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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 93 - (show annotations) (download) (as text)
Sat Feb 24 21:41:42 2007 UTC (7 years, 6 months ago) by nigel
File MIME type: text/html
File size: 77657 byte(s)
Load pcre-7.0 into code/trunk.

1 <html>
2 <head>
3 <title>pcrepattern specification</title>
4 </head>
5 <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
6 <h1>pcrepattern man page</h1>
7 <p>
8 Return to the <a href="index.html">PCRE index page</a>.
9 </p>
10 <p>
11 This page is part of the PCRE HTML documentation. It was generated automatically
12 from the original man page. If there is any nonsense in it, please consult the
13 man page, in case the conversion went wrong.
14 <br>
15 <ul>
16 <li><a name="TOC1" href="#SEC1">PCRE REGULAR EXPRESSION DETAILS</a>
17 <li><a name="TOC2" href="#SEC2">CHARACTERS AND METACHARACTERS</a>
18 <li><a name="TOC3" href="#SEC3">BACKSLASH</a>
19 <li><a name="TOC4" href="#SEC4">CIRCUMFLEX AND DOLLAR</a>
20 <li><a name="TOC5" href="#SEC5">FULL STOP (PERIOD, DOT)</a>
21 <li><a name="TOC6" href="#SEC6">MATCHING A SINGLE BYTE</a>
22 <li><a name="TOC7" href="#SEC7">SQUARE BRACKETS AND CHARACTER CLASSES</a>
23 <li><a name="TOC8" href="#SEC8">POSIX CHARACTER CLASSES</a>
24 <li><a name="TOC9" href="#SEC9">VERTICAL BAR</a>
25 <li><a name="TOC10" href="#SEC10">INTERNAL OPTION SETTING</a>
26 <li><a name="TOC11" href="#SEC11">SUBPATTERNS</a>
27 <li><a name="TOC12" href="#SEC12">NAMED SUBPATTERNS</a>
28 <li><a name="TOC13" href="#SEC13">REPETITION</a>
29 <li><a name="TOC14" href="#SEC14">ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS</a>
30 <li><a name="TOC15" href="#SEC15">BACK REFERENCES</a>
31 <li><a name="TOC16" href="#SEC16">ASSERTIONS</a>
32 <li><a name="TOC17" href="#SEC17">CONDITIONAL SUBPATTERNS</a>
33 <li><a name="TOC18" href="#SEC18">COMMENTS</a>
34 <li><a name="TOC19" href="#SEC19">RECURSIVE PATTERNS</a>
35 <li><a name="TOC20" href="#SEC20">SUBPATTERNS AS SUBROUTINES</a>
36 <li><a name="TOC21" href="#SEC21">CALLOUTS</a>
37 <li><a name="TOC22" href="#SEC22">SEE ALSO</a>
38 </ul>
39 <br><a name="SEC1" href="#TOC1">PCRE REGULAR EXPRESSION DETAILS</a><br>
40 <P>
41 The syntax and semantics of the regular expressions supported by PCRE are
42 described below. Regular expressions are also described in the Perl
43 documentation and in a number of books, some of which have copious examples.
44 Jeffrey Friedl's "Mastering Regular Expressions", published by O'Reilly, covers
45 regular expressions in great detail. This description of PCRE's regular
46 expressions is intended as reference material.
47 </P>
48 <P>
49 The original operation of PCRE was on strings of one-byte characters. However,
50 there is now also support for UTF-8 character strings. To use this, you must
51 build PCRE to include UTF-8 support, and then call <b>pcre_compile()</b> with
52 the PCRE_UTF8 option. How this affects pattern matching is mentioned in several
53 places below. There is also a summary of UTF-8 features in the
54 <a href="pcre.html#utf8support">section on UTF-8 support</a>
55 in the main
56 <a href="pcre.html"><b>pcre</b></a>
57 page.
58 </P>
59 <P>
60 The remainder of this document discusses the patterns that are supported by
61 PCRE when its main matching function, <b>pcre_exec()</b>, is used.
62 From release 6.0, PCRE offers a second matching function,
63 <b>pcre_dfa_exec()</b>, which matches using a different algorithm that is not
64 Perl-compatible. The advantages and disadvantages of the alternative function,
65 and how it differs from the normal function, are discussed in the
66 <a href="pcrematching.html"><b>pcrematching</b></a>
67 page.
68 </P>
69 <br><a name="SEC2" href="#TOC1">CHARACTERS AND METACHARACTERS</a><br>
70 <P>
71 A regular expression is a pattern that is matched against a subject string from
72 left to right. Most characters stand for themselves in a pattern, and match the
73 corresponding characters in the subject. As a trivial example, the pattern
74 <pre>
75 The quick brown fox
76 </pre>
77 matches a portion of a subject string that is identical to itself. When
78 caseless matching is specified (the PCRE_CASELESS option), letters are matched
79 independently of case. In UTF-8 mode, PCRE always understands the concept of
80 case for characters whose values are less than 128, so caseless matching is
81 always possible. For characters with higher values, the concept of case is
82 supported if PCRE is compiled with Unicode property support, but not otherwise.
83 If you want to use caseless matching for characters 128 and above, you must
84 ensure that PCRE is compiled with Unicode property support as well as with
85 UTF-8 support.
86 </P>
87 <P>
88 The power of regular expressions comes from the ability to include alternatives
89 and repetitions in the pattern. These are encoded in the pattern by the use of
90 <i>metacharacters</i>, which do not stand for themselves but instead are
91 interpreted in some special way.
92 </P>
93 <P>
94 There are two different sets of metacharacters: those that are recognized
95 anywhere in the pattern except within square brackets, and those that are
96 recognized within square brackets. Outside square brackets, the metacharacters
97 are as follows:
98 <pre>
99 \ general escape character with several uses
100 ^ assert start of string (or line, in multiline mode)
101 $ assert end of string (or line, in multiline mode)
102 . match any character except newline (by default)
103 [ start character class definition
104 | start of alternative branch
105 ( start subpattern
106 ) end subpattern
107 ? extends the meaning of (
108 also 0 or 1 quantifier
109 also quantifier minimizer
110 * 0 or more quantifier
111 + 1 or more quantifier
112 also "possessive quantifier"
113 { start min/max quantifier
114 </pre>
115 Part of a pattern that is in square brackets is called a "character class". In
116 a character class the only metacharacters are:
117 <pre>
118 \ general escape character
119 ^ negate the class, but only if the first character
120 - indicates character range
121 [ POSIX character class (only if followed by POSIX syntax)
122 ] terminates the character class
123 </pre>
124 The following sections describe the use of each of the metacharacters.
125 </P>
126 <br><a name="SEC3" href="#TOC1">BACKSLASH</a><br>
127 <P>
128 The backslash character has several uses. Firstly, if it is followed by a
129 non-alphanumeric character, it takes away any special meaning that character
130 may have. This use of backslash as an escape character applies both inside and
131 outside character classes.
132 </P>
133 <P>
134 For example, if you want to match a * character, you write \* in the pattern.
135 This escaping action applies whether or not the following character would
136 otherwise be interpreted as a metacharacter, so it is always safe to precede a
137 non-alphanumeric with backslash to specify that it stands for itself. In
138 particular, if you want to match a backslash, you write \\.
139 </P>
140 <P>
141 If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the
142 pattern (other than in a character class) and characters between a # outside
143 a character class and the next newline are ignored. An escaping backslash can
144 be used to include a whitespace or # character as part of the pattern.
145 </P>
146 <P>
147 If you want to remove the special meaning from a sequence of characters, you
148 can do so by putting them between \Q and \E. This is different from Perl in
149 that $ and @ are handled as literals in \Q...\E sequences in PCRE, whereas in
150 Perl, $ and @ cause variable interpolation. Note the following examples:
151 <pre>
152 Pattern PCRE matches Perl matches
153
154 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
155 \Qabc\$xyz\E abc\$xyz abc\$xyz
156 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
157 </pre>
158 The \Q...\E sequence is recognized both inside and outside character classes.
159 <a name="digitsafterbackslash"></a></P>
160 <br><b>
161 Non-printing characters
162 </b><br>
163 <P>
164 A second use of backslash provides a way of encoding non-printing characters
165 in patterns in a visible manner. There is no restriction on the appearance of
166 non-printing characters, apart from the binary zero that terminates a pattern,
167 but when a pattern is being prepared by text editing, it is usually easier to
168 use one of the following escape sequences than the binary character it
169 represents:
170 <pre>
171 \a alarm, that is, the BEL character (hex 07)
172 \cx "control-x", where x is any character
173 \e escape (hex 1B)
174 \f formfeed (hex 0C)
175 \n newline (hex 0A)
176 \r carriage return (hex 0D)
177 \t tab (hex 09)
178 \ddd character with octal code ddd, or backreference
179 \xhh character with hex code hh
180 \x{hhh..} character with hex code hhh..
181 </pre>
182 The precise effect of \cx is as follows: if x is a lower case letter, it
183 is converted to upper case. Then bit 6 of the character (hex 40) is inverted.
184 Thus \cz becomes hex 1A, but \c{ becomes hex 3B, while \c; becomes hex
185 7B.
186 </P>
187 <P>
188 After \x, from zero to two hexadecimal digits are read (letters can be in
189 upper or lower case). Any number of hexadecimal digits may appear between \x{
190 and }, but the value of the character code must be less than 256 in non-UTF-8
191 mode, and less than 2**31 in UTF-8 mode (that is, the maximum hexadecimal value
192 is 7FFFFFFF). If characters other than hexadecimal digits appear between \x{
193 and }, or if there is no terminating }, this form of escape is not recognized.
194 Instead, the initial \x will be interpreted as a basic hexadecimal escape,
195 with no following digits, giving a character whose value is zero.
196 </P>
197 <P>
198 Characters whose value is less than 256 can be defined by either of the two
199 syntaxes for \x. There is no difference in the way they are handled. For
200 example, \xdc is exactly the same as \x{dc}.
201 </P>
202 <P>
203 After \0 up to two further octal digits are read. If there are fewer than two
204 digits, just those that are present are used. Thus the sequence \0\x\07
205 specifies two binary zeros followed by a BEL character (code value 7). Make
206 sure you supply two digits after the initial zero if the pattern character that
207 follows is itself an octal digit.
208 </P>
209 <P>
210 The handling of a backslash followed by a digit other than 0 is complicated.
211 Outside a character class, PCRE reads it and any following digits as a decimal
212 number. If the number is less than 10, or if there have been at least that many
213 previous capturing left parentheses in the expression, the entire sequence is
214 taken as a <i>back reference</i>. A description of how this works is given
215 <a href="#backreferences">later,</a>
216 following the discussion of
217 <a href="#subpattern">parenthesized subpatterns.</a>
218 </P>
219 <P>
220 Inside a character class, or if the decimal number is greater than 9 and there
221 have not been that many capturing subpatterns, PCRE re-reads up to three octal
222 digits following the backslash, and uses them to generate a data character. Any
223 subsequent digits stand for themselves. In non-UTF-8 mode, the value of a
224 character specified in octal must be less than \400. In UTF-8 mode, values up
225 to \777 are permitted. For example:
226 <pre>
227 \040 is another way of writing a space
228 \40 is the same, provided there are fewer than 40 previous capturing subpatterns
229 \7 is always a back reference
230 \11 might be a back reference, or another way of writing a tab
231 \011 is always a tab
232 \0113 is a tab followed by the character "3"
233 \113 might be a back reference, otherwise the character with octal code 113
234 \377 might be a back reference, otherwise the byte consisting entirely of 1 bits
235 \81 is either a back reference, or a binary zero followed by the two characters "8" and "1"
236 </pre>
237 Note that octal values of 100 or greater must not be introduced by a leading
238 zero, because no more than three octal digits are ever read.
239 </P>
240 <P>
241 All the sequences that define a single character value can be used both inside
242 and outside character classes. In addition, inside a character class, the
243 sequence \b is interpreted as the backspace character (hex 08), and the
244 sequences \R and \X are interpreted as the characters "R" and "X",
245 respectively. Outside a character class, these sequences have different
246 meanings
247 <a href="#uniextseq">(see below).</a>
248 </P>
249 <br><b>
250 Absolute and relative back references
251 </b><br>
252 <P>
253 The sequence \g followed by a positive or negative number, optionally enclosed
254 in braces, is an absolute or relative back reference. Back references are
255 discussed
256 <a href="#backreferences">later,</a>
257 following the discussion of
258 <a href="#subpattern">parenthesized subpatterns.</a>
259 </P>
260 <br><b>
261 Generic character types
262 </b><br>
263 <P>
264 Another use of backslash is for specifying generic character types. The
265 following are always recognized:
266 <pre>
267 \d any decimal digit
268 \D any character that is not a decimal digit
269 \s any whitespace character
270 \S any character that is not a whitespace character
271 \w any "word" character
272 \W any "non-word" character
273 </pre>
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 <P>
278 These character type sequences can appear both inside and outside character
279 classes. They each match one character of the appropriate type. If the current
280 matching point is at the end of the subject string, all of them fail, since
281 there is no character to match.
282 </P>
283 <P>
284 For compatibility with Perl, \s does not match the VT character (code 11).
285 This makes it different from the the POSIX "space" class. The \s characters
286 are HT (9), LF (10), FF (12), CR (13), and space (32). (If "use locale;" is
287 included in a Perl script, \s may match the VT character. In PCRE, it never
288 does.)
289 </P>
290 <P>
291 A "word" character is an underscore or any character less than 256 that is a
292 letter or digit. The definition of letters and digits is controlled by PCRE's
293 low-valued character tables, and may vary if locale-specific matching is taking
294 place (see
295 <a href="pcreapi.html#localesupport">"Locale support"</a>
296 in the
297 <a href="pcreapi.html"><b>pcreapi</b></a>
298 page). For example, in the "fr_FR" (French) locale, some character codes
299 greater than 128 are used for accented letters, and these are matched by \w.
300 </P>
301 <P>
302 In UTF-8 mode, characters with values greater than 128 never match \d, \s, or
303 \w, and always match \D, \S, and \W. This is true even when Unicode
304 character property support is available. The use of locales with Unicode is
305 discouraged.
306 </P>
307 <br><b>
308 Newline sequences
309 </b><br>
310 <P>
311 Outside a character class, the escape sequence \R matches any Unicode newline
312 sequence. This is an extension to Perl. In non-UTF-8 mode \R is equivalent to
313 the following:
314 <pre>
315 (?&#62;\r\n|\n|\x0b|\f|\r|\x85)
316 </pre>
317 This is an example of an "atomic group", details of which are given
318 <a href="#atomicgroup">below.</a>
319 This particular group matches either the two-character sequence CR followed by
320 LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab,
321 U+000B), FF (formfeed, U+000C), CR (carriage return, U+000D), or NEL (next
322 line, U+0085). The two-character sequence is treated as a single unit that
323 cannot be split.
324 </P>
325 <P>
326 In UTF-8 mode, two additional characters whose codepoints are greater than 255
327 are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029).
328 Unicode character property support is not needed for these characters to be
329 recognized.
330 </P>
331 <P>
332 Inside a character class, \R matches the letter "R".
333 <a name="uniextseq"></a></P>
334 <br><b>
335 Unicode character properties
336 </b><br>
337 <P>
338 When PCRE is built with Unicode character property support, three additional
339 escape sequences to match character properties are available when UTF-8 mode
340 is selected. They are:
341 <pre>
342 \p{<i>xx</i>} a character with the <i>xx</i> property
343 \P{<i>xx</i>} a character without the <i>xx</i> property
344 \X an extended Unicode sequence
345 </pre>
346 The property names represented by <i>xx</i> above are limited to the Unicode
347 script names, the general category properties, and "Any", which matches any
348 character (including newline). Other properties such as "InMusicalSymbols" are
349 not currently supported by PCRE. Note that \P{Any} does not match any
350 characters, so always causes a match failure.
351 </P>
352 <P>
353 Sets of Unicode characters are defined as belonging to certain scripts. A
354 character from one of these sets can be matched using a script name. For
355 example:
356 <pre>
357 \p{Greek}
358 \P{Han}
359 </pre>
360 Those that are not part of an identified script are lumped together as
361 "Common". The current list of scripts is:
362 </P>
363 <P>
364 Arabic,
365 Armenian,
366 Balinese,
367 Bengali,
368 Bopomofo,
369 Braille,
370 Buginese,
371 Buhid,
372 Canadian_Aboriginal,
373 Cherokee,
374 Common,
375 Coptic,
376 Cuneiform,
377 Cypriot,
378 Cyrillic,
379 Deseret,
380 Devanagari,
381 Ethiopic,
382 Georgian,
383 Glagolitic,
384 Gothic,
385 Greek,
386 Gujarati,
387 Gurmukhi,
388 Han,
389 Hangul,
390 Hanunoo,
391 Hebrew,
392 Hiragana,
393 Inherited,
394 Kannada,
395 Katakana,
396 Kharoshthi,
397 Khmer,
398 Lao,
399 Latin,
400 Limbu,
401 Linear_B,
402 Malayalam,
403 Mongolian,
404 Myanmar,
405 New_Tai_Lue,
406 Nko,
407 Ogham,
408 Old_Italic,
409 Old_Persian,
410 Oriya,
411 Osmanya,
412 Phags_Pa,
413 Phoenician,
414 Runic,
415 Shavian,
416 Sinhala,
417 Syloti_Nagri,
418 Syriac,
419 Tagalog,
420 Tagbanwa,
421 Tai_Le,
422 Tamil,
423 Telugu,
424 Thaana,
425 Thai,
426 Tibetan,
427 Tifinagh,
428 Ugaritic,
429 Yi.
430 </P>
431 <P>
432 Each character has exactly one general category property, specified by a
433 two-letter abbreviation. For compatibility with Perl, negation can be specified
434 by including a circumflex between the opening brace and the property name. For
435 example, \p{^Lu} is the same as \P{Lu}.
436 </P>
437 <P>
438 If only one letter is specified with \p or \P, it includes all the general
439 category properties that start with that letter. In this case, in the absence
440 of negation, the curly brackets in the escape sequence are optional; these two
441 examples have the same effect:
442 <pre>
443 \p{L}
444 \pL
445 </pre>
446 The following general category property codes are supported:
447 <pre>
448 C Other
449 Cc Control
450 Cf Format
451 Cn Unassigned
452 Co Private use
453 Cs Surrogate
454
455 L Letter
456 Ll Lower case letter
457 Lm Modifier letter
458 Lo Other letter
459 Lt Title case letter
460 Lu Upper case letter
461
462 M Mark
463 Mc Spacing mark
464 Me Enclosing mark
465 Mn Non-spacing mark
466
467 N Number
468 Nd Decimal number
469 Nl Letter number
470 No Other number
471
472 P Punctuation
473 Pc Connector punctuation
474 Pd Dash punctuation
475 Pe Close punctuation
476 Pf Final punctuation
477 Pi Initial punctuation
478 Po Other punctuation
479 Ps Open punctuation
480
481 S Symbol
482 Sc Currency symbol
483 Sk Modifier symbol
484 Sm Mathematical symbol
485 So Other symbol
486
487 Z Separator
488 Zl Line separator
489 Zp Paragraph separator
490 Zs Space separator
491 </pre>
492 The special property L& is also supported: it matches a character that has
493 the Lu, Ll, or Lt property, in other words, a letter that is not classified as
494 a modifier or "other".
495 </P>
496 <P>
497 The long synonyms for these properties that Perl supports (such as \p{Letter})
498 are not supported by PCRE, nor is it permitted to prefix any of these
499 properties with "Is".
500 </P>
501 <P>
502 No character that is in the Unicode table has the Cn (unassigned) property.
503 Instead, this property is assumed for any code point that is not in the
504 Unicode table.
505 </P>
506 <P>
507 Specifying caseless matching does not affect these escape sequences. For
508 example, \p{Lu} always matches only upper case letters.
509 </P>
510 <P>
511 The \X escape matches any number of Unicode characters that form an extended
512 Unicode sequence. \X is equivalent to
513 <pre>
514 (?&#62;\PM\pM*)
515 </pre>
516 That is, it matches a character without the "mark" property, followed by zero
517 or more characters with the "mark" property, and treats the sequence as an
518 atomic group
519 <a href="#atomicgroup">(see below).</a>
520 Characters with the "mark" property are typically accents that affect the
521 preceding character.
522 </P>
523 <P>
524 Matching characters by Unicode property is not fast, because PCRE has to search
525 a structure that contains data for over fifteen thousand characters. That is
526 why the traditional escape sequences such as \d and \w do not use Unicode
527 properties in PCRE.
528 <a name="smallassertions"></a></P>
529 <br><b>
530 Simple assertions
531 </b><br>
532 <P>
533 The final use of backslash is for certain simple assertions. An assertion
534 specifies a condition that has to be met at a particular point in a match,
535 without consuming any characters from the subject string. The use of
536 subpatterns for more complicated assertions is described
537 <a href="#bigassertions">below.</a>
538 The backslashed assertions are:
539 <pre>
540 \b matches at a word boundary
541 \B matches when not at a word boundary
542 \A matches at the start of the subject
543 \Z matches at the end of the subject
544 also matches before a newline at the end of the subject
545 \z matches only at the end of the subject
546 \G matches at the first matching position in the subject
547 </pre>
548 These assertions may not appear in character classes (but note that \b has a
549 different meaning, namely the backspace character, inside a character class).
550 </P>
551 <P>
552 A word boundary is a position in the subject string where the current character
553 and the previous character do not both match \w or \W (i.e. one matches
554 \w and the other matches \W), or the start or end of the string if the
555 first or last character matches \w, respectively.
556 </P>
557 <P>
558 The \A, \Z, and \z assertions differ from the traditional circumflex and
559 dollar (described in the next section) in that they only ever match at the very
560 start and end of the subject string, whatever options are set. Thus, they are
561 independent of multiline mode. These three assertions are not affected by the
562 PCRE_NOTBOL or PCRE_NOTEOL options, which affect only the behaviour of the
563 circumflex and dollar metacharacters. However, if the <i>startoffset</i>
564 argument of <b>pcre_exec()</b> is non-zero, indicating that matching is to start
565 at a point other than the beginning of the subject, \A can never match. The
566 difference between \Z and \z is that \Z matches before a newline at the end
567 of the string as well as at the very end, whereas \z matches only at the end.
568 </P>
569 <P>
570 The \G assertion is true only when the current matching position is at the
571 start point of the match, as specified by the <i>startoffset</i> argument of
572 <b>pcre_exec()</b>. It differs from \A when the value of <i>startoffset</i> is
573 non-zero. By calling <b>pcre_exec()</b> multiple times with appropriate
574 arguments, you can mimic Perl's /g option, and it is in this kind of
575 implementation where \G can be useful.
576 </P>
577 <P>
578 Note, however, that PCRE's interpretation of \G, as the start of the current
579 match, is subtly different from Perl's, which defines it as the end of the
580 previous match. In Perl, these can be different when the previously matched
581 string was empty. Because PCRE does just one match at a time, it cannot
582 reproduce this behaviour.
583 </P>
584 <P>
585 If all the alternatives of a pattern begin with \G, the expression is anchored
586 to the starting match position, and the "anchored" flag is set in the compiled
587 regular expression.
588 </P>
589 <br><a name="SEC4" href="#TOC1">CIRCUMFLEX AND DOLLAR</a><br>
590 <P>
591 Outside a character class, in the default matching mode, the circumflex
592 character is an assertion that is true only if the current matching point is
593 at the start of the subject string. If the <i>startoffset</i> argument of
594 <b>pcre_exec()</b> is non-zero, circumflex can never match if the PCRE_MULTILINE
595 option is unset. Inside a character class, circumflex has an entirely different
596 meaning
597 <a href="#characterclass">(see below).</a>
598 </P>
599 <P>
600 Circumflex need not be the first character of the pattern if a number of
601 alternatives are involved, but it should be the first thing in each alternative
602 in which it appears if the pattern is ever to match that branch. If all
603 possible alternatives start with a circumflex, that is, if the pattern is
604 constrained to match only at the start of the subject, it is said to be an
605 "anchored" pattern. (There are also other constructs that can cause a pattern
606 to be anchored.)
607 </P>
608 <P>
609 A dollar character is an assertion that is true only if the current matching
610 point is at the end of the subject string, or immediately before a newline
611 at the end of the string (by default). Dollar need not be the last character of
612 the pattern if a number of alternatives are involved, but it should be the last
613 item in any branch in which it appears. Dollar has no special meaning in a
614 character class.
615 </P>
616 <P>
617 The meaning of dollar can be changed so that it matches only at the very end of
618 the string, by setting the PCRE_DOLLAR_ENDONLY option at compile time. This
619 does not affect the \Z assertion.
620 </P>
621 <P>
622 The meanings of the circumflex and dollar characters are changed if the
623 PCRE_MULTILINE option is set. When this is the case, a circumflex matches
624 immediately after internal newlines as well as at the start of the subject
625 string. It does not match after a newline that ends the string. A dollar
626 matches before any newlines in the string, as well as at the very end, when
627 PCRE_MULTILINE is set. When newline is specified as the two-character
628 sequence CRLF, isolated CR and LF characters do not indicate newlines.
629 </P>
630 <P>
631 For example, the pattern /^abc$/ matches the subject string "def\nabc" (where
632 \n represents a newline) in multiline mode, but not otherwise. Consequently,
633 patterns that are anchored in single line mode because all branches start with
634 ^ are not anchored in multiline mode, and a match for circumflex is possible
635 when the <i>startoffset</i> argument of <b>pcre_exec()</b> is non-zero. The
636 PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
637 </P>
638 <P>
639 Note that the sequences \A, \Z, and \z can be used to match the start and
640 end of the subject in both modes, and if all branches of a pattern start with
641 \A it is always anchored, whether or not PCRE_MULTILINE is set.
642 </P>
643 <br><a name="SEC5" href="#TOC1">FULL STOP (PERIOD, DOT)</a><br>
644 <P>
645 Outside a character class, a dot in the pattern matches any one character in
646 the subject string except (by default) a character that signifies the end of a
647 line. In UTF-8 mode, the matched character may be more than one byte long.
648 </P>
649 <P>
650 When a line ending is defined as a single character, dot never matches that
651 character; when the two-character sequence CRLF is used, dot does not match CR
652 if it is immediately followed by LF, but otherwise it matches all characters
653 (including isolated CRs and LFs). When any Unicode line endings are being
654 recognized, dot does not match CR or LF or any of the other line ending
655 characters.
656 </P>
657 <P>
658 The behaviour of dot with regard to newlines can be changed. If the PCRE_DOTALL
659 option is set, a dot matches any one character, without exception. If the
660 two-character sequence CRLF is present in the subject string, it takes two dots
661 to match it.
662 </P>
663 <P>
664 The handling of dot is entirely independent of the handling of circumflex and
665 dollar, the only relationship being that they both involve newlines. Dot has no
666 special meaning in a character class.
667 </P>
668 <br><a name="SEC6" href="#TOC1">MATCHING A SINGLE BYTE</a><br>
669 <P>
670 Outside a character class, the escape sequence \C matches any one byte, both
671 in and out of UTF-8 mode. Unlike a dot, it always matches any line-ending
672 characters. The feature is provided in Perl in order to match individual bytes
673 in UTF-8 mode. Because it breaks up UTF-8 characters into individual bytes,
674 what remains in the string may be a malformed UTF-8 string. For this reason,
675 the \C escape sequence is best avoided.
676 </P>
677 <P>
678 PCRE does not allow \C to appear in lookbehind assertions
679 <a href="#lookbehind">(described below),</a>
680 because in UTF-8 mode this would make it impossible to calculate the length of
681 the lookbehind.
682 <a name="characterclass"></a></P>
683 <br><a name="SEC7" href="#TOC1">SQUARE BRACKETS AND CHARACTER CLASSES</a><br>
684 <P>
685 An opening square bracket introduces a character class, terminated by a closing
686 square bracket. A closing square bracket on its own is not special. If a
687 closing square bracket is required as a member of the class, it should be the
688 first data character in the class (after an initial circumflex, if present) or
689 escaped with a backslash.
690 </P>
691 <P>
692 A character class matches a single character in the subject. In UTF-8 mode, the
693 character may occupy more than one byte. A matched character must be in the set
694 of characters defined by the class, unless the first character in the class
695 definition is a circumflex, in which case the subject character must not be in
696 the set defined by the class. If a circumflex is actually required as a member
697 of the class, ensure it is not the first character, or escape it with a
698 backslash.
699 </P>
700 <P>
701 For example, the character class [aeiou] matches any lower case vowel, while
702 [^aeiou] matches any character that is not a lower case vowel. Note that a
703 circumflex is just a convenient notation for specifying the characters that
704 are in the class by enumerating those that are not. A class that starts with a
705 circumflex is not an assertion: it still consumes a character from the subject
706 string, and therefore it fails if the current pointer is at the end of the
707 string.
708 </P>
709 <P>
710 In UTF-8 mode, characters with values greater than 255 can be included in a
711 class as a literal string of bytes, or by using the \x{ escaping mechanism.
712 </P>
713 <P>
714 When caseless matching is set, any letters in a class represent both their
715 upper case and lower case versions, so for example, a caseless [aeiou] matches
716 "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
717 caseful version would. In UTF-8 mode, PCRE always understands the concept of
718 case for characters whose values are less than 128, so caseless matching is
719 always possible. For characters with higher values, the concept of case is
720 supported if PCRE is compiled with Unicode property support, but not otherwise.
721 If you want to use caseless matching for characters 128 and above, you must
722 ensure that PCRE is compiled with Unicode property support as well as with
723 UTF-8 support.
724 </P>
725 <P>
726 Characters that might indicate line breaks are never treated in any special way
727 when matching character classes, whatever line-ending sequence is in use, and
728 whatever setting of the PCRE_DOTALL and PCRE_MULTILINE options is used. A class
729 such as [^a] always matches one of these characters.
730 </P>
731 <P>
732 The minus (hyphen) character can be used to specify a range of characters in a
733 character class. For example, [d-m] matches any letter between d and m,
734 inclusive. If a minus character is required in a class, it must be escaped with
735 a backslash or appear in a position where it cannot be interpreted as
736 indicating a range, typically as the first or last character in the class.
737 </P>
738 <P>
739 It is not possible to have the literal character "]" as the end character of a
740 range. A pattern such as [W-]46] is interpreted as a class of two characters
741 ("W" and "-") followed by a literal string "46]", so it would match "W46]" or
742 "-46]". However, if the "]" is escaped with a backslash it is interpreted as
743 the end of range, so [W-\]46] is interpreted as a class containing a range
744 followed by two other characters. The octal or hexadecimal representation of
745 "]" can also be used to end a range.
746 </P>
747 <P>
748 Ranges operate in the collating sequence of character values. They can also be
749 used for characters specified numerically, for example [\000-\037]. In UTF-8
750 mode, ranges can include characters whose values are greater than 255, for
751 example [\x{100}-\x{2ff}].
752 </P>
753 <P>
754 If a range that includes letters is used when caseless matching is set, it
755 matches the letters in either case. For example, [W-c] is equivalent to
756 [][\\^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if character
757 tables for the "fr_FR" locale are in use, [\xc8-\xcb] matches accented E
758 characters in both cases. In UTF-8 mode, PCRE supports the concept of case for
759 characters with values greater than 128 only when it is compiled with Unicode
760 property support.
761 </P>
762 <P>
763 The character types \d, \D, \p, \P, \s, \S, \w, and \W may also appear
764 in a character class, and add the characters that they match to the class. For
765 example, [\dABCDEF] matches any hexadecimal digit. A circumflex can
766 conveniently be used with the upper case character types to specify a more
767 restricted set of characters than the matching lower case type. For example,
768 the class [^\W_] matches any letter or digit, but not underscore.
769 </P>
770 <P>
771 The only metacharacters that are recognized in character classes are backslash,
772 hyphen (only where it can be interpreted as specifying a range), circumflex
773 (only at the start), opening square bracket (only when it can be interpreted as
774 introducing a POSIX class name - see the next section), and the terminating
775 closing square bracket. However, escaping other non-alphanumeric characters
776 does no harm.
777 </P>
778 <br><a name="SEC8" href="#TOC1">POSIX CHARACTER CLASSES</a><br>
779 <P>
780 Perl supports the POSIX notation for character classes. This uses names
781 enclosed by [: and :] within the enclosing square brackets. PCRE also supports
782 this notation. For example,
783 <pre>
784 [01[:alpha:]%]
785 </pre>
786 matches "0", "1", any alphabetic character, or "%". The supported class names
787 are
788 <pre>
789 alnum letters and digits
790 alpha letters
791 ascii character codes 0 - 127
792 blank space or tab only
793 cntrl control characters
794 digit decimal digits (same as \d)
795 graph printing characters, excluding space
796 lower lower case letters
797 print printing characters, including space
798 punct printing characters, excluding letters and digits
799 space white space (not quite the same as \s)
800 upper upper case letters
801 word "word" characters (same as \w)
802 xdigit hexadecimal digits
803 </pre>
804 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13), and
805 space (32). Notice that this list includes the VT character (code 11). This
806 makes "space" different to \s, which does not include VT (for Perl
807 compatibility).
808 </P>
809 <P>
810 The name "word" is a Perl extension, and "blank" is a GNU extension from Perl
811 5.8. Another Perl extension is negation, which is indicated by a ^ character
812 after the colon. For example,
813 <pre>
814 [12[:^digit:]]
815 </pre>
816 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX
817 syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
818 supported, and an error is given if they are encountered.
819 </P>
820 <P>
821 In UTF-8 mode, characters with values greater than 128 do not match any of
822 the POSIX character classes.
823 </P>
824 <br><a name="SEC9" href="#TOC1">VERTICAL BAR</a><br>
825 <P>
826 Vertical bar characters are used to separate alternative patterns. For example,
827 the pattern
828 <pre>
829 gilbert|sullivan
830 </pre>
831 matches either "gilbert" or "sullivan". Any number of alternatives may appear,
832 and an empty alternative is permitted (matching the empty string). The matching
833 process tries each alternative in turn, from left to right, and the first one
834 that succeeds is used. If the alternatives are within a subpattern
835 <a href="#subpattern">(defined below),</a>
836 "succeeds" means matching the rest of the main pattern as well as the
837 alternative in the subpattern.
838 </P>
839 <br><a name="SEC10" href="#TOC1">INTERNAL OPTION SETTING</a><br>
840 <P>
841 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
842 PCRE_EXTENDED options can be changed from within the pattern by a sequence of
843 Perl option letters enclosed between "(?" and ")". The option letters are
844 <pre>
845 i for PCRE_CASELESS
846 m for PCRE_MULTILINE
847 s for PCRE_DOTALL
848 x for PCRE_EXTENDED
849 </pre>
850 For example, (?im) sets caseless, multiline matching. It is also possible to
851 unset these options by preceding the letter with a hyphen, and a combined
852 setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
853 PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
854 permitted. If a letter appears both before and after the hyphen, the option is
855 unset.
856 </P>
857 <P>
858 When an option change occurs at top level (that is, not inside subpattern
859 parentheses), the change applies to the remainder of the pattern that follows.
860 If the change is placed right at the start of a pattern, PCRE extracts it into
861 the global options (and it will therefore show up in data extracted by the
862 <b>pcre_fullinfo()</b> function).
863 </P>
864 <P>
865 An option change within a subpattern (see below for a description of
866 subpatterns) affects only that part of the current pattern that follows it, so
867 <pre>
868 (a(?i)b)c
869 </pre>
870 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
871 By this means, options can be made to have different settings in different
872 parts of the pattern. Any changes made in one alternative do carry on
873 into subsequent branches within the same subpattern. For example,
874 <pre>
875 (a(?i)b|c)
876 </pre>
877 matches "ab", "aB", "c", and "C", even though when matching "C" the first
878 branch is abandoned before the option setting. This is because the effects of
879 option settings happen at compile time. There would be some very weird
880 behaviour otherwise.
881 </P>
882 <P>
883 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA can be
884 changed in the same way as the Perl-compatible options by using the characters
885 J, U and X respectively.
886 <a name="subpattern"></a></P>
887 <br><a name="SEC11" href="#TOC1">SUBPATTERNS</a><br>
888 <P>
889 Subpatterns are delimited by parentheses (round brackets), which can be nested.
890 Turning part of a pattern into a subpattern does two things:
891 <br>
892 <br>
893 1. It localizes a set of alternatives. For example, the pattern
894 <pre>
895 cat(aract|erpillar|)
896 </pre>
897 matches one of the words "cat", "cataract", or "caterpillar". Without the
898 parentheses, it would match "cataract", "erpillar" or an empty string.
899 <br>
900 <br>
901 2. It sets up the subpattern as a capturing subpattern. This means that, when
902 the whole pattern matches, that portion of the subject string that matched the
903 subpattern is passed back to the caller via the <i>ovector</i> argument of
904 <b>pcre_exec()</b>. Opening parentheses are counted from left to right (starting
905 from 1) to obtain numbers for the capturing subpatterns.
906 </P>
907 <P>
908 For example, if the string "the red king" is matched against the pattern
909 <pre>
910 the ((red|white) (king|queen))
911 </pre>
912 the captured substrings are "red king", "red", and "king", and are numbered 1,
913 2, and 3, respectively.
914 </P>
915 <P>
916 The fact that plain parentheses fulfil two functions is not always helpful.
917 There are often times when a grouping subpattern is required without a
918 capturing requirement. If an opening parenthesis is followed by a question mark
919 and a colon, the subpattern does not do any capturing, and is not counted when
920 computing the number of any subsequent capturing subpatterns. For example, if
921 the string "the white queen" is matched against the pattern
922 <pre>
923 the ((?:red|white) (king|queen))
924 </pre>
925 the captured substrings are "white queen" and "queen", and are numbered 1 and
926 2. The maximum number of capturing subpatterns is 65535.
927 </P>
928 <P>
929 As a convenient shorthand, if any option settings are required at the start of
930 a non-capturing subpattern, the option letters may appear between the "?" and
931 the ":". Thus the two patterns
932 <pre>
933 (?i:saturday|sunday)
934 (?:(?i)saturday|sunday)
935 </pre>
936 match exactly the same set of strings. Because alternative branches are tried
937 from left to right, and options are not reset until the end of the subpattern
938 is reached, an option setting in one branch does affect subsequent branches, so
939 the above patterns match "SUNDAY" as well as "Saturday".
940 </P>
941 <br><a name="SEC12" href="#TOC1">NAMED SUBPATTERNS</a><br>
942 <P>
943 Identifying capturing parentheses by number is simple, but it can be very hard
944 to keep track of the numbers in complicated regular expressions. Furthermore,
945 if an expression is modified, the numbers may change. To help with this
946 difficulty, PCRE supports the naming of subpatterns. This feature was not
947 added to Perl until release 5.10. Python had the feature earlier, and PCRE
948 introduced it at release 4.0, using the Python syntax. PCRE now supports both
949 the Perl and the Python syntax.
950 </P>
951 <P>
952 In PCRE, a subpattern can be named in one of three ways: (?&#60;name&#62;...) or
953 (?'name'...) as in Perl, or (?P&#60;name&#62;...) as in Python. References to capturing
954 parentheses from other parts of the pattern, such as
955 <a href="#backreferences">backreferences,</a>
956 <a href="#recursion">recursion,</a>
957 and
958 <a href="#conditions">conditions,</a>
959 can be made by name as well as by number.
960 </P>
961 <P>
962 Names consist of up to 32 alphanumeric characters and underscores. Named
963 capturing parentheses are still allocated numbers as well as names, exactly as
964 if the names were not present. The PCRE API provides function calls for
965 extracting the name-to-number translation table from a compiled pattern. There
966 is also a convenience function for extracting a captured substring by name.
967 </P>
968 <P>
969 By default, a name must be unique within a pattern, but it is possible to relax
970 this constraint by setting the PCRE_DUPNAMES option at compile time. This can
971 be useful for patterns where only one instance of the named parentheses can
972 match. Suppose you want to match the name of a weekday, either as a 3-letter
973 abbreviation or as the full name, and in both cases you want to extract the
974 abbreviation. This pattern (ignoring the line breaks) does the job:
975 <pre>
976 (?&#60;DN&#62;Mon|Fri|Sun)(?:day)?|
977 (?&#60;DN&#62;Tue)(?:sday)?|
978 (?&#60;DN&#62;Wed)(?:nesday)?|
979 (?&#60;DN&#62;Thu)(?:rsday)?|
980 (?&#60;DN&#62;Sat)(?:urday)?
981 </pre>
982 There are five capturing substrings, but only one is ever set after a match.
983 The convenience function for extracting the data by name returns the substring
984 for the first (and in this example, the only) subpattern of that name that
985 matched. This saves searching to find which numbered subpattern it was. If you
986 make a reference to a non-unique named subpattern from elsewhere in the
987 pattern, the one that corresponds to the lowest number is used. For further
988 details of the interfaces for handling named subpatterns, see the
989 <a href="pcreapi.html"><b>pcreapi</b></a>
990 documentation.
991 </P>
992 <br><a name="SEC13" href="#TOC1">REPETITION</a><br>
993 <P>
994 Repetition is specified by quantifiers, which can follow any of the following
995 items:
996 <pre>
997 a literal data character
998 the dot metacharacter
999 the \C escape sequence
1000 the \X escape sequence (in UTF-8 mode with Unicode properties)
1001 the \R escape sequence
1002 an escape such as \d that matches a single character
1003 a character class
1004 a back reference (see next section)
1005 a parenthesized subpattern (unless it is an assertion)
1006 </pre>
1007 The general repetition quantifier specifies a minimum and maximum number of
1008 permitted matches, by giving the two numbers in curly brackets (braces),
1009 separated by a comma. The numbers must be less than 65536, and the first must
1010 be less than or equal to the second. For example:
1011 <pre>
1012 z{2,4}
1013 </pre>
1014 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
1015 character. If the second number is omitted, but the comma is present, there is
1016 no upper limit; if the second number and the comma are both omitted, the
1017 quantifier specifies an exact number of required matches. Thus
1018 <pre>
1019 [aeiou]{3,}
1020 </pre>
1021 matches at least 3 successive vowels, but may match many more, while
1022 <pre>
1023 \d{8}
1024 </pre>
1025 matches exactly 8 digits. An opening curly bracket that appears in a position
1026 where a quantifier is not allowed, or one that does not match the syntax of a
1027 quantifier, is taken as a literal character. For example, {,6} is not a
1028 quantifier, but a literal string of four characters.
1029 </P>
1030 <P>
1031 In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to individual
1032 bytes. Thus, for example, \x{100}{2} matches two UTF-8 characters, each of
1033 which is represented by a two-byte sequence. Similarly, when Unicode property
1034 support is available, \X{3} matches three Unicode extended sequences, each of
1035 which may be several bytes long (and they may be of different lengths).
1036 </P>
1037 <P>
1038 The quantifier {0} is permitted, causing the expression to behave as if the
1039 previous item and the quantifier were not present.
1040 </P>
1041 <P>
1042 For convenience, the three most common quantifiers have single-character
1043 abbreviations:
1044 <pre>
1045 * is equivalent to {0,}
1046 + is equivalent to {1,}
1047 ? is equivalent to {0,1}
1048 </pre>
1049 It is possible to construct infinite loops by following a subpattern that can
1050 match no characters with a quantifier that has no upper limit, for example:
1051 <pre>
1052 (a?)*
1053 </pre>
1054 Earlier versions of Perl and PCRE used to give an error at compile time for
1055 such patterns. However, because there are cases where this can be useful, such
1056 patterns are now accepted, but if any repetition of the subpattern does in fact
1057 match no characters, the loop is forcibly broken.
1058 </P>
1059 <P>
1060 By default, the quantifiers are "greedy", that is, they match as much as
1061 possible (up to the maximum number of permitted times), without causing the
1062 rest of the pattern to fail. The classic example of where this gives problems
1063 is in trying to match comments in C programs. These appear between /* and */
1064 and within the comment, individual * and / characters may appear. An attempt to
1065 match C comments by applying the pattern
1066 <pre>
1067 /\*.*\*/
1068 </pre>
1069 to the string
1070 <pre>
1071 /* first comment */ not comment /* second comment */
1072 </pre>
1073 fails, because it matches the entire string owing to the greediness of the .*
1074 item.
1075 </P>
1076 <P>
1077 However, if a quantifier is followed by a question mark, it ceases to be
1078 greedy, and instead matches the minimum number of times possible, so the
1079 pattern
1080 <pre>
1081 /\*.*?\*/
1082 </pre>
1083 does the right thing with the C comments. The meaning of the various
1084 quantifiers is not otherwise changed, just the preferred number of matches.
1085 Do not confuse this use of question mark with its use as a quantifier in its
1086 own right. Because it has two uses, it can sometimes appear doubled, as in
1087 <pre>
1088 \d??\d
1089 </pre>
1090 which matches one digit by preference, but can match two if that is the only
1091 way the rest of the pattern matches.
1092 </P>
1093 <P>
1094 If the PCRE_UNGREEDY option is set (an option that is not available in Perl),
1095 the quantifiers are not greedy by default, but individual ones can be made
1096 greedy by following them with a question mark. In other words, it inverts the
1097 default behaviour.
1098 </P>
1099 <P>
1100 When a parenthesized subpattern is quantified with a minimum repeat count that
1101 is greater than 1 or with a limited maximum, more memory is required for the
1102 compiled pattern, in proportion to the size of the minimum or maximum.
1103 </P>
1104 <P>
1105 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent
1106 to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is
1107 implicitly anchored, because whatever follows will be tried against every
1108 character position in the subject string, so there is no point in retrying the
1109 overall match at any position after the first. PCRE normally treats such a
1110 pattern as though it were preceded by \A.
1111 </P>
1112 <P>
1113 In cases where it is known that the subject string contains no newlines, it is
1114 worth setting PCRE_DOTALL in order to obtain this optimization, or
1115 alternatively using ^ to indicate anchoring explicitly.
1116 </P>
1117 <P>
1118 However, there is one situation where the optimization cannot be used. When .*
1119 is inside capturing parentheses that are the subject of a backreference
1120 elsewhere in the pattern, a match at the start may fail where a later one
1121 succeeds. Consider, for example:
1122 <pre>
1123 (.*)abc\1
1124 </pre>
1125 If the subject is "xyz123abc123" the match point is the fourth character. For
1126 this reason, such a pattern is not implicitly anchored.
1127 </P>
1128 <P>
1129 When a capturing subpattern is repeated, the value captured is the substring
1130 that matched the final iteration. For example, after
1131 <pre>
1132 (tweedle[dume]{3}\s*)+
1133 </pre>
1134 has matched "tweedledum tweedledee" the value of the captured substring is
1135 "tweedledee". However, if there are nested capturing subpatterns, the
1136 corresponding captured values may have been set in previous iterations. For
1137 example, after
1138 <pre>
1139 /(a|(b))+/
1140 </pre>
1141 matches "aba" the value of the second captured substring is "b".
1142 <a name="atomicgroup"></a></P>
1143 <br><a name="SEC14" href="#TOC1">ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS</a><br>
1144 <P>
1145 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
1146 repetition, failure of what follows normally causes the repeated item to be
1147 re-evaluated to see if a different number of repeats allows the rest of the
1148 pattern to match. Sometimes it is useful to prevent this, either to change the
1149 nature of the match, or to cause it fail earlier than it otherwise might, when
1150 the author of the pattern knows there is no point in carrying on.
1151 </P>
1152 <P>
1153 Consider, for example, the pattern \d+foo when applied to the subject line
1154 <pre>
1155 123456bar
1156 </pre>
1157 After matching all 6 digits and then failing to match "foo", the normal
1158 action of the matcher is to try again with only 5 digits matching the \d+
1159 item, and then with 4, and so on, before ultimately failing. "Atomic grouping"
1160 (a term taken from Jeffrey Friedl's book) provides the means for specifying
1161 that once a subpattern has matched, it is not to be re-evaluated in this way.
1162 </P>
1163 <P>
1164 If we use atomic grouping for the previous example, the matcher gives up
1165 immediately on failing to match "foo" the first time. The notation is a kind of
1166 special parenthesis, starting with (?&#62; as in this example:
1167 <pre>
1168 (?&#62;\d+)foo
1169 </pre>
1170 This kind of parenthesis "locks up" the part of the pattern it contains once
1171 it has matched, and a failure further into the pattern is prevented from
1172 backtracking into it. Backtracking past it to previous items, however, works as
1173 normal.
1174 </P>
1175 <P>
1176 An alternative description is that a subpattern of this type matches the string
1177 of characters that an identical standalone pattern would match, if anchored at
1178 the current point in the subject string.
1179 </P>
1180 <P>
1181 Atomic grouping subpatterns are not capturing subpatterns. Simple cases such as
1182 the above example can be thought of as a maximizing repeat that must swallow
1183 everything it can. So, while both \d+ and \d+? are prepared to adjust the
1184 number of digits they match in order to make the rest of the pattern match,
1185 (?&#62;\d+) can only match an entire sequence of digits.
1186 </P>
1187 <P>
1188 Atomic groups in general can of course contain arbitrarily complicated
1189 subpatterns, and can be nested. However, when the subpattern for an atomic
1190 group is just a single repeated item, as in the example above, a simpler
1191 notation, called a "possessive quantifier" can be used. This consists of an
1192 additional + character following a quantifier. Using this notation, the
1193 previous example can be rewritten as
1194 <pre>
1195 \d++foo
1196 </pre>
1197 Possessive quantifiers are always greedy; the setting of the PCRE_UNGREEDY
1198 option is ignored. They are a convenient notation for the simpler forms of
1199 atomic group. However, there is no difference in the meaning of a possessive
1200 quantifier and the equivalent atomic group, though there may be a performance
1201 difference; possessive quantifiers should be slightly faster.
1202 </P>
1203 <P>
1204 The possessive quantifier syntax is an extension to the Perl 5.8 syntax.
1205 Jeffrey Friedl originated the idea (and the name) in the first edition of his
1206 book. Mike McCloskey liked it, so implemented it when he built Sun's Java
1207 package, and PCRE copied it from there. It ultimately found its way into Perl
1208 at release 5.10.
1209 </P>
1210 <P>
1211 PCRE has an optimization that automatically "possessifies" certain simple
1212 pattern constructs. For example, the sequence A+B is treated as A++B because
1213 there is no point in backtracking into a sequence of A's when B must follow.
1214 </P>
1215 <P>
1216 When a pattern contains an unlimited repeat inside a subpattern that can itself
1217 be repeated an unlimited number of times, the use of an atomic group is the
1218 only way to avoid some failing matches taking a very long time indeed. The
1219 pattern
1220 <pre>
1221 (\D+|&#60;\d+&#62;)*[!?]
1222 </pre>
1223 matches an unlimited number of substrings that either consist of non-digits, or
1224 digits enclosed in &#60;&#62;, followed by either ! or ?. When it matches, it runs
1225 quickly. However, if it is applied to
1226 <pre>
1227 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
1228 </pre>
1229 it takes a long time before reporting failure. This is because the string can
1230 be divided between the internal \D+ repeat and the external * repeat in a
1231 large number of ways, and all have to be tried. (The example uses [!?] rather
1232 than a single character at the end, because both PCRE and Perl have an
1233 optimization that allows for fast failure when a single character is used. They
1234 remember the last single character that is required for a match, and fail early
1235 if it is not present in the string.) If the pattern is changed so that it uses
1236 an atomic group, like this:
1237 <pre>
1238 ((?&#62;\D+)|&#60;\d+&#62;)*[!?]
1239 </pre>
1240 sequences of non-digits cannot be broken, and failure happens quickly.
1241 <a name="backreferences"></a></P>
1242 <br><a name="SEC15" href="#TOC1">BACK REFERENCES</a><br>
1243 <P>
1244 Outside a character class, a backslash followed by a digit greater than 0 (and
1245 possibly further digits) is a back reference to a capturing subpattern earlier
1246 (that is, to its left) in the pattern, provided there have been that many
1247 previous capturing left parentheses.
1248 </P>
1249 <P>
1250 However, if the decimal number following the backslash is less than 10, it is
1251 always taken as a back reference, and causes an error only if there are not
1252 that many capturing left parentheses in the entire pattern. In other words, the
1253 parentheses that are referenced need not be to the left of the reference for
1254 numbers less than 10. A "forward back reference" of this type can make sense
1255 when a repetition is involved and the subpattern to the right has participated
1256 in an earlier iteration.
1257 </P>
1258 <P>
1259 It is not possible to have a numerical "forward back reference" to a subpattern
1260 whose number is 10 or more using this syntax because a sequence such as \50 is
1261 interpreted as a character defined in octal. See the subsection entitled
1262 "Non-printing characters"
1263 <a href="#digitsafterbackslash">above</a>
1264 for further details of the handling of digits following a backslash. There is
1265 no such problem when named parentheses are used. A back reference to any
1266 subpattern is possible using named parentheses (see below).
1267 </P>
1268 <P>
1269 Another way of avoiding the ambiguity inherent in the use of digits following a
1270 backslash is to use the \g escape sequence, which is a feature introduced in
1271 Perl 5.10. This escape must be followed by a positive or a negative number,
1272 optionally enclosed in braces. These examples are all identical:
1273 <pre>
1274 (ring), \1
1275 (ring), \g1
1276 (ring), \g{1}
1277 </pre>
1278 A positive number specifies an absolute reference without the ambiguity that is
1279 present in the older syntax. It is also useful when literal digits follow the
1280 reference. A negative number is a relative reference. Consider this example:
1281 <pre>
1282 (abc(def)ghi)\g{-1}
1283 </pre>
1284 The sequence \g{-1} is a reference to the most recently started capturing
1285 subpattern before \g, that is, is it equivalent to \2. Similarly, \g{-2}
1286 would be equivalent to \1. The use of relative references can be helpful in
1287 long patterns, and also in patterns that are created by joining together
1288 fragments that contain references within themselves.
1289 </P>
1290 <P>
1291 A back reference matches whatever actually matched the capturing subpattern in
1292 the current subject string, rather than anything matching the subpattern
1293 itself (see
1294 <a href="#subpatternsassubroutines">"Subpatterns as subroutines"</a>
1295 below for a way of doing that). So the pattern
1296 <pre>
1297 (sens|respons)e and \1ibility
1298 </pre>
1299 matches "sense and sensibility" and "response and responsibility", but not
1300 "sense and responsibility". If caseful matching is in force at the time of the
1301 back reference, the case of letters is relevant. For example,
1302 <pre>
1303 ((?i)rah)\s+\1
1304 </pre>
1305 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
1306 capturing subpattern is matched caselessly.
1307 </P>
1308 <P>
1309 Back references to named subpatterns use the Perl syntax \k&#60;name&#62; or \k'name'
1310 or the Python syntax (?P=name). We could rewrite the above example in either of
1311 the following ways:
1312 <pre>
1313 (?&#60;p1&#62;(?i)rah)\s+\k&#60;p1&#62;
1314 (?P&#60;p1&#62;(?i)rah)\s+(?P=p1)
1315 </pre>
1316 A subpattern that is referenced by name may appear in the pattern before or
1317 after the reference.
1318 </P>
1319 <P>
1320 There may be more than one back reference to the same subpattern. If a
1321 subpattern has not actually been used in a particular match, any back
1322 references to it always fail. For example, the pattern
1323 <pre>
1324 (a|(bc))\2
1325 </pre>
1326 always fails if it starts to match "a" rather than "bc". Because there may be
1327 many capturing parentheses in a pattern, all digits following the backslash are
1328 taken as part of a potential back reference number. If the pattern continues
1329 with a digit character, some delimiter must be used to terminate the back
1330 reference. If the PCRE_EXTENDED option is set, this can be whitespace.
1331 Otherwise an empty comment (see
1332 <a href="#comments">"Comments"</a>
1333 below) can be used.
1334 </P>
1335 <P>
1336 A back reference that occurs inside the parentheses to which it refers fails
1337 when the subpattern is first used, so, for example, (a\1) never matches.
1338 However, such references can be useful inside repeated subpatterns. For
1339 example, the pattern
1340 <pre>
1341 (a|b\1)+
1342 </pre>
1343 matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
1344 the subpattern, the back reference matches the character string corresponding
1345 to the previous iteration. In order for this to work, the pattern must be such
1346 that the first iteration does not need to match the back reference. This can be
1347 done using alternation, as in the example above, or by a quantifier with a
1348 minimum of zero.
1349 <a name="bigassertions"></a></P>
1350 <br><a name="SEC16" href="#TOC1">ASSERTIONS</a><br>
1351 <P>
1352 An assertion is a test on the characters following or preceding the current
1353 matching point that does not actually consume any characters. The simple
1354 assertions coded as \b, \B, \A, \G, \Z, \z, ^ and $ are described
1355 <a href="#smallassertions">above.</a>
1356 </P>
1357 <P>
1358 More complicated assertions are coded as subpatterns. There are two kinds:
1359 those that look ahead of the current position in the subject string, and those
1360 that look behind it. An assertion subpattern is matched in the normal way,
1361 except that it does not cause the current matching position to be changed.
1362 </P>
1363 <P>
1364 Assertion subpatterns are not capturing subpatterns, and may not be repeated,
1365 because it makes no sense to assert the same thing several times. If any kind
1366 of assertion contains capturing subpatterns within it, these are counted for
1367 the purposes of numbering the capturing subpatterns in the whole pattern.
1368 However, substring capturing is carried out only for positive assertions,
1369 because it does not make sense for negative assertions.
1370 </P>
1371 <br><b>
1372 Lookahead assertions
1373 </b><br>
1374 <P>
1375 Lookahead assertions start with (?= for positive assertions and (?! for
1376 negative assertions. For example,
1377 <pre>
1378 \w+(?=;)
1379 </pre>
1380 matches a word followed by a semicolon, but does not include the semicolon in
1381 the match, and
1382 <pre>
1383 foo(?!bar)
1384 </pre>
1385 matches any occurrence of "foo" that is not followed by "bar". Note that the
1386 apparently similar pattern
1387 <pre>
1388 (?!foo)bar
1389 </pre>
1390 does not find an occurrence of "bar" that is preceded by something other than
1391 "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
1392 (?!foo) is always true when the next three characters are "bar". A
1393 lookbehind assertion is needed to achieve the other effect.
1394 </P>
1395 <P>
1396 If you want to force a matching failure at some point in a pattern, the most
1397 convenient way to do it is with (?!) because an empty string always matches, so
1398 an assertion that requires there not to be an empty string must always fail.
1399 <a name="lookbehind"></a></P>
1400 <br><b>
1401 Lookbehind assertions
1402 </b><br>
1403 <P>
1404 Lookbehind assertions start with (?&#60;= for positive assertions and (?&#60;! for
1405 negative assertions. For example,
1406 <pre>
1407 (?&#60;!foo)bar
1408 </pre>
1409 does find an occurrence of "bar" that is not preceded by "foo". The contents of
1410 a lookbehind assertion are restricted such that all the strings it matches must
1411 have a fixed length. However, if there are several top-level alternatives, they
1412 do not all have to have the same fixed length. Thus
1413 <pre>
1414 (?&#60;=bullock|donkey)
1415 </pre>
1416 is permitted, but
1417 <pre>
1418 (?&#60;!dogs?|cats?)
1419 </pre>
1420 causes an error at compile time. Branches that match different length strings
1421 are permitted only at the top level of a lookbehind assertion. This is an
1422 extension compared with Perl (at least for 5.8), which requires all branches to
1423 match the same length of string. An assertion such as
1424 <pre>
1425 (?&#60;=ab(c|de))
1426 </pre>
1427 is not permitted, because its single top-level branch can match two different
1428 lengths, but it is acceptable if rewritten to use two top-level branches:
1429 <pre>
1430 (?&#60;=abc|abde)
1431 </pre>
1432 The implementation of lookbehind assertions is, for each alternative, to
1433 temporarily move the current position back by the fixed length and then try to
1434 match. If there are insufficient characters before the current position, the
1435 assertion fails.
1436 </P>
1437 <P>
1438 PCRE does not allow the \C escape (which matches a single byte in UTF-8 mode)
1439 to appear in lookbehind assertions, because it makes it impossible to calculate
1440 the length of the lookbehind. The \X and \R escapes, which can match
1441 different numbers of bytes, are also not permitted.
1442 </P>
1443 <P>
1444 Possessive quantifiers can be used in conjunction with lookbehind assertions to
1445 specify efficient matching at the end of the subject string. Consider a simple
1446 pattern such as
1447 <pre>
1448 abcd$
1449 </pre>
1450 when applied to a long string that does not match. Because matching proceeds
1451 from left to right, PCRE will look for each "a" in the subject and then see if
1452 what follows matches the rest of the pattern. If the pattern is specified as
1453 <pre>
1454 ^.*abcd$
1455 </pre>
1456 the initial .* matches the entire string at first, but when this fails (because
1457 there is no following "a"), it backtracks to match all but the last character,
1458 then all but the last two characters, and so on. Once again the search for "a"
1459 covers the entire string, from right to left, so we are no better off. However,
1460 if the pattern is written as
1461 <pre>
1462 ^.*+(?&#60;=abcd)
1463 </pre>
1464 there can be no backtracking for the .*+ item; it can match only the entire
1465 string. The subsequent lookbehind assertion does a single test on the last four
1466 characters. If it fails, the match fails immediately. For long strings, this
1467 approach makes a significant difference to the processing time.
1468 </P>
1469 <br><b>
1470 Using multiple assertions
1471 </b><br>
1472 <P>
1473 Several assertions (of any sort) may occur in succession. For example,
1474 <pre>
1475 (?&#60;=\d{3})(?&#60;!999)foo
1476 </pre>
1477 matches "foo" preceded by three digits that are not "999". Notice that each of
1478 the assertions is applied independently at the same point in the subject
1479 string. First there is a check that the previous three characters are all
1480 digits, and then there is a check that the same three characters are not "999".
1481 This pattern does <i>not</i> match "foo" preceded by six characters, the first
1482 of which are digits and the last three of which are not "999". For example, it
1483 doesn't match "123abcfoo". A pattern to do that is
1484 <pre>
1485 (?&#60;=\d{3}...)(?&#60;!999)foo
1486 </pre>
1487 This time the first assertion looks at the preceding six characters, checking
1488 that the first three are digits, and then the second assertion checks that the
1489 preceding three characters are not "999".
1490 </P>
1491 <P>
1492 Assertions can be nested in any combination. For example,
1493 <pre>
1494 (?&#60;=(?&#60;!foo)bar)baz
1495 </pre>
1496 matches an occurrence of "baz" that is preceded by "bar" which in turn is not
1497 preceded by "foo", while
1498 <pre>
1499 (?&#60;=\d{3}(?!999)...)foo
1500 </pre>
1501 is another pattern that matches "foo" preceded by three digits and any three
1502 characters that are not "999".
1503 <a name="conditions"></a></P>
1504 <br><a name="SEC17" href="#TOC1">CONDITIONAL SUBPATTERNS</a><br>
1505 <P>
1506 It is possible to cause the matching process to obey a subpattern
1507 conditionally or to choose between two alternative subpatterns, depending on
1508 the result of an assertion, or whether a previous capturing subpattern matched
1509 or not. The two possible forms of conditional subpattern are
1510 <pre>
1511 (?(condition)yes-pattern)
1512 (?(condition)yes-pattern|no-pattern)
1513 </pre>
1514 If the condition is satisfied, the yes-pattern is used; otherwise the
1515 no-pattern (if present) is used. If there are more than two alternatives in the
1516 subpattern, a compile-time error occurs.
1517 </P>
1518 <P>
1519 There are four kinds of condition: references to subpatterns, references to
1520 recursion, a pseudo-condition called DEFINE, and assertions.
1521 </P>
1522 <br><b>
1523 Checking for a used subpattern by number
1524 </b><br>
1525 <P>
1526 If the text between the parentheses consists of a sequence of digits, the
1527 condition is true if the capturing subpattern of that number has previously
1528 matched.
1529 </P>
1530 <P>
1531 Consider the following pattern, which contains non-significant white space to
1532 make it more readable (assume the PCRE_EXTENDED option) and to divide it into
1533 three parts for ease of discussion:
1534 <pre>
1535 ( \( )? [^()]+ (?(1) \) )
1536 </pre>
1537 The first part matches an optional opening parenthesis, and if that
1538 character is present, sets it as the first captured substring. The second part
1539 matches one or more characters that are not parentheses. The third part is a
1540 conditional subpattern that tests whether the first set of parentheses matched
1541 or not. If they did, that is, if subject started with an opening parenthesis,
1542 the condition is true, and so the yes-pattern is executed and a closing
1543 parenthesis is required. Otherwise, since no-pattern is not present, the
1544 subpattern matches nothing. In other words, this pattern matches a sequence of
1545 non-parentheses, optionally enclosed in parentheses.
1546 </P>
1547 <br><b>
1548 Checking for a used subpattern by name
1549 </b><br>
1550 <P>
1551 Perl uses the syntax (?(&#60;name&#62;)...) or (?('name')...) to test for a used
1552 subpattern by name. For compatibility with earlier versions of PCRE, which had
1553 this facility before Perl, the syntax (?(name)...) is also recognized. However,
1554 there is a possible ambiguity with this syntax, because subpattern names may
1555 consist entirely of digits. PCRE looks first for a named subpattern; if it
1556 cannot find one and the name consists entirely of digits, PCRE looks for a
1557 subpattern of that number, which must be greater than zero. Using subpattern
1558 names that consist entirely of digits is not recommended.
1559 </P>
1560 <P>
1561 Rewriting the above example to use a named subpattern gives this:
1562 <pre>
1563 (?&#60;OPEN&#62; \( )? [^()]+ (?(&#60;OPEN&#62;) \) )
1564
1565 </PRE>
1566 </P>
1567 <br><b>
1568 Checking for pattern recursion
1569 </b><br>
1570 <P>
1571 If the condition is the string (R), and there is no subpattern with the name R,
1572 the condition is true if a recursive call to the whole pattern or any
1573 subpattern has been made. If digits or a name preceded by ampersand follow the
1574 letter R, for example:
1575 <pre>
1576 (?(R3)...) or (?(R&name)...)
1577 </pre>
1578 the condition is true if the most recent recursion is into the subpattern whose
1579 number or name is given. This condition does not check the entire recursion
1580 stack.
1581 </P>
1582 <P>
1583 At "top level", all these recursion test conditions are false. Recursive
1584 patterns are described below.
1585 </P>
1586 <br><b>
1587 Defining subpatterns for use by reference only
1588 </b><br>
1589 <P>
1590 If the condition is the string (DEFINE), and there is no subpattern with the
1591 name DEFINE, the condition is always false. In this case, there may be only one
1592 alternative in the subpattern. It is always skipped if control reaches this
1593 point in the pattern; the idea of DEFINE is that it can be used to define
1594 "subroutines" that can be referenced from elsewhere. (The use of "subroutines"
1595 is described below.) For example, a pattern to match an IPv4 address could be
1596 written like this (ignore whitespace and line breaks):
1597 <pre>
1598 (?(DEFINE) (?&#60;byte&#62; 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
1599 \b (?&byte) (\.(?&byte)){3} \b
1600 </pre>
1601 The first part of the pattern is a DEFINE group inside which a another group
1602 named "byte" is defined. This matches an individual component of an IPv4
1603 address (a number less than 256). When matching takes place, this part of the
1604 pattern is skipped because DEFINE acts like a false condition.
1605 </P>
1606 <P>
1607 The rest of the pattern uses references to the named group to match the four
1608 dot-separated components of an IPv4 address, insisting on a word boundary at
1609 each end.
1610 </P>
1611 <br><b>
1612 Assertion conditions
1613 </b><br>
1614 <P>
1615 If the condition is not in any of the above formats, it must be an assertion.
1616 This may be a positive or negative lookahead or lookbehind assertion. Consider
1617 this pattern, again containing non-significant white space, and with the two
1618 alternatives on the second line:
1619 <pre>
1620 (?(?=[^a-z]*[a-z])
1621 \d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} )
1622 </pre>
1623 The condition is a positive lookahead assertion that matches an optional
1624 sequence of non-letters followed by a letter. In other words, it tests for the
1625 presence of at least one letter in the subject. If a letter is found, the
1626 subject is matched against the first alternative; otherwise it is matched
1627 against the second. This pattern matches strings in one of the two forms
1628 dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
1629 <a name="comments"></a></P>
1630 <br><a name="SEC18" href="#TOC1">COMMENTS</a><br>
1631 <P>
1632 The sequence (?# marks the start of a comment that continues up to the next
1633 closing parenthesis. Nested parentheses are not permitted. The characters
1634 that make up a comment play no part in the pattern matching at all.
1635 </P>
1636 <P>
1637 If the PCRE_EXTENDED option is set, an unescaped # character outside a
1638 character class introduces a comment that continues to immediately after the
1639 next newline in the pattern.
1640 <a name="recursion"></a></P>
1641 <br><a name="SEC19" href="#TOC1">RECURSIVE PATTERNS</a><br>
1642 <P>
1643 Consider the problem of matching a string in parentheses, allowing for
1644 unlimited nested parentheses. Without the use of recursion, the best that can
1645 be done is to use a pattern that matches up to some fixed depth of nesting. It
1646 is not possible to handle an arbitrary nesting depth.
1647 </P>
1648 <P>
1649 For some time, Perl has provided a facility that allows regular expressions to
1650 recurse (amongst other things). It does this by interpolating Perl code in the
1651 expression at run time, and the code can refer to the expression itself. A Perl
1652 pattern using code interpolation to solve the parentheses problem can be
1653 created like this:
1654 <pre>
1655 $re = qr{\( (?: (?&#62;[^()]+) | (?p{$re}) )* \)}x;
1656 </pre>
1657 The (?p{...}) item interpolates Perl code at run time, and in this case refers
1658 recursively to the pattern in which it appears.
1659 </P>
1660 <P>
1661 Obviously, PCRE cannot support the interpolation of Perl code. Instead, it
1662 supports special syntax for recursion of the entire pattern, and also for
1663 individual subpattern recursion. After its introduction in PCRE and Python,
1664 this kind of recursion was introduced into Perl at release 5.10.
1665 </P>
1666 <P>
1667 A special item that consists of (? followed by a number greater than zero and a
1668 closing parenthesis is a recursive call of the subpattern of the given number,
1669 provided that it occurs inside that subpattern. (If not, it is a "subroutine"
1670 call, which is described in the next section.) The special item (?R) or (?0) is
1671 a recursive call of the entire regular expression.
1672 </P>
1673 <P>
1674 In PCRE (like Python, but unlike Perl), a recursive subpattern call is always
1675 treated as an atomic group. That is, once it has matched some of the subject
1676 string, it is never re-entered, even if it contains untried alternatives and
1677 there is a subsequent matching failure.
1678 </P>
1679 <P>
1680 This PCRE pattern solves the nested parentheses problem (assume the
1681 PCRE_EXTENDED option is set so that white space is ignored):
1682 <pre>
1683 \( ( (?&#62;[^()]+) | (?R) )* \)
1684 </pre>
1685 First it matches an opening parenthesis. Then it matches any number of
1686 substrings which can either be a sequence of non-parentheses, or a recursive
1687 match of the pattern itself (that is, a correctly parenthesized substring).
1688 Finally there is a closing parenthesis.
1689 </P>
1690 <P>
1691 If this were part of a larger pattern, you would not want to recurse the entire
1692 pattern, so instead you could use this:
1693 <pre>
1694 ( \( ( (?&#62;[^()]+) | (?1) )* \) )
1695 </pre>
1696 We have put the pattern into parentheses, and caused the recursion to refer to
1697 them instead of the whole pattern. In a larger pattern, keeping track of
1698 parenthesis numbers can be tricky. It may be more convenient to use named
1699 parentheses instead. The Perl syntax for this is (?&name); PCRE's earlier
1700 syntax (?P&#62;name) is also supported. We could rewrite the above example as
1701 follows:
1702 <pre>
1703 (?&#60;pn&#62; \( ( (?&#62;[^()]+) | (?&pn) )* \) )
1704 </pre>
1705 If there is more than one subpattern with the same name, the earliest one is
1706 used. This particular example pattern contains nested unlimited repeats, and so
1707 the use of atomic grouping for matching strings of non-parentheses is important
1708 when applying the pattern to strings that do not match. For example, when this
1709 pattern is applied to
1710 <pre>
1711 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
1712 </pre>
1713 it yields "no match" quickly. However, if atomic grouping is not used,
1714 the match runs for a very long time indeed because there are so many different
1715 ways the + and * repeats can carve up the subject, and all have to be tested
1716 before failure can be reported.
1717 </P>
1718 <P>
1719 At the end of a match, the values set for any capturing subpatterns are those
1720 from the outermost level of the recursion at which the subpattern value is set.
1721 If you want to obtain intermediate values, a callout function can be used (see
1722 below and the
1723 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1724 documentation). If the pattern above is matched against
1725 <pre>
1726 (ab(cd)ef)
1727 </pre>
1728 the value for the capturing parentheses is "ef", which is the last value taken
1729 on at the top level. If additional parentheses are added, giving
1730 <pre>
1731 \( ( ( (?&#62;[^()]+) | (?R) )* ) \)
1732 ^ ^
1733 ^ ^
1734 </pre>
1735 the string they capture is "ab(cd)ef", the contents of the top level
1736 parentheses. If there are more than 15 capturing parentheses in a pattern, PCRE
1737 has to obtain extra memory to store data during a recursion, which it does by
1738 using <b>pcre_malloc</b>, freeing it via <b>pcre_free</b> afterwards. If no
1739 memory can be obtained, the match fails with the PCRE_ERROR_NOMEMORY error.
1740 </P>
1741 <P>
1742 Do not confuse the (?R) item with the condition (R), which tests for recursion.
1743 Consider this pattern, which matches text in angle brackets, allowing for
1744 arbitrary nesting. Only digits are allowed in nested brackets (that is, when
1745 recursing), whereas any characters are permitted at the outer level.
1746 <pre>
1747 &#60; (?: (?(R) \d++ | [^&#60;&#62;]*+) | (?R)) * &#62;
1748 </pre>
1749 In this pattern, (?(R) is the start of a conditional subpattern, with two
1750 different alternatives for the recursive and non-recursive cases. The (?R) item
1751 is the actual recursive call.
1752 <a name="subpatternsassubroutines"></a></P>
1753 <br><a name="SEC20" href="#TOC1">SUBPATTERNS AS SUBROUTINES</a><br>
1754 <P>
1755 If the syntax for a recursive subpattern reference (either by number or by
1756 name) is used outside the parentheses to which it refers, it operates like a
1757 subroutine in a programming language. The "called" subpattern may be defined
1758 before or after the reference. An earlier example pointed out that the pattern
1759 <pre>
1760 (sens|respons)e and \1ibility
1761 </pre>
1762 matches "sense and sensibility" and "response and responsibility", but not
1763 "sense and responsibility". If instead the pattern
1764 <pre>
1765 (sens|respons)e and (?1)ibility
1766 </pre>
1767 is used, it does match "sense and responsibility" as well as the other two
1768 strings. Another example is given in the discussion of DEFINE above.
1769 </P>
1770 <P>
1771 Like recursive subpatterns, a "subroutine" call is always treated as an atomic
1772 group. That is, once it has matched some of the subject string, it is never
1773 re-entered, even if it contains untried alternatives and there is a subsequent
1774 matching failure.
1775 </P>
1776 <P>
1777 When a subpattern is used as a subroutine, processing options such as
1778 case-independence are fixed when the subpattern is defined. They cannot be
1779 changed for different calls. For example, consider this pattern:
1780 <pre>
1781 (abc)(?i:(?1))
1782 </pre>
1783 It matches "abcabc". It does not match "abcABC" because the change of
1784 processing option does not affect the called subpattern.
1785 </P>
1786 <br><a name="SEC21" href="#TOC1">CALLOUTS</a><br>
1787 <P>
1788 Perl has a feature whereby using the sequence (?{...}) causes arbitrary Perl
1789 code to be obeyed in the middle of matching a regular expression. This makes it
1790 possible, amongst other things, to extract different substrings that match the
1791 same pair of parentheses when there is a repetition.
1792 </P>
1793 <P>
1794 PCRE provides a similar feature, but of course it cannot obey arbitrary Perl
1795 code. The feature is called "callout". The caller of PCRE provides an external
1796 function by putting its entry point in the global variable <i>pcre_callout</i>.
1797 By default, this variable contains NULL, which disables all calling out.
1798 </P>
1799 <P>
1800 Within a regular expression, (?C) indicates the points at which the external
1801 function is to be called. If you want to identify different callout points, you
1802 can put a number less than 256 after the letter C. The default value is zero.
1803 For example, this pattern has two callout points:
1804 <pre>
1805 (?C1)\dabc(?C2)def
1806 </pre>
1807 If the PCRE_AUTO_CALLOUT flag is passed to <b>pcre_compile()</b>, callouts are
1808 automatically installed before each item in the pattern. They are all numbered
1809 255.
1810 </P>
1811 <P>
1812 During matching, when PCRE reaches a callout point (and <i>pcre_callout</i> is
1813 set), the external function is called. It is provided with the number of the
1814 callout, the position in the pattern, and, optionally, one item of data
1815 originally supplied by the caller of <b>pcre_exec()</b>. The callout function
1816 may cause matching to proceed, to backtrack, or to fail altogether. A complete
1817 description of the interface to the callout function is given in the
1818 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1819 documentation.
1820 </P>
1821 <br><a name="SEC22" href="#TOC1">SEE ALSO</a><br>
1822 <P>
1823 <b>pcreapi</b>(3), <b>pcrecallout</b>(3), <b>pcrematching</b>(3), <b>pcre</b>(3).
1824 </P>
1825 <P>
1826 Last updated: 06 December 2006
1827 <br>
1828 Copyright &copy; 1997-2006 University of Cambridge.
1829 <p>
1830 Return to the <a href="index.html">PCRE index page</a>.
1831 </p>

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12