/[pcre]/code/trunk/doc/pcrepattern.3
ViewVC logotype

Contents of /code/trunk/doc/pcrepattern.3

Parent Directory Parent Directory | Revision Log Revision Log


Revision 182 - (show annotations) (download)
Wed Jun 13 15:09:54 2007 UTC (7 years, 6 months ago) by ph10
File size: 80398 byte(s)
More document tidies, pre-release.

1 .TH PCREPATTERN 3
2 .SH NAME
3 PCRE - Perl-compatible regular expressions
4 .SH "PCRE REGULAR EXPRESSION DETAILS"
5 .rs
6 .sp
7 The syntax and semantics of the regular expressions supported by PCRE are
8 described below. Regular expressions are also described in the Perl
9 documentation and in a number of books, some of which have copious examples.
10 Jeffrey Friedl's "Mastering Regular Expressions", published by O'Reilly, covers
11 regular expressions in great detail. This description of PCRE's regular
12 expressions is intended as reference material.
13 .P
14 The original operation of PCRE was on strings of one-byte characters. However,
15 there is now also support for UTF-8 character strings. To use this, you must
16 build PCRE to include UTF-8 support, and then call \fBpcre_compile()\fP with
17 the PCRE_UTF8 option. How this affects pattern matching is mentioned in several
18 places below. There is also a summary of UTF-8 features in the
19 .\" HTML <a href="pcre.html#utf8support">
20 .\" </a>
21 section on UTF-8 support
22 .\"
23 in the main
24 .\" HREF
25 \fBpcre\fP
26 .\"
27 page.
28 .P
29 The remainder of this document discusses the patterns that are supported by
30 PCRE when its main matching function, \fBpcre_exec()\fP, is used.
31 From release 6.0, PCRE offers a second matching function,
32 \fBpcre_dfa_exec()\fP, which matches using a different algorithm that is not
33 Perl-compatible. Some of the features discussed below are not available when
34 \fBpcre_dfa_exec()\fP is used. The advantages and disadvantages of the
35 alternative function, and how it differs from the normal function, are
36 discussed in the
37 .\" HREF
38 \fBpcrematching\fP
39 .\"
40 page.
41 .
42 .
43 .SH "CHARACTERS AND METACHARACTERS"
44 .rs
45 .sp
46 A regular expression is a pattern that is matched against a subject string from
47 left to right. Most characters stand for themselves in a pattern, and match the
48 corresponding characters in the subject. As a trivial example, the pattern
49 .sp
50 The quick brown fox
51 .sp
52 matches a portion of a subject string that is identical to itself. When
53 caseless matching is specified (the PCRE_CASELESS option), letters are matched
54 independently of case. In UTF-8 mode, PCRE always understands the concept of
55 case for characters whose values are less than 128, so caseless matching is
56 always possible. For characters with higher values, the concept of case is
57 supported if PCRE is compiled with Unicode property support, but not otherwise.
58 If you want to use caseless matching for characters 128 and above, you must
59 ensure that PCRE is compiled with Unicode property support as well as with
60 UTF-8 support.
61 .P
62 The power of regular expressions comes from the ability to include alternatives
63 and repetitions in the pattern. These are encoded in the pattern by the use of
64 \fImetacharacters\fP, which do not stand for themselves but instead are
65 interpreted in some special way.
66 .P
67 There are two different sets of metacharacters: those that are recognized
68 anywhere in the pattern except within square brackets, and those that are
69 recognized within square brackets. Outside square brackets, the metacharacters
70 are as follows:
71 .sp
72 \e general escape character with several uses
73 ^ assert start of string (or line, in multiline mode)
74 $ assert end of string (or line, in multiline mode)
75 . match any character except newline (by default)
76 [ start character class definition
77 | start of alternative branch
78 ( start subpattern
79 ) end subpattern
80 ? extends the meaning of (
81 also 0 or 1 quantifier
82 also quantifier minimizer
83 * 0 or more quantifier
84 + 1 or more quantifier
85 also "possessive quantifier"
86 { start min/max quantifier
87 .sp
88 Part of a pattern that is in square brackets is called a "character class". In
89 a character class the only metacharacters are:
90 .sp
91 \e general escape character
92 ^ negate the class, but only if the first character
93 - indicates character range
94 .\" JOIN
95 [ POSIX character class (only if followed by POSIX
96 syntax)
97 ] terminates the character class
98 .sp
99 The following sections describe the use of each of the metacharacters.
100 .
101 .
102 .SH BACKSLASH
103 .rs
104 .sp
105 The backslash character has several uses. Firstly, if it is followed by a
106 non-alphanumeric character, it takes away any special meaning that character
107 may have. This use of backslash as an escape character applies both inside and
108 outside character classes.
109 .P
110 For example, if you want to match a * character, you write \e* in the pattern.
111 This escaping action applies whether or not the following character would
112 otherwise be interpreted as a metacharacter, so it is always safe to precede a
113 non-alphanumeric with backslash to specify that it stands for itself. In
114 particular, if you want to match a backslash, you write \e\e.
115 .P
116 If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the
117 pattern (other than in a character class) and characters between a # outside
118 a character class and the next newline are ignored. An escaping backslash can
119 be used to include a whitespace or # character as part of the pattern.
120 .P
121 If you want to remove the special meaning from a sequence of characters, you
122 can do so by putting them between \eQ and \eE. This is different from Perl in
123 that $ and @ are handled as literals in \eQ...\eE sequences in PCRE, whereas in
124 Perl, $ and @ cause variable interpolation. Note the following examples:
125 .sp
126 Pattern PCRE matches Perl matches
127 .sp
128 .\" JOIN
129 \eQabc$xyz\eE abc$xyz abc followed by the
130 contents of $xyz
131 \eQabc\e$xyz\eE abc\e$xyz abc\e$xyz
132 \eQabc\eE\e$\eQxyz\eE abc$xyz abc$xyz
133 .sp
134 The \eQ...\eE sequence is recognized both inside and outside character classes.
135 .
136 .
137 .\" HTML <a name="digitsafterbackslash"></a>
138 .SS "Non-printing characters"
139 .rs
140 .sp
141 A second use of backslash provides a way of encoding non-printing characters
142 in patterns in a visible manner. There is no restriction on the appearance of
143 non-printing characters, apart from the binary zero that terminates a pattern,
144 but when a pattern is being prepared by text editing, it is usually easier to
145 use one of the following escape sequences than the binary character it
146 represents:
147 .sp
148 \ea alarm, that is, the BEL character (hex 07)
149 \ecx "control-x", where x is any character
150 \ee escape (hex 1B)
151 \ef formfeed (hex 0C)
152 \en newline (hex 0A)
153 \er carriage return (hex 0D)
154 \et tab (hex 09)
155 \eddd character with octal code ddd, or backreference
156 \exhh character with hex code hh
157 \ex{hhh..} character with hex code hhh..
158 .sp
159 The precise effect of \ecx is as follows: if x is a lower case letter, it
160 is converted to upper case. Then bit 6 of the character (hex 40) is inverted.
161 Thus \ecz becomes hex 1A, but \ec{ becomes hex 3B, while \ec; becomes hex
162 7B.
163 .P
164 After \ex, from zero to two hexadecimal digits are read (letters can be in
165 upper or lower case). Any number of hexadecimal digits may appear between \ex{
166 and }, but the value of the character code must be less than 256 in non-UTF-8
167 mode, and less than 2**31 in UTF-8 mode (that is, the maximum hexadecimal value
168 is 7FFFFFFF). If characters other than hexadecimal digits appear between \ex{
169 and }, or if there is no terminating }, this form of escape is not recognized.
170 Instead, the initial \ex will be interpreted as a basic hexadecimal escape,
171 with no following digits, giving a character whose value is zero.
172 .P
173 Characters whose value is less than 256 can be defined by either of the two
174 syntaxes for \ex. There is no difference in the way they are handled. For
175 example, \exdc is exactly the same as \ex{dc}.
176 .P
177 After \e0 up to two further octal digits are read. If there are fewer than two
178 digits, just those that are present are used. Thus the sequence \e0\ex\e07
179 specifies two binary zeros followed by a BEL character (code value 7). Make
180 sure you supply two digits after the initial zero if the pattern character that
181 follows is itself an octal digit.
182 .P
183 The handling of a backslash followed by a digit other than 0 is complicated.
184 Outside a character class, PCRE reads it and any following digits as a decimal
185 number. If the number is less than 10, or if there have been at least that many
186 previous capturing left parentheses in the expression, the entire sequence is
187 taken as a \fIback reference\fP. A description of how this works is given
188 .\" HTML <a href="#backreferences">
189 .\" </a>
190 later,
191 .\"
192 following the discussion of
193 .\" HTML <a href="#subpattern">
194 .\" </a>
195 parenthesized subpatterns.
196 .\"
197 .P
198 Inside a character class, or if the decimal number is greater than 9 and there
199 have not been that many capturing subpatterns, PCRE re-reads up to three octal
200 digits following the backslash, and uses them to generate a data character. Any
201 subsequent digits stand for themselves. In non-UTF-8 mode, the value of a
202 character specified in octal must be less than \e400. In UTF-8 mode, values up
203 to \e777 are permitted. For example:
204 .sp
205 \e040 is another way of writing a space
206 .\" JOIN
207 \e40 is the same, provided there are fewer than 40
208 previous capturing subpatterns
209 \e7 is always a back reference
210 .\" JOIN
211 \e11 might be a back reference, or another way of
212 writing a tab
213 \e011 is always a tab
214 \e0113 is a tab followed by the character "3"
215 .\" JOIN
216 \e113 might be a back reference, otherwise the
217 character with octal code 113
218 .\" JOIN
219 \e377 might be a back reference, otherwise
220 the byte consisting entirely of 1 bits
221 .\" JOIN
222 \e81 is either a back reference, or a binary zero
223 followed by the two characters "8" and "1"
224 .sp
225 Note that octal values of 100 or greater must not be introduced by a leading
226 zero, because no more than three octal digits are ever read.
227 .P
228 All the sequences that define a single character value can be used both inside
229 and outside character classes. In addition, inside a character class, the
230 sequence \eb is interpreted as the backspace character (hex 08), and the
231 sequences \eR and \eX are interpreted as the characters "R" and "X",
232 respectively. Outside a character class, these sequences have different
233 meanings
234 .\" HTML <a href="#uniextseq">
235 .\" </a>
236 (see below).
237 .\"
238 .
239 .
240 .SS "Absolute and relative back references"
241 .rs
242 .sp
243 The sequence \eg followed by a positive or negative number, optionally enclosed
244 in braces, is an absolute or relative back reference. A named back reference
245 can be coded as \eg{name}. Back references are discussed
246 .\" HTML <a href="#backreferences">
247 .\" </a>
248 later,
249 .\"
250 following the discussion of
251 .\" HTML <a href="#subpattern">
252 .\" </a>
253 parenthesized subpatterns.
254 .\"
255 .
256 .
257 .SS "Generic character types"
258 .rs
259 .sp
260 Another use of backslash is for specifying generic character types. The
261 following are always recognized:
262 .sp
263 \ed any decimal digit
264 \eD any character that is not a decimal digit
265 \eh any horizontal whitespace character
266 \eH any character that is not a horizontal whitespace character
267 \es any whitespace character
268 \eS any character that is not a whitespace character
269 \ev any vertical whitespace character
270 \eV any character that is not a vertical whitespace character
271 \ew any "word" character
272 \eW any "non-word" character
273 .sp
274 Each pair of escape sequences partitions the complete set of characters into
275 two disjoint sets. Any given character matches one, and only one, of each pair.
276 .P
277 These character type sequences can appear both inside and outside character
278 classes. They each match one character of the appropriate type. If the current
279 matching point is at the end of the subject string, all of them fail, since
280 there is no character to match.
281 .P
282 For compatibility with Perl, \es does not match the VT character (code 11).
283 This makes it different from the the POSIX "space" class. The \es characters
284 are HT (9), LF (10), FF (12), CR (13), and space (32). If "use locale;" is
285 included in a Perl script, \es may match the VT character. In PCRE, it never
286 does.
287 .P
288 In UTF-8 mode, characters with values greater than 128 never match \ed, \es, or
289 \ew, and always match \eD, \eS, and \eW. This is true even when Unicode
290 character property support is available. These sequences retain their original
291 meanings from before UTF-8 support was available, mainly for efficiency
292 reasons.
293 .P
294 The sequences \eh, \eH, \ev, and \eV are Perl 5.10 features. In contrast to the
295 other sequences, these do match certain high-valued codepoints in UTF-8 mode.
296 The horizontal space characters are:
297 .sp
298 U+0009 Horizontal tab
299 U+0020 Space
300 U+00A0 Non-break space
301 U+1680 Ogham space mark
302 U+180E Mongolian vowel separator
303 U+2000 En quad
304 U+2001 Em quad
305 U+2002 En space
306 U+2003 Em space
307 U+2004 Three-per-em space
308 U+2005 Four-per-em space
309 U+2006 Six-per-em space
310 U+2007 Figure space
311 U+2008 Punctuation space
312 U+2009 Thin space
313 U+200A Hair space
314 U+202F Narrow no-break space
315 U+205F Medium mathematical space
316 U+3000 Ideographic space
317 .sp
318 The vertical space characters are:
319 .sp
320 U+000A Linefeed
321 U+000B Vertical tab
322 U+000C Formfeed
323 U+000D Carriage return
324 U+0085 Next line
325 U+2028 Line separator
326 U+2029 Paragraph separator
327 .P
328 A "word" character is an underscore or any character less than 256 that is a
329 letter or digit. The definition of letters and digits is controlled by PCRE's
330 low-valued character tables, and may vary if locale-specific matching is taking
331 place (see
332 .\" HTML <a href="pcreapi.html#localesupport">
333 .\" </a>
334 "Locale support"
335 .\"
336 in the
337 .\" HREF
338 \fBpcreapi\fP
339 .\"
340 page). For example, in a French locale such as "fr_FR" in Unix-like systems,
341 or "french" in Windows, some character codes greater than 128 are used for
342 accented letters, and these are matched by \ew. The use of locales with Unicode
343 is discouraged.
344 .
345 .
346 .SS "Newline sequences"
347 .rs
348 .sp
349 Outside a character class, the escape sequence \eR matches any Unicode newline
350 sequence. This is a Perl 5.10 feature. In non-UTF-8 mode \eR is equivalent to
351 the following:
352 .sp
353 (?>\er\en|\en|\ex0b|\ef|\er|\ex85)
354 .sp
355 This is an example of an "atomic group", details of which are given
356 .\" HTML <a href="#atomicgroup">
357 .\" </a>
358 below.
359 .\"
360 This particular group matches either the two-character sequence CR followed by
361 LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab,
362 U+000B), FF (formfeed, U+000C), CR (carriage return, U+000D), or NEL (next
363 line, U+0085). The two-character sequence is treated as a single unit that
364 cannot be split.
365 .P
366 In UTF-8 mode, two additional characters whose codepoints are greater than 255
367 are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029).
368 Unicode character property support is not needed for these characters to be
369 recognized.
370 .P
371 Inside a character class, \eR matches the letter "R".
372 .
373 .
374 .\" HTML <a name="uniextseq"></a>
375 .SS Unicode character properties
376 .rs
377 .sp
378 When PCRE is built with Unicode character property support, three additional
379 escape sequences to match character properties are available when UTF-8 mode
380 is selected. They are:
381 .sp
382 \ep{\fIxx\fP} a character with the \fIxx\fP property
383 \eP{\fIxx\fP} a character without the \fIxx\fP property
384 \eX an extended Unicode sequence
385 .sp
386 The property names represented by \fIxx\fP above are limited to the Unicode
387 script names, the general category properties, and "Any", which matches any
388 character (including newline). Other properties such as "InMusicalSymbols" are
389 not currently supported by PCRE. Note that \eP{Any} does not match any
390 characters, so always causes a match failure.
391 .P
392 Sets of Unicode characters are defined as belonging to certain scripts. A
393 character from one of these sets can be matched using a script name. For
394 example:
395 .sp
396 \ep{Greek}
397 \eP{Han}
398 .sp
399 Those that are not part of an identified script are lumped together as
400 "Common". The current list of scripts is:
401 .P
402 Arabic,
403 Armenian,
404 Balinese,
405 Bengali,
406 Bopomofo,
407 Braille,
408 Buginese,
409 Buhid,
410 Canadian_Aboriginal,
411 Cherokee,
412 Common,
413 Coptic,
414 Cuneiform,
415 Cypriot,
416 Cyrillic,
417 Deseret,
418 Devanagari,
419 Ethiopic,
420 Georgian,
421 Glagolitic,
422 Gothic,
423 Greek,
424 Gujarati,
425 Gurmukhi,
426 Han,
427 Hangul,
428 Hanunoo,
429 Hebrew,
430 Hiragana,
431 Inherited,
432 Kannada,
433 Katakana,
434 Kharoshthi,
435 Khmer,
436 Lao,
437 Latin,
438 Limbu,
439 Linear_B,
440 Malayalam,
441 Mongolian,
442 Myanmar,
443 New_Tai_Lue,
444 Nko,
445 Ogham,
446 Old_Italic,
447 Old_Persian,
448 Oriya,
449 Osmanya,
450 Phags_Pa,
451 Phoenician,
452 Runic,
453 Shavian,
454 Sinhala,
455 Syloti_Nagri,
456 Syriac,
457 Tagalog,
458 Tagbanwa,
459 Tai_Le,
460 Tamil,
461 Telugu,
462 Thaana,
463 Thai,
464 Tibetan,
465 Tifinagh,
466 Ugaritic,
467 Yi.
468 .P
469 Each character has exactly one general category property, specified by a
470 two-letter abbreviation. For compatibility with Perl, negation can be specified
471 by including a circumflex between the opening brace and the property name. For
472 example, \ep{^Lu} is the same as \eP{Lu}.
473 .P
474 If only one letter is specified with \ep or \eP, it includes all the general
475 category properties that start with that letter. In this case, in the absence
476 of negation, the curly brackets in the escape sequence are optional; these two
477 examples have the same effect:
478 .sp
479 \ep{L}
480 \epL
481 .sp
482 The following general category property codes are supported:
483 .sp
484 C Other
485 Cc Control
486 Cf Format
487 Cn Unassigned
488 Co Private use
489 Cs Surrogate
490 .sp
491 L Letter
492 Ll Lower case letter
493 Lm Modifier letter
494 Lo Other letter
495 Lt Title case letter
496 Lu Upper case letter
497 .sp
498 M Mark
499 Mc Spacing mark
500 Me Enclosing mark
501 Mn Non-spacing mark
502 .sp
503 N Number
504 Nd Decimal number
505 Nl Letter number
506 No Other number
507 .sp
508 P Punctuation
509 Pc Connector punctuation
510 Pd Dash punctuation
511 Pe Close punctuation
512 Pf Final punctuation
513 Pi Initial punctuation
514 Po Other punctuation
515 Ps Open punctuation
516 .sp
517 S Symbol
518 Sc Currency symbol
519 Sk Modifier symbol
520 Sm Mathematical symbol
521 So Other symbol
522 .sp
523 Z Separator
524 Zl Line separator
525 Zp Paragraph separator
526 Zs Space separator
527 .sp
528 The special property L& is also supported: it matches a character that has
529 the Lu, Ll, or Lt property, in other words, a letter that is not classified as
530 a modifier or "other".
531 .P
532 The long synonyms for these properties that Perl supports (such as \ep{Letter})
533 are not supported by PCRE, nor is it permitted to prefix any of these
534 properties with "Is".
535 .P
536 No character that is in the Unicode table has the Cn (unassigned) property.
537 Instead, this property is assumed for any code point that is not in the
538 Unicode table.
539 .P
540 Specifying caseless matching does not affect these escape sequences. For
541 example, \ep{Lu} always matches only upper case letters.
542 .P
543 The \eX escape matches any number of Unicode characters that form an extended
544 Unicode sequence. \eX is equivalent to
545 .sp
546 (?>\ePM\epM*)
547 .sp
548 That is, it matches a character without the "mark" property, followed by zero
549 or more characters with the "mark" property, and treats the sequence as an
550 atomic group
551 .\" HTML <a href="#atomicgroup">
552 .\" </a>
553 (see below).
554 .\"
555 Characters with the "mark" property are typically accents that affect the
556 preceding character.
557 .P
558 Matching characters by Unicode property is not fast, because PCRE has to search
559 a structure that contains data for over fifteen thousand characters. That is
560 why the traditional escape sequences such as \ed and \ew do not use Unicode
561 properties in PCRE.
562 .
563 .
564 .\" HTML <a name="resetmatchstart"></a>
565 .SS "Resetting the match start"
566 .rs
567 .sp
568 The escape sequence \eK, which is a Perl 5.10 feature, causes any previously
569 matched characters not to be included in the final matched sequence. For
570 example, the pattern:
571 .sp
572 foo\eKbar
573 .sp
574 matches "foobar", but reports that it has matched "bar". This feature is
575 similar to a lookbehind assertion
576 .\" HTML <a href="#lookbehind">
577 .\" </a>
578 (described below).
579 .\"
580 However, in this case, the part of the subject before the real match does not
581 have to be of fixed length, as lookbehind assertions do. The use of \eK does
582 not interfere with the setting of
583 .\" HTML <a href="#subpattern">
584 .\" </a>
585 captured substrings.
586 .\"
587 For example, when the pattern
588 .sp
589 (foo)\eKbar
590 .sp
591 matches "foobar", the first substring is still set to "foo".
592 .
593 .
594 .\" HTML <a name="smallassertions"></a>
595 .SS "Simple assertions"
596 .rs
597 .sp
598 The final use of backslash is for certain simple assertions. An assertion
599 specifies a condition that has to be met at a particular point in a match,
600 without consuming any characters from the subject string. The use of
601 subpatterns for more complicated assertions is described
602 .\" HTML <a href="#bigassertions">
603 .\" </a>
604 below.
605 .\"
606 The backslashed assertions are:
607 .sp
608 \eb matches at a word boundary
609 \eB matches when not at a word boundary
610 \eA matches at the start of the subject
611 \eZ matches at the end of the subject
612 also matches before a newline at the end of the subject
613 \ez matches only at the end of the subject
614 \eG matches at the first matching position in the subject
615 .sp
616 These assertions may not appear in character classes (but note that \eb has a
617 different meaning, namely the backspace character, inside a character class).
618 .P
619 A word boundary is a position in the subject string where the current character
620 and the previous character do not both match \ew or \eW (i.e. one matches
621 \ew and the other matches \eW), or the start or end of the string if the
622 first or last character matches \ew, respectively.
623 .P
624 The \eA, \eZ, and \ez assertions differ from the traditional circumflex and
625 dollar (described in the next section) in that they only ever match at the very
626 start and end of the subject string, whatever options are set. Thus, they are
627 independent of multiline mode. These three assertions are not affected by the
628 PCRE_NOTBOL or PCRE_NOTEOL options, which affect only the behaviour of the
629 circumflex and dollar metacharacters. However, if the \fIstartoffset\fP
630 argument of \fBpcre_exec()\fP is non-zero, indicating that matching is to start
631 at a point other than the beginning of the subject, \eA can never match. The
632 difference between \eZ and \ez is that \eZ matches before a newline at the end
633 of the string as well as at the very end, whereas \ez matches only at the end.
634 .P
635 The \eG assertion is true only when the current matching position is at the
636 start point of the match, as specified by the \fIstartoffset\fP argument of
637 \fBpcre_exec()\fP. It differs from \eA when the value of \fIstartoffset\fP is
638 non-zero. By calling \fBpcre_exec()\fP multiple times with appropriate
639 arguments, you can mimic Perl's /g option, and it is in this kind of
640 implementation where \eG can be useful.
641 .P
642 Note, however, that PCRE's interpretation of \eG, as the start of the current
643 match, is subtly different from Perl's, which defines it as the end of the
644 previous match. In Perl, these can be different when the previously matched
645 string was empty. Because PCRE does just one match at a time, it cannot
646 reproduce this behaviour.
647 .P
648 If all the alternatives of a pattern begin with \eG, the expression is anchored
649 to the starting match position, and the "anchored" flag is set in the compiled
650 regular expression.
651 .
652 .
653 .SH "CIRCUMFLEX AND DOLLAR"
654 .rs
655 .sp
656 Outside a character class, in the default matching mode, the circumflex
657 character is an assertion that is true only if the current matching point is
658 at the start of the subject string. If the \fIstartoffset\fP argument of
659 \fBpcre_exec()\fP is non-zero, circumflex can never match if the PCRE_MULTILINE
660 option is unset. Inside a character class, circumflex has an entirely different
661 meaning
662 .\" HTML <a href="#characterclass">
663 .\" </a>
664 (see below).
665 .\"
666 .P
667 Circumflex need not be the first character of the pattern if a number of
668 alternatives are involved, but it should be the first thing in each alternative
669 in which it appears if the pattern is ever to match that branch. If all
670 possible alternatives start with a circumflex, that is, if the pattern is
671 constrained to match only at the start of the subject, it is said to be an
672 "anchored" pattern. (There are also other constructs that can cause a pattern
673 to be anchored.)
674 .P
675 A dollar character is an assertion that is true only if the current matching
676 point is at the end of the subject string, or immediately before a newline
677 at the end of the string (by default). Dollar need not be the last character of
678 the pattern if a number of alternatives are involved, but it should be the last
679 item in any branch in which it appears. Dollar has no special meaning in a
680 character class.
681 .P
682 The meaning of dollar can be changed so that it matches only at the very end of
683 the string, by setting the PCRE_DOLLAR_ENDONLY option at compile time. This
684 does not affect the \eZ assertion.
685 .P
686 The meanings of the circumflex and dollar characters are changed if the
687 PCRE_MULTILINE option is set. When this is the case, a circumflex matches
688 immediately after internal newlines as well as at the start of the subject
689 string. It does not match after a newline that ends the string. A dollar
690 matches before any newlines in the string, as well as at the very end, when
691 PCRE_MULTILINE is set. When newline is specified as the two-character
692 sequence CRLF, isolated CR and LF characters do not indicate newlines.
693 .P
694 For example, the pattern /^abc$/ matches the subject string "def\enabc" (where
695 \en represents a newline) in multiline mode, but not otherwise. Consequently,
696 patterns that are anchored in single line mode because all branches start with
697 ^ are not anchored in multiline mode, and a match for circumflex is possible
698 when the \fIstartoffset\fP argument of \fBpcre_exec()\fP is non-zero. The
699 PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
700 .P
701 Note that the sequences \eA, \eZ, and \ez can be used to match the start and
702 end of the subject in both modes, and if all branches of a pattern start with
703 \eA it is always anchored, whether or not PCRE_MULTILINE is set.
704 .
705 .
706 .SH "FULL STOP (PERIOD, DOT)"
707 .rs
708 .sp
709 Outside a character class, a dot in the pattern matches any one character in
710 the subject string except (by default) a character that signifies the end of a
711 line. In UTF-8 mode, the matched character may be more than one byte long.
712 .P
713 When a line ending is defined as a single character, dot never matches that
714 character; when the two-character sequence CRLF is used, dot does not match CR
715 if it is immediately followed by LF, but otherwise it matches all characters
716 (including isolated CRs and LFs). When any Unicode line endings are being
717 recognized, dot does not match CR or LF or any of the other line ending
718 characters.
719 .P
720 The behaviour of dot with regard to newlines can be changed. If the PCRE_DOTALL
721 option is set, a dot matches any one character, without exception. If the
722 two-character sequence CRLF is present in the subject string, it takes two dots
723 to match it.
724 .P
725 The handling of dot is entirely independent of the handling of circumflex and
726 dollar, the only relationship being that they both involve newlines. Dot has no
727 special meaning in a character class.
728 .
729 .
730 .SH "MATCHING A SINGLE BYTE"
731 .rs
732 .sp
733 Outside a character class, the escape sequence \eC matches any one byte, both
734 in and out of UTF-8 mode. Unlike a dot, it always matches any line-ending
735 characters. The feature is provided in Perl in order to match individual bytes
736 in UTF-8 mode. Because it breaks up UTF-8 characters into individual bytes,
737 what remains in the string may be a malformed UTF-8 string. For this reason,
738 the \eC escape sequence is best avoided.
739 .P
740 PCRE does not allow \eC to appear in lookbehind assertions
741 .\" HTML <a href="#lookbehind">
742 .\" </a>
743 (described below),
744 .\"
745 because in UTF-8 mode this would make it impossible to calculate the length of
746 the lookbehind.
747 .
748 .
749 .\" HTML <a name="characterclass"></a>
750 .SH "SQUARE BRACKETS AND CHARACTER CLASSES"
751 .rs
752 .sp
753 An opening square bracket introduces a character class, terminated by a closing
754 square bracket. A closing square bracket on its own is not special. If a
755 closing square bracket is required as a member of the class, it should be the
756 first data character in the class (after an initial circumflex, if present) or
757 escaped with a backslash.
758 .P
759 A character class matches a single character in the subject. In UTF-8 mode, the
760 character may occupy more than one byte. A matched character must be in the set
761 of characters defined by the class, unless the first character in the class
762 definition is a circumflex, in which case the subject character must not be in
763 the set defined by the class. If a circumflex is actually required as a member
764 of the class, ensure it is not the first character, or escape it with a
765 backslash.
766 .P
767 For example, the character class [aeiou] matches any lower case vowel, while
768 [^aeiou] matches any character that is not a lower case vowel. Note that a
769 circumflex is just a convenient notation for specifying the characters that
770 are in the class by enumerating those that are not. A class that starts with a
771 circumflex is not an assertion: it still consumes a character from the subject
772 string, and therefore it fails if the current pointer is at the end of the
773 string.
774 .P
775 In UTF-8 mode, characters with values greater than 255 can be included in a
776 class as a literal string of bytes, or by using the \ex{ escaping mechanism.
777 .P
778 When caseless matching is set, any letters in a class represent both their
779 upper case and lower case versions, so for example, a caseless [aeiou] matches
780 "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
781 caseful version would. In UTF-8 mode, PCRE always understands the concept of
782 case for characters whose values are less than 128, so caseless matching is
783 always possible. For characters with higher values, the concept of case is
784 supported if PCRE is compiled with Unicode property support, but not otherwise.
785 If you want to use caseless matching for characters 128 and above, you must
786 ensure that PCRE is compiled with Unicode property support as well as with
787 UTF-8 support.
788 .P
789 Characters that might indicate line breaks are never treated in any special way
790 when matching character classes, whatever line-ending sequence is in use, and
791 whatever setting of the PCRE_DOTALL and PCRE_MULTILINE options is used. A class
792 such as [^a] always matches one of these characters.
793 .P
794 The minus (hyphen) character can be used to specify a range of characters in a
795 character class. For example, [d-m] matches any letter between d and m,
796 inclusive. If a minus character is required in a class, it must be escaped with
797 a backslash or appear in a position where it cannot be interpreted as
798 indicating a range, typically as the first or last character in the class.
799 .P
800 It is not possible to have the literal character "]" as the end character of a
801 range. A pattern such as [W-]46] is interpreted as a class of two characters
802 ("W" and "-") followed by a literal string "46]", so it would match "W46]" or
803 "-46]". However, if the "]" is escaped with a backslash it is interpreted as
804 the end of range, so [W-\e]46] is interpreted as a class containing a range
805 followed by two other characters. The octal or hexadecimal representation of
806 "]" can also be used to end a range.
807 .P
808 Ranges operate in the collating sequence of character values. They can also be
809 used for characters specified numerically, for example [\e000-\e037]. In UTF-8
810 mode, ranges can include characters whose values are greater than 255, for
811 example [\ex{100}-\ex{2ff}].
812 .P
813 If a range that includes letters is used when caseless matching is set, it
814 matches the letters in either case. For example, [W-c] is equivalent to
815 [][\e\e^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if character
816 tables for a French locale are in use, [\exc8-\excb] matches accented E
817 characters in both cases. In UTF-8 mode, PCRE supports the concept of case for
818 characters with values greater than 128 only when it is compiled with Unicode
819 property support.
820 .P
821 The character types \ed, \eD, \ep, \eP, \es, \eS, \ew, and \eW may also appear
822 in a character class, and add the characters that they match to the class. For
823 example, [\edABCDEF] matches any hexadecimal digit. A circumflex can
824 conveniently be used with the upper case character types to specify a more
825 restricted set of characters than the matching lower case type. For example,
826 the class [^\eW_] matches any letter or digit, but not underscore.
827 .P
828 The only metacharacters that are recognized in character classes are backslash,
829 hyphen (only where it can be interpreted as specifying a range), circumflex
830 (only at the start), opening square bracket (only when it can be interpreted as
831 introducing a POSIX class name - see the next section), and the terminating
832 closing square bracket. However, escaping other non-alphanumeric characters
833 does no harm.
834 .
835 .
836 .SH "POSIX CHARACTER CLASSES"
837 .rs
838 .sp
839 Perl supports the POSIX notation for character classes. This uses names
840 enclosed by [: and :] within the enclosing square brackets. PCRE also supports
841 this notation. For example,
842 .sp
843 [01[:alpha:]%]
844 .sp
845 matches "0", "1", any alphabetic character, or "%". The supported class names
846 are
847 .sp
848 alnum letters and digits
849 alpha letters
850 ascii character codes 0 - 127
851 blank space or tab only
852 cntrl control characters
853 digit decimal digits (same as \ed)
854 graph printing characters, excluding space
855 lower lower case letters
856 print printing characters, including space
857 punct printing characters, excluding letters and digits
858 space white space (not quite the same as \es)
859 upper upper case letters
860 word "word" characters (same as \ew)
861 xdigit hexadecimal digits
862 .sp
863 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13), and
864 space (32). Notice that this list includes the VT character (code 11). This
865 makes "space" different to \es, which does not include VT (for Perl
866 compatibility).
867 .P
868 The name "word" is a Perl extension, and "blank" is a GNU extension from Perl
869 5.8. Another Perl extension is negation, which is indicated by a ^ character
870 after the colon. For example,
871 .sp
872 [12[:^digit:]]
873 .sp
874 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX
875 syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
876 supported, and an error is given if they are encountered.
877 .P
878 In UTF-8 mode, characters with values greater than 128 do not match any of
879 the POSIX character classes.
880 .
881 .
882 .SH "VERTICAL BAR"
883 .rs
884 .sp
885 Vertical bar characters are used to separate alternative patterns. For example,
886 the pattern
887 .sp
888 gilbert|sullivan
889 .sp
890 matches either "gilbert" or "sullivan". Any number of alternatives may appear,
891 and an empty alternative is permitted (matching the empty string). The matching
892 process tries each alternative in turn, from left to right, and the first one
893 that succeeds is used. If the alternatives are within a subpattern
894 .\" HTML <a href="#subpattern">
895 .\" </a>
896 (defined below),
897 .\"
898 "succeeds" means matching the rest of the main pattern as well as the
899 alternative in the subpattern.
900 .
901 .
902 .SH "INTERNAL OPTION SETTING"
903 .rs
904 .sp
905 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
906 PCRE_EXTENDED options can be changed from within the pattern by a sequence of
907 Perl option letters enclosed between "(?" and ")". The option letters are
908 .sp
909 i for PCRE_CASELESS
910 m for PCRE_MULTILINE
911 s for PCRE_DOTALL
912 x for PCRE_EXTENDED
913 .sp
914 For example, (?im) sets caseless, multiline matching. It is also possible to
915 unset these options by preceding the letter with a hyphen, and a combined
916 setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
917 PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
918 permitted. If a letter appears both before and after the hyphen, the option is
919 unset.
920 .P
921 When an option change occurs at top level (that is, not inside subpattern
922 parentheses), the change applies to the remainder of the pattern that follows.
923 If the change is placed right at the start of a pattern, PCRE extracts it into
924 the global options (and it will therefore show up in data extracted by the
925 \fBpcre_fullinfo()\fP function).
926 .P
927 An option change within a subpattern (see below for a description of
928 subpatterns) affects only that part of the current pattern that follows it, so
929 .sp
930 (a(?i)b)c
931 .sp
932 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
933 By this means, options can be made to have different settings in different
934 parts of the pattern. Any changes made in one alternative do carry on
935 into subsequent branches within the same subpattern. For example,
936 .sp
937 (a(?i)b|c)
938 .sp
939 matches "ab", "aB", "c", and "C", even though when matching "C" the first
940 branch is abandoned before the option setting. This is because the effects of
941 option settings happen at compile time. There would be some very weird
942 behaviour otherwise.
943 .P
944 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA can be
945 changed in the same way as the Perl-compatible options by using the characters
946 J, U and X respectively.
947 .
948 .
949 .\" HTML <a name="subpattern"></a>
950 .SH SUBPATTERNS
951 .rs
952 .sp
953 Subpatterns are delimited by parentheses (round brackets), which can be nested.
954 Turning part of a pattern into a subpattern does two things:
955 .sp
956 1. It localizes a set of alternatives. For example, the pattern
957 .sp
958 cat(aract|erpillar|)
959 .sp
960 matches one of the words "cat", "cataract", or "caterpillar". Without the
961 parentheses, it would match "cataract", "erpillar" or an empty string.
962 .sp
963 2. It sets up the subpattern as a capturing subpattern. This means that, when
964 the whole pattern matches, that portion of the subject string that matched the
965 subpattern is passed back to the caller via the \fIovector\fP argument of
966 \fBpcre_exec()\fP. Opening parentheses are counted from left to right (starting
967 from 1) to obtain numbers for the capturing subpatterns.
968 .P
969 For example, if the string "the red king" is matched against the pattern
970 .sp
971 the ((red|white) (king|queen))
972 .sp
973 the captured substrings are "red king", "red", and "king", and are numbered 1,
974 2, and 3, respectively.
975 .P
976 The fact that plain parentheses fulfil two functions is not always helpful.
977 There are often times when a grouping subpattern is required without a
978 capturing requirement. If an opening parenthesis is followed by a question mark
979 and a colon, the subpattern does not do any capturing, and is not counted when
980 computing the number of any subsequent capturing subpatterns. For example, if
981 the string "the white queen" is matched against the pattern
982 .sp
983 the ((?:red|white) (king|queen))
984 .sp
985 the captured substrings are "white queen" and "queen", and are numbered 1 and
986 2. The maximum number of capturing subpatterns is 65535.
987 .P
988 As a convenient shorthand, if any option settings are required at the start of
989 a non-capturing subpattern, the option letters may appear between the "?" and
990 the ":". Thus the two patterns
991 .sp
992 (?i:saturday|sunday)
993 (?:(?i)saturday|sunday)
994 .sp
995 match exactly the same set of strings. Because alternative branches are tried
996 from left to right, and options are not reset until the end of the subpattern
997 is reached, an option setting in one branch does affect subsequent branches, so
998 the above patterns match "SUNDAY" as well as "Saturday".
999 .
1000 .
1001 .SH "DUPLICATE SUBPATTERN NUMBERS"
1002 .rs
1003 .sp
1004 Perl 5.10 introduced a feature whereby each alternative in a subpattern uses
1005 the same numbers for its capturing parentheses. Such a subpattern starts with
1006 (?| and is itself a non-capturing subpattern. For example, consider this
1007 pattern:
1008 .sp
1009 (?|(Sat)ur|(Sun))day
1010 .sp
1011 Because the two alternatives are inside a (?| group, both sets of capturing
1012 parentheses are numbered one. Thus, when the pattern matches, you can look
1013 at captured substring number one, whichever alternative matched. This construct
1014 is useful when you want to capture part, but not all, of one of a number of
1015 alternatives. Inside a (?| group, parentheses are numbered as usual, but the
1016 number is reset at the start of each branch. The numbers of any capturing
1017 buffers that follow the subpattern start after the highest number used in any
1018 branch. The following example is taken from the Perl documentation.
1019 The numbers underneath show in which buffer the captured content will be
1020 stored.
1021 .sp
1022 # before ---------------branch-reset----------- after
1023 / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
1024 # 1 2 2 3 2 3 4
1025 .sp
1026 A backreference or a recursive call to a numbered subpattern always refers to
1027 the first one in the pattern with the given number.
1028 .P
1029 An alternative approach to using this "branch reset" feature is to use
1030 duplicate named subpatterns, as described in the next section.
1031 .
1032 .
1033 .SH "NAMED SUBPATTERNS"
1034 .rs
1035 .sp
1036 Identifying capturing parentheses by number is simple, but it can be very hard
1037 to keep track of the numbers in complicated regular expressions. Furthermore,
1038 if an expression is modified, the numbers may change. To help with this
1039 difficulty, PCRE supports the naming of subpatterns. This feature was not
1040 added to Perl until release 5.10. Python had the feature earlier, and PCRE
1041 introduced it at release 4.0, using the Python syntax. PCRE now supports both
1042 the Perl and the Python syntax.
1043 .P
1044 In PCRE, a subpattern can be named in one of three ways: (?<name>...) or
1045 (?'name'...) as in Perl, or (?P<name>...) as in Python. References to capturing
1046 parentheses from other parts of the pattern, such as
1047 .\" HTML <a href="#backreferences">
1048 .\" </a>
1049 backreferences,
1050 .\"
1051 .\" HTML <a href="#recursion">
1052 .\" </a>
1053 recursion,
1054 .\"
1055 and
1056 .\" HTML <a href="#conditions">
1057 .\" </a>
1058 conditions,
1059 .\"
1060 can be made by name as well as by number.
1061 .P
1062 Names consist of up to 32 alphanumeric characters and underscores. Named
1063 capturing parentheses are still allocated numbers as well as names, exactly as
1064 if the names were not present. The PCRE API provides function calls for
1065 extracting the name-to-number translation table from a compiled pattern. There
1066 is also a convenience function for extracting a captured substring by name.
1067 .P
1068 By default, a name must be unique within a pattern, but it is possible to relax
1069 this constraint by setting the PCRE_DUPNAMES option at compile time. This can
1070 be useful for patterns where only one instance of the named parentheses can
1071 match. Suppose you want to match the name of a weekday, either as a 3-letter
1072 abbreviation or as the full name, and in both cases you want to extract the
1073 abbreviation. This pattern (ignoring the line breaks) does the job:
1074 .sp
1075 (?<DN>Mon|Fri|Sun)(?:day)?|
1076 (?<DN>Tue)(?:sday)?|
1077 (?<DN>Wed)(?:nesday)?|
1078 (?<DN>Thu)(?:rsday)?|
1079 (?<DN>Sat)(?:urday)?
1080 .sp
1081 There are five capturing substrings, but only one is ever set after a match.
1082 (An alternative way of solving this problem is to use a "branch reset"
1083 subpattern, as described in the previous section.)
1084 .P
1085 The convenience function for extracting the data by name returns the substring
1086 for the first (and in this example, the only) subpattern of that name that
1087 matched. This saves searching to find which numbered subpattern it was. If you
1088 make a reference to a non-unique named subpattern from elsewhere in the
1089 pattern, the one that corresponds to the lowest number is used. For further
1090 details of the interfaces for handling named subpatterns, see the
1091 .\" HREF
1092 \fBpcreapi\fP
1093 .\"
1094 documentation.
1095 .
1096 .
1097 .SH REPETITION
1098 .rs
1099 .sp
1100 Repetition is specified by quantifiers, which can follow any of the following
1101 items:
1102 .sp
1103 a literal data character
1104 the dot metacharacter
1105 the \eC escape sequence
1106 the \eX escape sequence (in UTF-8 mode with Unicode properties)
1107 the \eR escape sequence
1108 an escape such as \ed that matches a single character
1109 a character class
1110 a back reference (see next section)
1111 a parenthesized subpattern (unless it is an assertion)
1112 .sp
1113 The general repetition quantifier specifies a minimum and maximum number of
1114 permitted matches, by giving the two numbers in curly brackets (braces),
1115 separated by a comma. The numbers must be less than 65536, and the first must
1116 be less than or equal to the second. For example:
1117 .sp
1118 z{2,4}
1119 .sp
1120 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
1121 character. If the second number is omitted, but the comma is present, there is
1122 no upper limit; if the second number and the comma are both omitted, the
1123 quantifier specifies an exact number of required matches. Thus
1124 .sp
1125 [aeiou]{3,}
1126 .sp
1127 matches at least 3 successive vowels, but may match many more, while
1128 .sp
1129 \ed{8}
1130 .sp
1131 matches exactly 8 digits. An opening curly bracket that appears in a position
1132 where a quantifier is not allowed, or one that does not match the syntax of a
1133 quantifier, is taken as a literal character. For example, {,6} is not a
1134 quantifier, but a literal string of four characters.
1135 .P
1136 In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to individual
1137 bytes. Thus, for example, \ex{100}{2} matches two UTF-8 characters, each of
1138 which is represented by a two-byte sequence. Similarly, when Unicode property
1139 support is available, \eX{3} matches three Unicode extended sequences, each of
1140 which may be several bytes long (and they may be of different lengths).
1141 .P
1142 The quantifier {0} is permitted, causing the expression to behave as if the
1143 previous item and the quantifier were not present.
1144 .P
1145 For convenience, the three most common quantifiers have single-character
1146 abbreviations:
1147 .sp
1148 * is equivalent to {0,}
1149 + is equivalent to {1,}
1150 ? is equivalent to {0,1}
1151 .sp
1152 It is possible to construct infinite loops by following a subpattern that can
1153 match no characters with a quantifier that has no upper limit, for example:
1154 .sp
1155 (a?)*
1156 .sp
1157 Earlier versions of Perl and PCRE used to give an error at compile time for
1158 such patterns. However, because there are cases where this can be useful, such
1159 patterns are now accepted, but if any repetition of the subpattern does in fact
1160 match no characters, the loop is forcibly broken.
1161 .P
1162 By default, the quantifiers are "greedy", that is, they match as much as
1163 possible (up to the maximum number of permitted times), without causing the
1164 rest of the pattern to fail. The classic example of where this gives problems
1165 is in trying to match comments in C programs. These appear between /* and */
1166 and within the comment, individual * and / characters may appear. An attempt to
1167 match C comments by applying the pattern
1168 .sp
1169 /\e*.*\e*/
1170 .sp
1171 to the string
1172 .sp
1173 /* first comment */ not comment /* second comment */
1174 .sp
1175 fails, because it matches the entire string owing to the greediness of the .*
1176 item.
1177 .P
1178 However, if a quantifier is followed by a question mark, it ceases to be
1179 greedy, and instead matches the minimum number of times possible, so the
1180 pattern
1181 .sp
1182 /\e*.*?\e*/
1183 .sp
1184 does the right thing with the C comments. The meaning of the various
1185 quantifiers is not otherwise changed, just the preferred number of matches.
1186 Do not confuse this use of question mark with its use as a quantifier in its
1187 own right. Because it has two uses, it can sometimes appear doubled, as in
1188 .sp
1189 \ed??\ed
1190 .sp
1191 which matches one digit by preference, but can match two if that is the only
1192 way the rest of the pattern matches.
1193 .P
1194 If the PCRE_UNGREEDY option is set (an option that is not available in Perl),
1195 the quantifiers are not greedy by default, but individual ones can be made
1196 greedy by following them with a question mark. In other words, it inverts the
1197 default behaviour.
1198 .P
1199 When a parenthesized subpattern is quantified with a minimum repeat count that
1200 is greater than 1 or with a limited maximum, more memory is required for the
1201 compiled pattern, in proportion to the size of the minimum or maximum.
1202 .P
1203 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent
1204 to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is
1205 implicitly anchored, because whatever follows will be tried against every
1206 character position in the subject string, so there is no point in retrying the
1207 overall match at any position after the first. PCRE normally treats such a
1208 pattern as though it were preceded by \eA.
1209 .P
1210 In cases where it is known that the subject string contains no newlines, it is
1211 worth setting PCRE_DOTALL in order to obtain this optimization, or
1212 alternatively using ^ to indicate anchoring explicitly.
1213 .P
1214 However, there is one situation where the optimization cannot be used. When .*
1215 is inside capturing parentheses that are the subject of a backreference
1216 elsewhere in the pattern, a match at the start may fail where a later one
1217 succeeds. Consider, for example:
1218 .sp
1219 (.*)abc\e1
1220 .sp
1221 If the subject is "xyz123abc123" the match point is the fourth character. For
1222 this reason, such a pattern is not implicitly anchored.
1223 .P
1224 When a capturing subpattern is repeated, the value captured is the substring
1225 that matched the final iteration. For example, after
1226 .sp
1227 (tweedle[dume]{3}\es*)+
1228 .sp
1229 has matched "tweedledum tweedledee" the value of the captured substring is
1230 "tweedledee". However, if there are nested capturing subpatterns, the
1231 corresponding captured values may have been set in previous iterations. For
1232 example, after
1233 .sp
1234 /(a|(b))+/
1235 .sp
1236 matches "aba" the value of the second captured substring is "b".
1237 .
1238 .
1239 .\" HTML <a name="atomicgroup"></a>
1240 .SH "ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS"
1241 .rs
1242 .sp
1243 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
1244 repetition, failure of what follows normally causes the repeated item to be
1245 re-evaluated to see if a different number of repeats allows the rest of the
1246 pattern to match. Sometimes it is useful to prevent this, either to change the
1247 nature of the match, or to cause it fail earlier than it otherwise might, when
1248 the author of the pattern knows there is no point in carrying on.
1249 .P
1250 Consider, for example, the pattern \ed+foo when applied to the subject line
1251 .sp
1252 123456bar
1253 .sp
1254 After matching all 6 digits and then failing to match "foo", the normal
1255 action of the matcher is to try again with only 5 digits matching the \ed+
1256 item, and then with 4, and so on, before ultimately failing. "Atomic grouping"
1257 (a term taken from Jeffrey Friedl's book) provides the means for specifying
1258 that once a subpattern has matched, it is not to be re-evaluated in this way.
1259 .P
1260 If we use atomic grouping for the previous example, the matcher gives up
1261 immediately on failing to match "foo" the first time. The notation is a kind of
1262 special parenthesis, starting with (?> as in this example:
1263 .sp
1264 (?>\ed+)foo
1265 .sp
1266 This kind of parenthesis "locks up" the part of the pattern it contains once
1267 it has matched, and a failure further into the pattern is prevented from
1268 backtracking into it. Backtracking past it to previous items, however, works as
1269 normal.
1270 .P
1271 An alternative description is that a subpattern of this type matches the string
1272 of characters that an identical standalone pattern would match, if anchored at
1273 the current point in the subject string.
1274 .P
1275 Atomic grouping subpatterns are not capturing subpatterns. Simple cases such as
1276 the above example can be thought of as a maximizing repeat that must swallow
1277 everything it can. So, while both \ed+ and \ed+? are prepared to adjust the
1278 number of digits they match in order to make the rest of the pattern match,
1279 (?>\ed+) can only match an entire sequence of digits.
1280 .P
1281 Atomic groups in general can of course contain arbitrarily complicated
1282 subpatterns, and can be nested. However, when the subpattern for an atomic
1283 group is just a single repeated item, as in the example above, a simpler
1284 notation, called a "possessive quantifier" can be used. This consists of an
1285 additional + character following a quantifier. Using this notation, the
1286 previous example can be rewritten as
1287 .sp
1288 \ed++foo
1289 .sp
1290 Possessive quantifiers are always greedy; the setting of the PCRE_UNGREEDY
1291 option is ignored. They are a convenient notation for the simpler forms of
1292 atomic group. However, there is no difference in the meaning of a possessive
1293 quantifier and the equivalent atomic group, though there may be a performance
1294 difference; possessive quantifiers should be slightly faster.
1295 .P
1296 The possessive quantifier syntax is an extension to the Perl 5.8 syntax.
1297 Jeffrey Friedl originated the idea (and the name) in the first edition of his
1298 book. Mike McCloskey liked it, so implemented it when he built Sun's Java
1299 package, and PCRE copied it from there. It ultimately found its way into Perl
1300 at release 5.10.
1301 .P
1302 PCRE has an optimization that automatically "possessifies" certain simple
1303 pattern constructs. For example, the sequence A+B is treated as A++B because
1304 there is no point in backtracking into a sequence of A's when B must follow.
1305 .P
1306 When a pattern contains an unlimited repeat inside a subpattern that can itself
1307 be repeated an unlimited number of times, the use of an atomic group is the
1308 only way to avoid some failing matches taking a very long time indeed. The
1309 pattern
1310 .sp
1311 (\eD+|<\ed+>)*[!?]
1312 .sp
1313 matches an unlimited number of substrings that either consist of non-digits, or
1314 digits enclosed in <>, followed by either ! or ?. When it matches, it runs
1315 quickly. However, if it is applied to
1316 .sp
1317 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
1318 .sp
1319 it takes a long time before reporting failure. This is because the string can
1320 be divided between the internal \eD+ repeat and the external * repeat in a
1321 large number of ways, and all have to be tried. (The example uses [!?] rather
1322 than a single character at the end, because both PCRE and Perl have an
1323 optimization that allows for fast failure when a single character is used. They
1324 remember the last single character that is required for a match, and fail early
1325 if it is not present in the string.) If the pattern is changed so that it uses
1326 an atomic group, like this:
1327 .sp
1328 ((?>\eD+)|<\ed+>)*[!?]
1329 .sp
1330 sequences of non-digits cannot be broken, and failure happens quickly.
1331 .
1332 .
1333 .\" HTML <a name="backreferences"></a>
1334 .SH "BACK REFERENCES"
1335 .rs
1336 .sp
1337 Outside a character class, a backslash followed by a digit greater than 0 (and
1338 possibly further digits) is a back reference to a capturing subpattern earlier
1339 (that is, to its left) in the pattern, provided there have been that many
1340 previous capturing left parentheses.
1341 .P
1342 However, if the decimal number following the backslash is less than 10, it is
1343 always taken as a back reference, and causes an error only if there are not
1344 that many capturing left parentheses in the entire pattern. In other words, the
1345 parentheses that are referenced need not be to the left of the reference for
1346 numbers less than 10. A "forward back reference" of this type can make sense
1347 when a repetition is involved and the subpattern to the right has participated
1348 in an earlier iteration.
1349 .P
1350 It is not possible to have a numerical "forward back reference" to a subpattern
1351 whose number is 10 or more using this syntax because a sequence such as \e50 is
1352 interpreted as a character defined in octal. See the subsection entitled
1353 "Non-printing characters"
1354 .\" HTML <a href="#digitsafterbackslash">
1355 .\" </a>
1356 above
1357 .\"
1358 for further details of the handling of digits following a backslash. There is
1359 no such problem when named parentheses are used. A back reference to any
1360 subpattern is possible using named parentheses (see below).
1361 .P
1362 Another way of avoiding the ambiguity inherent in the use of digits following a
1363 backslash is to use the \eg escape sequence, which is a feature introduced in
1364 Perl 5.10. This escape must be followed by a positive or a negative number,
1365 optionally enclosed in braces. These examples are all identical:
1366 .sp
1367 (ring), \e1
1368 (ring), \eg1
1369 (ring), \eg{1}
1370 .sp
1371 A positive number specifies an absolute reference without the ambiguity that is
1372 present in the older syntax. It is also useful when literal digits follow the
1373 reference. A negative number is a relative reference. Consider this example:
1374 .sp
1375 (abc(def)ghi)\eg{-1}
1376 .sp
1377 The sequence \eg{-1} is a reference to the most recently started capturing
1378 subpattern before \eg, that is, is it equivalent to \e2. Similarly, \eg{-2}
1379 would be equivalent to \e1. The use of relative references can be helpful in
1380 long patterns, and also in patterns that are created by joining together
1381 fragments that contain references within themselves.
1382 .P
1383 A back reference matches whatever actually matched the capturing subpattern in
1384 the current subject string, rather than anything matching the subpattern
1385 itself (see
1386 .\" HTML <a href="#subpatternsassubroutines">
1387 .\" </a>
1388 "Subpatterns as subroutines"
1389 .\"
1390 below for a way of doing that). So the pattern
1391 .sp
1392 (sens|respons)e and \e1ibility
1393 .sp
1394 matches "sense and sensibility" and "response and responsibility", but not
1395 "sense and responsibility". If caseful matching is in force at the time of the
1396 back reference, the case of letters is relevant. For example,
1397 .sp
1398 ((?i)rah)\es+\e1
1399 .sp
1400 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
1401 capturing subpattern is matched caselessly.
1402 .P
1403 There are several different ways of writing back references to named
1404 subpatterns. The .NET syntax \ek{name} and the Perl syntax \ek<name> or
1405 \ek'name' are supported, as is the Python syntax (?P=name). Perl 5.10's unified
1406 back reference syntax, in which \eg can be used for both numeric and named
1407 references, is also supported. We could rewrite the above example in any of
1408 the following ways:
1409 .sp
1410 (?<p1>(?i)rah)\es+\ek<p1>
1411 (?'p1'(?i)rah)\es+\ek{p1}
1412 (?P<p1>(?i)rah)\es+(?P=p1)
1413 (?<p1>(?i)rah)\es+\eg{p1}
1414 .sp
1415 A subpattern that is referenced by name may appear in the pattern before or
1416 after the reference.
1417 .P
1418 There may be more than one back reference to the same subpattern. If a
1419 subpattern has not actually been used in a particular match, any back
1420 references to it always fail. For example, the pattern
1421 .sp
1422 (a|(bc))\e2
1423 .sp
1424 always fails if it starts to match "a" rather than "bc". Because there may be
1425 many capturing parentheses in a pattern, all digits following the backslash are
1426 taken as part of a potential back reference number. If the pattern continues
1427 with a digit character, some delimiter must be used to terminate the back
1428 reference. If the PCRE_EXTENDED option is set, this can be whitespace.
1429 Otherwise an empty comment (see
1430 .\" HTML <a href="#comments">
1431 .\" </a>
1432 "Comments"
1433 .\"
1434 below) can be used.
1435 .P
1436 A back reference that occurs inside the parentheses to which it refers fails
1437 when the subpattern is first used, so, for example, (a\e1) never matches.
1438 However, such references can be useful inside repeated subpatterns. For
1439 example, the pattern
1440 .sp
1441 (a|b\e1)+
1442 .sp
1443 matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
1444 the subpattern, the back reference matches the character string corresponding
1445 to the previous iteration. In order for this to work, the pattern must be such
1446 that the first iteration does not need to match the back reference. This can be
1447 done using alternation, as in the example above, or by a quantifier with a
1448 minimum of zero.
1449 .
1450 .
1451 .\" HTML <a name="bigassertions"></a>
1452 .SH ASSERTIONS
1453 .rs
1454 .sp
1455 An assertion is a test on the characters following or preceding the current
1456 matching point that does not actually consume any characters. The simple
1457 assertions coded as \eb, \eB, \eA, \eG, \eZ, \ez, ^ and $ are described
1458 .\" HTML <a href="#smallassertions">
1459 .\" </a>
1460 above.
1461 .\"
1462 .P
1463 More complicated assertions are coded as subpatterns. There are two kinds:
1464 those that look ahead of the current position in the subject string, and those
1465 that look behind it. An assertion subpattern is matched in the normal way,
1466 except that it does not cause the current matching position to be changed.
1467 .P
1468 Assertion subpatterns are not capturing subpatterns, and may not be repeated,
1469 because it makes no sense to assert the same thing several times. If any kind
1470 of assertion contains capturing subpatterns within it, these are counted for
1471 the purposes of numbering the capturing subpatterns in the whole pattern.
1472 However, substring capturing is carried out only for positive assertions,
1473 because it does not make sense for negative assertions.
1474 .
1475 .
1476 .SS "Lookahead assertions"
1477 .rs
1478 .sp
1479 Lookahead assertions start with (?= for positive assertions and (?! for
1480 negative assertions. For example,
1481 .sp
1482 \ew+(?=;)
1483 .sp
1484 matches a word followed by a semicolon, but does not include the semicolon in
1485 the match, and
1486 .sp
1487 foo(?!bar)
1488 .sp
1489 matches any occurrence of "foo" that is not followed by "bar". Note that the
1490 apparently similar pattern
1491 .sp
1492 (?!foo)bar
1493 .sp
1494 does not find an occurrence of "bar" that is preceded by something other than
1495 "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
1496 (?!foo) is always true when the next three characters are "bar". A
1497 lookbehind assertion is needed to achieve the other effect.
1498 .P
1499 If you want to force a matching failure at some point in a pattern, the most
1500 convenient way to do it is with (?!) because an empty string always matches, so
1501 an assertion that requires there not to be an empty string must always fail.
1502 .
1503 .
1504 .\" HTML <a name="lookbehind"></a>
1505 .SS "Lookbehind assertions"
1506 .rs
1507 .sp
1508 Lookbehind assertions start with (?<= for positive assertions and (?<! for
1509 negative assertions. For example,
1510 .sp
1511 (?<!foo)bar
1512 .sp
1513 does find an occurrence of "bar" that is not preceded by "foo". The contents of
1514 a lookbehind assertion are restricted such that all the strings it matches must
1515 have a fixed length. However, if there are several top-level alternatives, they
1516 do not all have to have the same fixed length. Thus
1517 .sp
1518 (?<=bullock|donkey)
1519 .sp
1520 is permitted, but
1521 .sp
1522 (?<!dogs?|cats?)
1523 .sp
1524 causes an error at compile time. Branches that match different length strings
1525 are permitted only at the top level of a lookbehind assertion. This is an
1526 extension compared with Perl (at least for 5.8), which requires all branches to
1527 match the same length of string. An assertion such as
1528 .sp
1529 (?<=ab(c|de))
1530 .sp
1531 is not permitted, because its single top-level branch can match two different
1532 lengths, but it is acceptable if rewritten to use two top-level branches:
1533 .sp
1534 (?<=abc|abde)
1535 .sp
1536 In some cases, the Perl 5.10 escape sequence \eK
1537 .\" HTML <a href="#resetmatchstart">
1538 .\" </a>
1539 (see above)
1540 .\"
1541 can be used instead of a lookbehind assertion; this is not restricted to a
1542 fixed-length.
1543 .P
1544 The implementation of lookbehind assertions is, for each alternative, to
1545 temporarily move the current position back by the fixed length and then try to
1546 match. If there are insufficient characters before the current position, the
1547 assertion fails.
1548 .P
1549 PCRE does not allow the \eC escape (which matches a single byte in UTF-8 mode)
1550 to appear in lookbehind assertions, because it makes it impossible to calculate
1551 the length of the lookbehind. The \eX and \eR escapes, which can match
1552 different numbers of bytes, are also not permitted.
1553 .P
1554 Possessive quantifiers can be used in conjunction with lookbehind assertions to
1555 specify efficient matching at the end of the subject string. Consider a simple
1556 pattern such as
1557 .sp
1558 abcd$
1559 .sp
1560 when applied to a long string that does not match. Because matching proceeds
1561 from left to right, PCRE will look for each "a" in the subject and then see if
1562 what follows matches the rest of the pattern. If the pattern is specified as
1563 .sp
1564 ^.*abcd$
1565 .sp
1566 the initial .* matches the entire string at first, but when this fails (because
1567 there is no following "a"), it backtracks to match all but the last character,
1568 then all but the last two characters, and so on. Once again the search for "a"
1569 covers the entire string, from right to left, so we are no better off. However,
1570 if the pattern is written as
1571 .sp
1572 ^.*+(?<=abcd)
1573 .sp
1574 there can be no backtracking for the .*+ item; it can match only the entire
1575 string. The subsequent lookbehind assertion does a single test on the last four
1576 characters. If it fails, the match fails immediately. For long strings, this
1577 approach makes a significant difference to the processing time.
1578 .
1579 .
1580 .SS "Using multiple assertions"
1581 .rs
1582 .sp
1583 Several assertions (of any sort) may occur in succession. For example,
1584 .sp
1585 (?<=\ed{3})(?<!999)foo
1586 .sp
1587 matches "foo" preceded by three digits that are not "999". Notice that each of
1588 the assertions is applied independently at the same point in the subject
1589 string. First there is a check that the previous three characters are all
1590 digits, and then there is a check that the same three characters are not "999".
1591 This pattern does \fInot\fP match "foo" preceded by six characters, the first
1592 of which are digits and the last three of which are not "999". For example, it
1593 doesn't match "123abcfoo". A pattern to do that is
1594 .sp
1595 (?<=\ed{3}...)(?<!999)foo
1596 .sp
1597 This time the first assertion looks at the preceding six characters, checking
1598 that the first three are digits, and then the second assertion checks that the
1599 preceding three characters are not "999".
1600 .P
1601 Assertions can be nested in any combination. For example,
1602 .sp
1603 (?<=(?<!foo)bar)baz
1604 .sp
1605 matches an occurrence of "baz" that is preceded by "bar" which in turn is not
1606 preceded by "foo", while
1607 .sp
1608 (?<=\ed{3}(?!999)...)foo
1609 .sp
1610 is another pattern that matches "foo" preceded by three digits and any three
1611 characters that are not "999".
1612 .
1613 .
1614 .\" HTML <a name="conditions"></a>
1615 .SH "CONDITIONAL SUBPATTERNS"
1616 .rs
1617 .sp
1618 It is possible to cause the matching process to obey a subpattern
1619 conditionally or to choose between two alternative subpatterns, depending on
1620 the result of an assertion, or whether a previous capturing subpattern matched
1621 or not. The two possible forms of conditional subpattern are
1622 .sp
1623 (?(condition)yes-pattern)
1624 (?(condition)yes-pattern|no-pattern)
1625 .sp
1626 If the condition is satisfied, the yes-pattern is used; otherwise the
1627 no-pattern (if present) is used. If there are more than two alternatives in the
1628 subpattern, a compile-time error occurs.
1629 .P
1630 There are four kinds of condition: references to subpatterns, references to
1631 recursion, a pseudo-condition called DEFINE, and assertions.
1632 .
1633 .SS "Checking for a used subpattern by number"
1634 .rs
1635 .sp
1636 If the text between the parentheses consists of a sequence of digits, the
1637 condition is true if the capturing subpattern of that number has previously
1638 matched. An alternative notation is to precede the digits with a plus or minus
1639 sign. In this case, the subpattern number is relative rather than absolute.
1640 The most recently opened parentheses can be referenced by (?(-1), the next most
1641 recent by (?(-2), and so on. In looping constructs it can also make sense to
1642 refer to subsequent groups with constructs such as (?(+2).
1643 .P
1644 Consider the following pattern, which contains non-significant white space to
1645 make it more readable (assume the PCRE_EXTENDED option) and to divide it into
1646 three parts for ease of discussion:
1647 .sp
1648 ( \e( )? [^()]+ (?(1) \e) )
1649 .sp
1650 The first part matches an optional opening parenthesis, and if that
1651 character is present, sets it as the first captured substring. The second part
1652 matches one or more characters that are not parentheses. The third part is a
1653 conditional subpattern that tests whether the first set of parentheses matched
1654 or not. If they did, that is, if subject started with an opening parenthesis,
1655 the condition is true, and so the yes-pattern is executed and a closing
1656 parenthesis is required. Otherwise, since no-pattern is not present, the
1657 subpattern matches nothing. In other words, this pattern matches a sequence of
1658 non-parentheses, optionally enclosed in parentheses.
1659 .P
1660 If you were embedding this pattern in a larger one, you could use a relative
1661 reference:
1662 .sp
1663 ...other stuff... ( \e( )? [^()]+ (?(-1) \e) ) ...
1664 .sp
1665 This makes the fragment independent of the parentheses in the larger pattern.
1666 .
1667 .SS "Checking for a used subpattern by name"
1668 .rs
1669 .sp
1670 Perl uses the syntax (?(<name>)...) or (?('name')...) to test for a used
1671 subpattern by name. For compatibility with earlier versions of PCRE, which had
1672 this facility before Perl, the syntax (?(name)...) is also recognized. However,
1673 there is a possible ambiguity with this syntax, because subpattern names may
1674 consist entirely of digits. PCRE looks first for a named subpattern; if it
1675 cannot find one and the name consists entirely of digits, PCRE looks for a
1676 subpattern of that number, which must be greater than zero. Using subpattern
1677 names that consist entirely of digits is not recommended.
1678 .P
1679 Rewriting the above example to use a named subpattern gives this:
1680 .sp
1681 (?<OPEN> \e( )? [^()]+ (?(<OPEN>) \e) )
1682 .sp
1683 .
1684 .SS "Checking for pattern recursion"
1685 .rs
1686 .sp
1687 If the condition is the string (R), and there is no subpattern with the name R,
1688 the condition is true if a recursive call to the whole pattern or any
1689 subpattern has been made. If digits or a name preceded by ampersand follow the
1690 letter R, for example:
1691 .sp
1692 (?(R3)...) or (?(R&name)...)
1693 .sp
1694 the condition is true if the most recent recursion is into the subpattern whose
1695 number or name is given. This condition does not check the entire recursion
1696 stack.
1697 .P
1698 At "top level", all these recursion test conditions are false. Recursive
1699 patterns are described below.
1700 .
1701 .SS "Defining subpatterns for use by reference only"
1702 .rs
1703 .sp
1704 If the condition is the string (DEFINE), and there is no subpattern with the
1705 name DEFINE, the condition is always false. In this case, there may be only one
1706 alternative in the subpattern. It is always skipped if control reaches this
1707 point in the pattern; the idea of DEFINE is that it can be used to define
1708 "subroutines" that can be referenced from elsewhere. (The use of "subroutines"
1709 is described below.) For example, a pattern to match an IPv4 address could be
1710 written like this (ignore whitespace and line breaks):
1711 .sp
1712 (?(DEFINE) (?<byte> 2[0-4]\ed | 25[0-5] | 1\ed\ed | [1-9]?\ed) )
1713 \eb (?&byte) (\e.(?&byte)){3} \eb
1714 .sp
1715 The first part of the pattern is a DEFINE group inside which a another group
1716 named "byte" is defined. This matches an individual component of an IPv4
1717 address (a number less than 256). When matching takes place, this part of the
1718 pattern is skipped because DEFINE acts like a false condition.
1719 .P
1720 The rest of the pattern uses references to the named group to match the four
1721 dot-separated components of an IPv4 address, insisting on a word boundary at
1722 each end.
1723 .
1724 .SS "Assertion conditions"
1725 .rs
1726 .sp
1727 If the condition is not in any of the above formats, it must be an assertion.
1728 This may be a positive or negative lookahead or lookbehind assertion. Consider
1729 this pattern, again containing non-significant white space, and with the two
1730 alternatives on the second line:
1731 .sp
1732 (?(?=[^a-z]*[a-z])
1733 \ed{2}-[a-z]{3}-\ed{2} | \ed{2}-\ed{2}-\ed{2} )
1734 .sp
1735 The condition is a positive lookahead assertion that matches an optional
1736 sequence of non-letters followed by a letter. In other words, it tests for the
1737 presence of at least one letter in the subject. If a letter is found, the
1738 subject is matched against the first alternative; otherwise it is matched
1739 against the second. This pattern matches strings in one of the two forms
1740 dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
1741 .
1742 .
1743 .\" HTML <a name="comments"></a>
1744 .SH COMMENTS
1745 .rs
1746 .sp
1747 The sequence (?# marks the start of a comment that continues up to the next
1748 closing parenthesis. Nested parentheses are not permitted. The characters
1749 that make up a comment play no part in the pattern matching at all.
1750 .P
1751 If the PCRE_EXTENDED option is set, an unescaped # character outside a
1752 character class introduces a comment that continues to immediately after the
1753 next newline in the pattern.
1754 .
1755 .
1756 .\" HTML <a name="recursion"></a>
1757 .SH "RECURSIVE PATTERNS"
1758 .rs
1759 .sp
1760 Consider the problem of matching a string in parentheses, allowing for
1761 unlimited nested parentheses. Without the use of recursion, the best that can
1762 be done is to use a pattern that matches up to some fixed depth of nesting. It
1763 is not possible to handle an arbitrary nesting depth.
1764 .P
1765 For some time, Perl has provided a facility that allows regular expressions to
1766 recurse (amongst other things). It does this by interpolating Perl code in the
1767 expression at run time, and the code can refer to the expression itself. A Perl
1768 pattern using code interpolation to solve the parentheses problem can be
1769 created like this:
1770 .sp
1771 $re = qr{\e( (?: (?>[^()]+) | (?p{$re}) )* \e)}x;
1772 .sp
1773 The (?p{...}) item interpolates Perl code at run time, and in this case refers
1774 recursively to the pattern in which it appears.
1775 .P
1776 Obviously, PCRE cannot support the interpolation of Perl code. Instead, it
1777 supports special syntax for recursion of the entire pattern, and also for
1778 individual subpattern recursion. After its introduction in PCRE and Python,
1779 this kind of recursion was introduced into Perl at release 5.10.
1780 .P
1781 A special item that consists of (? followed by a number greater than zero and a
1782 closing parenthesis is a recursive call of the subpattern of the given number,
1783 provided that it occurs inside that subpattern. (If not, it is a "subroutine"
1784 call, which is described in the next section.) The special item (?R) or (?0) is
1785 a recursive call of the entire regular expression.
1786 .P
1787 In PCRE (like Python, but unlike Perl), a recursive subpattern call is always
1788 treated as an atomic group. That is, once it has matched some of the subject
1789 string, it is never re-entered, even if it contains untried alternatives and
1790 there is a subsequent matching failure.
1791 .P
1792 This PCRE pattern solves the nested parentheses problem (assume the
1793 PCRE_EXTENDED option is set so that white space is ignored):
1794 .sp
1795 \e( ( (?>[^()]+) | (?R) )* \e)
1796 .sp
1797 First it matches an opening parenthesis. Then it matches any number of
1798 substrings which can either be a sequence of non-parentheses, or a recursive
1799 match of the pattern itself (that is, a correctly parenthesized substring).
1800 Finally there is a closing parenthesis.
1801 .P
1802 If this were part of a larger pattern, you would not want to recurse the entire
1803 pattern, so instead you could use this:
1804 .sp
1805 ( \e( ( (?>[^()]+) | (?1) )* \e) )
1806 .sp
1807 We have put the pattern into parentheses, and caused the recursion to refer to
1808 them instead of the whole pattern.
1809 .P
1810 In a larger pattern, keeping track of parenthesis numbers can be tricky. This
1811 is made easier by the use of relative references. (A Perl 5.10 feature.)
1812 Instead of (?1) in the pattern above you can write (?-2) to refer to the second
1813 most recently opened parentheses preceding the recursion. In other words, a
1814 negative number counts capturing parentheses leftwards from the point at which
1815 it is encountered.
1816 .P
1817 It is also possible to refer to subsequently opened parentheses, by writing
1818 references such as (?+2). However, these cannot be recursive because the
1819 reference is not inside the parentheses that are referenced. They are always
1820 "subroutine" calls, as described in the next section.
1821 .P
1822 An alternative approach is to use named parentheses instead. The Perl syntax
1823 for this is (?&name); PCRE's earlier syntax (?P>name) is also supported. We
1824 could rewrite the above example as follows:
1825 .sp
1826 (?<pn> \e( ( (?>[^()]+) | (?&pn) )* \e) )
1827 .sp
1828 If there is more than one subpattern with the same name, the earliest one is
1829 used.
1830 .P
1831 This particular example pattern that we have been looking at contains nested
1832 unlimited repeats, and so the use of atomic grouping for matching strings of
1833 non-parentheses is important when applying the pattern to strings that do not
1834 match. For example, when this pattern is applied to
1835 .sp
1836 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
1837 .sp
1838 it yields "no match" quickly. However, if atomic grouping is not used,
1839 the match runs for a very long time indeed because there are so many different
1840 ways the + and * repeats can carve up the subject, and all have to be tested
1841 before failure can be reported.
1842 .P
1843 At the end of a match, the values set for any capturing subpatterns are those
1844 from the outermost level of the recursion at which the subpattern value is set.
1845 If you want to obtain intermediate values, a callout function can be used (see
1846 below and the
1847 .\" HREF
1848 \fBpcrecallout\fP
1849 .\"
1850 documentation). If the pattern above is matched against
1851 .sp
1852 (ab(cd)ef)
1853 .sp
1854 the value for the capturing parentheses is "ef", which is the last value taken
1855 on at the top level. If additional parentheses are added, giving
1856 .sp
1857 \e( ( ( (?>[^()]+) | (?R) )* ) \e)
1858 ^ ^
1859 ^ ^
1860 .sp
1861 the string they capture is "ab(cd)ef", the contents of the top level
1862 parentheses. If there are more than 15 capturing parentheses in a pattern, PCRE
1863 has to obtain extra memory to store data during a recursion, which it does by
1864 using \fBpcre_malloc\fP, freeing it via \fBpcre_free\fP afterwards. If no
1865 memory can be obtained, the match fails with the PCRE_ERROR_NOMEMORY error.
1866 .P
1867 Do not confuse the (?R) item with the condition (R), which tests for recursion.
1868 Consider this pattern, which matches text in angle brackets, allowing for
1869 arbitrary nesting. Only digits are allowed in nested brackets (that is, when
1870 recursing), whereas any characters are permitted at the outer level.
1871 .sp
1872 < (?: (?(R) \ed++ | [^<>]*+) | (?R)) * >
1873 .sp
1874 In this pattern, (?(R) is the start of a conditional subpattern, with two
1875 different alternatives for the recursive and non-recursive cases. The (?R) item
1876 is the actual recursive call.
1877 .
1878 .
1879 .\" HTML <a name="subpatternsassubroutines"></a>
1880 .SH "SUBPATTERNS AS SUBROUTINES"
1881 .rs
1882 .sp
1883 If the syntax for a recursive subpattern reference (either by number or by
1884 name) is used outside the parentheses to which it refers, it operates like a
1885 subroutine in a programming language. The "called" subpattern may be defined
1886 before or after the reference. A numbered reference can be absolute or
1887 relative, as in these examples:
1888 .sp
1889 (...(absolute)...)...(?2)...
1890 (...(relative)...)...(?-1)...
1891 (...(?+1)...(relative)...
1892 .sp
1893 An earlier example pointed out that the pattern
1894 .sp
1895 (sens|respons)e and \e1ibility
1896 .sp
1897 matches "sense and sensibility" and "response and responsibility", but not
1898 "sense and responsibility". If instead the pattern
1899 .sp
1900 (sens|respons)e and (?1)ibility
1901 .sp
1902 is used, it does match "sense and responsibility" as well as the other two
1903 strings. Another example is given in the discussion of DEFINE above.
1904 .P
1905 Like recursive subpatterns, a "subroutine" call is always treated as an atomic
1906 group. That is, once it has matched some of the subject string, it is never
1907 re-entered, even if it contains untried alternatives and there is a subsequent
1908 matching failure.
1909 .P
1910 When a subpattern is used as a subroutine, processing options such as
1911 case-independence are fixed when the subpattern is defined. They cannot be
1912 changed for different calls. For example, consider this pattern:
1913 .sp
1914 (abc)(?i:(?-1))
1915 .sp
1916 It matches "abcabc". It does not match "abcABC" because the change of
1917 processing option does not affect the called subpattern.
1918 .
1919 .
1920 .SH CALLOUTS
1921 .rs
1922 .sp
1923 Perl has a feature whereby using the sequence (?{...}) causes arbitrary Perl
1924 code to be obeyed in the middle of matching a regular expression. This makes it
1925 possible, amongst other things, to extract different substrings that match the
1926 same pair of parentheses when there is a repetition.
1927 .P
1928 PCRE provides a similar feature, but of course it cannot obey arbitrary Perl
1929 code. The feature is called "callout". The caller of PCRE provides an external
1930 function by putting its entry point in the global variable \fIpcre_callout\fP.
1931 By default, this variable contains NULL, which disables all calling out.
1932 .P
1933 Within a regular expression, (?C) indicates the points at which the external
1934 function is to be called. If you want to identify different callout points, you
1935 can put a number less than 256 after the letter C. The default value is zero.
1936 For example, this pattern has two callout points:
1937 .sp
1938 (?C1)abc(?C2)def
1939 .sp
1940 If the PCRE_AUTO_CALLOUT flag is passed to \fBpcre_compile()\fP, callouts are
1941 automatically installed before each item in the pattern. They are all numbered
1942 255.
1943 .P
1944 During matching, when PCRE reaches a callout point (and \fIpcre_callout\fP is
1945 set), the external function is called. It is provided with the number of the
1946 callout, the position in the pattern, and, optionally, one item of data
1947 originally supplied by the caller of \fBpcre_exec()\fP. The callout function
1948 may cause matching to proceed, to backtrack, or to fail altogether. A complete
1949 description of the interface to the callout function is given in the
1950 .\" HREF
1951 \fBpcrecallout\fP
1952 .\"
1953 documentation.
1954 .
1955 .
1956 .SH "SEE ALSO"
1957 .rs
1958 .sp
1959 \fBpcreapi\fP(3), \fBpcrecallout\fP(3), \fBpcrematching\fP(3), \fBpcre\fP(3).
1960 .
1961 .
1962 .SH AUTHOR
1963 .rs
1964 .sp
1965 .nf
1966 Philip Hazel
1967 University Computing Service
1968 Cambridge CB2 3QH, England.
1969 .fi
1970 .
1971 .
1972 .SH REVISION
1973 .rs
1974 .sp
1975 .nf
1976 Last updated: 13 June 2007
1977 Copyright (c) 1997-2007 University of Cambridge.
1978 .fi

Properties

Name Value
svn:eol-style native
svn:keywords "Author Date Id Revision Url"

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12