \p{Z} matches, plus HT, LF, FF, CR
3557 \w any character that \p{L} or \p{N} matches, plus underscore
3558
3559 The upper case escapes match the inverse sets of characters. Note that
3560 \d matches only decimal digits, whereas \w matches any Unicode digit,
3561 as well as any Unicode letter, and underscore. Note also that PCRE_UCP
3562 affects \b, and \B because they are defined in terms of \w and \W.
3563 Matching these sequences is noticeably slower when PCRE_UCP is set.
3564
3565 The sequences \h, \H, \v, and \V are features that were added to Perl
3566 at release 5.10. In contrast to the other sequences, which match only
3567 ASCII characters by default, these always match certain high-valued
3568 codepoints in UTF-8 mode, whether or not PCRE_UCP is set. The horizon-
3569 tal space characters are:
3570
3571 U+0009 Horizontal tab
3572 U+0020 Space
3573 U+00A0 Non-break space
3574 U+1680 Ogham space mark
3575 U+180E Mongolian vowel separator
3576 U+2000 En quad
3577 U+2001 Em quad
3578 U+2002 En space
3579 U+2003 Em space
3580 U+2004 Three-per-em space
3581 U+2005 Four-per-em space
3582 U+2006 Six-per-em space
3583 U+2007 Figure space
3584 U+2008 Punctuation space
3585 U+2009 Thin space
3586 U+200A Hair space
3587 U+202F Narrow no-break space
3588 U+205F Medium mathematical space
3589 U+3000 Ideographic space
3590
3591 The vertical space characters are:
3592
3593 U+000A Linefeed
3594 U+000B Vertical tab
3595 U+000C Formfeed
3596 U+000D Carriage return
3597 U+0085 Next line
3598 U+2028 Line separator
3599 U+2029 Paragraph separator
3600
3601 Newline sequences
3602
3603 Outside a character class, by default, the escape sequence \R matches
3604 any Unicode newline sequence. In non-UTF-8 mode \R is equivalent to the
3605 following:
3606
3607 (?>\r\n|\n|\x0b|\f|\r|\x85)
3608
3609 This is an example of an "atomic group", details of which are given
3610 below. This particular group matches either the two-character sequence
3611 CR followed by LF, or one of the single characters LF (linefeed,
3612 U+000A), VT (vertical tab, U+000B), FF (formfeed, U+000C), CR (carriage
3613 return, U+000D), or NEL (next line, U+0085). The two-character sequence
3614 is treated as a single unit that cannot be split.
3615
3616 In UTF-8 mode, two additional characters whose codepoints are greater
3617 than 255 are added: LS (line separator, U+2028) and PS (paragraph sepa-
3618 rator, U+2029). Unicode character property support is not needed for
3619 these characters to be recognized.
3620
3621 It is possible to restrict \R to match only CR, LF, or CRLF (instead of
3622 the complete set of Unicode line endings) by setting the option
3623 PCRE_BSR_ANYCRLF either at compile time or when the pattern is matched.
3624 (BSR is an abbrevation for "backslash R".) This can be made the default
3625 when PCRE is built; if this is the case, the other behaviour can be
3626 requested via the PCRE_BSR_UNICODE option. It is also possible to
3627 specify these settings by starting a pattern string with one of the
3628 following sequences:
3629
3630 (*BSR_ANYCRLF) CR, LF, or CRLF only
3631 (*BSR_UNICODE) any Unicode newline sequence
3632
3633 These override the default and the options given to pcre_compile() or
3634 pcre_compile2(), but they can be overridden by options given to
3635 pcre_exec() or pcre_dfa_exec(). Note that these special settings, which
3636 are not Perl-compatible, are recognized only at the very start of a
3637 pattern, and that they must be in upper case. If more than one of them
3638 is present, the last one is used. They can be combined with a change of
3639 newline convention; for example, a pattern can start with:
3640
3641 (*ANY)(*BSR_ANYCRLF)
3642
3643 They can also be combined with the (*UTF8) or (*UCP) special sequences.
3644 Inside a character class, \R is treated as an unrecognized escape
3645 sequence, and so matches the letter "R" by default, but causes an error
3646 if PCRE_EXTRA is set.
3647
3648 Unicode character properties
3649
3650 When PCRE is built with Unicode character property support, three addi-
3651 tional escape sequences that match characters with specific properties
3652 are available. When not in UTF-8 mode, these sequences are of course
3653 limited to testing characters whose codepoints are less than 256, but
3654 they do work in this mode. The extra escape sequences are:
3655
3656 \p{xx} a character with the xx property
3657 \P{xx} a character without the xx property
3658 \X an extended Unicode sequence
3659
3660 The property names represented by xx above are limited to the Unicode
3661 script names, the general category properties, "Any", which matches any
3662 character (including newline), and some special PCRE properties
3663 (described in the next section). Other Perl properties such as "InMu-
3664 sicalSymbols" are not currently supported by PCRE. Note that \P{Any}
3665 does not match any characters, so always causes a match failure.
3666
3667 Sets of Unicode characters are defined as belonging to certain scripts.
3668 A character from one of these sets can be matched using a script name.
3669 For example:
3670
3671 \p{Greek}
3672 \P{Han}
3673
3674 Those that are not part of an identified script are lumped together as
3675 "Common". The current list of scripts is:
3676
3677 Arabic, Armenian, Avestan, Balinese, Bamum, Bengali, Bopomofo, Braille,
3678 Buginese, Buhid, Canadian_Aboriginal, Carian, Cham, Cherokee, Common,
3679 Coptic, Cuneiform, Cypriot, Cyrillic, Deseret, Devanagari, Egyp-
3680 tian_Hieroglyphs, Ethiopic, Georgian, Glagolitic, Gothic, Greek,
3681 Gujarati, Gurmukhi, Han, Hangul, Hanunoo, Hebrew, Hiragana, Impe-
3682 rial_Aramaic, Inherited, Inscriptional_Pahlavi, Inscriptional_Parthian,
3683 Javanese, Kaithi, Kannada, Katakana, Kayah_Li, Kharoshthi, Khmer, Lao,
3684 Latin, Lepcha, Limbu, Linear_B, Lisu, Lycian, Lydian, Malayalam,
3685 Meetei_Mayek, Mongolian, Myanmar, New_Tai_Lue, Nko, Ogham, Old_Italic,
3686 Old_Persian, Old_South_Arabian, Old_Turkic, Ol_Chiki, Oriya, Osmanya,
3687 Phags_Pa, Phoenician, Rejang, Runic, Samaritan, Saurashtra, Shavian,
3688 Sinhala, Sundanese, Syloti_Nagri, Syriac, Tagalog, Tagbanwa, Tai_Le,
3689 Tai_Tham, Tai_Viet, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh,
3690 Ugaritic, Vai, Yi.
3691
3692 Each character has exactly one Unicode general category property, spec-
3693 ified by a two-letter abbreviation. For compatibility with Perl, nega-
3694 tion can be specified by including a circumflex between the opening
3695 brace and the property name. For example, \p{^Lu} is the same as
3696 \P{Lu}.
3697
3698 If only one letter is specified with \p or \P, it includes all the gen-
3699 eral category properties that start with that letter. In this case, in
3700 the absence of negation, the curly brackets in the escape sequence are
3701 optional; these two examples have the same effect:
3702
3703 \p{L}
3704 \pL
3705
3706 The following general category property codes are supported:
3707
3708 C Other
3709 Cc Control
3710 Cf Format
3711 Cn Unassigned
3712 Co Private use
3713 Cs Surrogate
3714
3715 L Letter
3716 Ll Lower case letter
3717 Lm Modifier letter
3718 Lo Other letter
3719 Lt Title case letter
3720 Lu Upper case letter
3721
3722 M Mark
3723 Mc Spacing mark
3724 Me Enclosing mark
3725 Mn Non-spacing mark
3726
3727 N Number
3728 Nd Decimal number
3729 Nl Letter number
3730 No Other number
3731
3732 P Punctuation
3733 Pc Connector punctuation
3734 Pd Dash punctuation
3735 Pe Close punctuation
3736 Pf Final punctuation
3737 Pi Initial punctuation
3738 Po Other punctuation
3739 Ps Open punctuation
3740
3741 S Symbol
3742 Sc Currency symbol
3743 Sk Modifier symbol
3744 Sm Mathematical symbol
3745 So Other symbol
3746
3747 Z Separator
3748 Zl Line separator
3749 Zp Paragraph separator
3750 Zs Space separator
3751
3752 The special property L& is also supported: it matches a character that
3753 has the Lu, Ll, or Lt property, in other words, a letter that is not
3754 classified as a modifier or "other".
3755
3756 The Cs (Surrogate) property applies only to characters in the range
3757 U+D800 to U+DFFF. Such characters are not valid in UTF-8 strings (see
3758 RFC 3629) and so cannot be tested by PCRE, unless UTF-8 validity check-
3759 ing has been turned off (see the discussion of PCRE_NO_UTF8_CHECK in
3760 the pcreapi page). Perl does not support the Cs property.
3761
3762 The long synonyms for property names that Perl supports (such as
3763 \p{Letter}) are not supported by PCRE, nor is it permitted to prefix
3764 any of these properties with "Is".
3765
3766 No character that is in the Unicode table has the Cn (unassigned) prop-
3767 erty. Instead, this property is assumed for any code point that is not
3768 in the Unicode table.
3769
3770 Specifying caseless matching does not affect these escape sequences.
3771 For example, \p{Lu} always matches only upper case letters.
3772
3773 The \X escape matches any number of Unicode characters that form an
3774 extended Unicode sequence. \X is equivalent to
3775
3776 (?>\PM\pM*)
3777
3778 That is, it matches a character without the "mark" property, followed
3779 by zero or more characters with the "mark" property, and treats the
3780 sequence as an atomic group (see below). Characters with the "mark"
3781 property are typically accents that affect the preceding character.
3782 None of them have codepoints less than 256, so in non-UTF-8 mode \X
3783 matches any one character.
3784
3785 Matching characters by Unicode property is not fast, because PCRE has
3786 to search a structure that contains data for over fifteen thousand
3787 characters. That is why the traditional escape sequences such as \d and
3788 \w do not use Unicode properties in PCRE by default, though you can
3789 make them do so by setting the PCRE_UCP option for pcre_compile() or by
3790 starting the pattern with (*UCP).
3791
3792 PCRE's additional properties
3793
3794 As well as the standard Unicode properties described in the previous
3795 section, PCRE supports four more that make it possible to convert tra-
3796 ditional escape sequences such as \w and \s and POSIX character classes
3797 to use Unicode properties. PCRE uses these non-standard, non-Perl prop-
3798 erties internally when PCRE_UCP is set. They are:
3799
3800 Xan Any alphanumeric character
3801 Xps Any POSIX space character
3802 Xsp Any Perl space character
3803 Xwd Any Perl "word" character
3804
3805 Xan matches characters that have either the L (letter) or the N (num-
3806 ber) property. Xps matches the characters tab, linefeed, vertical tab,
3807 formfeed, or carriage return, and any other character that has the Z
3808 (separator) property. Xsp is the same as Xps, except that vertical tab
3809 is excluded. Xwd matches the same characters as Xan, plus underscore.
3810
3811 Resetting the match start
3812
3813 The escape sequence \K causes any previously matched characters not to
3814 be included in the final matched sequence. For example, the pattern:
3815
3816 foo\Kbar
3817
3818 matches "foobar", but reports that it has matched "bar". This feature
3819 is similar to a lookbehind assertion (described below). However, in
3820 this case, the part of the subject before the real match does not have
3821 to be of fixed length, as lookbehind assertions do. The use of \K does
3822 not interfere with the setting of captured substrings. For example,
3823 when the pattern
3824
3825 (foo)\Kbar
3826
3827 matches "foobar", the first substring is still set to "foo".
3828
3829 Perl documents that the use of \K within assertions is "not well
3830 defined". In PCRE, \K is acted upon when it occurs inside positive
3831 assertions, but is ignored in negative assertions.
3832
3833 Simple assertions
3834
3835 The final use of backslash is for certain simple assertions. An asser-
3836 tion specifies a condition that has to be met at a particular point in
3837 a match, without consuming any characters from the subject string. The
3838 use of subpatterns for more complicated assertions is described below.
3839 The backslashed assertions are:
3840
3841 \b matches at a word boundary
3842 \B matches when not at a word boundary
3843 \A matches at the start of the subject
3844 \Z matches at the end of the subject
3845 also matches before a newline at the end of the subject
3846 \z matches only at the end of the subject
3847 \G matches at the first matching position in the subject
3848
3849 Inside a character class, \b has a different meaning; it matches the
3850 backspace character. If any other of these assertions appears in a
3851 character class, by default it matches the corresponding literal char-
3852 acter (for example, \B matches the letter B). However, if the
3853 PCRE_EXTRA option is set, an "invalid escape sequence" error is gener-
3854 ated instead.
3855
3856 A word boundary is a position in the subject string where the current
3857 character and the previous character do not both match \w or \W (i.e.
3858 one matches \w and the other matches \W), or the start or end of the
3859 string if the first or last character matches \w, respectively. In
3860 UTF-8 mode, the meanings of \w and \W can be changed by setting the
3861 PCRE_UCP option. When this is done, it also affects \b and \B. Neither
3862 PCRE nor Perl has a separate "start of word" or "end of word" metase-
3863 quence. However, whatever follows \b normally determines which it is.
3864 For example, the fragment \ba matches "a" at the start of a word.
3865
3866 The \A, \Z, and \z assertions differ from the traditional circumflex
3867 and dollar (described in the next section) in that they only ever match
3868 at the very start and end of the subject string, whatever options are
3869 set. Thus, they are independent of multiline mode. These three asser-
3870 tions are not affected by the PCRE_NOTBOL or PCRE_NOTEOL options, which
3871 affect only the behaviour of the circumflex and dollar metacharacters.
3872 However, if the startoffset argument of pcre_exec() is non-zero, indi-
3873 cating that matching is to start at a point other than the beginning of
3874 the subject, \A can never match. The difference between \Z and \z is
3875 that \Z matches before a newline at the end of the string as well as at
3876 the very end, whereas \z matches only at the end.
3877
3878 The \G assertion is true only when the current matching position is at
3879 the start point of the match, as specified by the startoffset argument
3880 of pcre_exec(). It differs from \A when the value of startoffset is
3881 non-zero. By calling pcre_exec() multiple times with appropriate argu-
3882 ments, you can mimic Perl's /g option, and it is in this kind of imple-
3883 mentation where \G can be useful.
3884
3885 Note, however, that PCRE's interpretation of \G, as the start of the
3886 current match, is subtly different from Perl's, which defines it as the
3887 end of the previous match. In Perl, these can be different when the
3888 previously matched string was empty. Because PCRE does just one match
3889 at a time, it cannot reproduce this behaviour.
3890
3891 If all the alternatives of a pattern begin with \G, the expression is
3892 anchored to the starting match position, and the "anchored" flag is set
3893 in the compiled regular expression.
3894
3895
3896 CIRCUMFLEX AND DOLLAR
3897
3898 Outside a character class, in the default matching mode, the circumflex
3899 character is an assertion that is true only if the current matching
3900 point is at the start of the subject string. If the startoffset argu-
3901 ment of pcre_exec() is non-zero, circumflex can never match if the
3902 PCRE_MULTILINE option is unset. Inside a character class, circumflex
3903 has an entirely different meaning (see below).
3904
3905 Circumflex need not be the first character of the pattern if a number
3906 of alternatives are involved, but it should be the first thing in each
3907 alternative in which it appears if the pattern is ever to match that
3908 branch. If all possible alternatives start with a circumflex, that is,
3909 if the pattern is constrained to match only at the start of the sub-
3910 ject, it is said to be an "anchored" pattern. (There are also other
3911 constructs that can cause a pattern to be anchored.)
3912
3913 A dollar character is an assertion that is true only if the current
3914 matching point is at the end of the subject string, or immediately
3915 before a newline at the end of the string (by default). Dollar need not
3916 be the last character of the pattern if a number of alternatives are
3917 involved, but it should be the last item in any branch in which it
3918 appears. Dollar has no special meaning in a character class.
3919
3920 The meaning of dollar can be changed so that it matches only at the
3921 very end of the string, by setting the PCRE_DOLLAR_ENDONLY option at
3922 compile time. This does not affect the \Z assertion.
3923
3924 The meanings of the circumflex and dollar characters are changed if the
3925 PCRE_MULTILINE option is set. When this is the case, a circumflex
3926 matches immediately after internal newlines as well as at the start of
3927 the subject string. It does not match after a newline that ends the
3928 string. A dollar matches before any newlines in the string, as well as
3929 at the very end, when PCRE_MULTILINE is set. When newline is specified
3930 as the two-character sequence CRLF, isolated CR and LF characters do
3931 not indicate newlines.
3932
3933 For example, the pattern /^abc$/ matches the subject string "def\nabc"
3934 (where \n represents a newline) in multiline mode, but not otherwise.
3935 Consequently, patterns that are anchored in single line mode because
3936 all branches start with ^ are not anchored in multiline mode, and a
3937 match for circumflex is possible when the startoffset argument of
3938 pcre_exec() is non-zero. The PCRE_DOLLAR_ENDONLY option is ignored if
3939 PCRE_MULTILINE is set.
3940
3941 Note that the sequences \A, \Z, and \z can be used to match the start
3942 and end of the subject in both modes, and if all branches of a pattern
3943 start with \A it is always anchored, whether or not PCRE_MULTILINE is
3944 set.
3945
3946
3947 FULL STOP (PERIOD, DOT) AND \N
3948
3949 Outside a character class, a dot in the pattern matches any one charac-
3950 ter in the subject string except (by default) a character that signi-
3951 fies the end of a line. In UTF-8 mode, the matched character may be
3952 more than one byte long.
3953
3954 When a line ending is defined as a single character, dot never matches
3955 that character; when the two-character sequence CRLF is used, dot does
3956 not match CR if it is immediately followed by LF, but otherwise it
3957 matches all characters (including isolated CRs and LFs). When any Uni-
3958 code line endings are being recognized, dot does not match CR or LF or
3959 any of the other line ending characters.
3960
3961 The behaviour of dot with regard to newlines can be changed. If the
3962 PCRE_DOTALL option is set, a dot matches any one character, without
3963 exception. If the two-character sequence CRLF is present in the subject
3964 string, it takes two dots to match it.
3965
3966 The handling of dot is entirely independent of the handling of circum-
3967 flex and dollar, the only relationship being that they both involve
3968 newlines. Dot has no special meaning in a character class.
3969
3970 The escape sequence \N behaves like a dot, except that it is not
3971 affected by the PCRE_DOTALL option. In other words, it matches any
3972 character except one that signifies the end of a line.
3973
3974
3975 MATCHING A SINGLE BYTE
3976
3977 Outside a character class, the escape sequence \C matches any one byte,
3978 both in and out of UTF-8 mode. Unlike a dot, it always matches any
3979 line-ending characters. The feature is provided in Perl in order to
3980 match individual bytes in UTF-8 mode. Because it breaks up UTF-8 char-
3981 acters into individual bytes, the rest of the string may start with a
3982 malformed UTF-8 character. For this reason, the \C escape sequence is
3983 best avoided.
3984
3985 PCRE does not allow \C to appear in lookbehind assertions (described
3986 below), because in UTF-8 mode this would make it impossible to calcu-
3987 late the length of the lookbehind.
3988
3989
3990 SQUARE BRACKETS AND CHARACTER CLASSES
3991
3992 An opening square bracket introduces a character class, terminated by a
3993 closing square bracket. A closing square bracket on its own is not spe-
3994 cial by default. However, if the PCRE_JAVASCRIPT_COMPAT option is set,
3995 a lone closing square bracket causes a compile-time error. If a closing
3996 square bracket is required as a member of the class, it should be the
3997 first data character in the class (after an initial circumflex, if
3998 present) or escaped with a backslash.
3999
4000 A character class matches a single character in the subject. In UTF-8
4001 mode, the character may be more than one byte long. A matched character
4002 must be in the set of characters defined by the class, unless the first
4003 character in the class definition is a circumflex, in which case the
4004 subject character must not be in the set defined by the class. If a
4005 circumflex is actually required as a member of the class, ensure it is
4006 not the first character, or escape it with a backslash.
4007
4008 For example, the character class [aeiou] matches any lower case vowel,
4009 while [^aeiou] matches any character that is not a lower case vowel.
4010 Note that a circumflex is just a convenient notation for specifying the
4011 characters that are in the class by enumerating those that are not. A
4012 class that starts with a circumflex is not an assertion; it still con-
4013 sumes a character from the subject string, and therefore it fails if
4014 the current pointer is at the end of the string.
4015
4016 In UTF-8 mode, characters with values greater than 255 can be included
4017 in a class as a literal string of bytes, or by using the \x{ escaping
4018 mechanism.
4019
4020 When caseless matching is set, any letters in a class represent both
4021 their upper case and lower case versions, so for example, a caseless
4022 [aeiou] matches "A" as well as "a", and a caseless [^aeiou] does not
4023 match "A", whereas a caseful version would. In UTF-8 mode, PCRE always
4024 understands the concept of case for characters whose values are less
4025 than 128, so caseless matching is always possible. For characters with
4026 higher values, the concept of case is supported if PCRE is compiled
4027 with Unicode property support, but not otherwise. If you want to use
4028 caseless matching in UTF8-mode for characters 128 and above, you must
4029 ensure that PCRE is compiled with Unicode property support as well as
4030 with UTF-8 support.
4031
4032 Characters that might indicate line breaks are never treated in any
4033 special way when matching character classes, whatever line-ending
4034 sequence is in use, and whatever setting of the PCRE_DOTALL and
4035 PCRE_MULTILINE options is used. A class such as [^a] always matches one
4036 of these characters.
4037
4038 The minus (hyphen) character can be used to specify a range of charac-
4039 ters in a character class. For example, [d-m] matches any letter
4040 between d and m, inclusive. If a minus character is required in a
4041 class, it must be escaped with a backslash or appear in a position
4042 where it cannot be interpreted as indicating a range, typically as the
4043 first or last character in the class.
4044
4045 It is not possible to have the literal character "]" as the end charac-
4046 ter of a range. A pattern such as [W-]46] is interpreted as a class of
4047 two characters ("W" and "-") followed by a literal string "46]", so it
4048 would match "W46]" or "-46]". However, if the "]" is escaped with a
4049 backslash it is interpreted as the end of range, so [W-\]46] is inter-
4050 preted as a class containing a range followed by two other characters.
4051 The octal or hexadecimal representation of "]" can also be used to end
4052 a range.
4053
4054 Ranges operate in the collating sequence of character values. They can
4055 also be used for characters specified numerically, for example
4056 [\000-\037]. In UTF-8 mode, ranges can include characters whose values
4057 are greater than 255, for example [\x{100}-\x{2ff}].
4058
4059 If a range that includes letters is used when caseless matching is set,
4060 it matches the letters in either case. For example, [W-c] is equivalent
4061 to [][\\^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if
4062 character tables for a French locale are in use, [\xc8-\xcb] matches
4063 accented E characters in both cases. In UTF-8 mode, PCRE supports the
4064 concept of case for characters with values greater than 128 only when
4065 it is compiled with Unicode property support.
4066
4067 The character types \d, \D, \h, \H, \p, \P, \s, \S, \v, \V, \w, and \W
4068 may also appear in a character class, and add the characters that they
4069 match to the class. For example, [\dABCDEF] matches any hexadecimal
4070 digit. A circumflex can conveniently be used with the upper case char-
4071 acter types to specify a more restricted set of characters than the
4072 matching lower case type. For example, the class [^\W_] matches any
4073 letter or digit, but not underscore.
4074
4075 The only metacharacters that are recognized in character classes are
4076 backslash, hyphen (only where it can be interpreted as specifying a
4077 range), circumflex (only at the start), opening square bracket (only
4078 when it can be interpreted as introducing a POSIX class name - see the
4079 next section), and the terminating closing square bracket. However,
4080 escaping other non-alphanumeric characters does no harm.
4081
4082
4083 POSIX CHARACTER CLASSES
4084
4085 Perl supports the POSIX notation for character classes. This uses names
4086 enclosed by [: and :] within the enclosing square brackets. PCRE also
4087 supports this notation. For example,
4088
4089 [01[:alpha:]%]
4090
4091 matches "0", "1", any alphabetic character, or "%". The supported class
4092 names are:
4093
4094 alnum letters and digits
4095 alpha letters
4096 ascii character codes 0 - 127
4097 blank space or tab only
4098 cntrl control characters
4099 digit decimal digits (same as \d)
4100 graph printing characters, excluding space
4101 lower lower case letters
4102 print printing characters, including space
4103 punct printing characters, excluding letters and digits and space
4104 space white space (not quite the same as \s)
4105 upper upper case letters
4106 word "word" characters (same as \w)
4107 xdigit hexadecimal digits
4108
4109 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13),
4110 and space (32). Notice that this list includes the VT character (code
4111 11). This makes "space" different to \s, which does not include VT (for
4112 Perl compatibility).
4113
4114 The name "word" is a Perl extension, and "blank" is a GNU extension
4115 from Perl 5.8. Another Perl extension is negation, which is indicated
4116 by a ^ character after the colon. For example,
4117
4118 [12[:^digit:]]
4119
4120 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the
4121 POSIX syntax [.ch.] and [=ch=] where "ch" is a "collating element", but
4122 these are not supported, and an error is given if they are encountered.
4123
4124 By default, in UTF-8 mode, characters with values greater than 128 do
4125 not match any of the POSIX character classes. However, if the PCRE_UCP
4126 option is passed to pcre_compile(), some of the classes are changed so
4127 that Unicode character properties are used. This is achieved by replac-
4128 ing the POSIX classes by other sequences, as follows:
4129
4130 [:alnum:] becomes \p{Xan}
4131 [:alpha:] becomes \p{L}
4132 [:blank:] becomes \h
4133 [:digit:] becomes \p{Nd}
4134 [:lower:] becomes \p{Ll}
4135 [:space:] becomes \p{Xps}
4136 [:upper:] becomes \p{Lu}
4137 [:word:] becomes \p{Xwd}
4138
4139 Negated versions, such as [:^alpha:] use \P instead of \p. The other
4140 POSIX classes are unchanged, and match only characters with code points
4141 less than 128.
4142
4143
4144 VERTICAL BAR
4145
4146 Vertical bar characters are used to separate alternative patterns. For
4147 example, the pattern
4148
4149 gilbert|sullivan
4150
4151 matches either "gilbert" or "sullivan". Any number of alternatives may
4152 appear, and an empty alternative is permitted (matching the empty
4153 string). The matching process tries each alternative in turn, from left
4154 to right, and the first one that succeeds is used. If the alternatives
4155 are within a subpattern (defined below), "succeeds" means matching the
4156 rest of the main pattern as well as the alternative in the subpattern.
4157
4158
4159 INTERNAL OPTION SETTING
4160
4161 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
4162 PCRE_EXTENDED options (which are Perl-compatible) can be changed from
4163 within the pattern by a sequence of Perl option letters enclosed
4164 between "(?" and ")". The option letters are
4165
4166 i for PCRE_CASELESS
4167 m for PCRE_MULTILINE
4168 s for PCRE_DOTALL
4169 x for PCRE_EXTENDED
4170
4171 For example, (?im) sets caseless, multiline matching. It is also possi-
4172 ble to unset these options by preceding the letter with a hyphen, and a
4173 combined setting and unsetting such as (?im-sx), which sets PCRE_CASE-
4174 LESS and PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED,
4175 is also permitted. If a letter appears both before and after the
4176 hyphen, the option is unset.
4177
4178 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA
4179 can be changed in the same way as the Perl-compatible options by using
4180 the characters J, U and X respectively.
4181
4182 When one of these option changes occurs at top level (that is, not
4183 inside subpattern parentheses), the change applies to the remainder of
4184 the pattern that follows. If the change is placed right at the start of
4185 a pattern, PCRE extracts it into the global options (and it will there-
4186 fore show up in data extracted by the pcre_fullinfo() function).
4187
4188 An option change within a subpattern (see below for a description of
4189 subpatterns) affects only that part of the subpattern that follows it,
4190 so
4191
4192 (a(?i)b)c
4193
4194 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not
4195 used). By this means, options can be made to have different settings
4196 in different parts of the pattern. Any changes made in one alternative
4197 do carry on into subsequent branches within the same subpattern. For
4198 example,
4199
4200 (a(?i)b|c)
4201
4202 matches "ab", "aB", "c", and "C", even though when matching "C" the
4203 first branch is abandoned before the option setting. This is because
4204 the effects of option settings happen at compile time. There would be
4205 some very weird behaviour otherwise.
4206
4207 Note: There are other PCRE-specific options that can be set by the
4208 application when the compile or match functions are called. In some
4209 cases the pattern can contain special leading sequences such as (*CRLF)
4210 to override what the application has set or what has been defaulted.
4211 Details are given in the section entitled "Newline sequences" above.
4212 There are also the (*UTF8) and (*UCP) leading sequences that can be
4213 used to set UTF-8 and Unicode property modes; they are equivalent to
4214 setting the PCRE_UTF8 and the PCRE_UCP options, respectively.
4215
4216
4217 SUBPATTERNS
4218
4219 Subpatterns are delimited by parentheses (round brackets), which can be
4220 nested. Turning part of a pattern into a subpattern does two things:
4221
4222 1. It localizes a set of alternatives. For example, the pattern
4223
4224 cat(aract|erpillar|)
4225
4226 matches "cataract", "caterpillar", or "cat". Without the parentheses,
4227 it would match "cataract", "erpillar" or an empty string.
4228
4229 2. It sets up the subpattern as a capturing subpattern. This means
4230 that, when the whole pattern matches, that portion of the subject
4231 string that matched the subpattern is passed back to the caller via the
4232 ovector argument of pcre_exec(). Opening parentheses are counted from
4233 left to right (starting from 1) to obtain numbers for the capturing
4234 subpatterns. For example, if the string "the red king" is matched
4235 against the pattern
4236
4237 the ((red|white) (king|queen))
4238
4239 the captured substrings are "red king", "red", and "king", and are num-
4240 bered 1, 2, and 3, respectively.
4241
4242 The fact that plain parentheses fulfil two functions is not always
4243 helpful. There are often times when a grouping subpattern is required
4244 without a capturing requirement. If an opening parenthesis is followed
4245 by a question mark and a colon, the subpattern does not do any captur-
4246 ing, and is not counted when computing the number of any subsequent
4247 capturing subpatterns. For example, if the string "the white queen" is
4248 matched against the pattern
4249
4250 the ((?:red|white) (king|queen))
4251
4252 the captured substrings are "white queen" and "queen", and are numbered
4253 1 and 2. The maximum number of capturing subpatterns is 65535.
4254
4255 As a convenient shorthand, if any option settings are required at the
4256 start of a non-capturing subpattern, the option letters may appear
4257 between the "?" and the ":". Thus the two patterns
4258
4259 (?i:saturday|sunday)
4260 (?:(?i)saturday|sunday)
4261
4262 match exactly the same set of strings. Because alternative branches are
4263 tried from left to right, and options are not reset until the end of
4264 the subpattern is reached, an option setting in one branch does affect
4265 subsequent branches, so the above patterns match "SUNDAY" as well as
4266 "Saturday".
4267
4268
4269 DUPLICATE SUBPATTERN NUMBERS
4270
4271 Perl 5.10 introduced a feature whereby each alternative in a subpattern
4272 uses the same numbers for its capturing parentheses. Such a subpattern
4273 starts with (?| and is itself a non-capturing subpattern. For example,
4274 consider this pattern:
4275
4276 (?|(Sat)ur|(Sun))day
4277
4278 Because the two alternatives are inside a (?| group, both sets of cap-
4279 turing parentheses are numbered one. Thus, when the pattern matches,
4280 you can look at captured substring number one, whichever alternative
4281 matched. This construct is useful when you want to capture part, but
4282 not all, of one of a number of alternatives. Inside a (?| group, paren-
4283 theses are numbered as usual, but the number is reset at the start of
4284 each branch. The numbers of any capturing parentheses that follow the
4285 subpattern start after the highest number used in any branch. The fol-
4286 lowing example is taken from the Perl documentation. The numbers under-
4287 neath show in which buffer the captured content will be stored.
4288
4289 # before ---------------branch-reset----------- after
4290 / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
4291 # 1 2 2 3 2 3 4
4292
4293 A back reference to a numbered subpattern uses the most recent value
4294 that is set for that number by any subpattern. The following pattern
4295 matches "abcabc" or "defdef":
4296
4297 /(?|(abc)|(def))\1/
4298
4299 In contrast, a recursive or "subroutine" call to a numbered subpattern
4300 always refers to the first one in the pattern with the given number.
4301 The following pattern matches "abcabc" or "defabc":
4302
4303 /(?|(abc)|(def))(?1)/
4304
4305 If a condition test for a subpattern's having matched refers to a non-
4306 unique number, the test is true if any of the subpatterns of that num-
4307 ber have matched.
4308
4309 An alternative approach to using this "branch reset" feature is to use
4310 duplicate named subpatterns, as described in the next section.
4311
4312
4313 NAMED SUBPATTERNS
4314
4315 Identifying capturing parentheses by number is simple, but it can be
4316 very hard to keep track of the numbers in complicated regular expres-
4317 sions. Furthermore, if an expression is modified, the numbers may
4318 change. To help with this difficulty, PCRE supports the naming of sub-
4319 patterns. This feature was not added to Perl until release 5.10. Python
4320 had the feature earlier, and PCRE introduced it at release 4.0, using
4321 the Python syntax. PCRE now supports both the Perl and the Python syn-
4322 tax. Perl allows identically numbered subpatterns to have different
4323 names, but PCRE does not.
4324
4325 In PCRE, a subpattern can be named in one of three ways: (?<name>...)
4326 or (?'name'...) as in Perl, or (?P<name>...) as in Python. References
4327 to capturing parentheses from other parts of the pattern, such as back
4328 references, recursion, and conditions, can be made by name as well as
4329 by number.
4330
4331 Names consist of up to 32 alphanumeric characters and underscores.
4332 Named capturing parentheses are still allocated numbers as well as
4333 names, exactly as if the names were not present. The PCRE API provides
4334 function calls for extracting the name-to-number translation table from
4335 a compiled pattern. There is also a convenience function for extracting
4336 a captured substring by name.
4337
4338 By default, a name must be unique within a pattern, but it is possible
4339 to relax this constraint by setting the PCRE_DUPNAMES option at compile
4340 time. (Duplicate names are also always permitted for subpatterns with
4341 the same number, set up as described in the previous section.) Dupli-
4342 cate names can be useful for patterns where only one instance of the
4343 named parentheses can match. Suppose you want to match the name of a
4344 weekday, either as a 3-letter abbreviation or as the full name, and in
4345 both cases you want to extract the abbreviation. This pattern (ignoring
4346 the line breaks) does the job:
4347
4348 (?<DN>Mon|Fri|Sun)(?:day)?|
4349 (?<DN>Tue)(?:sday)?|
4350 (?<DN>Wed)(?:nesday)?|
4351 (?<DN>Thu)(?:rsday)?|
4352 (?<DN>Sat)(?:urday)?
4353
4354 There are five capturing substrings, but only one is ever set after a
4355 match. (An alternative way of solving this problem is to use a "branch
4356 reset" subpattern, as described in the previous section.)
4357
4358 The convenience function for extracting the data by name returns the
4359 substring for the first (and in this example, the only) subpattern of
4360 that name that matched. This saves searching to find which numbered
4361 subpattern it was.
4362
4363 If you make a back reference to a non-unique named subpattern from
4364 elsewhere in the pattern, the one that corresponds to the first occur-
4365 rence of the name is used. In the absence of duplicate numbers (see the
4366 previous section) this is the one with the lowest number. If you use a
4367 named reference in a condition test (see the section about conditions
4368 below), either to check whether a subpattern has matched, or to check
4369 for recursion, all subpatterns with the same name are tested. If the
4370 condition is true for any one of them, the overall condition is true.
4371 This is the same behaviour as testing by number. For further details of
4372 the interfaces for handling named subpatterns, see the pcreapi documen-
4373 tation.
4374
4375 Warning: You cannot use different names to distinguish between two sub-
4376 patterns with the same number because PCRE uses only the numbers when
4377 matching. For this reason, an error is given at compile time if differ-
4378 ent names are given to subpatterns with the same number. However, you
4379 can give the same name to subpatterns with the same number, even when
4380 PCRE_DUPNAMES is not set.
4381
4382
4383 REPETITION
4384
4385 Repetition is specified by quantifiers, which can follow any of the
4386 following items:
4387
4388 a literal data character
4389 the dot metacharacter
4390 the \C escape sequence
4391 the \X escape sequence (in UTF-8 mode with Unicode properties)
4392 the \R escape sequence
4393 an escape such as \d or \pL that matches a single character
4394 a character class
4395 a back reference (see next section)
4396 a parenthesized subpattern (unless it is an assertion)
4397 a recursive or "subroutine" call to a subpattern
4398
4399 The general repetition quantifier specifies a minimum and maximum num-
4400 ber of permitted matches, by giving the two numbers in curly brackets
4401 (braces), separated by a comma. The numbers must be less than 65536,
4402 and the first must be less than or equal to the second. For example:
4403
4404 z{2,4}
4405
4406 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a
4407 special character. If the second number is omitted, but the comma is
4408 present, there is no upper limit; if the second number and the comma
4409 are both omitted, the quantifier specifies an exact number of required
4410 matches. Thus
4411
4412 [aeiou]{3,}
4413
4414 matches at least 3 successive vowels, but may match many more, while
4415
4416 \d{8}
4417
4418 matches exactly 8 digits. An opening curly bracket that appears in a
4419 position where a quantifier is not allowed, or one that does not match
4420 the syntax of a quantifier, is taken as a literal character. For exam-
4421 ple, {,6} is not a quantifier, but a literal string of four characters.
4422
4423 In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to
4424 individual bytes. Thus, for example, \x{100}{2} matches two UTF-8 char-
4425 acters, each of which is represented by a two-byte sequence. Similarly,
4426 when Unicode property support is available, \X{3} matches three Unicode
4427 extended sequences, each of which may be several bytes long (and they
4428 may be of different lengths).
4429
4430 The quantifier {0} is permitted, causing the expression to behave as if
4431 the previous item and the quantifier were not present. This may be use-
4432 ful for subpatterns that are referenced as subroutines from elsewhere
4433 in the pattern (but see also the section entitled "Defining subpatterns
4434 for use by reference only" below). Items other than subpatterns that
4435 have a {0} quantifier are omitted from the compiled pattern.
4436
4437 For convenience, the three most common quantifiers have single-charac-
4438 ter abbreviations:
4439
4440 * is equivalent to {0,}
4441 + is equivalent to {1,}
4442 ? is equivalent to {0,1}
4443
4444 It is possible to construct infinite loops by following a subpattern
4445 that can match no characters with a quantifier that has no upper limit,
4446 for example:
4447
4448 (a?)*
4449
4450 Earlier versions of Perl and PCRE used to give an error at compile time
4451 for such patterns. However, because there are cases where this can be
4452 useful, such patterns are now accepted, but if any repetition of the
4453 subpattern does in fact match no characters, the loop is forcibly bro-
4454 ken.
4455
4456 By default, the quantifiers are "greedy", that is, they match as much
4457 as possible (up to the maximum number of permitted times), without
4458 causing the rest of the pattern to fail. The classic example of where
4459 this gives problems is in trying to match comments in C programs. These
4460 appear between /* and */ and within the comment, individual * and /
4461 characters may appear. An attempt to match C comments by applying the
4462 pattern
4463
4464 /\*.*\*/
4465
4466 to the string
4467
4468 /* first comment */ not comment /* second comment */
4469
4470 fails, because it matches the entire string owing to the greediness of
4471 the .* item.
4472
4473 However, if a quantifier is followed by a question mark, it ceases to
4474 be greedy, and instead matches the minimum number of times possible, so
4475 the pattern
4476
4477 /\*.*?\*/
4478
4479 does the right thing with the C comments. The meaning of the various
4480 quantifiers is not otherwise changed, just the preferred number of
4481 matches. Do not confuse this use of question mark with its use as a
4482 quantifier in its own right. Because it has two uses, it can sometimes
4483 appear doubled, as in
4484
4485 \d??\d
4486
4487 which matches one digit by preference, but can match two if that is the
4488 only way the rest of the pattern matches.
4489
4490 If the PCRE_UNGREEDY option is set (an option that is not available in
4491 Perl), the quantifiers are not greedy by default, but individual ones
4492 can be made greedy by following them with a question mark. In other
4493 words, it inverts the default behaviour.
4494
4495 When a parenthesized subpattern is quantified with a minimum repeat
4496 count that is greater than 1 or with a limited maximum, more memory is
4497 required for the compiled pattern, in proportion to the size of the
4498 minimum or maximum.
4499
4500 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equiv-
4501 alent to Perl's /s) is set, thus allowing the dot to match newlines,
4502 the pattern is implicitly anchored, because whatever follows will be
4503 tried against every character position in the subject string, so there
4504 is no point in retrying the overall match at any position after the
4505 first. PCRE normally treats such a pattern as though it were preceded
4506 by \A.
4507
4508 In cases where it is known that the subject string contains no new-
4509 lines, it is worth setting PCRE_DOTALL in order to obtain this opti-
4510 mization, or alternatively using ^ to indicate anchoring explicitly.
4511
4512 However, there is one situation where the optimization cannot be used.
4513 When .* is inside capturing parentheses that are the subject of a back
4514 reference elsewhere in the pattern, a match at the start may fail where
4515 a later one succeeds. Consider, for example:
4516
4517 (.*)abc\1
4518
4519 If the subject is "xyz123abc123" the match point is the fourth charac-
4520 ter. For this reason, such a pattern is not implicitly anchored.
4521
4522 When a capturing subpattern is repeated, the value captured is the sub-
4523 string that matched the final iteration. For example, after
4524
4525 (tweedle[dume]{3}\s*)+
4526
4527 has matched "tweedledum tweedledee" the value of the captured substring
4528 is "tweedledee". However, if there are nested capturing subpatterns,
4529 the corresponding captured values may have been set in previous itera-
4530 tions. For example, after
4531
4532 /(a|(b))+/
4533
4534 matches "aba" the value of the second captured substring is "b".
4535
4536
4537 ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS
4538
4539 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
4540 repetition, failure of what follows normally causes the repeated item
4541 to be re-evaluated to see if a different number of repeats allows the
4542 rest of the pattern to match. Sometimes it is useful to prevent this,
4543 either to change the nature of the match, or to cause it fail earlier
4544 than it otherwise might, when the author of the pattern knows there is
4545 no point in carrying on.
4546
4547 Consider, for example, the pattern \d+foo when applied to the subject
4548 line
4549
4550 123456bar
4551
4552 After matching all 6 digits and then failing to match "foo", the normal
4553 action of the matcher is to try again with only 5 digits matching the
4554 \d+ item, and then with 4, and so on, before ultimately failing.
4555 "Atomic grouping" (a term taken from Jeffrey Friedl's book) provides
4556 the means for specifying that once a subpattern has matched, it is not
4557 to be re-evaluated in this way.
4558
4559 If we use atomic grouping for the previous example, the matcher gives
4560 up immediately on failing to match "foo" the first time. The notation
4561 is a kind of special parenthesis, starting with (?> as in this example:
4562
4563 (?>\d+)foo
4564
4565 This kind of parenthesis "locks up" the part of the pattern it con-
4566 tains once it has matched, and a failure further into the pattern is
4567 prevented from backtracking into it. Backtracking past it to previous
4568 items, however, works as normal.
4569
4570 An alternative description is that a subpattern of this type matches
4571 the string of characters that an identical standalone pattern would
4572 match, if anchored at the current point in the subject string.
4573
4574 Atomic grouping subpatterns are not capturing subpatterns. Simple cases
4575 such as the above example can be thought of as a maximizing repeat that
4576 must swallow everything it can. So, while both \d+ and \d+? are pre-
4577 pared to adjust the number of digits they match in order to make the
4578 rest of the pattern match, (?>\d+) can only match an entire sequence of
4579 digits.
4580
4581 Atomic groups in general can of course contain arbitrarily complicated
4582 subpatterns, and can be nested. However, when the subpattern for an
4583 atomic group is just a single repeated item, as in the example above, a
4584 simpler notation, called a "possessive quantifier" can be used. This
4585 consists of an additional + character following a quantifier. Using
4586 this notation, the previous example can be rewritten as
4587
4588 \d++foo
4589
4590 Note that a possessive quantifier can be used with an entire group, for
4591 example:
4592
4593 (abc|xyz){2,3}+
4594
4595 Possessive quantifiers are always greedy; the setting of the
4596 PCRE_UNGREEDY option is ignored. They are a convenient notation for the
4597 simpler forms of atomic group. However, there is no difference in the
4598 meaning of a possessive quantifier and the equivalent atomic group,
4599 though there may be a performance difference; possessive quantifiers
4600 should be slightly faster.
4601
4602 The possessive quantifier syntax is an extension to the Perl 5.8 syn-
4603 tax. Jeffrey Friedl originated the idea (and the name) in the first
4604 edition of his book. Mike McCloskey liked it, so implemented it when he
4605 built Sun's Java package, and PCRE copied it from there. It ultimately
4606 found its way into Perl at release 5.10.
4607
4608 PCRE has an optimization that automatically "possessifies" certain sim-
4609 ple pattern constructs. For example, the sequence A+B is treated as
4610 A++B because there is no point in backtracking into a sequence of A's
4611 when B must follow.
4612
4613 When a pattern contains an unlimited repeat inside a subpattern that
4614 can itself be repeated an unlimited number of times, the use of an
4615 atomic group is the only way to avoid some failing matches taking a
4616 very long time indeed. The pattern
4617
4618 (\D+|<\d+>)*[!?]
4619
4620 matches an unlimited number of substrings that either consist of non-
4621 digits, or digits enclosed in <>, followed by either ! or ?. When it
4622 matches, it runs quickly. However, if it is applied to
4623
4624 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
4625
4626 it takes a long time before reporting failure. This is because the
4627 string can be divided between the internal \D+ repeat and the external
4628 * repeat in a large number of ways, and all have to be tried. (The
4629 example uses [!?] rather than a single character at the end, because
4630 both PCRE and Perl have an optimization that allows for fast failure
4631 when a single character is used. They remember the last single charac-
4632 ter that is required for a match, and fail early if it is not present
4633 in the string.) If the pattern is changed so that it uses an atomic
4634 group, like this:
4635
4636 ((?>\D+)|<\d+>)*[!?]
4637
4638 sequences of non-digits cannot be broken, and failure happens quickly.
4639
4640
4641 BACK REFERENCES
4642
4643 Outside a character class, a backslash followed by a digit greater than
4644 0 (and possibly further digits) is a back reference to a capturing sub-
4645 pattern earlier (that is, to its left) in the pattern, provided there
4646 have been that many previous capturing left parentheses.
4647
4648 However, if the decimal number following the backslash is less than 10,
4649 it is always taken as a back reference, and causes an error only if
4650 there are not that many capturing left parentheses in the entire pat-
4651 tern. In other words, the parentheses that are referenced need not be
4652 to the left of the reference for numbers less than 10. A "forward back
4653 reference" of this type can make sense when a repetition is involved
4654 and the subpattern to the right has participated in an earlier itera-
4655 tion.
4656
4657 It is not possible to have a numerical "forward back reference" to a
4658 subpattern whose number is 10 or more using this syntax because a
4659 sequence such as \50 is interpreted as a character defined in octal.
4660 See the subsection entitled "Non-printing characters" above for further
4661 details of the handling of digits following a backslash. There is no
4662 such problem when named parentheses are used. A back reference to any
4663 subpattern is possible using named parentheses (see below).
4664
4665 Another way of avoiding the ambiguity inherent in the use of digits
4666 following a backslash is to use the \g escape sequence. This escape
4667 must be followed by an unsigned number or a negative number, optionally
4668 enclosed in braces. These examples are all identical:
4669
4670 (ring), \1
4671 (ring), \g1
4672 (ring), \g{1}
4673
4674 An unsigned number specifies an absolute reference without the ambigu-
4675 ity that is present in the older syntax. It is also useful when literal
4676 digits follow the reference. A negative number is a relative reference.
4677 Consider this example:
4678
4679 (abc(def)ghi)\g{-1}
4680
4681 The sequence \g{-1} is a reference to the most recently started captur-
4682 ing subpattern before \g, that is, is it equivalent to \2. Similarly,
4683 \g{-2} would be equivalent to \1. The use of relative references can be
4684 helpful in long patterns, and also in patterns that are created by
4685 joining together fragments that contain references within themselves.
4686
4687 A back reference matches whatever actually matched the capturing sub-
4688 pattern in the current subject string, rather than anything matching
4689 the subpattern itself (see "Subpatterns as subroutines" below for a way
4690 of doing that). So the pattern
4691
4692 (sens|respons)e and \1ibility
4693
4694 matches "sense and sensibility" and "response and responsibility", but
4695 not "sense and responsibility". If caseful matching is in force at the
4696 time of the back reference, the case of letters is relevant. For exam-
4697 ple,
4698
4699 ((?i)rah)\s+\1
4700
4701 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the
4702 original capturing subpattern is matched caselessly.
4703
4704 There are several different ways of writing back references to named
4705 subpatterns. The .NET syntax \k{name} and the Perl syntax \k<name> or
4706 \k'name' are supported, as is the Python syntax (?P=name). Perl 5.10's
4707 unified back reference syntax, in which \g can be used for both numeric
4708 and named references, is also supported. We could rewrite the above
4709 example in any of the following ways:
4710
4711 (?<p1>(?i)rah)\s+\k<p1>
4712 (?'p1'(?i)rah)\s+\k{p1}
4713 (?P<p1>(?i)rah)\s+(?P=p1)
4714 (?<p1>(?i)rah)\s+\g{p1}
4715
4716 A subpattern that is referenced by name may appear in the pattern
4717 before or after the reference.
4718
4719 There may be more than one back reference to the same subpattern. If a
4720 subpattern has not actually been used in a particular match, any back
4721 references to it always fail by default. For example, the pattern
4722
4723 (a|(bc))\2
4724
4725 always fails if it starts to match "a" rather than "bc". However, if
4726 the PCRE_JAVASCRIPT_COMPAT option is set at compile time, a back refer-
4727 ence to an unset value matches an empty string.
4728
4729 Because there may be many capturing parentheses in a pattern, all dig-
4730 its following a backslash are taken as part of a potential back refer-
4731 ence number. If the pattern continues with a digit character, some
4732 delimiter must be used to terminate the back reference. If the
4733 PCRE_EXTENDED option is set, this can be whitespace. Otherwise, the \g{
4734 syntax or an empty comment (see "Comments" below) can be used.
4735
4736 Recursive back references
4737
4738 A back reference that occurs inside the parentheses to which it refers
4739 fails when the subpattern is first used, so, for example, (a\1) never
4740 matches. However, such references can be useful inside repeated sub-
4741 patterns. For example, the pattern
4742
4743 (a|b\1)+
4744
4745 matches any number of "a"s and also "aba", "ababbaa" etc. At each iter-
4746 ation of the subpattern, the back reference matches the character
4747 string corresponding to the previous iteration. In order for this to
4748 work, the pattern must be such that the first iteration does not need
4749 to match the back reference. This can be done using alternation, as in
4750 the example above, or by a quantifier with a minimum of zero.
4751
4752 Back references of this type cause the group that they reference to be
4753 treated as an atomic group. Once the whole group has been matched, a
4754 subsequent matching failure cannot cause backtracking into the middle
4755 of the group.
4756
4757
4758 ASSERTIONS
4759
4760 An assertion is a test on the characters following or preceding the
4761 current matching point that does not actually consume any characters.
4762 The simple assertions coded as \b, \B, \A, \G, \Z, \z, ^ and $ are
4763 described above.
4764
4765 More complicated assertions are coded as subpatterns. There are two
4766 kinds: those that look ahead of the current position in the subject
4767 string, and those that look behind it. An assertion subpattern is
4768 matched in the normal way, except that it does not cause the current
4769 matching position to be changed.
4770
4771 Assertion subpatterns are not capturing subpatterns, and may not be
4772 repeated, because it makes no sense to assert the same thing several
4773 times. If any kind of assertion contains capturing subpatterns within
4774 it, these are counted for the purposes of numbering the capturing sub-
4775 patterns in the whole pattern. However, substring capturing is carried
4776 out only for positive assertions, because it does not make sense for
4777 negative assertions.
4778
4779 Lookahead assertions
4780
4781 Lookahead assertions start with (?= for positive assertions and (?! for
4782 negative assertions. For example,
4783
4784 \w+(?=;)
4785
4786 matches a word followed by a semicolon, but does not include the semi-
4787 colon in the match, and
4788
4789 foo(?!bar)
4790
4791 matches any occurrence of "foo" that is not followed by "bar". Note
4792 that the apparently similar pattern
4793
4794 (?!foo)bar
4795
4796 does not find an occurrence of "bar" that is preceded by something
4797 other than "foo"; it finds any occurrence of "bar" whatsoever, because
4798 the assertion (?!foo) is always true when the next three characters are
4799 "bar". A lookbehind assertion is needed to achieve the other effect.
4800
4801 If you want to force a matching failure at some point in a pattern, the
4802 most convenient way to do it is with (?!) because an empty string
4803 always matches, so an assertion that requires there not to be an empty
4804 string must always fail. The backtracking control verb (*FAIL) or (*F)
4805 is essentially a synonym for (?!).
4806
4807 Lookbehind assertions
4808
4809 Lookbehind assertions start with (?<= for positive assertions and (?<!
4810 for negative assertions. For example,
4811
4812 (?<!foo)bar
4813
4814 does find an occurrence of "bar" that is not preceded by "foo". The
4815 contents of a lookbehind assertion are restricted such that all the
4816 strings it matches must have a fixed length. However, if there are sev-
4817 eral top-level alternatives, they do not all have to have the same
4818 fixed length. Thus
4819
4820 (?<=bullock|donkey)
4821
4822 is permitted, but
4823
4824 (?<!dogs?|cats?)
4825
4826 causes an error at compile time. Branches that match different length
4827 strings are permitted only at the top level of a lookbehind assertion.
4828 This is an extension compared with Perl, which requires all branches to
4829 match the same length of string. An assertion such as
4830
4831 (?<=ab(c|de))
4832
4833 is not permitted, because its single top-level branch can match two
4834 different lengths, but it is acceptable to PCRE if rewritten to use two
4835 top-level branches:
4836
4837 (?<=abc|abde)
4838
4839 In some cases, the escape sequence \K (see above) can be used instead
4840 of a lookbehind assertion to get round the fixed-length restriction.
4841
4842 The implementation of lookbehind assertions is, for each alternative,
4843 to temporarily move the current position back by the fixed length and
4844 then try to match. If there are insufficient characters before the cur-
4845 rent position, the assertion fails.
4846
4847 PCRE does not allow the \C escape (which matches a single byte in UTF-8
4848 mode) to appear in lookbehind assertions, because it makes it impossi-
4849 ble to calculate the length of the lookbehind. The \X and \R escapes,
4850 which can match different numbers of bytes, are also not permitted.
4851
4852 "Subroutine" calls (see below) such as (?2) or (?&X) are permitted in
4853 lookbehinds, as long as the subpattern matches a fixed-length string.
4854 Recursion, however, is not supported.
4855
4856 Possessive quantifiers can be used in conjunction with lookbehind
4857 assertions to specify efficient matching of fixed-length strings at the
4858 end of subject strings. Consider a simple pattern such as
4859
4860 abcd$
4861
4862 when applied to a long string that does not match. Because matching
4863 proceeds from left to right, PCRE will look for each "a" in the subject
4864 and then see if what follows matches the rest of the pattern. If the
4865 pattern is specified as
4866
4867 ^.*abcd$
4868
4869 the initial .* matches the entire string at first, but when this fails
4870 (because there is no following "a"), it backtracks to match all but the
4871 last character, then all but the last two characters, and so on. Once
4872 again the search for "a" covers the entire string, from right to left,
4873 so we are no better off. However, if the pattern is written as
4874
4875 ^.*+(?<=abcd)
4876
4877 there can be no backtracking for the .*+ item; it can match only the
4878 entire string. The subsequent lookbehind assertion does a single test
4879 on the last four characters. If it fails, the match fails immediately.
4880 For long strings, this approach makes a significant difference to the
4881 processing time.
4882
4883 Using multiple assertions
4884
4885 Several assertions (of any sort) may occur in succession. For example,
4886
4887 (?<=\d{3})(?<!999)foo
4888
4889 matches "foo" preceded by three digits that are not "999". Notice that
4890 each of the assertions is applied independently at the same point in
4891 the subject string. First there is a check that the previous three
4892 characters are all digits, and then there is a check that the same
4893 three characters are not "999". This pattern does not match "foo" pre-
4894 ceded by six characters, the first of which are digits and the last
4895 three of which are not "999". For example, it doesn't match "123abc-
4896 foo". A pattern to do that is
4897
4898 (?<=\d{3}...)(?<!999)foo
4899
4900 This time the first assertion looks at the preceding six characters,
4901 checking that the first three are digits, and then the second assertion
4902 checks that the preceding three characters are not "999".
4903
4904 Assertions can be nested in any combination. For example,
4905
4906 (?<=(?<!foo)bar)baz
4907
4908 matches an occurrence of "baz" that is preceded by "bar" which in turn
4909 is not preceded by "foo", while
4910
4911 (?<=\d{3}(?!999)...)foo
4912
4913 is another pattern that matches "foo" preceded by three digits and any
4914 three characters that are not "999".
4915
4916
4917 CONDITIONAL SUBPATTERNS
4918
4919 It is possible to cause the matching process to obey a subpattern con-
4920 ditionally or to choose between two alternative subpatterns, depending
4921 on the result of an assertion, or whether a specific capturing subpat-
4922 tern has already been matched. The two possible forms of conditional
4923 subpattern are:
4924
4925 (?(condition)yes-pattern)
4926 (?(condition)yes-pattern|no-pattern)
4927
4928 If the condition is satisfied, the yes-pattern is used; otherwise the
4929 no-pattern (if present) is used. If there are more than two alterna-
4930 tives in the subpattern, a compile-time error occurs. Each of the two
4931 alternatives may itself contain nested subpatterns of any form, includ-
4932 ing conditional subpatterns; the restriction to two alternatives
4933 applies only at the level of the condition. This pattern fragment is an
4934 example where the alternatives are complex:
4935
4936 (?(1) (A|B|C) | (D | (?(2)E|F) | E) )
4937
4938
4939 There are four kinds of condition: references to subpatterns, refer-
4940 ences to recursion, a pseudo-condition called DEFINE, and assertions.
4941
4942 Checking for a used subpattern by number
4943
4944 If the text between the parentheses consists of a sequence of digits,
4945 the condition is true if a capturing subpattern of that number has pre-
4946 viously matched. If there is more than one capturing subpattern with
4947 the same number (see the earlier section about duplicate subpattern
4948 numbers), the condition is true if any of them have been set. An alter-
4949 native notation is to precede the digits with a plus or minus sign. In
4950 this case, the subpattern number is relative rather than absolute. The
4951 most recently opened parentheses can be referenced by (?(-1), the next
4952 most recent by (?(-2), and so on. In looping constructs it can also
4953 make sense to refer to subsequent groups with constructs such as
4954 (?(+2).
4955
4956 Consider the following pattern, which contains non-significant white
4957 space to make it more readable (assume the PCRE_EXTENDED option) and to
4958 divide it into three parts for ease of discussion:
4959
4960 ( \( )? [^()]+ (?(1) \) )
4961
4962 The first part matches an optional opening parenthesis, and if that
4963 character is present, sets it as the first captured substring. The sec-
4964 ond part matches one or more characters that are not parentheses. The
4965 third part is a conditional subpattern that tests whether the first set
4966 of parentheses matched or not. If they did, that is, if subject started
4967 with an opening parenthesis, the condition is true, and so the yes-pat-
4968 tern is executed and a closing parenthesis is required. Otherwise,
4969 since no-pattern is not present, the subpattern matches nothing. In
4970 other words, this pattern matches a sequence of non-parentheses,
4971 optionally enclosed in parentheses.
4972
4973 If you were embedding this pattern in a larger one, you could use a
4974 relative reference:
4975
4976 ...other stuff... ( \( )? [^()]+ (?(-1) \) ) ...
4977
4978 This makes the fragment independent of the parentheses in the larger
4979 pattern.
4980
4981 Checking for a used subpattern by name
4982
4983 Perl uses the syntax (?(<name>)...) or (?('name')...) to test for a
4984 used subpattern by name. For compatibility with earlier versions of
4985 PCRE, which had this facility before Perl, the syntax (?(name)...) is
4986 also recognized. However, there is a possible ambiguity with this syn-
4987 tax, because subpattern names may consist entirely of digits. PCRE
4988 looks first for a named subpattern; if it cannot find one and the name
4989 consists entirely of digits, PCRE looks for a subpattern of that num-
4990 ber, which must be greater than zero. Using subpattern names that con-
4991 sist entirely of digits is not recommended.
4992
4993 Rewriting the above example to use a named subpattern gives this:
4994
4995 (?<OPEN> \( )? [^()]+ (?(<OPEN>) \) )
4996
4997 If the name used in a condition of this kind is a duplicate, the test
4998 is applied to all subpatterns of the same name, and is true if any one
4999 of them has matched.
5000
5001 Checking for pattern recursion
5002
5003 If the condition is the string (R), and there is no subpattern with the
5004 name R, the condition is true if a recursive call to the whole pattern
5005 or any subpattern has been made. If digits or a name preceded by amper-
5006 sand follow the letter R, for example:
5007
5008 (?(R3)...) or (?(R&name)...)
5009
5010 the condition is true if the most recent recursion is into a subpattern
5011 whose number or name is given. This condition does not check the entire
5012 recursion stack. If the name used in a condition of this kind is a
5013 duplicate, the test is applied to all subpatterns of the same name, and
5014 is true if any one of them is the most recent recursion.
5015
5016 At "top level", all these recursion test conditions are false. The
5017 syntax for recursive patterns is described below.
5018
5019 Defining subpatterns for use by reference only
5020
5021 If the condition is the string (DEFINE), and there is no subpattern
5022 with the name DEFINE, the condition is always false. In this case,
5023 there may be only one alternative in the subpattern. It is always
5024 skipped if control reaches this point in the pattern; the idea of
5025 DEFINE is that it can be used to define "subroutines" that can be ref-
5026 erenced from elsewhere. (The use of "subroutines" is described below.)
5027 For example, a pattern to match an IPv4 address could be written like
5028 this (ignore whitespace and line breaks):
5029
5030 (?(DEFINE) (?<byte> 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
5031 \b (?&byte) (\.(?&byte)){3} \b
5032
5033 The first part of the pattern is a DEFINE group inside which a another
5034 group named "byte" is defined. This matches an individual component of
5035 an IPv4 address (a number less than 256). When matching takes place,
5036 this part of the pattern is skipped because DEFINE acts like a false
5037 condition. The rest of the pattern uses references to the named group
5038 to match the four dot-separated components of an IPv4 address, insist-
5039 ing on a word boundary at each end.
5040
5041 Assertion conditions
5042
5043 If the condition is not in any of the above formats, it must be an
5044 assertion. This may be a positive or negative lookahead or lookbehind
5045 assertion. Consider this pattern, again containing non-significant
5046 white space, and with the two alternatives on the second line:
5047
5048 (?(?=[^a-z]*[a-z])
5049 \d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} )
5050
5051 The condition is a positive lookahead assertion that matches an
5052 optional sequence of non-letters followed by a letter. In other words,
5053 it tests for the presence of at least one letter in the subject. If a
5054 letter is found, the subject is matched against the first alternative;
5055 otherwise it is matched against the second. This pattern matches
5056 strings in one of the two forms dd-aaa-dd or dd-dd-dd, where aaa are
5057 letters and dd are digits.
5058
5059
5060 COMMENTS
5061
5062 There are two ways of including comments in patterns that are processed
5063 by PCRE. In both cases, the start of the comment must not be in a char-
5064 acter class, nor in the middle of any other sequence of related charac-
5065 ters such as (?: or a subpattern name or number. The characters that
5066 make up a comment play no part in the pattern matching.
5067
5068 The sequence (?# marks the start of a comment that continues up to the
5069 next closing parenthesis. Nested parentheses are not permitted. If the
5070 PCRE_EXTENDED option is set, an unescaped # character also introduces a
5071 comment, which in this case continues to immediately after the next
5072 newline character or character sequence in the pattern. Which charac-
5073 ters are interpreted as newlines is controlled by the options passed to
5074 pcre_compile() or by a special sequence at the start of the pattern, as
5075 described in the section entitled "Newline conventions" above. Note
5076 that the end of this type of comment is a literal newline sequence in
5077 the pattern; escape sequences that happen to represent a newline do not
5078 count. For example, consider this pattern when PCRE_EXTENDED is set,
5079 and the default newline convention is in force:
5080
5081 abc #comment \n still comment
5082
5083 On encountering the # character, pcre_compile() skips along, looking
5084 for a newline in the pattern. The sequence \n is still literal at this
5085 stage, so it does not terminate the comment. Only an actual character
5086 with the code value 0x0a (the default newline) does so.
5087
5088
5089 RECURSIVE PATTERNS
5090
5091 Consider the problem of matching a string in parentheses, allowing for
5092 unlimited nested parentheses. Without the use of recursion, the best
5093 that can be done is to use a pattern that matches up to some fixed
5094 depth of nesting. It is not possible to handle an arbitrary nesting
5095 depth.
5096
5097 For some time, Perl has provided a facility that allows regular expres-
5098 sions to recurse (amongst other things). It does this by interpolating
5099 Perl code in the expression at run time, and the code can refer to the
5100 expression itself. A Perl pattern using code interpolation to solve the
5101 parentheses problem can be created like this:
5102
5103 $re = qr{\( (?: (?>[^()]+) | (?p{$re}) )* \)}x;
5104
5105 The (?p{...}) item interpolates Perl code at run time, and in this case
5106 refers recursively to the pattern in which it appears.
5107
5108 Obviously, PCRE cannot support the interpolation of Perl code. Instead,
5109 it supports special syntax for recursion of the entire pattern, and
5110 also for individual subpattern recursion. After its introduction in
5111 PCRE and Python, this kind of recursion was subsequently introduced
5112 into Perl at release 5.10.
5113
5114 A special item that consists of (? followed by a number greater than
5115 zero and a closing parenthesis is a recursive call of the subpattern of
5116 the given number, provided that it occurs inside that subpattern. (If
5117 not, it is a "subroutine" call, which is described in the next sec-
5118 tion.) The special item (?R) or (?0) is a recursive call of the entire
5119 regular expression.
5120
5121 This PCRE pattern solves the nested parentheses problem (assume the
5122 PCRE_EXTENDED option is set so that white space is ignored):
5123
5124 \( ( [^()]++ | (?R) )* \)
5125
5126 First it matches an opening parenthesis. Then it matches any number of
5127 substrings which can either be a sequence of non-parentheses, or a
5128 recursive match of the pattern itself (that is, a correctly parenthe-
5129 sized substring). Finally there is a closing parenthesis. Note the use
5130 of a possessive quantifier to avoid backtracking into sequences of non-
5131 parentheses.
5132
5133 If this were part of a larger pattern, you would not want to recurse
5134 the entire pattern, so instead you could use this:
5135
5136 ( \( ( [^()]++ | (?1) )* \) )
5137
5138 We have put the pattern into parentheses, and caused the recursion to
5139 refer to them instead of the whole pattern.
5140
5141 In a larger pattern, keeping track of parenthesis numbers can be
5142 tricky. This is made easier by the use of relative references. Instead
5143 of (?1) in the pattern above you can write (?-2) to refer to the second
5144 most recently opened parentheses preceding the recursion. In other
5145 words, a negative number counts capturing parentheses leftwards from
5146 the point at which it is encountered.
5147
5148 It is also possible to refer to subsequently opened parentheses, by
5149 writing references such as (?+2). However, these cannot be recursive
5150 because the reference is not inside the parentheses that are refer-
5151 enced. They are always "subroutine" calls, as described in the next
5152 section.
5153
5154 An alternative approach is to use named parentheses instead. The Perl
5155 syntax for this is (?&name); PCRE's earlier syntax (?P>name) is also
5156 supported. We could rewrite the above example as follows:
5157
5158 (?<pn> \( ( [^()]++ | (?&pn) )* \) )
5159
5160 If there is more than one subpattern with the same name, the earliest
5161 one is used.
5162
5163 This particular example pattern that we have been looking at contains
5164 nested unlimited repeats, and so the use of a possessive quantifier for
5165 matching strings of non-parentheses is important when applying the pat-
5166 tern to strings that do not match. For example, when this pattern is
5167 applied to
5168
5169 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
5170
5171 it yields "no match" quickly. However, if a possessive quantifier is
5172 not used, the match runs for a very long time indeed because there are
5173 so many different ways the + and * repeats can carve up the subject,
5174 and all have to be tested before failure can be reported.
5175
5176 At the end of a match, the values of capturing parentheses are those
5177 from the outermost level. If you want to obtain intermediate values, a
5178 callout function can be used (see below and the pcrecallout documenta-
5179 tion). If the pattern above is matched against
5180
5181 (ab(cd)ef)
5182
5183 the value for the inner capturing parentheses (numbered 2) is "ef",
5184 which is the last value taken on at the top level. If a capturing sub-
5185 pattern is not matched at the top level, its final value is unset, even
5186 if it is (temporarily) set at a deeper level.
5187
5188 If there are more than 15 capturing parentheses in a pattern, PCRE has
5189 to obtain extra memory to store data during a recursion, which it does
5190 by using pcre_malloc, freeing it via pcre_free afterwards. If no memory
5191 can be obtained, the match fails with the PCRE_ERROR_NOMEMORY error.
5192
5193 Do not confuse the (?R) item with the condition (R), which tests for
5194 recursion. Consider this pattern, which matches text in angle brack-
5195 ets, allowing for arbitrary nesting. Only digits are allowed in nested
5196 brackets (that is, when recursing), whereas any characters are permit-
5197 ted at the outer level.
5198
5199 < (?: (?(R) \d++ | [^<>]*+) | (?R)) * >
5200
5201 In this pattern, (?(R) is the start of a conditional subpattern, with
5202 two different alternatives for the recursive and non-recursive cases.
5203 The (?R) item is the actual recursive call.
5204
5205 Recursion difference from Perl
5206
5207 In PCRE (like Python, but unlike Perl), a recursive subpattern call is
5208 always treated as an atomic group. That is, once it has matched some of
5209 the subject string, it is never re-entered, even if it contains untried
5210 alternatives and there is a subsequent matching failure. This can be
5211 illustrated by the following pattern, which purports to match a palin-
5212 dromic string that contains an odd number of characters (for example,
5213 "a", "aba", "abcba", "abcdcba"):
5214
5215 ^(.|(.)(?1)\2)$
5216
5217 The idea is that it either matches a single character, or two identical
5218 characters surrounding a sub-palindrome. In Perl, this pattern works;
5219 in PCRE it does not if the pattern is longer than three characters.
5220 Consider the subject string "abcba":
5221
5222 At the top level, the first character is matched, but as it is not at
5223 the end of the string, the first alternative fails; the second alterna-
5224 tive is taken and the recursion kicks in. The recursive call to subpat-
5225 tern 1 successfully matches the next character ("b"). (Note that the
5226 beginning and end of line tests are not part of the recursion).
5227
5228 Back at the top level, the next character ("c") is compared with what
5229 subpattern 2 matched, which was "a". This fails. Because the recursion
5230 is treated as an atomic group, there are now no backtracking points,
5231 and so the entire match fails. (Perl is able, at this point, to re-
5232 enter the recursion and try the second alternative.) However, if the
5233 pattern is written with the alternatives in the other order, things are
5234 different:
5235
5236 ^((.)(?1)\2|.)$
5237
5238 This time, the recursing alternative is tried first, and continues to
5239 recurse until it runs out of characters, at which point the recursion
5240 fails. But this time we do have another alternative to try at the
5241 higher level. That is the big difference: in the previous case the
5242 remaining alternative is at a deeper recursion level, which PCRE cannot
5243 use.
5244
5245 To change the pattern so that matches all palindromic strings, not just
5246 those with an odd number of characters, it is tempting to change the
5247 pattern to this:
5248
5249 ^((.)(?1)\2|.?)$
5250
5251 Again, this works in Perl, but not in PCRE, and for the same reason.
5252 When a deeper recursion has matched a single character, it cannot be
5253 entered again in order to match an empty string. The solution is to
5254 separate the two cases, and write out the odd and even cases as alter-
5255 natives at the higher level:
5256
5257 ^(?:((.)(?1)\2|)|((.)(?3)\4|.))
5258
5259 If you want to match typical palindromic phrases, the pattern has to
5260 ignore all non-word characters, which can be done like this:
5261
5262 ^\W*+(?:((.)\W*+(?1)\W*+\2|)|((.)\W*+(?3)\W*+\4|\W*+.\W*+))\W*+$
5263
5264 If run with the PCRE_CASELESS option, this pattern matches phrases such
5265 as "A man, a plan, a canal: Panama!" and it works well in both PCRE and
5266 Perl. Note the use of the possessive quantifier *+ to avoid backtrack-
5267 ing into sequences of non-word characters. Without this, PCRE takes a
5268 great deal longer (ten times or more) to match typical phrases, and
5269 Perl takes so long that you think it has gone into a loop.
5270
5271 WARNING: The palindrome-matching patterns above work only if the sub-
5272 ject string does not start with a palindrome that is shorter than the
5273 entire string. For example, although "abcba" is correctly matched, if
5274 the subject is "ababa", PCRE finds the palindrome "aba" at the start,
5275 then fails at top level because the end of the string does not follow.
5276 Once again, it cannot jump back into the recursion to try other alter-
5277 natives, so the entire match fails.
5278
5279
5280 SUBPATTERNS AS SUBROUTINES
5281
5282 If the syntax for a recursive subpattern reference (either by number or
5283 by name) is used outside the parentheses to which it refers, it oper-
5284 ates like a subroutine in a programming language. The "called" subpat-
5285 tern may be defined before or after the reference. A numbered reference
5286 can be absolute or relative, as in these examples:
5287
5288 (...(absolute)...)...(?2)...
5289 (...(relative)...)...(?-1)...
5290 (...(?+1)...(relative)...
5291
5292 An earlier example pointed out that the pattern
5293
5294 (sens|respons)e and \1ibility
5295
5296 matches "sense and sensibility" and "response and responsibility", but
5297 not "sense and responsibility". If instead the pattern
5298
5299 (sens|respons)e and (?1)ibility
5300
5301 is used, it does match "sense and responsibility" as well as the other
5302 two strings. Another example is given in the discussion of DEFINE
5303 above.
5304
5305 Like recursive subpatterns, a subroutine call is always treated as an
5306 atomic group. That is, once it has matched some of the subject string,
5307 it is never re-entered, even if it contains untried alternatives and
5308 there is a subsequent matching failure. Any capturing parentheses that
5309 are set during the subroutine call revert to their previous values
5310 afterwards.
5311
5312 When a subpattern is used as a subroutine, processing options such as
5313 case-independence are fixed when the subpattern is defined. They cannot
5314 be changed for different calls. For example, consider this pattern:
5315
5316 (abc)(?i:(?-1))
5317
5318 It matches "abcabc". It does not match "abcABC" because the change of
5319 processing option does not affect the called subpattern.
5320
5321
5322 ONIGURUMA SUBROUTINE SYNTAX
5323
5324 For compatibility with Oniguruma, the non-Perl syntax \g followed by a
5325 name or a number enclosed either in angle brackets or single quotes, is
5326 an alternative syntax for referencing a subpattern as a subroutine,
5327 possibly recursively. Here are two of the examples used above, rewrit-
5328 ten using this syntax:
5329
5330 (?<pn> \( ( (?>[^()]+) | \g<pn> )* \) )
5331 (sens|respons)e and \g'1'ibility
5332
5333 PCRE supports an extension to Oniguruma: if a number is preceded by a
5334 plus or a minus sign it is taken as a relative reference. For example:
5335
5336 (abc)(?i:\g<-1>)
5337
5338 Note that \g{...} (Perl syntax) and \g<...> (Oniguruma syntax) are not
5339 synonymous. The former is a back reference; the latter is a subroutine
5340 call.
5341
5342
5343 CALLOUTS
5344
5345 Perl has a feature whereby using the sequence (?{...}) causes arbitrary
5346 Perl code to be obeyed in the middle of matching a regular expression.
5347 This makes it possible, amongst other things, to extract different sub-
5348 strings that match the same pair of parentheses when there is a repeti-
5349 tion.
5350
5351 PCRE provides a similar feature, but of course it cannot obey arbitrary
5352 Perl code. The feature is called "callout". The caller of PCRE provides
5353 an external function by putting its entry point in the global variable
5354 pcre_callout. By default, this variable contains NULL, which disables
5355 all calling out.
5356
5357 Within a regular expression, (?C) indicates the points at which the
5358 external function is to be called. If you want to identify different
5359 callout points, you can put a number less than 256 after the letter C.
5360 The default value is zero. For example, this pattern has two callout
5361 points:
5362
5363 (?C1)abc(?C2)def
5364
5365 If the PCRE_AUTO_CALLOUT flag is passed to pcre_compile(), callouts are
5366 automatically installed before each item in the pattern. They are all
5367 numbered 255.
5368
5369 During matching, when PCRE reaches a callout point (and pcre_callout is
5370 set), the external function is called. It is provided with the number
5371 of the callout, the position in the pattern, and, optionally, one item
5372 of data originally supplied by the caller of pcre_exec(). The callout
5373 function may cause matching to proceed, to backtrack, or to fail alto-
5374 gether. A complete description of the interface to the callout function
5375 is given in the pcrecallout documentation.
5376
5377
5378 BACKTRACKING CONTROL
5379
5380 Perl 5.10 introduced a number of "Special Backtracking Control Verbs",
5381 which are described in the Perl documentation as "experimental and sub-
5382 ject to change or removal in a future version of Perl". It goes on to
5383 say: "Their usage in production code should be noted to avoid problems
5384 during upgrades." The same remarks apply to the PCRE features described
5385 in this section.
5386
5387 Since these verbs are specifically related to backtracking, most of
5388 them can be used only when the pattern is to be matched using
5389 pcre_exec(), which uses a backtracking algorithm. With the exception of
5390 (*FAIL), which behaves like a failing negative assertion, they cause an
5391 error if encountered by pcre_dfa_exec().
5392
5393 If any of these verbs are used in an assertion or subroutine subpattern
5394 (including recursive subpatterns), their effect is confined to that
5395 subpattern; it does not extend to the surrounding pattern. Note that
5396 such subpatterns are processed as anchored at the point where they are
5397 tested.
5398
5399 The new verbs make use of what was previously invalid syntax: an open-
5400 ing parenthesis followed by an asterisk. They are generally of the form
5401 (*VERB) or (*VERB:NAME). Some may take either form, with differing be-
5402 haviour, depending on whether or not an argument is present. An name is
5403 a sequence of letters, digits, and underscores. If the name is empty,
5404 that is, if the closing parenthesis immediately follows the colon, the
5405 effect is as if the colon were not there. Any number of these verbs may
5406 occur in a pattern.
5407
5408 PCRE contains some optimizations that are used to speed up matching by
5409 running some checks at the start of each match attempt. For example, it
5410 may know the minimum length of matching subject, or that a particular
5411 character must be present. When one of these optimizations suppresses
5412 the running of a match, any included backtracking verbs will not, of
5413 course, be processed. You can suppress the start-of-match optimizations
5414 by setting the PCRE_NO_START_OPTIMIZE option when calling pcre_exec().
5415
5416 Verbs that act immediately
5417
5418 The following verbs act as soon as they are encountered. They may not
5419 be followed by a name.
5420
5421 (*ACCEPT)
5422
5423 This verb causes the match to end successfully, skipping the remainder
5424 of the pattern. When inside a recursion, only the innermost pattern is
5425 ended immediately. If (*ACCEPT) is inside capturing parentheses, the
5426 data so far is captured. (This feature was added to PCRE at release
5427 8.00.) For example:
5428
5429 A((?:A|B(*ACCEPT)|C)D)
5430
5431 This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is cap-
5432 tured by the outer parentheses.
5433
5434 (*FAIL) or (*F)
5435
5436 This verb causes the match to fail, forcing backtracking to occur. It
5437 is equivalent to (?!) but easier to read. The Perl documentation notes
5438 that it is probably useful only when combined with (?{}) or (??{}).
5439 Those are, of course, Perl features that are not present in PCRE. The
5440 nearest equivalent is the callout feature, as for example in this pat-
5441 tern:
5442
5443 a+(?C)(*FAIL)
5444
5445 A match with the string "aaaa" always fails, but the callout is taken
5446 before each backtrack happens (in this example, 10 times).
5447
5448 Recording which path was taken
5449
5450 There is one verb whose main purpose is to track how a match was
5451 arrived at, though it also has a secondary use in conjunction with
5452 advancing the match starting point (see (*SKIP) below).
5453
5454 (*MARK:NAME) or (*:NAME)
5455
5456 A name is always required with this verb. There may be as many
5457 instances of (*MARK) as you like in a pattern, and their names do not
5458 have to be unique.
5459
5460 When a match succeeds, the name of the last-encountered (*MARK) is
5461 passed back to the caller via the pcre_extra data structure, as
5462 described in the section on pcre_extra in the pcreapi documentation. No
5463 data is returned for a partial match. Here is an example of pcretest
5464 output, where the /K modifier requests the retrieval and outputting of
5465 (*MARK) data:
5466
5467 /X(*MARK:A)Y|X(*MARK:B)Z/K
5468 XY
5469 0: XY
5470 MK: A
5471 XZ
5472 0: XZ
5473 MK: B
5474
5475 The (*MARK) name is tagged with "MK:" in this output, and in this exam-
5476 ple it indicates which of the two alternatives matched. This is a more
5477 efficient way of obtaining this information than putting each alterna-
5478 tive in its own capturing parentheses.
5479
5480 A name may also be returned after a failed match if the final path
5481 through the pattern involves (*MARK). However, unless (*MARK) used in
5482 conjunction with (*COMMIT), this is unlikely to happen for an unan-
5483 chored pattern because, as the starting point for matching is advanced,
5484 the final check is often with an empty string, causing a failure before
5485 (*MARK) is reached. For example:
5486
5487 /X(*MARK:A)Y|X(*MARK:B)Z/K
5488 XP
5489 No match
5490
5491 There are three potential starting points for this match (starting with
5492 X, starting with P, and with an empty string). If the pattern is
5493 anchored, the result is different:
5494
5495 /^X(*MARK:A)Y|^X(*MARK:B)Z/K
5496 XP
5497 No match, mark = B
5498
5499 PCRE's start-of-match optimizations can also interfere with this. For
5500 example, if, as a result of a call to pcre_study(), it knows the mini-
5501 mum subject length for a match, a shorter subject will not be scanned
5502 at all.
5503
5504 Note that similar anomalies (though different in detail) exist in Perl,
5505 no doubt for the same reasons. The use of (*MARK) data after a failed
5506 match of an unanchored pattern is not recommended, unless (*COMMIT) is
5507 involved.
5508
5509 Verbs that act after backtracking
5510
5511 The following verbs do nothing when they are encountered. Matching con-
5512 tinues with what follows, but if there is no subsequent match, causing
5513 a backtrack to the verb, a failure is forced. That is, backtracking
5514 cannot pass to the left of the verb. However, when one of these verbs
5515 appears inside an atomic group, its effect is confined to that group,
5516 because once the group has been matched, there is never any backtrack-
5517 ing into it. In this situation, backtracking can "jump back" to the
5518 left of the entire atomic group. (Remember also, as stated above, that
5519 this localization also applies in subroutine calls and assertions.)
5520
5521 These verbs differ in exactly what kind of failure occurs when back-
5522 tracking reaches them.
5523
5524 (*COMMIT)
5525
5526 This verb, which may not be followed by a name, causes the whole match
5527 to fail outright if the rest of the pattern does not match. Even if the
5528 pattern is unanchored, no further attempts to find a match by advancing
5529 the starting point take place. Once (*COMMIT) has been passed,
5530 pcre_exec() is committed to finding a match at the current starting
5531 point, or not at all. For example:
5532
5533 a+(*COMMIT)b
5534
5535 This matches "xxaab" but not "aacaab". It can be thought of as a kind
5536 of dynamic anchor, or "I've started, so I must finish." The name of the
5537 most recently passed (*MARK) in the path is passed back when (*COMMIT)
5538 forces a match failure.
5539
5540 Note that (*COMMIT) at the start of a pattern is not the same as an
5541 anchor, unless PCRE's start-of-match optimizations are turned off, as
5542 shown in this pcretest example:
5543
5544 /(*COMMIT)abc/
5545 xyzabc
5546 0: abc
5547 xyzabc\Y
5548 No match
5549
5550 PCRE knows that any match must start with "a", so the optimization
5551 skips along the subject to "a" before running the first match attempt,
5552 which succeeds. When the optimization is disabled by the \Y escape in
5553 the second subject, the match starts at "x" and so the (*COMMIT) causes
5554 it to fail without trying any other starting points.
5555
5556 (*PRUNE) or (*PRUNE:NAME)
5557
5558 This verb causes the match to fail at the current starting position in
5559 the subject if the rest of the pattern does not match. If the pattern
5560 is unanchored, the normal "bumpalong" advance to the next starting
5561 character then happens. Backtracking can occur as usual to the left of
5562 (*PRUNE), before it is reached, or when matching to the right of
5563 (*PRUNE), but if there is no match to the right, backtracking cannot
5564 cross (*PRUNE). In simple cases, the use of (*PRUNE) is just an alter-
5565 native to an atomic group or possessive quantifier, but there are some
5566 uses of (*PRUNE) that cannot be expressed in any other way. The behav-
5567 iour of (*PRUNE:NAME) is the same as (*MARK:NAME)(*PRUNE) when the
5568 match fails completely; the name is passed back if this is the final
5569 attempt. (*PRUNE:NAME) does not pass back a name if the match suc-
5570 ceeds. In an anchored pattern (*PRUNE) has the same effect as (*COM-
5571 MIT).
5572
5573 (*SKIP)
5574
5575 This verb, when given without a name, is like (*PRUNE), except that if
5576 the pattern is unanchored, the "bumpalong" advance is not to the next
5577 character, but to the position in the subject where (*SKIP) was encoun-
5578 tered. (*SKIP) signifies that whatever text was matched leading up to
5579 it cannot be part of a successful match. Consider:
5580
5581 a+(*SKIP)b
5582
5583 If the subject is "aaaac...", after the first match attempt fails
5584 (starting at the first character in the string), the starting point
5585 skips on to start the next attempt at "c". Note that a possessive quan-
5586 tifer does not have the same effect as this example; although it would
5587 suppress backtracking during the first match attempt, the second
5588 attempt would start at the second character instead of skipping on to
5589 "c".
5590
5591 (*SKIP:NAME)
5592
5593 When (*SKIP) has an associated name, its behaviour is modified. If the
5594 following pattern fails to match, the previous path through the pattern
5595 is searched for the most recent (*MARK) that has the same name. If one
5596 is found, the "bumpalong" advance is to the subject position that cor-
5597 responds to that (*MARK) instead of to where (*SKIP) was encountered.
5598 If no (*MARK) with a matching name is found, normal "bumpalong" of one
5599 character happens (the (*SKIP) is ignored).
5600
5601 (*THEN) or (*THEN:NAME)
5602
5603 This verb causes a skip to the next alternation in the innermost
5604 enclosing group if the rest of the pattern does not match. That is, it
5605 cancels pending backtracking, but only within the current alternation.
5606 Its name comes from the observation that it can be used for a pattern-
5607 based if-then-else block:
5608
5609 ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
5610
5611 If the COND1 pattern matches, FOO is tried (and possibly further items
5612 after the end of the group if FOO succeeds); on failure the matcher
5613 skips to the second alternative and tries COND2, without backtracking
5614 into COND1. The behaviour of (*THEN:NAME) is exactly the same as
5615 (*MARK:NAME)(*THEN) if the overall match fails. If (*THEN) is not
5616 directly inside an alternation, it acts like (*PRUNE).
5617
5618 The above verbs provide four different "strengths" of control when sub-
5619 sequent matching fails. (*THEN) is the weakest, carrying on the match
5620 at the next alternation. (*PRUNE) comes next, failing the match at the
5621 current starting position, but allowing an advance to the next charac-
5622 ter (for an unanchored pattern). (*SKIP) is similar, except that the
5623 advance may be more than one character. (*COMMIT) is the strongest,
5624 causing the entire match to fail.
5625
5626 If more than one is present in a pattern, the "stongest" one wins. For
5627 example, consider this pattern, where A, B, etc. are complex pattern
5628 fragments:
5629
5630 (A(*COMMIT)B(*THEN)C|D)
5631
5632 Once A has matched, PCRE is committed to this match, at the current
5633 starting position. If subsequently B matches, but C does not, the nor-
5634 mal (*THEN) action of trying the next alternation (that is, D) does not
5635 happen because (*COMMIT) overrides.
5636
5637
5638 SEE ALSO
5639
5640 pcreapi(3), pcrecallout(3), pcrematching(3), pcresyntax(3), pcre(3).
5641
5642
5643 AUTHOR
5644
5645 Philip Hazel
5646 University Computing Service
5647 Cambridge CB2 3QH, England.
5648
5649
5650 REVISION
5651
5652 Last updated: 17 November 2010
5653 Copyright (c) 1997-2010 University of Cambridge.
5654 ------------------------------------------------------------------------------
5655
5656
5657 PCRESYNTAX(3) PCRESYNTAX(3)
5658
5659
5660 NAME
5661 PCRE - Perl-compatible regular expressions
5662
5663
5664 PCRE REGULAR EXPRESSION SYNTAX SUMMARY
5665
5666 The full syntax and semantics of the regular expressions that are sup-
5667 ported by PCRE are described in the pcrepattern documentation. This
5668 document contains just a quick-reference summary of the syntax.
5669
5670
5671 QUOTING
5672
5673 \x where x is non-alphanumeric is a literal x
5674 \Q...\E treat enclosed characters as literal
5675
5676
5677 CHARACTERS
5678
5679 \a alarm, that is, the BEL character (hex 07)
5680 \cx "control-x", where x is any character
5681 \e escape (hex 1B)
5682 \f formfeed (hex 0C)
5683 \n newline (hex 0A)
5684 \r carriage return (hex 0D)
5685 \t tab (hex 09)
5686 \ddd character with octal code ddd, or backreference
5687 \xhh character with hex code hh
5688 \x{hhh..} character with hex code hhh..
5689
5690
5691 CHARACTER TYPES
5692
5693 . any character except newline;
5694 in dotall mode, any character whatsoever
5695 \C one byte, even in UTF-8 mode (best avoided)
5696 \d a decimal digit
5697 \D a character that is not a decimal digit
5698 \h a horizontal whitespace character
5699 \H a character that is not a horizontal whitespace character
5700 \N a character that is not a newline
5701 \p{xx} a character with the xx property
5702 \P{xx} a character without the xx property
5703 \R a newline sequence
5704 \s a whitespace character
5705 \S a character that is not a whitespace character
5706 \v a vertical whitespace character
5707 \V a character that is not a vertical whitespace character
5708 \w a "word" character
5709 \W a "non-word" character
5710 \X an extended Unicode sequence
5711
5712 In PCRE, by default, \d, \D, \s, \S, \w, and \W recognize only ASCII
5713 characters, even in UTF-8 mode. However, this can be changed by setting
5714 the PCRE_UCP option.
5715
5716
5717 GENERAL CATEGORY PROPERTIES FOR \p and \P
5718
5719 C Other
5720 Cc Control
5721 Cf Format
5722 Cn Unassigned
5723 Co Private use
5724 Cs Surrogate
5725
5726 L Letter
5727 Ll Lower case letter
5728 Lm Modifier letter
5729 Lo Other letter
5730 Lt Title case letter
5731 Lu Upper case letter
5732 L& Ll, Lu, or Lt
5733
5734 M Mark
5735 Mc Spacing mark
5736 Me Enclosing mark
5737 Mn Non-spacing mark
5738
5739 N Number
5740 Nd Decimal number
5741 Nl Letter number
5742 No Other number
5743
5744 P Punctuation
5745 Pc Connector punctuation
5746 Pd Dash punctuation
5747 Pe Close punctuation
5748 Pf Final punctuation
5749 Pi Initial punctuation
5750 Po Other punctuation
5751 Ps Open punctuation
5752
5753 S Symbol
5754 Sc Currency symbol
5755 Sk Modifier symbol
5756 Sm Mathematical symbol
5757 So Other symbol
5758
5759 Z Separator
5760 Zl Line separator
5761 Zp Paragraph separator
5762 Zs Space separator
5763
5764
5765 PCRE SPECIAL CATEGORY PROPERTIES FOR \p and \P
5766
5767 Xan Alphanumeric: union of properties L and N
5768 Xps POSIX space: property Z or tab, NL, VT, FF, CR
5769 Xsp Perl space: property Z or tab, NL, FF, CR
5770 Xwd Perl word: property Xan or underscore
5771
5772
5773 SCRIPT NAMES FOR \p AND \P
5774
5775 Arabic, Armenian, Avestan, Balinese, Bamum, Bengali, Bopomofo, Braille,
5776 Buginese, Buhid, Canadian_Aboriginal, Carian, Cham, Cherokee, Common,
5777 Coptic, Cuneiform, Cypriot, Cyrillic, Deseret, Devanagari, Egyp-
5778 tian_Hieroglyphs, Ethiopic, Georgian, Glagolitic, Gothic, Greek,
5779 Gujarati, Gurmukhi, Han, Hangul, Hanunoo, Hebrew, Hiragana, Impe-
5780 rial_Aramaic, Inherited, Inscriptional_Pahlavi, Inscriptional_Parthian,
5781 Javanese, Kaithi, Kannada, Katakana, Kayah_Li, Kharoshthi, Khmer, Lao,
5782 Latin, Lepcha, Limbu, Linear_B, Lisu, Lycian, Lydian, Malayalam,
5783 Meetei_Mayek, Mongolian, Myanmar, New_Tai_Lue, Nko, Ogham, Old_Italic,
5784 Old_Persian, Old_South_Arabian, Old_Turkic, Ol_Chiki, Oriya, Osmanya,
5785 Phags_Pa, Phoenician, Rejang, Runic, Samaritan, Saurashtra, Shavian,
5786 Sinhala, Sundanese, Syloti_Nagri, Syriac, Tagalog, Tagbanwa, Tai_Le,
5787 Tai_Tham, Tai_Viet, Tamil, Telugu, Thaana, Thai, Tibetan, Tifinagh,
5788 Ugaritic, Vai, Yi.
5789
5790
5791 CHARACTER CLASSES
5792
5793 [...] positive character class
5794 [^...] negative character class
5795 [x-y] range (can be used for hex characters)
5796 [[:xxx:]] positive POSIX named set
5797 [[:^xxx:]] negative POSIX named set
5798
5799 alnum alphanumeric
5800 alpha alphabetic
5801 ascii 0-127
5802 blank space or tab
5803 cntrl control character
5804 digit decimal digit
5805 graph printing, excluding space
5806 lower lower case letter
5807 print printing, including space
5808 punct printing, excluding alphanumeric
5809 space whitespace
5810 upper upper case letter
5811 word same as \w
5812 xdigit hexadecimal digit
5813
5814 In PCRE, POSIX character set names recognize only ASCII characters by
5815 default, but some of them use Unicode properties if PCRE_UCP is set.
5816 You can use \Q...\E inside a character class.
5817
5818
5819 QUANTIFIERS
5820
5821 ? 0 or 1, greedy
5822 ?+ 0 or 1, possessive
5823 ?? 0 or 1, lazy
5824 * 0 or more, greedy
5825 *+ 0 or more, possessive
5826 *? 0 or more, lazy
5827 + 1 or more, greedy
5828 ++ 1 or more, possessive
5829 +? 1 or more, lazy
5830 {n} exactly n
5831 {n,m} at least n, no more than m, greedy
5832 {n,m}+ at least n, no more than m, possessive
5833 {n,m}? at least n, no more than m, lazy
5834 {n,} n or more, greedy
5835 {n,}+ n or more, possessive
5836 {n,}? n or more, lazy
5837
5838
5839 ANCHORS AND SIMPLE ASSERTIONS
5840
5841 \b word boundary
5842 \B not a word boundary
5843 ^ start of subject
5844 also after internal newline in multiline mode
5845 \A start of subject
5846 $ end of subject
5847 also before newline at end of subject
5848 also before internal newline in multiline mode
5849 \Z end of subject
5850 also before newline at end of subject
5851 \z end of subject
5852 \G first matching position in subject
5853
5854
5855 MATCH POINT RESET
5856
5857 \K reset start of match
5858
5859
5860 ALTERNATION
5861
5862 expr|expr|expr...
5863
5864
5865 CAPTURING
5866
5867 (...) capturing group
5868 (?<name>...) named capturing group (Perl)
5869 (?'name'...) named capturing group (Perl)
5870 (?P<name>...) named capturing group (Python)
5871 (?:...) non-capturing group
5872 (?|...) non-capturing group; reset group numbers for
5873 capturing groups in each alternative
5874
5875
5876 ATOMIC GROUPS
5877
5878 (?>...) atomic, non-capturing group
5879
5880
5881 COMMENT
5882
5883 (?#....) comment (not nestable)
5884
5885
5886 OPTION SETTING
5887
5888 (?i) caseless
5889 (?J) allow duplicate names
5890 (?m) multiline
5891 (?s) single line (dotall)
5892 (?U) default ungreedy (lazy)
5893 (?x) extended (ignore white space)
5894 (?-...) unset option(s)
5895
5896 The following are recognized only at the start of a pattern or after
5897 one of the newline-setting options with similar syntax:
5898
5899 (*UTF8) set UTF-8 mode (PCRE_UTF8)
5900 (*UCP) set PCRE_UCP (use Unicode properties for \d etc)
5901
5902
5903 LOOKAHEAD AND LOOKBEHIND ASSERTIONS
5904
5905 (?=...) positive look ahead
5906 (?!...) negative look ahead
5907 (?<=...) positive look behind
5908 (?<!...) negative look behind
5909
5910 Each top-level branch of a look behind must be of a fixed length.
5911
5912
5913 BACKREFERENCES
5914
5915 \n reference by number (can be ambiguous)
5916 \gn reference by number
5917 \g{n} reference by number
5918 \g{-n} relative reference by number
5919 \k<name> reference by name (Perl)
5920 \k'name' reference by name (Perl)
5921 \g{name} reference by name (Perl)
5922 \k{name} reference by name (.NET)
5923 (?P=name) reference by name (Python)
5924
5925
5926 SUBROUTINE REFERENCES (POSSIBLY RECURSIVE)
5927
5928 (?R) recurse whole pattern
5929 (?n) call subpattern by absolute number
5930 (?+n) call subpattern by relative number
5931 (?-n) call subpattern by relative number
5932 (?&name) call subpattern by name (Perl)
5933 (?P>name) call subpattern by name (Python)
5934 \g<name> call subpattern by name (Oniguruma)
5935 \g'name' call subpattern by name (Oniguruma)
5936 \g<n> call subpattern by absolute number (Oniguruma)
5937 \g'n' call subpattern by absolute number (Oniguruma)
5938 \g<+n> call subpattern by relative number (PCRE extension)
5939 \g'+n' call subpattern by relative number (PCRE extension)
5940 \g<-n> call subpattern by relative number (PCRE extension)
5941 \g'-n' call subpattern by relative number (PCRE extension)
5942
5943
5944 CONDITIONAL PATTERNS
5945
5946 (?(condition)yes-pattern)
5947 (?(condition)yes-pattern|no-pattern)
5948
5949 (?(n)... absolute reference condition
5950 (?(+n)... relative reference condition
5951 (?(-n)... relative reference condition
5952 (?(<name>)... named reference condition (Perl)
5953 (?('name')... named reference condition (Perl)
5954 (?(name)... named reference condition (PCRE)
5955 (?(R)... overall recursion condition
5956 (?(Rn)... specific group recursion condition
5957 (?(R&name)... specific recursion condition
5958 (?(DEFINE)... define subpattern for reference
5959 (?(assert)... assertion condition
5960
5961
5962 BACKTRACKING CONTROL
5963
5964 The following act immediately they are reached:
5965
5966 (*ACCEPT) force successful match
5967 (*FAIL) force backtrack; synonym (*F)
5968
5969 The following act only when a subsequent match failure causes a back-
5970 track to reach them. They all force a match failure, but they differ in
5971 what happens afterwards. Those that advance the start-of-match point do
5972 so only if the pattern is not anchored.
5973
5974 (*COMMIT) overall failure, no advance of starting point
5975 (*PRUNE) advance to next starting character
5976 (*SKIP) advance start to current matching position
5977 (*THEN) local failure, backtrack to next alternation
5978
5979
5980 NEWLINE CONVENTIONS
5981
5982 These are recognized only at the very start of the pattern or after a
5983 (*BSR_...) or (*UTF8) or (*UCP) option.
5984
5985 (*CR) carriage return only
5986 (*LF) linefeed only
5987 (*CRLF) carriage return followed by linefeed
5988 (*ANYCRLF) all three of the above
5989 (*ANY) any Unicode newline sequence
5990
5991
5992 WHAT \R MATCHES
5993
5994 These are recognized only at the very start of the pattern or after a
5995 (*...) option that sets the newline convention or UTF-8 or UCP mode.
5996
5997 (*BSR_ANYCRLF) CR, LF, or CRLF
5998 (*BSR_UNICODE) any Unicode newline sequence
5999
6000
6001 CALLOUTS
6002
6003 (?C) callout
6004 (?Cn) callout with data n
6005
6006
6007 SEE ALSO
6008
6009 pcrepattern(3), pcreapi(3), pcrecallout(3), pcrematching(3), pcre(3).
6010
6011
6012 AUTHOR
6013
6014 Philip Hazel
6015 University Computing Service
6016 Cambridge CB2 3QH, England.
6017
6018
6019 REVISION
6020
6021 Last updated: 12 May 2010
6022 Copyright (c) 1997-2010 University of Cambridge.
6023 ------------------------------------------------------------------------------
6024
6025
6026 PCREPARTIAL(3) PCREPARTIAL(3)
6027
6028
6029 NAME
6030 PCRE - Perl-compatible regular expressions
6031
6032
6033 PARTIAL MATCHING IN PCRE
6034
6035 In normal use of PCRE, if the subject string that is passed to
6036 pcre_exec() or pcre_dfa_exec() matches as far as it goes, but is too
6037 short to match the entire pattern, PCRE_ERROR_NOMATCH is returned.
6038 There are circumstances where it might be helpful to distinguish this
6039 case from other cases in which there is no match.
6040
6041 Consider, for example, an application where a human is required to type
6042 in data for a field with specific formatting requirements. An example
6043