ViewVC logotype

Contents of /code/branches/pcre16/doc/html/pcrepattern.html

Parent Directory Parent Directory | Revision Log Revision Log

Revision 801 - (show annotations)
Mon Dec 12 16:23:37 2011 UTC (9 years, 4 months ago) by ph10
File MIME type: text/html
File size: 123388 byte(s)
Error occurred while calculating annotation data.
Merge changes from trunk r755 to r800 into the 16-bit branch.
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">NEWLINE CONVENTIONS</a>
18 <li><a name="TOC3" href="#SEC3">CHARACTERS AND METACHARACTERS</a>
19 <li><a name="TOC4" href="#SEC4">BACKSLASH</a>
20 <li><a name="TOC5" href="#SEC5">CIRCUMFLEX AND DOLLAR</a>
21 <li><a name="TOC6" href="#SEC6">FULL STOP (PERIOD, DOT) AND \N</a>
22 <li><a name="TOC7" href="#SEC7">MATCHING A SINGLE BYTE</a>
24 <li><a name="TOC9" href="#SEC9">POSIX CHARACTER CLASSES</a>
25 <li><a name="TOC10" href="#SEC10">VERTICAL BAR</a>
26 <li><a name="TOC11" href="#SEC11">INTERNAL OPTION SETTING</a>
27 <li><a name="TOC12" href="#SEC12">SUBPATTERNS</a>
28 <li><a name="TOC13" href="#SEC13">DUPLICATE SUBPATTERN NUMBERS</a>
29 <li><a name="TOC14" href="#SEC14">NAMED SUBPATTERNS</a>
30 <li><a name="TOC15" href="#SEC15">REPETITION</a>
32 <li><a name="TOC17" href="#SEC17">BACK REFERENCES</a>
33 <li><a name="TOC18" href="#SEC18">ASSERTIONS</a>
34 <li><a name="TOC19" href="#SEC19">CONDITIONAL SUBPATTERNS</a>
35 <li><a name="TOC20" href="#SEC20">COMMENTS</a>
36 <li><a name="TOC21" href="#SEC21">RECURSIVE PATTERNS</a>
37 <li><a name="TOC22" href="#SEC22">SUBPATTERNS AS SUBROUTINES</a>
38 <li><a name="TOC23" href="#SEC23">ONIGURUMA SUBROUTINE SYNTAX</a>
39 <li><a name="TOC24" href="#SEC24">CALLOUTS</a>
40 <li><a name="TOC25" href="#SEC25">BACKTRACKING CONTROL</a>
41 <li><a name="TOC26" href="#SEC26">SEE ALSO</a>
42 <li><a name="TOC27" href="#SEC27">AUTHOR</a>
43 <li><a name="TOC28" href="#SEC28">REVISION</a>
44 </ul>
45 <br><a name="SEC1" href="#TOC1">PCRE REGULAR EXPRESSION DETAILS</a><br>
46 <P>
47 The syntax and semantics of the regular expressions that are supported by PCRE
48 are described in detail below. There is a quick-reference syntax summary in the
49 <a href="pcresyntax.html"><b>pcresyntax</b></a>
50 page. PCRE tries to match Perl syntax and semantics as closely as it can. PCRE
51 also supports some alternative regular expression syntax (which does not
52 conflict with the Perl syntax) in order to provide some compatibility with
53 regular expressions in Python, .NET, and Oniguruma.
54 </P>
55 <P>
56 Perl's regular expressions are described in its own documentation, and
57 regular expressions in general are covered in a number of books, some of which
58 have copious examples. Jeffrey Friedl's "Mastering Regular Expressions",
59 published by O'Reilly, covers regular expressions in great detail. This
60 description of PCRE's regular expressions is intended as reference material.
61 </P>
62 <P>
63 The original operation of PCRE was on strings of one-byte characters. However,
64 there is now also support for UTF-8 character strings. To use this,
65 PCRE must be built to include UTF-8 support, and you must call
66 <b>pcre_compile()</b> or <b>pcre_compile2()</b> with the PCRE_UTF8 option. There
67 is also a special sequence that can be given at the start of a pattern:
68 <pre>
69 (*UTF8)
70 </pre>
71 Starting a pattern with this sequence is equivalent to setting the PCRE_UTF8
72 option. This feature is not Perl-compatible. How setting UTF-8 mode affects
73 pattern matching is mentioned in several places below. There is also a summary
74 of UTF-8 features in the
75 <a href="pcreunicode.html"><b>pcreunicode</b></a>
76 page.
77 </P>
78 <P>
79 Another special sequence that may appear at the start of a pattern or in
80 combination with (*UTF8) is:
81 <pre>
82 (*UCP)
83 </pre>
84 This has the same effect as setting the PCRE_UCP option: it causes sequences
85 such as \d and \w to use Unicode properties to determine character types,
86 instead of recognizing only characters with codes less than 128 via a lookup
87 table.
88 </P>
89 <P>
90 If a pattern starts with (*NO_START_OPT), it has the same effect as setting the
91 PCRE_NO_START_OPTIMIZE option either at compile or matching time. There are
92 also some more of these special sequences that are concerned with the handling
93 of newlines; they are described below.
94 </P>
95 <P>
96 The remainder of this document discusses the patterns that are supported by
97 PCRE when its main matching function, <b>pcre_exec()</b>, is used.
98 From release 6.0, PCRE offers a second matching function,
99 <b>pcre_dfa_exec()</b>, which matches using a different algorithm that is not
100 Perl-compatible. Some of the features discussed below are not available when
101 <b>pcre_dfa_exec()</b> is used. The advantages and disadvantages of the
102 alternative function, and how it differs from the normal function, are
103 discussed in the
104 <a href="pcrematching.html"><b>pcrematching</b></a>
105 page.
106 <a name="newlines"></a></P>
107 <br><a name="SEC2" href="#TOC1">NEWLINE CONVENTIONS</a><br>
108 <P>
109 PCRE supports five different conventions for indicating line breaks in
110 strings: a single CR (carriage return) character, a single LF (linefeed)
111 character, the two-character sequence CRLF, any of the three preceding, or any
112 Unicode newline sequence. The
113 <a href="pcreapi.html"><b>pcreapi</b></a>
114 page has
115 <a href="pcreapi.html#newlines">further discussion</a>
116 about newlines, and shows how to set the newline convention in the
117 <i>options</i> arguments for the compiling and matching functions.
118 </P>
119 <P>
120 It is also possible to specify a newline convention by starting a pattern
121 string with one of the following five sequences:
122 <pre>
123 (*CR) carriage return
124 (*LF) linefeed
125 (*CRLF) carriage return, followed by linefeed
126 (*ANYCRLF) any of the three above
127 (*ANY) all Unicode newline sequences
128 </pre>
129 These override the default and the options given to <b>pcre_compile()</b> or
130 <b>pcre_compile2()</b>. For example, on a Unix system where LF is the default
131 newline sequence, the pattern
132 <pre>
133 (*CR)a.b
134 </pre>
135 changes the convention to CR. That pattern matches "a\nb" because LF is no
136 longer a newline. Note that these special settings, which are not
137 Perl-compatible, are recognized only at the very start of a pattern, and that
138 they must be in upper case. If more than one of them is present, the last one
139 is used.
140 </P>
141 <P>
142 The newline convention affects the interpretation of the dot metacharacter when
143 PCRE_DOTALL is not set, and also the behaviour of \N. However, it does not
144 affect what the \R escape sequence matches. By default, this is any Unicode
145 newline sequence, for Perl compatibility. However, this can be changed; see the
146 description of \R in the section entitled
147 <a href="#newlineseq">"Newline sequences"</a>
148 below. A change of \R setting can be combined with a change of newline
149 convention.
150 </P>
151 <br><a name="SEC3" href="#TOC1">CHARACTERS AND METACHARACTERS</a><br>
152 <P>
153 A regular expression is a pattern that is matched against a subject string from
154 left to right. Most characters stand for themselves in a pattern, and match the
155 corresponding characters in the subject. As a trivial example, the pattern
156 <pre>
157 The quick brown fox
158 </pre>
159 matches a portion of a subject string that is identical to itself. When
160 caseless matching is specified (the PCRE_CASELESS option), letters are matched
161 independently of case. In UTF-8 mode, PCRE always understands the concept of
162 case for characters whose values are less than 128, so caseless matching is
163 always possible. For characters with higher values, the concept of case is
164 supported if PCRE is compiled with Unicode property support, but not otherwise.
165 If you want to use caseless matching for characters 128 and above, you must
166 ensure that PCRE is compiled with Unicode property support as well as with
167 UTF-8 support.
168 </P>
169 <P>
170 The power of regular expressions comes from the ability to include alternatives
171 and repetitions in the pattern. These are encoded in the pattern by the use of
172 <i>metacharacters</i>, which do not stand for themselves but instead are
173 interpreted in some special way.
174 </P>
175 <P>
176 There are two different sets of metacharacters: those that are recognized
177 anywhere in the pattern except within square brackets, and those that are
178 recognized within square brackets. Outside square brackets, the metacharacters
179 are as follows:
180 <pre>
181 \ general escape character with several uses
182 ^ assert start of string (or line, in multiline mode)
183 $ assert end of string (or line, in multiline mode)
184 . match any character except newline (by default)
185 [ start character class definition
186 | start of alternative branch
187 ( start subpattern
188 ) end subpattern
189 ? extends the meaning of (
190 also 0 or 1 quantifier
191 also quantifier minimizer
192 * 0 or more quantifier
193 + 1 or more quantifier
194 also "possessive quantifier"
195 { start min/max quantifier
196 </pre>
197 Part of a pattern that is in square brackets is called a "character class". In
198 a character class the only metacharacters are:
199 <pre>
200 \ general escape character
201 ^ negate the class, but only if the first character
202 - indicates character range
203 [ POSIX character class (only if followed by POSIX syntax)
204 ] terminates the character class
205 </pre>
206 The following sections describe the use of each of the metacharacters.
207 </P>
208 <br><a name="SEC4" href="#TOC1">BACKSLASH</a><br>
209 <P>
210 The backslash character has several uses. Firstly, if it is followed by a
211 character that is not a number or a letter, it takes away any special meaning
212 that character may have. This use of backslash as an escape character applies
213 both inside and outside character classes.
214 </P>
215 <P>
216 For example, if you want to match a * character, you write \* in the pattern.
217 This escaping action applies whether or not the following character would
218 otherwise be interpreted as a metacharacter, so it is always safe to precede a
219 non-alphanumeric with backslash to specify that it stands for itself. In
220 particular, if you want to match a backslash, you write \\.
221 </P>
222 <P>
223 In UTF-8 mode, only ASCII numbers and letters have any special meaning after a
224 backslash. All other characters (in particular, those whose codepoints are
225 greater than 127) are treated as literals.
226 </P>
227 <P>
228 If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the
229 pattern (other than in a character class) and characters between a # outside
230 a character class and the next newline are ignored. An escaping backslash can
231 be used to include a whitespace or # character as part of the pattern.
232 </P>
233 <P>
234 If you want to remove the special meaning from a sequence of characters, you
235 can do so by putting them between \Q and \E. This is different from Perl in
236 that $ and @ are handled as literals in \Q...\E sequences in PCRE, whereas in
237 Perl, $ and @ cause variable interpolation. Note the following examples:
238 <pre>
239 Pattern PCRE matches Perl matches
241 \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
242 \Qabc\$xyz\E abc\$xyz abc\$xyz
243 \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
244 </pre>
245 The \Q...\E sequence is recognized both inside and outside character classes.
246 An isolated \E that is not preceded by \Q is ignored. If \Q is not followed
247 by \E later in the pattern, the literal interpretation continues to the end of
248 the pattern (that is, \E is assumed at the end). If the isolated \Q is inside
249 a character class, this causes an error, because the character class is not
250 terminated.
251 <a name="digitsafterbackslash"></a></P>
252 <br><b>
253 Non-printing characters
254 </b><br>
255 <P>
256 A second use of backslash provides a way of encoding non-printing characters
257 in patterns in a visible manner. There is no restriction on the appearance of
258 non-printing characters, apart from the binary zero that terminates a pattern,
259 but when a pattern is being prepared by text editing, it is often easier to use
260 one of the following escape sequences than the binary character it represents:
261 <pre>
262 \a alarm, that is, the BEL character (hex 07)
263 \cx "control-x", where x is any ASCII character
264 \e escape (hex 1B)
265 \f formfeed (hex 0C)
266 \n linefeed (hex 0A)
267 \r carriage return (hex 0D)
268 \t tab (hex 09)
269 \ddd character with octal code ddd, or back reference
270 \xhh character with hex code hh
271 \x{hhh..} character with hex code hhh.. (non-JavaScript mode)
272 \uhhhh character with hex code hhhh (JavaScript mode only)
273 </pre>
274 The precise effect of \cx is as follows: if x is a lower case letter, it
275 is converted to upper case. Then bit 6 of the character (hex 40) is inverted.
276 Thus \cz becomes hex 1A (z is 7A), but \c{ becomes hex 3B ({ is 7B), while
277 \c; becomes hex 7B (; is 3B). If the byte following \c has a value greater
278 than 127, a compile-time error occurs. This locks out non-ASCII characters in
279 both byte mode and UTF-8 mode. (When PCRE is compiled in EBCDIC mode, all byte
280 values are valid. A lower case letter is converted to upper case, and then the
281 0xc0 bits are flipped.)
282 </P>
283 <P>
284 By default, after \x, from zero to two hexadecimal digits are read (letters
285 can be in upper or lower case). Any number of hexadecimal digits may appear
286 between \x{ and }, but the value of the character code must be less than 256
287 in non-UTF-8 mode, and less than 2**31 in UTF-8 mode. That is, the maximum
288 value in hexadecimal is 7FFFFFFF. Note that this is bigger than the largest
289 Unicode code point, which is 10FFFF.
290 </P>
291 <P>
292 If characters other than hexadecimal digits appear between \x{ and }, or if
293 there is no terminating }, this form of escape is not recognized. Instead, the
294 initial \x will be interpreted as a basic hexadecimal escape, with no
295 following digits, giving a character whose value is zero.
296 </P>
297 <P>
298 If the PCRE_JAVASCRIPT_COMPAT option is set, the interpretation of \x is
299 as just described only when it is followed by two hexadecimal digits.
300 Otherwise, it matches a literal "x" character. In JavaScript mode, support for
301 code points greater than 256 is provided by \u, which must be followed by
302 four hexadecimal digits; otherwise it matches a literal "u" character.
303 </P>
304 <P>
305 Characters whose value is less than 256 can be defined by either of the two
306 syntaxes for \x (or by \u in JavaScript mode). There is no difference in the
307 way they are handled. For example, \xdc is exactly the same as \x{dc} (or
308 \u00dc in JavaScript mode).
309 </P>
310 <P>
311 After \0 up to two further octal digits are read. If there are fewer than two
312 digits, just those that are present are used. Thus the sequence \0\x\07
313 specifies two binary zeros followed by a BEL character (code value 7). Make
314 sure you supply two digits after the initial zero if the pattern character that
315 follows is itself an octal digit.
316 </P>
317 <P>
318 The handling of a backslash followed by a digit other than 0 is complicated.
319 Outside a character class, PCRE reads it and any following digits as a decimal
320 number. If the number is less than 10, or if there have been at least that many
321 previous capturing left parentheses in the expression, the entire sequence is
322 taken as a <i>back reference</i>. A description of how this works is given
323 <a href="#backreferences">later,</a>
324 following the discussion of
325 <a href="#subpattern">parenthesized subpatterns.</a>
326 </P>
327 <P>
328 Inside a character class, or if the decimal number is greater than 9 and there
329 have not been that many capturing subpatterns, PCRE re-reads up to three octal
330 digits following the backslash, and uses them to generate a data character. Any
331 subsequent digits stand for themselves. In non-UTF-8 mode, the value of a
332 character specified in octal must be less than \400. In UTF-8 mode, values up
333 to \777 are permitted. For example:
334 <pre>
335 \040 is another way of writing a space
336 \40 is the same, provided there are fewer than 40 previous capturing subpatterns
337 \7 is always a back reference
338 \11 might be a back reference, or another way of writing a tab
339 \011 is always a tab
340 \0113 is a tab followed by the character "3"
341 \113 might be a back reference, otherwise the character with octal code 113
342 \377 might be a back reference, otherwise the byte consisting entirely of 1 bits
343 \81 is either a back reference, or a binary zero followed by the two characters "8" and "1"
344 </pre>
345 Note that octal values of 100 or greater must not be introduced by a leading
346 zero, because no more than three octal digits are ever read.
347 </P>
348 <P>
349 All the sequences that define a single character value can be used both inside
350 and outside character classes. In addition, inside a character class, \b is
351 interpreted as the backspace character (hex 08).
352 </P>
353 <P>
354 \N is not allowed in a character class. \B, \R, and \X are not special
355 inside a character class. Like other unrecognized escape sequences, they are
356 treated as the literal characters "B", "R", and "X" by default, but cause an
357 error if the PCRE_EXTRA option is set. Outside a character class, these
358 sequences have different meanings.
359 </P>
360 <br><b>
361 Unsupported escape sequences
362 </b><br>
363 <P>
364 In Perl, the sequences \l, \L, \u, and \U are recognized by its string
365 handler and used to modify the case of following characters. By default, PCRE
366 does not support these escape sequences. However, if the PCRE_JAVASCRIPT_COMPAT
367 option is set, \U matches a "U" character, and \u can be used to define a
368 character by code point, as described in the previous section.
369 </P>
370 <br><b>
371 Absolute and relative back references
372 </b><br>
373 <P>
374 The sequence \g followed by an unsigned or a negative number, optionally
375 enclosed in braces, is an absolute or relative back reference. A named back
376 reference can be coded as \g{name}. Back references are discussed
377 <a href="#backreferences">later,</a>
378 following the discussion of
379 <a href="#subpattern">parenthesized subpatterns.</a>
380 </P>
381 <br><b>
382 Absolute and relative subroutine calls
383 </b><br>
384 <P>
385 For compatibility with Oniguruma, the non-Perl syntax \g followed by a name or
386 a number enclosed either in angle brackets or single quotes, is an alternative
387 syntax for referencing a subpattern as a "subroutine". Details are discussed
388 <a href="#onigurumasubroutines">later.</a>
389 Note that \g{...} (Perl syntax) and \g&#60;...&#62; (Oniguruma syntax) are <i>not</i>
390 synonymous. The former is a back reference; the latter is a
391 <a href="#subpatternsassubroutines">subroutine</a>
392 call.
393 <a name="genericchartypes"></a></P>
394 <br><b>
395 Generic character types
396 </b><br>
397 <P>
398 Another use of backslash is for specifying generic character types:
399 <pre>
400 \d any decimal digit
401 \D any character that is not a decimal digit
402 \h any horizontal whitespace character
403 \H any character that is not a horizontal whitespace character
404 \s any whitespace character
405 \S any character that is not a whitespace character
406 \v any vertical whitespace character
407 \V any character that is not a vertical whitespace character
408 \w any "word" character
409 \W any "non-word" character
410 </pre>
411 There is also the single sequence \N, which matches a non-newline character.
412 This is the same as
413 <a href="#fullstopdot">the "." metacharacter</a>
414 when PCRE_DOTALL is not set. Perl also uses \N to match characters by name;
415 PCRE does not support this.
416 </P>
417 <P>
418 Each pair of lower and upper case escape sequences partitions the complete set
419 of characters into two disjoint sets. Any given character matches one, and only
420 one, of each pair. The sequences can appear both inside and outside character
421 classes. They each match one character of the appropriate type. If the current
422 matching point is at the end of the subject string, all of them fail, because
423 there is no character to match.
424 </P>
425 <P>
426 For compatibility with Perl, \s does not match the VT character (code 11).
427 This makes it different from the the POSIX "space" class. The \s characters
428 are HT (9), LF (10), FF (12), CR (13), and space (32). If "use locale;" is
429 included in a Perl script, \s may match the VT character. In PCRE, it never
430 does.
431 </P>
432 <P>
433 A "word" character is an underscore or any character that is a letter or digit.
434 By default, the definition of letters and digits is controlled by PCRE's
435 low-valued character tables, and may vary if locale-specific matching is taking
436 place (see
437 <a href="pcreapi.html#localesupport">"Locale support"</a>
438 in the
439 <a href="pcreapi.html"><b>pcreapi</b></a>
440 page). For example, in a French locale such as "fr_FR" in Unix-like systems,
441 or "french" in Windows, some character codes greater than 128 are used for
442 accented letters, and these are then matched by \w. The use of locales with
443 Unicode is discouraged.
444 </P>
445 <P>
446 By default, in UTF-8 mode, characters with values greater than 128 never match
447 \d, \s, or \w, and always match \D, \S, and \W. These sequences retain
448 their original meanings from before UTF-8 support was available, mainly for
449 efficiency reasons. However, if PCRE is compiled with Unicode property support,
450 and the PCRE_UCP option is set, the behaviour is changed so that Unicode
451 properties are used to determine character types, as follows:
452 <pre>
453 \d any character that \p{Nd} matches (decimal digit)
454 \s any character that \p{Z} matches, plus HT, LF, FF, CR
455 \w any character that \p{L} or \p{N} matches, plus underscore
456 </pre>
457 The upper case escapes match the inverse sets of characters. Note that \d
458 matches only decimal digits, whereas \w matches any Unicode digit, as well as
459 any Unicode letter, and underscore. Note also that PCRE_UCP affects \b, and
460 \B because they are defined in terms of \w and \W. Matching these sequences
461 is noticeably slower when PCRE_UCP is set.
462 </P>
463 <P>
464 The sequences \h, \H, \v, and \V are features that were added to Perl at
465 release 5.10. In contrast to the other sequences, which match only ASCII
466 characters by default, these always match certain high-valued codepoints in
467 UTF-8 mode, whether or not PCRE_UCP is set. The horizontal space characters
468 are:
469 <pre>
470 U+0009 Horizontal tab
471 U+0020 Space
472 U+00A0 Non-break space
473 U+1680 Ogham space mark
474 U+180E Mongolian vowel separator
475 U+2000 En quad
476 U+2001 Em quad
477 U+2002 En space
478 U+2003 Em space
479 U+2004 Three-per-em space
480 U+2005 Four-per-em space
481 U+2006 Six-per-em space
482 U+2007 Figure space
483 U+2008 Punctuation space
484 U+2009 Thin space
485 U+200A Hair space
486 U+202F Narrow no-break space
487 U+205F Medium mathematical space
488 U+3000 Ideographic space
489 </pre>
490 The vertical space characters are:
491 <pre>
492 U+000A Linefeed
493 U+000B Vertical tab
494 U+000C Formfeed
495 U+000D Carriage return
496 U+0085 Next line
497 U+2028 Line separator
498 U+2029 Paragraph separator
499 <a name="newlineseq"></a></PRE>
500 </P>
501 <br><b>
502 Newline sequences
503 </b><br>
504 <P>
505 Outside a character class, by default, the escape sequence \R matches any
506 Unicode newline sequence. In non-UTF-8 mode \R is equivalent to the following:
507 <pre>
508 (?&#62;\r\n|\n|\x0b|\f|\r|\x85)
509 </pre>
510 This is an example of an "atomic group", details of which are given
511 <a href="#atomicgroup">below.</a>
512 This particular group matches either the two-character sequence CR followed by
513 LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab,
514 U+000B), FF (formfeed, U+000C), CR (carriage return, U+000D), or NEL (next
515 line, U+0085). The two-character sequence is treated as a single unit that
516 cannot be split.
517 </P>
518 <P>
519 In UTF-8 mode, two additional characters whose codepoints are greater than 255
520 are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029).
521 Unicode character property support is not needed for these characters to be
522 recognized.
523 </P>
524 <P>
525 It is possible to restrict \R to match only CR, LF, or CRLF (instead of the
526 complete set of Unicode line endings) by setting the option PCRE_BSR_ANYCRLF
527 either at compile time or when the pattern is matched. (BSR is an abbrevation
528 for "backslash R".) This can be made the default when PCRE is built; if this is
529 the case, the other behaviour can be requested via the PCRE_BSR_UNICODE option.
530 It is also possible to specify these settings by starting a pattern string with
531 one of the following sequences:
532 <pre>
533 (*BSR_ANYCRLF) CR, LF, or CRLF only
534 (*BSR_UNICODE) any Unicode newline sequence
535 </pre>
536 These override the default and the options given to <b>pcre_compile()</b> or
537 <b>pcre_compile2()</b>, but they can be overridden by options given to
538 <b>pcre_exec()</b> or <b>pcre_dfa_exec()</b>. Note that these special settings,
539 which are not Perl-compatible, are recognized only at the very start of a
540 pattern, and that they must be in upper case. If more than one of them is
541 present, the last one is used. They can be combined with a change of newline
542 convention; for example, a pattern can start with:
543 <pre>
545 </pre>
546 They can also be combined with the (*UTF8) or (*UCP) special sequences. Inside
547 a character class, \R is treated as an unrecognized escape sequence, and so
548 matches the letter "R" by default, but causes an error if PCRE_EXTRA is set.
549 <a name="uniextseq"></a></P>
550 <br><b>
551 Unicode character properties
552 </b><br>
553 <P>
554 When PCRE is built with Unicode character property support, three additional
555 escape sequences that match characters with specific properties are available.
556 When not in UTF-8 mode, these sequences are of course limited to testing
557 characters whose codepoints are less than 256, but they do work in this mode.
558 The extra escape sequences are:
559 <pre>
560 \p{<i>xx</i>} a character with the <i>xx</i> property
561 \P{<i>xx</i>} a character without the <i>xx</i> property
562 \X an extended Unicode sequence
563 </pre>
564 The property names represented by <i>xx</i> above are limited to the Unicode
565 script names, the general category properties, "Any", which matches any
566 character (including newline), and some special PCRE properties (described
567 in the
568 <a href="#extraprops">next section).</a>
569 Other Perl properties such as "InMusicalSymbols" are not currently supported by
570 PCRE. Note that \P{Any} does not match any characters, so always causes a
571 match failure.
572 </P>
573 <P>
574 Sets of Unicode characters are defined as belonging to certain scripts. A
575 character from one of these sets can be matched using a script name. For
576 example:
577 <pre>
578 \p{Greek}
579 \P{Han}
580 </pre>
581 Those that are not part of an identified script are lumped together as
582 "Common". The current list of scripts is:
583 </P>
584 <P>
585 Arabic,
586 Armenian,
587 Avestan,
588 Balinese,
589 Bamum,
590 Bengali,
591 Bopomofo,
592 Braille,
593 Buginese,
594 Buhid,
595 Canadian_Aboriginal,
596 Carian,
597 Cham,
598 Cherokee,
599 Common,
600 Coptic,
601 Cuneiform,
602 Cypriot,
603 Cyrillic,
604 Deseret,
605 Devanagari,
606 Egyptian_Hieroglyphs,
607 Ethiopic,
608 Georgian,
609 Glagolitic,
610 Gothic,
611 Greek,
612 Gujarati,
613 Gurmukhi,
614 Han,
615 Hangul,
616 Hanunoo,
617 Hebrew,
618 Hiragana,
619 Imperial_Aramaic,
620 Inherited,
621 Inscriptional_Pahlavi,
622 Inscriptional_Parthian,
623 Javanese,
624 Kaithi,
625 Kannada,
626 Katakana,
627 Kayah_Li,
628 Kharoshthi,
629 Khmer,
630 Lao,
631 Latin,
632 Lepcha,
633 Limbu,
634 Linear_B,
635 Lisu,
636 Lycian,
637 Lydian,
638 Malayalam,
639 Meetei_Mayek,
640 Mongolian,
641 Myanmar,
642 New_Tai_Lue,
643 Nko,
644 Ogham,
645 Old_Italic,
646 Old_Persian,
647 Old_South_Arabian,
648 Old_Turkic,
649 Ol_Chiki,
650 Oriya,
651 Osmanya,
652 Phags_Pa,
653 Phoenician,
654 Rejang,
655 Runic,
656 Samaritan,
657 Saurashtra,
658 Shavian,
659 Sinhala,
660 Sundanese,
661 Syloti_Nagri,
662 Syriac,
663 Tagalog,
664 Tagbanwa,
665 Tai_Le,
666 Tai_Tham,
667 Tai_Viet,
668 Tamil,
669 Telugu,
670 Thaana,
671 Thai,
672 Tibetan,
673 Tifinagh,
674 Ugaritic,
675 Vai,
676 Yi.
677 </P>
678 <P>
679 Each character has exactly one Unicode general category property, specified by
680 a two-letter abbreviation. For compatibility with Perl, negation can be
681 specified by including a circumflex between the opening brace and the property
682 name. For example, \p{^Lu} is the same as \P{Lu}.
683 </P>
684 <P>
685 If only one letter is specified with \p or \P, it includes all the general
686 category properties that start with that letter. In this case, in the absence
687 of negation, the curly brackets in the escape sequence are optional; these two
688 examples have the same effect:
689 <pre>
690 \p{L}
691 \pL
692 </pre>
693 The following general category property codes are supported:
694 <pre>
695 C Other
696 Cc Control
697 Cf Format
698 Cn Unassigned
699 Co Private use
700 Cs Surrogate
702 L Letter
703 Ll Lower case letter
704 Lm Modifier letter
705 Lo Other letter
706 Lt Title case letter
707 Lu Upper case letter
709 M Mark
710 Mc Spacing mark
711 Me Enclosing mark
712 Mn Non-spacing mark
714 N Number
715 Nd Decimal number
716 Nl Letter number
717 No Other number
719 P Punctuation
720 Pc Connector punctuation
721 Pd Dash punctuation
722 Pe Close punctuation
723 Pf Final punctuation
724 Pi Initial punctuation
725 Po Other punctuation
726 Ps Open punctuation
728 S Symbol
729 Sc Currency symbol
730 Sk Modifier symbol
731 Sm Mathematical symbol
732 So Other symbol
734 Z Separator
735 Zl Line separator
736 Zp Paragraph separator
737 Zs Space separator
738 </pre>
739 The special property L& is also supported: it matches a character that has
740 the Lu, Ll, or Lt property, in other words, a letter that is not classified as
741 a modifier or "other".
742 </P>
743 <P>
744 The Cs (Surrogate) property applies only to characters in the range U+D800 to
745 U+DFFF. Such characters are not valid in UTF-8 strings (see RFC 3629) and so
746 cannot be tested by PCRE, unless UTF-8 validity checking has been turned off
747 (see the discussion of PCRE_NO_UTF8_CHECK in the
748 <a href="pcreapi.html"><b>pcreapi</b></a>
749 page). Perl does not support the Cs property.
750 </P>
751 <P>
752 The long synonyms for property names that Perl supports (such as \p{Letter})
753 are not supported by PCRE, nor is it permitted to prefix any of these
754 properties with "Is".
755 </P>
756 <P>
757 No character that is in the Unicode table has the Cn (unassigned) property.
758 Instead, this property is assumed for any code point that is not in the
759 Unicode table.
760 </P>
761 <P>
762 Specifying caseless matching does not affect these escape sequences. For
763 example, \p{Lu} always matches only upper case letters.
764 </P>
765 <P>
766 The \X escape matches any number of Unicode characters that form an extended
767 Unicode sequence. \X is equivalent to
768 <pre>
769 (?&#62;\PM\pM*)
770 </pre>
771 That is, it matches a character without the "mark" property, followed by zero
772 or more characters with the "mark" property, and treats the sequence as an
773 atomic group
774 <a href="#atomicgroup">(see below).</a>
775 Characters with the "mark" property are typically accents that affect the
776 preceding character. None of them have codepoints less than 256, so in
777 non-UTF-8 mode \X matches any one character.
778 </P>
779 <P>
780 Note that recent versions of Perl have changed \X to match what Unicode calls
781 an "extended grapheme cluster", which has a more complicated definition.
782 </P>
783 <P>
784 Matching characters by Unicode property is not fast, because PCRE has to search
785 a structure that contains data for over fifteen thousand characters. That is
786 why the traditional escape sequences such as \d and \w do not use Unicode
787 properties in PCRE by default, though you can make them do so by setting the
788 PCRE_UCP option for <b>pcre_compile()</b> or by starting the pattern with
789 (*UCP).
790 <a name="extraprops"></a></P>
791 <br><b>
792 PCRE's additional properties
793 </b><br>
794 <P>
795 As well as the standard Unicode properties described in the previous
796 section, PCRE supports four more that make it possible to convert traditional
797 escape sequences such as \w and \s and POSIX character classes to use Unicode
798 properties. PCRE uses these non-standard, non-Perl properties internally when
799 PCRE_UCP is set. They are:
800 <pre>
801 Xan Any alphanumeric character
802 Xps Any POSIX space character
803 Xsp Any Perl space character
804 Xwd Any Perl "word" character
805 </pre>
806 Xan matches characters that have either the L (letter) or the N (number)
807 property. Xps matches the characters tab, linefeed, vertical tab, formfeed, or
808 carriage return, and any other character that has the Z (separator) property.
809 Xsp is the same as Xps, except that vertical tab is excluded. Xwd matches the
810 same characters as Xan, plus underscore.
811 <a name="resetmatchstart"></a></P>
812 <br><b>
813 Resetting the match start
814 </b><br>
815 <P>
816 The escape sequence \K causes any previously matched characters not to be
817 included in the final matched sequence. For example, the pattern:
818 <pre>
819 foo\Kbar
820 </pre>
821 matches "foobar", but reports that it has matched "bar". This feature is
822 similar to a lookbehind assertion
823 <a href="#lookbehind">(described below).</a>
824 However, in this case, the part of the subject before the real match does not
825 have to be of fixed length, as lookbehind assertions do. The use of \K does
826 not interfere with the setting of
827 <a href="#subpattern">captured substrings.</a>
828 For example, when the pattern
829 <pre>
830 (foo)\Kbar
831 </pre>
832 matches "foobar", the first substring is still set to "foo".
833 </P>
834 <P>
835 Perl documents that the use of \K within assertions is "not well defined". In
836 PCRE, \K is acted upon when it occurs inside positive assertions, but is
837 ignored in negative assertions.
838 <a name="smallassertions"></a></P>
839 <br><b>
840 Simple assertions
841 </b><br>
842 <P>
843 The final use of backslash is for certain simple assertions. An assertion
844 specifies a condition that has to be met at a particular point in a match,
845 without consuming any characters from the subject string. The use of
846 subpatterns for more complicated assertions is described
847 <a href="#bigassertions">below.</a>
848 The backslashed assertions are:
849 <pre>
850 \b matches at a word boundary
851 \B matches when not at a word boundary
852 \A matches at the start of the subject
853 \Z matches at the end of the subject
854 also matches before a newline at the end of the subject
855 \z matches only at the end of the subject
856 \G matches at the first matching position in the subject
857 </pre>
858 Inside a character class, \b has a different meaning; it matches the backspace
859 character. If any other of these assertions appears in a character class, by
860 default it matches the corresponding literal character (for example, \B
861 matches the letter B). However, if the PCRE_EXTRA option is set, an "invalid
862 escape sequence" error is generated instead.
863 </P>
864 <P>
865 A word boundary is a position in the subject string where the current character
866 and the previous character do not both match \w or \W (i.e. one matches
867 \w and the other matches \W), or the start or end of the string if the
868 first or last character matches \w, respectively. In UTF-8 mode, the meanings
869 of \w and \W can be changed by setting the PCRE_UCP option. When this is
870 done, it also affects \b and \B. Neither PCRE nor Perl has a separate "start
871 of word" or "end of word" metasequence. However, whatever follows \b normally
872 determines which it is. For example, the fragment \ba matches "a" at the start
873 of a word.
874 </P>
875 <P>
876 The \A, \Z, and \z assertions differ from the traditional circumflex and
877 dollar (described in the next section) in that they only ever match at the very
878 start and end of the subject string, whatever options are set. Thus, they are
879 independent of multiline mode. These three assertions are not affected by the
880 PCRE_NOTBOL or PCRE_NOTEOL options, which affect only the behaviour of the
881 circumflex and dollar metacharacters. However, if the <i>startoffset</i>
882 argument of <b>pcre_exec()</b> is non-zero, indicating that matching is to start
883 at a point other than the beginning of the subject, \A can never match. The
884 difference between \Z and \z is that \Z matches before a newline at the end
885 of the string as well as at the very end, whereas \z matches only at the end.
886 </P>
887 <P>
888 The \G assertion is true only when the current matching position is at the
889 start point of the match, as specified by the <i>startoffset</i> argument of
890 <b>pcre_exec()</b>. It differs from \A when the value of <i>startoffset</i> is
891 non-zero. By calling <b>pcre_exec()</b> multiple times with appropriate
892 arguments, you can mimic Perl's /g option, and it is in this kind of
893 implementation where \G can be useful.
894 </P>
895 <P>
896 Note, however, that PCRE's interpretation of \G, as the start of the current
897 match, is subtly different from Perl's, which defines it as the end of the
898 previous match. In Perl, these can be different when the previously matched
899 string was empty. Because PCRE does just one match at a time, it cannot
900 reproduce this behaviour.
901 </P>
902 <P>
903 If all the alternatives of a pattern begin with \G, the expression is anchored
904 to the starting match position, and the "anchored" flag is set in the compiled
905 regular expression.
906 </P>
907 <br><a name="SEC5" href="#TOC1">CIRCUMFLEX AND DOLLAR</a><br>
908 <P>
909 Outside a character class, in the default matching mode, the circumflex
910 character is an assertion that is true only if the current matching point is
911 at the start of the subject string. If the <i>startoffset</i> argument of
912 <b>pcre_exec()</b> is non-zero, circumflex can never match if the PCRE_MULTILINE
913 option is unset. Inside a character class, circumflex has an entirely different
914 meaning
915 <a href="#characterclass">(see below).</a>
916 </P>
917 <P>
918 Circumflex need not be the first character of the pattern if a number of
919 alternatives are involved, but it should be the first thing in each alternative
920 in which it appears if the pattern is ever to match that branch. If all
921 possible alternatives start with a circumflex, that is, if the pattern is
922 constrained to match only at the start of the subject, it is said to be an
923 "anchored" pattern. (There are also other constructs that can cause a pattern
924 to be anchored.)
925 </P>
926 <P>
927 A dollar character is an assertion that is true only if the current matching
928 point is at the end of the subject string, or immediately before a newline
929 at the end of the string (by default). Dollar need not be the last character of
930 the pattern if a number of alternatives are involved, but it should be the last
931 item in any branch in which it appears. Dollar has no special meaning in a
932 character class.
933 </P>
934 <P>
935 The meaning of dollar can be changed so that it matches only at the very end of
936 the string, by setting the PCRE_DOLLAR_ENDONLY option at compile time. This
937 does not affect the \Z assertion.
938 </P>
939 <P>
940 The meanings of the circumflex and dollar characters are changed if the
941 PCRE_MULTILINE option is set. When this is the case, a circumflex matches
942 immediately after internal newlines as well as at the start of the subject
943 string. It does not match after a newline that ends the string. A dollar
944 matches before any newlines in the string, as well as at the very end, when
945 PCRE_MULTILINE is set. When newline is specified as the two-character
946 sequence CRLF, isolated CR and LF characters do not indicate newlines.
947 </P>
948 <P>
949 For example, the pattern /^abc$/ matches the subject string "def\nabc" (where
950 \n represents a newline) in multiline mode, but not otherwise. Consequently,
951 patterns that are anchored in single line mode because all branches start with
952 ^ are not anchored in multiline mode, and a match for circumflex is possible
953 when the <i>startoffset</i> argument of <b>pcre_exec()</b> is non-zero. The
954 PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
955 </P>
956 <P>
957 Note that the sequences \A, \Z, and \z can be used to match the start and
958 end of the subject in both modes, and if all branches of a pattern start with
959 \A it is always anchored, whether or not PCRE_MULTILINE is set.
960 <a name="fullstopdot"></a></P>
961 <br><a name="SEC6" href="#TOC1">FULL STOP (PERIOD, DOT) AND \N</a><br>
962 <P>
963 Outside a character class, a dot in the pattern matches any one character in
964 the subject string except (by default) a character that signifies the end of a
965 line. In UTF-8 mode, the matched character may be more than one byte long.
966 </P>
967 <P>
968 When a line ending is defined as a single character, dot never matches that
969 character; when the two-character sequence CRLF is used, dot does not match CR
970 if it is immediately followed by LF, but otherwise it matches all characters
971 (including isolated CRs and LFs). When any Unicode line endings are being
972 recognized, dot does not match CR or LF or any of the other line ending
973 characters.
974 </P>
975 <P>
976 The behaviour of dot with regard to newlines can be changed. If the PCRE_DOTALL
977 option is set, a dot matches any one character, without exception. If the
978 two-character sequence CRLF is present in the subject string, it takes two dots
979 to match it.
980 </P>
981 <P>
982 The handling of dot is entirely independent of the handling of circumflex and
983 dollar, the only relationship being that they both involve newlines. Dot has no
984 special meaning in a character class.
985 </P>
986 <P>
987 The escape sequence \N behaves like a dot, except that it is not affected by
988 the PCRE_DOTALL option. In other words, it matches any character except one
989 that signifies the end of a line. Perl also uses \N to match characters by
990 name; PCRE does not support this.
991 </P>
992 <br><a name="SEC7" href="#TOC1">MATCHING A SINGLE BYTE</a><br>
993 <P>
994 Outside a character class, the escape sequence \C matches any one byte, both
995 in and out of UTF-8 mode. Unlike a dot, it always matches line-ending
996 characters. The feature is provided in Perl in order to match individual bytes
997 in UTF-8 mode, but it is unclear how it can usefully be used. Because \C
998 breaks up characters into individual bytes, matching one byte with \C in UTF-8
999 mode means that the rest of the string may start with a malformed UTF-8
1000 character. This has undefined results, because PCRE assumes that it is dealing
1001 with valid UTF-8 strings (and by default it checks this at the start of
1002 processing unless the PCRE_NO_UTF8_CHECK option is used).
1003 </P>
1004 <P>
1005 PCRE does not allow \C to appear in lookbehind assertions
1006 <a href="#lookbehind">(described below)</a>
1007 in UTF-8 mode, because this would make it impossible to calculate the length of
1008 the lookbehind.
1009 </P>
1010 <P>
1011 In general, the \C escape sequence is best avoided in UTF-8 mode. However, one
1012 way of using it that avoids the problem of malformed UTF-8 characters is to
1013 use a lookahead to check the length of the next character, as in this pattern
1014 (ignore white space and line breaks):
1015 <pre>
1016 (?| (?=[\x00-\x7f])(\C) |
1017 (?=[\x80-\x{7ff}])(\C)(\C) |
1018 (?=[\x{800}-\x{ffff}])(\C)(\C)(\C) |
1019 (?=[\x{10000}-\x{1fffff}])(\C)(\C)(\C)(\C))
1020 </pre>
1021 A group that starts with (?| resets the capturing parentheses numbers in each
1022 alternative (see
1023 <a href="#dupsubpatternnumber">"Duplicate Subpattern Numbers"</a>
1024 below). The assertions at the start of each branch check the next UTF-8
1025 character for values whose encoding uses 1, 2, 3, or 4 bytes, respectively. The
1026 character's individual bytes are then captured by the appropriate number of
1027 groups.
1028 <a name="characterclass"></a></P>
1029 <br><a name="SEC8" href="#TOC1">SQUARE BRACKETS AND CHARACTER CLASSES</a><br>
1030 <P>
1031 An opening square bracket introduces a character class, terminated by a closing
1032 square bracket. A closing square bracket on its own is not special by default.
1033 However, if the PCRE_JAVASCRIPT_COMPAT option is set, a lone closing square
1034 bracket causes a compile-time error. If a closing square bracket is required as
1035 a member of the class, it should be the first data character in the class
1036 (after an initial circumflex, if present) or escaped with a backslash.
1037 </P>
1038 <P>
1039 A character class matches a single character in the subject. In UTF-8 mode, the
1040 character may be more than one byte long. A matched character must be in the
1041 set of characters defined by the class, unless the first character in the class
1042 definition is a circumflex, in which case the subject character must not be in
1043 the set defined by the class. If a circumflex is actually required as a member
1044 of the class, ensure it is not the first character, or escape it with a
1045 backslash.
1046 </P>
1047 <P>
1048 For example, the character class [aeiou] matches any lower case vowel, while
1049 [^aeiou] matches any character that is not a lower case vowel. Note that a
1050 circumflex is just a convenient notation for specifying the characters that
1051 are in the class by enumerating those that are not. A class that starts with a
1052 circumflex is not an assertion; it still consumes a character from the subject
1053 string, and therefore it fails if the current pointer is at the end of the
1054 string.
1055 </P>
1056 <P>
1057 In UTF-8 mode, characters with values greater than 255 can be included in a
1058 class as a literal string of bytes, or by using the \x{ escaping mechanism.
1059 </P>
1060 <P>
1061 When caseless matching is set, any letters in a class represent both their
1062 upper case and lower case versions, so for example, a caseless [aeiou] matches
1063 "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
1064 caseful version would. In UTF-8 mode, PCRE always understands the concept of
1065 case for characters whose values are less than 128, so caseless matching is
1066 always possible. For characters with higher values, the concept of case is
1067 supported if PCRE is compiled with Unicode property support, but not otherwise.
1068 If you want to use caseless matching in UTF8-mode for characters 128 and above,
1069 you must ensure that PCRE is compiled with Unicode property support as well as
1070 with UTF-8 support.
1071 </P>
1072 <P>
1073 Characters that might indicate line breaks are never treated in any special way
1074 when matching character classes, whatever line-ending sequence is in use, and
1075 whatever setting of the PCRE_DOTALL and PCRE_MULTILINE options is used. A class
1076 such as [^a] always matches one of these characters.
1077 </P>
1078 <P>
1079 The minus (hyphen) character can be used to specify a range of characters in a
1080 character class. For example, [d-m] matches any letter between d and m,
1081 inclusive. If a minus character is required in a class, it must be escaped with
1082 a backslash or appear in a position where it cannot be interpreted as
1083 indicating a range, typically as the first or last character in the class.
1084 </P>
1085 <P>
1086 It is not possible to have the literal character "]" as the end character of a
1087 range. A pattern such as [W-]46] is interpreted as a class of two characters
1088 ("W" and "-") followed by a literal string "46]", so it would match "W46]" or
1089 "-46]". However, if the "]" is escaped with a backslash it is interpreted as
1090 the end of range, so [W-\]46] is interpreted as a class containing a range
1091 followed by two other characters. The octal or hexadecimal representation of
1092 "]" can also be used to end a range.
1093 </P>
1094 <P>
1095 Ranges operate in the collating sequence of character values. They can also be
1096 used for characters specified numerically, for example [\000-\037]. In UTF-8
1097 mode, ranges can include characters whose values are greater than 255, for
1098 example [\x{100}-\x{2ff}].
1099 </P>
1100 <P>
1101 If a range that includes letters is used when caseless matching is set, it
1102 matches the letters in either case. For example, [W-c] is equivalent to
1103 [][\\^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if character
1104 tables for a French locale are in use, [\xc8-\xcb] matches accented E
1105 characters in both cases. In UTF-8 mode, PCRE supports the concept of case for
1106 characters with values greater than 128 only when it is compiled with Unicode
1107 property support.
1108 </P>
1109 <P>
1110 The character escape sequences \d, \D, \h, \H, \p, \P, \s, \S, \v,
1111 \V, \w, and \W may appear in a character class, and add the characters that
1112 they match to the class. For example, [\dABCDEF] matches any hexadecimal
1113 digit. In UTF-8 mode, the PCRE_UCP option affects the meanings of \d, \s, \w
1114 and their upper case partners, just as it does when they appear outside a
1115 character class, as described in the section entitled
1116 <a href="#genericchartypes">"Generic character types"</a>
1117 above. The escape sequence \b has a different meaning inside a character
1118 class; it matches the backspace character. The sequences \B, \N, \R, and \X
1119 are not special inside a character class. Like any other unrecognized escape
1120 sequences, they are treated as the literal characters "B", "N", "R", and "X" by
1121 default, but cause an error if the PCRE_EXTRA option is set.
1122 </P>
1123 <P>
1124 A circumflex can conveniently be used with the upper case character types to
1125 specify a more restricted set of characters than the matching lower case type.
1126 For example, the class [^\W_] matches any letter or digit, but not underscore,
1127 whereas [\w] includes underscore. A positive character class should be read as
1128 "something OR something OR ..." and a negative class as "NOT something AND NOT
1129 something AND NOT ...".
1130 </P>
1131 <P>
1132 The only metacharacters that are recognized in character classes are backslash,
1133 hyphen (only where it can be interpreted as specifying a range), circumflex
1134 (only at the start), opening square bracket (only when it can be interpreted as
1135 introducing a POSIX class name - see the next section), and the terminating
1136 closing square bracket. However, escaping other non-alphanumeric characters
1137 does no harm.
1138 </P>
1139 <br><a name="SEC9" href="#TOC1">POSIX CHARACTER CLASSES</a><br>
1140 <P>
1141 Perl supports the POSIX notation for character classes. This uses names
1142 enclosed by [: and :] within the enclosing square brackets. PCRE also supports
1143 this notation. For example,
1144 <pre>
1145 [01[:alpha:]%]
1146 </pre>
1147 matches "0", "1", any alphabetic character, or "%". The supported class names
1148 are:
1149 <pre>
1150 alnum letters and digits
1151 alpha letters
1152 ascii character codes 0 - 127
1153 blank space or tab only
1154 cntrl control characters
1155 digit decimal digits (same as \d)
1156 graph printing characters, excluding space
1157 lower lower case letters
1158 print printing characters, including space
1159 punct printing characters, excluding letters and digits and space
1160 space white space (not quite the same as \s)
1161 upper upper case letters
1162 word "word" characters (same as \w)
1163 xdigit hexadecimal digits
1164 </pre>
1165 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13), and
1166 space (32). Notice that this list includes the VT character (code 11). This
1167 makes "space" different to \s, which does not include VT (for Perl
1168 compatibility).
1169 </P>
1170 <P>
1171 The name "word" is a Perl extension, and "blank" is a GNU extension from Perl
1172 5.8. Another Perl extension is negation, which is indicated by a ^ character
1173 after the colon. For example,
1174 <pre>
1175 [12[:^digit:]]
1176 </pre>
1177 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX
1178 syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
1179 supported, and an error is given if they are encountered.
1180 </P>
1181 <P>
1182 By default, in UTF-8 mode, characters with values greater than 128 do not match
1183 any of the POSIX character classes. However, if the PCRE_UCP option is passed
1184 to <b>pcre_compile()</b>, some of the classes are changed so that Unicode
1185 character properties are used. This is achieved by replacing the POSIX classes
1186 by other sequences, as follows:
1187 <pre>
1188 [:alnum:] becomes \p{Xan}
1189 [:alpha:] becomes \p{L}
1190 [:blank:] becomes \h
1191 [:digit:] becomes \p{Nd}
1192 [:lower:] becomes \p{Ll}
1193 [:space:] becomes \p{Xps}
1194 [:upper:] becomes \p{Lu}
1195 [:word:] becomes \p{Xwd}
1196 </pre>
1197 Negated versions, such as [:^alpha:] use \P instead of \p. The other POSIX
1198 classes are unchanged, and match only characters with code points less than
1199 128.
1200 </P>
1201 <br><a name="SEC10" href="#TOC1">VERTICAL BAR</a><br>
1202 <P>
1203 Vertical bar characters are used to separate alternative patterns. For example,
1204 the pattern
1205 <pre>
1206 gilbert|sullivan
1207 </pre>
1208 matches either "gilbert" or "sullivan". Any number of alternatives may appear,
1209 and an empty alternative is permitted (matching the empty string). The matching
1210 process tries each alternative in turn, from left to right, and the first one
1211 that succeeds is used. If the alternatives are within a subpattern
1212 <a href="#subpattern">(defined below),</a>
1213 "succeeds" means matching the rest of the main pattern as well as the
1214 alternative in the subpattern.
1215 </P>
1216 <br><a name="SEC11" href="#TOC1">INTERNAL OPTION SETTING</a><br>
1217 <P>
1218 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
1219 PCRE_EXTENDED options (which are Perl-compatible) can be changed from within
1220 the pattern by a sequence of Perl option letters enclosed between "(?" and ")".
1221 The option letters are
1222 <pre>
1223 i for PCRE_CASELESS
1225 s for PCRE_DOTALL
1226 x for PCRE_EXTENDED
1227 </pre>
1228 For example, (?im) sets caseless, multiline matching. It is also possible to
1229 unset these options by preceding the letter with a hyphen, and a combined
1230 setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
1231 PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
1232 permitted. If a letter appears both before and after the hyphen, the option is
1233 unset.
1234 </P>
1235 <P>
1236 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA can be
1237 changed in the same way as the Perl-compatible options by using the characters
1238 J, U and X respectively.
1239 </P>
1240 <P>
1241 When one of these option changes occurs at top level (that is, not inside
1242 subpattern parentheses), the change applies to the remainder of the pattern
1243 that follows. If the change is placed right at the start of a pattern, PCRE
1244 extracts it into the global options (and it will therefore show up in data
1245 extracted by the <b>pcre_fullinfo()</b> function).
1246 </P>
1247 <P>
1248 An option change within a subpattern (see below for a description of
1249 subpatterns) affects only that part of the subpattern that follows it, so
1250 <pre>
1251 (a(?i)b)c
1252 </pre>
1253 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
1254 By this means, options can be made to have different settings in different
1255 parts of the pattern. Any changes made in one alternative do carry on
1256 into subsequent branches within the same subpattern. For example,
1257 <pre>
1258 (a(?i)b|c)
1259 </pre>
1260 matches "ab", "aB", "c", and "C", even though when matching "C" the first
1261 branch is abandoned before the option setting. This is because the effects of
1262 option settings happen at compile time. There would be some very weird
1263 behaviour otherwise.
1264 </P>
1265 <P>
1266 <b>Note:</b> There are other PCRE-specific options that can be set by the
1267 application when the compile or match functions are called. In some cases the
1268 pattern can contain special leading sequences such as (*CRLF) to override what
1269 the application has set or what has been defaulted. Details are given in the
1270 section entitled
1271 <a href="#newlineseq">"Newline sequences"</a>
1272 above. There are also the (*UTF8) and (*UCP) leading sequences that can be used
1273 to set UTF-8 and Unicode property modes; they are equivalent to setting the
1274 PCRE_UTF8 and the PCRE_UCP options, respectively.
1275 <a name="subpattern"></a></P>
1276 <br><a name="SEC12" href="#TOC1">SUBPATTERNS</a><br>
1277 <P>
1278 Subpatterns are delimited by parentheses (round brackets), which can be nested.
1279 Turning part of a pattern into a subpattern does two things:
1280 <br>
1281 <br>
1282 1. It localizes a set of alternatives. For example, the pattern
1283 <pre>
1284 cat(aract|erpillar|)
1285 </pre>
1286 matches "cataract", "caterpillar", or "cat". Without the parentheses, it would
1287 match "cataract", "erpillar" or an empty string.
1288 <br>
1289 <br>
1290 2. It sets up the subpattern as a capturing subpattern. This means that, when
1291 the whole pattern matches, that portion of the subject string that matched the
1292 subpattern is passed back to the caller via the <i>ovector</i> argument of
1293 <b>pcre_exec()</b>. Opening parentheses are counted from left to right (starting
1294 from 1) to obtain numbers for the capturing subpatterns. For example, if the
1295 string "the red king" is matched against the pattern
1296 <pre>
1297 the ((red|white) (king|queen))
1298 </pre>
1299 the captured substrings are "red king", "red", and "king", and are numbered 1,
1300 2, and 3, respectively.
1301 </P>
1302 <P>
1303 The fact that plain parentheses fulfil two functions is not always helpful.
1304 There are often times when a grouping subpattern is required without a
1305 capturing requirement. If an opening parenthesis is followed by a question mark
1306 and a colon, the subpattern does not do any capturing, and is not counted when
1307 computing the number of any subsequent capturing subpatterns. For example, if
1308 the string "the white queen" is matched against the pattern
1309 <pre>
1310 the ((?:red|white) (king|queen))
1311 </pre>
1312 the captured substrings are "white queen" and "queen", and are numbered 1 and
1313 2. The maximum number of capturing subpatterns is 65535.
1314 </P>
1315 <P>
1316 As a convenient shorthand, if any option settings are required at the start of
1317 a non-capturing subpattern, the option letters may appear between the "?" and
1318 the ":". Thus the two patterns
1319 <pre>
1320 (?i:saturday|sunday)
1321 (?:(?i)saturday|sunday)
1322 </pre>
1323 match exactly the same set of strings. Because alternative branches are tried
1324 from left to right, and options are not reset until the end of the subpattern
1325 is reached, an option setting in one branch does affect subsequent branches, so
1326 the above patterns match "SUNDAY" as well as "Saturday".
1327 <a name="dupsubpatternnumber"></a></P>
1328 <br><a name="SEC13" href="#TOC1">DUPLICATE SUBPATTERN NUMBERS</a><br>
1329 <P>
1330 Perl 5.10 introduced a feature whereby each alternative in a subpattern uses
1331 the same numbers for its capturing parentheses. Such a subpattern starts with
1332 (?| and is itself a non-capturing subpattern. For example, consider this
1333 pattern:
1334 <pre>
1335 (?|(Sat)ur|(Sun))day
1336 </pre>
1337 Because the two alternatives are inside a (?| group, both sets of capturing
1338 parentheses are numbered one. Thus, when the pattern matches, you can look
1339 at captured substring number one, whichever alternative matched. This construct
1340 is useful when you want to capture part, but not all, of one of a number of
1341 alternatives. Inside a (?| group, parentheses are numbered as usual, but the
1342 number is reset at the start of each branch. The numbers of any capturing
1343 parentheses that follow the subpattern start after the highest number used in
1344 any branch. The following example is taken from the Perl documentation. The
1345 numbers underneath show in which buffer the captured content will be stored.
1346 <pre>
1347 # before ---------------branch-reset----------- after
1348 / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
1349 # 1 2 2 3 2 3 4
1350 </pre>
1351 A back reference to a numbered subpattern uses the most recent value that is
1352 set for that number by any subpattern. The following pattern matches "abcabc"
1353 or "defdef":
1354 <pre>
1355 /(?|(abc)|(def))\1/
1356 </pre>
1357 In contrast, a subroutine call to a numbered subpattern always refers to the
1358 first one in the pattern with the given number. The following pattern matches
1359 "abcabc" or "defabc":
1360 <pre>
1361 /(?|(abc)|(def))(?1)/
1362 </pre>
1363 If a
1364 <a href="#conditions">condition test</a>
1365 for a subpattern's having matched refers to a non-unique number, the test is
1366 true if any of the subpatterns of that number have matched.
1367 </P>
1368 <P>
1369 An alternative approach to using this "branch reset" feature is to use
1370 duplicate named subpatterns, as described in the next section.
1371 </P>
1372 <br><a name="SEC14" href="#TOC1">NAMED SUBPATTERNS</a><br>
1373 <P>
1374 Identifying capturing parentheses by number is simple, but it can be very hard
1375 to keep track of the numbers in complicated regular expressions. Furthermore,
1376 if an expression is modified, the numbers may change. To help with this
1377 difficulty, PCRE supports the naming of subpatterns. This feature was not
1378 added to Perl until release 5.10. Python had the feature earlier, and PCRE
1379 introduced it at release 4.0, using the Python syntax. PCRE now supports both
1380 the Perl and the Python syntax. Perl allows identically numbered subpatterns to
1381 have different names, but PCRE does not.
1382 </P>
1383 <P>
1384 In PCRE, a subpattern can be named in one of three ways: (?&#60;name&#62;...) or
1385 (?'name'...) as in Perl, or (?P&#60;name&#62;...) as in Python. References to capturing
1386 parentheses from other parts of the pattern, such as
1387 <a href="#backreferences">back references,</a>
1388 <a href="#recursion">recursion,</a>
1389 and
1390 <a href="#conditions">conditions,</a>
1391 can be made by name as well as by number.
1392 </P>
1393 <P>
1394 Names consist of up to 32 alphanumeric characters and underscores. Named
1395 capturing parentheses are still allocated numbers as well as names, exactly as
1396 if the names were not present. The PCRE API provides function calls for
1397 extracting the name-to-number translation table from a compiled pattern. There
1398 is also a convenience function for extracting a captured substring by name.
1399 </P>
1400 <P>
1401 By default, a name must be unique within a pattern, but it is possible to relax
1402 this constraint by setting the PCRE_DUPNAMES option at compile time. (Duplicate
1403 names are also always permitted for subpatterns with the same number, set up as
1404 described in the previous section.) Duplicate names can be useful for patterns
1405 where only one instance of the named parentheses can match. Suppose you want to
1406 match the name of a weekday, either as a 3-letter abbreviation or as the full
1407 name, and in both cases you want to extract the abbreviation. This pattern
1408 (ignoring the line breaks) does the job:
1409 <pre>
1410 (?&#60;DN&#62;Mon|Fri|Sun)(?:day)?|
1411 (?&#60;DN&#62;Tue)(?:sday)?|
1412 (?&#60;DN&#62;Wed)(?:nesday)?|
1413 (?&#60;DN&#62;Thu)(?:rsday)?|
1414 (?&#60;DN&#62;Sat)(?:urday)?
1415 </pre>
1416 There are five capturing substrings, but only one is ever set after a match.
1417 (An alternative way of solving this problem is to use a "branch reset"
1418 subpattern, as described in the previous section.)
1419 </P>
1420 <P>
1421 The convenience function for extracting the data by name returns the substring
1422 for the first (and in this example, the only) subpattern of that name that
1423 matched. This saves searching to find which numbered subpattern it was.
1424 </P>
1425 <P>
1426 If you make a back reference to a non-unique named subpattern from elsewhere in
1427 the pattern, the one that corresponds to the first occurrence of the name is
1428 used. In the absence of duplicate numbers (see the previous section) this is
1429 the one with the lowest number. If you use a named reference in a condition
1430 test (see the
1431 <a href="#conditions">section about conditions</a>
1432 below), either to check whether a subpattern has matched, or to check for
1433 recursion, all subpatterns with the same name are tested. If the condition is
1434 true for any one of them, the overall condition is true. This is the same
1435 behaviour as testing by number. For further details of the interfaces for
1436 handling named subpatterns, see the
1437 <a href="pcreapi.html"><b>pcreapi</b></a>
1438 documentation.
1439 </P>
1440 <P>
1441 <b>Warning:</b> You cannot use different names to distinguish between two
1442 subpatterns with the same number because PCRE uses only the numbers when
1443 matching. For this reason, an error is given at compile time if different names
1444 are given to subpatterns with the same number. However, you can give the same
1445 name to subpatterns with the same number, even when PCRE_DUPNAMES is not set.
1446 </P>
1447 <br><a name="SEC15" href="#TOC1">REPETITION</a><br>
1448 <P>
1449 Repetition is specified by quantifiers, which can follow any of the following
1450 items:
1451 <pre>
1452 a literal data character
1453 the dot metacharacter
1454 the \C escape sequence
1455 the \X escape sequence (in UTF-8 mode with Unicode properties)
1456 the \R escape sequence
1457 an escape such as \d or \pL that matches a single character
1458 a character class
1459 a back reference (see next section)
1460 a parenthesized subpattern (including assertions)
1461 a subroutine call to a subpattern (recursive or otherwise)
1462 </pre>
1463 The general repetition quantifier specifies a minimum and maximum number of
1464 permitted matches, by giving the two numbers in curly brackets (braces),
1465 separated by a comma. The numbers must be less than 65536, and the first must
1466 be less than or equal to the second. For example:
1467 <pre>
1468 z{2,4}
1469 </pre>
1470 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
1471 character. If the second number is omitted, but the comma is present, there is
1472 no upper limit; if the second number and the comma are both omitted, the
1473 quantifier specifies an exact number of required matches. Thus
1474 <pre>
1475 [aeiou]{3,}
1476 </pre>
1477 matches at least 3 successive vowels, but may match many more, while
1478 <pre>
1479 \d{8}
1480 </pre>
1481 matches exactly 8 digits. An opening curly bracket that appears in a position
1482 where a quantifier is not allowed, or one that does not match the syntax of a
1483 quantifier, is taken as a literal character. For example, {,6} is not a
1484 quantifier, but a literal string of four characters.
1485 </P>
1486 <P>
1487 In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to individual
1488 bytes. Thus, for example, \x{100}{2} matches two UTF-8 characters, each of
1489 which is represented by a two-byte sequence. Similarly, when Unicode property
1490 support is available, \X{3} matches three Unicode extended sequences, each of
1491 which may be several bytes long (and they may be of different lengths).
1492 </P>
1493 <P>
1494 The quantifier {0} is permitted, causing the expression to behave as if the
1495 previous item and the quantifier were not present. This may be useful for
1496 subpatterns that are referenced as
1497 <a href="#subpatternsassubroutines">subroutines</a>
1498 from elsewhere in the pattern (but see also the section entitled
1499 <a href="#subdefine">"Defining subpatterns for use by reference only"</a>
1500 below). Items other than subpatterns that have a {0} quantifier are omitted
1501 from the compiled pattern.
1502 </P>
1503 <P>
1504 For convenience, the three most common quantifiers have single-character
1505 abbreviations:
1506 <pre>
1507 * is equivalent to {0,}
1508 + is equivalent to {1,}
1509 ? is equivalent to {0,1}
1510 </pre>
1511 It is possible to construct infinite loops by following a subpattern that can
1512 match no characters with a quantifier that has no upper limit, for example:
1513 <pre>
1514 (a?)*
1515 </pre>
1516 Earlier versions of Perl and PCRE used to give an error at compile time for
1517 such patterns. However, because there are cases where this can be useful, such
1518 patterns are now accepted, but if any repetition of the subpattern does in fact
1519 match no characters, the loop is forcibly broken.
1520 </P>
1521 <P>
1522 By default, the quantifiers are "greedy", that is, they match as much as
1523 possible (up to the maximum number of permitted times), without causing the
1524 rest of the pattern to fail. The classic example of where this gives problems
1525 is in trying to match comments in C programs. These appear between /* and */
1526 and within the comment, individual * and / characters may appear. An attempt to
1527 match C comments by applying the pattern
1528 <pre>
1529 /\*.*\*/
1530 </pre>
1531 to the string
1532 <pre>
1533 /* first comment */ not comment /* second comment */
1534 </pre>
1535 fails, because it matches the entire string owing to the greediness of the .*
1536 item.
1537 </P>
1538 <P>
1539 However, if a quantifier is followed by a question mark, it ceases to be
1540 greedy, and instead matches the minimum number of times possible, so the
1541 pattern
1542 <pre>
1543 /\*.*?\*/
1544 </pre>
1545 does the right thing with the C comments. The meaning of the various
1546 quantifiers is not otherwise changed, just the preferred number of matches.
1547 Do not confuse this use of question mark with its use as a quantifier in its
1548 own right. Because it has two uses, it can sometimes appear doubled, as in
1549 <pre>
1550 \d??\d
1551 </pre>
1552 which matches one digit by preference, but can match two if that is the only
1553 way the rest of the pattern matches.
1554 </P>
1555 <P>
1556 If the PCRE_UNGREEDY option is set (an option that is not available in Perl),
1557 the quantifiers are not greedy by default, but individual ones can be made
1558 greedy by following them with a question mark. In other words, it inverts the
1559 default behaviour.
1560 </P>
1561 <P>
1562 When a parenthesized subpattern is quantified with a minimum repeat count that
1563 is greater than 1 or with a limited maximum, more memory is required for the
1564 compiled pattern, in proportion to the size of the minimum or maximum.
1565 </P>
1566 <P>
1567 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent
1568 to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is
1569 implicitly anchored, because whatever follows will be tried against every
1570 character position in the subject string, so there is no point in retrying the
1571 overall match at any position after the first. PCRE normally treats such a
1572 pattern as though it were preceded by \A.
1573 </P>
1574 <P>
1575 In cases where it is known that the subject string contains no newlines, it is
1576 worth setting PCRE_DOTALL in order to obtain this optimization, or
1577 alternatively using ^ to indicate anchoring explicitly.
1578 </P>
1579 <P>
1580 However, there is one situation where the optimization cannot be used. When .*
1581 is inside capturing parentheses that are the subject of a back reference
1582 elsewhere in the pattern, a match at the start may fail where a later one
1583 succeeds. Consider, for example:
1584 <pre>
1585 (.*)abc\1
1586 </pre>
1587 If the subject is "xyz123abc123" the match point is the fourth character. For
1588 this reason, such a pattern is not implicitly anchored.
1589 </P>
1590 <P>
1591 When a capturing subpattern is repeated, the value captured is the substring
1592 that matched the final iteration. For example, after
1593 <pre>
1594 (tweedle[dume]{3}\s*)+
1595 </pre>
1596 has matched "tweedledum tweedledee" the value of the captured substring is
1597 "tweedledee". However, if there are nested capturing subpatterns, the
1598 corresponding captured values may have been set in previous iterations. For
1599 example, after
1600 <pre>
1601 /(a|(b))+/
1602 </pre>
1603 matches "aba" the value of the second captured substring is "b".
1604 <a name="atomicgroup"></a></P>
1605 <br><a name="SEC16" href="#TOC1">ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS</a><br>
1606 <P>
1607 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
1608 repetition, failure of what follows normally causes the repeated item to be
1609 re-evaluated to see if a different number of repeats allows the rest of the
1610 pattern to match. Sometimes it is useful to prevent this, either to change the
1611 nature of the match, or to cause it fail earlier than it otherwise might, when
1612 the author of the pattern knows there is no point in carrying on.
1613 </P>
1614 <P>
1615 Consider, for example, the pattern \d+foo when applied to the subject line
1616 <pre>
1617 123456bar
1618 </pre>
1619 After matching all 6 digits and then failing to match "foo", the normal
1620 action of the matcher is to try again with only 5 digits matching the \d+
1621 item, and then with 4, and so on, before ultimately failing. "Atomic grouping"
1622 (a term taken from Jeffrey Friedl's book) provides the means for specifying
1623 that once a subpattern has matched, it is not to be re-evaluated in this way.
1624 </P>
1625 <P>
1626 If we use atomic grouping for the previous example, the matcher gives up
1627 immediately on failing to match "foo" the first time. The notation is a kind of
1628 special parenthesis, starting with (?&#62; as in this example:
1629 <pre>
1630 (?&#62;\d+)foo
1631 </pre>
1632 This kind of parenthesis "locks up" the part of the pattern it contains once
1633 it has matched, and a failure further into the pattern is prevented from
1634 backtracking into it. Backtracking past it to previous items, however, works as
1635 normal.
1636 </P>
1637 <P>
1638 An alternative description is that a subpattern of this type matches the string
1639 of characters that an identical standalone pattern would match, if anchored at
1640 the current point in the subject string.
1641 </P>
1642 <P>
1643 Atomic grouping subpatterns are not capturing subpatterns. Simple cases such as
1644 the above example can be thought of as a maximizing repeat that must swallow
1645 everything it can. So, while both \d+ and \d+? are prepared to adjust the
1646 number of digits they match in order to make the rest of the pattern match,
1647 (?&#62;\d+) can only match an entire sequence of digits.
1648 </P>
1649 <P>
1650 Atomic groups in general can of course contain arbitrarily complicated
1651 subpatterns, and can be nested. However, when the subpattern for an atomic
1652 group is just a single repeated item, as in the example above, a simpler
1653 notation, called a "possessive quantifier" can be used. This consists of an
1654 additional + character following a quantifier. Using this notation, the
1655 previous example can be rewritten as
1656 <pre>
1657 \d++foo
1658 </pre>
1659 Note that a possessive quantifier can be used with an entire group, for
1660 example:
1661 <pre>
1662 (abc|xyz){2,3}+
1663 </pre>
1664 Possessive quantifiers are always greedy; the setting of the PCRE_UNGREEDY
1665 option is ignored. They are a convenient notation for the simpler forms of
1666 atomic group. However, there is no difference in the meaning of a possessive
1667 quantifier and the equivalent atomic group, though there may be a performance
1668 difference; possessive quantifiers should be slightly faster.
1669 </P>
1670 <P>
1671 The possessive quantifier syntax is an extension to the Perl 5.8 syntax.
1672 Jeffrey Friedl originated the idea (and the name) in the first edition of his
1673 book. Mike McCloskey liked it, so implemented it when he built Sun's Java
1674 package, and PCRE copied it from there. It ultimately found its way into Perl
1675 at release 5.10.
1676 </P>
1677 <P>
1678 PCRE has an optimization that automatically "possessifies" certain simple
1679 pattern constructs. For example, the sequence A+B is treated as A++B because
1680 there is no point in backtracking into a sequence of A's when B must follow.
1681 </P>
1682 <P>
1683 When a pattern contains an unlimited repeat inside a subpattern that can itself
1684 be repeated an unlimited number of times, the use of an atomic group is the
1685 only way to avoid some failing matches taking a very long time indeed. The
1686 pattern
1687 <pre>
1688 (\D+|&#60;\d+&#62;)*[!?]
1689 </pre>
1690 matches an unlimited number of substrings that either consist of non-digits, or
1691 digits enclosed in &#60;&#62;, followed by either ! or ?. When it matches, it runs
1692 quickly. However, if it is applied to
1693 <pre>
1694 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
1695 </pre>
1696 it takes a long time before reporting failure. This is because the string can
1697 be divided between the internal \D+ repeat and the external * repeat in a
1698 large number of ways, and all have to be tried. (The example uses [!?] rather
1699 than a single character at the end, because both PCRE and Perl have an
1700 optimization that allows for fast failure when a single character is used. They
1701 remember the last single character that is required for a match, and fail early
1702 if it is not present in the string.) If the pattern is changed so that it uses
1703 an atomic group, like this:
1704 <pre>
1705 ((?&#62;\D+)|&#60;\d+&#62;)*[!?]
1706 </pre>
1707 sequences of non-digits cannot be broken, and failure happens quickly.
1708 <a name="backreferences"></a></P>
1709 <br><a name="SEC17" href="#TOC1">BACK REFERENCES</a><br>
1710 <P>
1711 Outside a character class, a backslash followed by a digit greater than 0 (and
1712 possibly further digits) is a back reference to a capturing subpattern earlier
1713 (that is, to its left) in the pattern, provided there have been that many
1714 previous capturing left parentheses.
1715 </P>
1716 <P>
1717 However, if the decimal number following the backslash is less than 10, it is
1718 always taken as a back reference, and causes an error only if there are not
1719 that many capturing left parentheses in the entire pattern. In other words, the
1720 parentheses that are referenced need not be to the left of the reference for
1721 numbers less than 10. A "forward back reference" of this type can make sense
1722 when a repetition is involved and the subpattern to the right has participated
1723 in an earlier iteration.
1724 </P>
1725 <P>
1726 It is not possible to have a numerical "forward back reference" to a subpattern
1727 whose number is 10 or more using this syntax because a sequence such as \50 is
1728 interpreted as a character defined in octal. See the subsection entitled
1729 "Non-printing characters"
1730 <a href="#digitsafterbackslash">above</a>
1731 for further details of the handling of digits following a backslash. There is
1732 no such problem when named parentheses are used. A back reference to any
1733 subpattern is possible using named parentheses (see below).
1734 </P>
1735 <P>
1736 Another way of avoiding the ambiguity inherent in the use of digits following a
1737 backslash is to use the \g escape sequence. This escape must be followed by an
1738 unsigned number or a negative number, optionally enclosed in braces. These
1739 examples are all identical:
1740 <pre>
1741 (ring), \1
1742 (ring), \g1
1743 (ring), \g{1}
1744 </pre>
1745 An unsigned number specifies an absolute reference without the ambiguity that
1746 is present in the older syntax. It is also useful when literal digits follow
1747 the reference. A negative number is a relative reference. Consider this
1748 example:
1749 <pre>
1750 (abc(def)ghi)\g{-1}
1751 </pre>
1752 The sequence \g{-1} is a reference to the most recently started capturing
1753 subpattern before \g, that is, is it equivalent to \2 in this example.
1754 Similarly, \g{-2} would be equivalent to \1. The use of relative references
1755 can be helpful in long patterns, and also in patterns that are created by
1756 joining together fragments that contain references within themselves.
1757 </P>
1758 <P>
1759 A back reference matches whatever actually matched the capturing subpattern in
1760 the current subject string, rather than anything matching the subpattern
1761 itself (see
1762 <a href="#subpatternsassubroutines">"Subpatterns as subroutines"</a>
1763 below for a way of doing that). So the pattern
1764 <pre>
1765 (sens|respons)e and \1ibility
1766 </pre>
1767 matches "sense and sensibility" and "response and responsibility", but not
1768 "sense and responsibility". If caseful matching is in force at the time of the
1769 back reference, the case of letters is relevant. For example,
1770 <pre>
1771 ((?i)rah)\s+\1
1772 </pre>
1773 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
1774 capturing subpattern is matched caselessly.
1775 </P>
1776 <P>
1777 There are several different ways of writing back references to named
1778 subpatterns. The .NET syntax \k{name} and the Perl syntax \k&#60;name&#62; or
1779 \k'name' are supported, as is the Python syntax (?P=name). Perl 5.10's unified
1780 back reference syntax, in which \g can be used for both numeric and named
1781 references, is also supported. We could rewrite the above example in any of
1782 the following ways:
1783 <pre>
1784 (?&#60;p1&#62;(?i)rah)\s+\k&#60;p1&#62;
1785 (?'p1'(?i)rah)\s+\k{p1}
1786 (?P&#60;p1&#62;(?i)rah)\s+(?P=p1)
1787 (?&#60;p1&#62;(?i)rah)\s+\g{p1}
1788 </pre>
1789 A subpattern that is referenced by name may appear in the pattern before or
1790 after the reference.
1791 </P>
1792 <P>
1793 There may be more than one back reference to the same subpattern. If a
1794 subpattern has not actually been used in a particular match, any back
1795 references to it always fail by default. For example, the pattern
1796 <pre>
1797 (a|(bc))\2
1798 </pre>
1799 always fails if it starts to match "a" rather than "bc". However, if the
1800 PCRE_JAVASCRIPT_COMPAT option is set at compile time, a back reference to an
1801 unset value matches an empty string.
1802 </P>
1803 <P>
1804 Because there may be many capturing parentheses in a pattern, all digits
1805 following a backslash are taken as part of a potential back reference number.
1806 If the pattern continues with a digit character, some delimiter must be used to
1807 terminate the back reference. If the PCRE_EXTENDED option is set, this can be
1808 whitespace. Otherwise, the \g{ syntax or an empty comment (see
1809 <a href="#comments">"Comments"</a>
1810 below) can be used.
1811 </P>
1812 <br><b>
1813 Recursive back references
1814 </b><br>
1815 <P>
1816 A back reference that occurs inside the parentheses to which it refers fails
1817 when the subpattern is first used, so, for example, (a\1) never matches.
1818 However, such references can be useful inside repeated subpatterns. For
1819 example, the pattern
1820 <pre>
1821 (a|b\1)+
1822 </pre>
1823 matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
1824 the subpattern, the back reference matches the character string corresponding
1825 to the previous iteration. In order for this to work, the pattern must be such
1826 that the first iteration does not need to match the back reference. This can be
1827 done using alternation, as in the example above, or by a quantifier with a
1828 minimum of zero.
1829 </P>
1830 <P>
1831 Back references of this type cause the group that they reference to be treated
1832 as an
1833 <a href="#atomicgroup">atomic group.</a>
1834 Once the whole group has been matched, a subsequent matching failure cannot
1835 cause backtracking into the middle of the group.
1836 <a name="bigassertions"></a></P>
1837 <br><a name="SEC18" href="#TOC1">ASSERTIONS</a><br>
1838 <P>
1839 An assertion is a test on the characters following or preceding the current
1840 matching point that does not actually consume any characters. The simple
1841 assertions coded as \b, \B, \A, \G, \Z, \z, ^ and $ are described
1842 <a href="#smallassertions">above.</a>
1843 </P>
1844 <P>
1845 More complicated assertions are coded as subpatterns. There are two kinds:
1846 those that look ahead of the current position in the subject string, and those
1847 that look behind it. An assertion subpattern is matched in the normal way,
1848 except that it does not cause the current matching position to be changed.
1849 </P>
1850 <P>
1851 Assertion subpatterns are not capturing subpatterns. If such an assertion
1852 contains capturing subpatterns within it, these are counted for the purposes of
1853 numbering the capturing subpatterns in the whole pattern. However, substring
1854 capturing is carried out only for positive assertions, because it does not make
1855 sense for negative assertions.
1856 </P>
1857 <P>
1858 For compatibility with Perl, assertion subpatterns may be repeated; though
1859 it makes no sense to assert the same thing several times, the side effect of
1860 capturing parentheses may occasionally be useful. In practice, there only three
1861 cases:
1862 <br>
1863 <br>
1864 (1) If the quantifier is {0}, the assertion is never obeyed during matching.
1865 However, it may contain internal capturing parenthesized groups that are called
1866 from elsewhere via the
1867 <a href="#subpatternsassubroutines">subroutine mechanism.</a>
1868 <br>
1869 <br>
1870 (2) If quantifier is {0,n} where n is greater than zero, it is treated as if it
1871 were {0,1}. At run time, the rest of the pattern match is tried with and
1872 without the assertion, the order depending on the greediness of the quantifier.
1873 <br>
1874 <br>
1875 (3) If the minimum repetition is greater than zero, the quantifier is ignored.
1876 The assertion is obeyed just once when encountered during matching.
1877 </P>
1878 <br><b>
1879 Lookahead assertions
1880 </b><br>
1881 <P>
1882 Lookahead assertions start with (?= for positive assertions and (?! for
1883 negative assertions. For example,
1884 <pre>
1885 \w+(?=;)
1886 </pre>
1887 matches a word followed by a semicolon, but does not include the semicolon in
1888 the match, and
1889 <pre>
1890 foo(?!bar)
1891 </pre>
1892 matches any occurrence of "foo" that is not followed by "bar". Note that the
1893 apparently similar pattern
1894 <pre>
1895 (?!foo)bar
1896 </pre>
1897 does not find an occurrence of "bar" that is preceded by something other than
1898 "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
1899 (?!foo) is always true when the next three characters are "bar". A
1900 lookbehind assertion is needed to achieve the other effect.
1901 </P>
1902 <P>
1903 If you want to force a matching failure at some point in a pattern, the most
1904 convenient way to do it is with (?!) because an empty string always matches, so
1905 an assertion that requires there not to be an empty string must always fail.
1906 The backtracking control verb (*FAIL) or (*F) is a synonym for (?!).
1907 <a name="lookbehind"></a></P>
1908 <br><b>
1909 Lookbehind assertions
1910 </b><br>
1911 <P>
1912 Lookbehind assertions start with (?&#60;= for positive assertions and (?&#60;! for
1913 negative assertions. For example,
1914 <pre>
1915 (?&#60;!foo)bar
1916 </pre>
1917 does find an occurrence of "bar" that is not preceded by "foo". The contents of
1918 a lookbehind assertion are restricted such that all the strings it matches must
1919 have a fixed length. However, if there are several top-level alternatives, they
1920 do not all have to have the same fixed length. Thus
1921 <pre>
1922 (?&#60;=bullock|donkey)
1923 </pre>
1924 is permitted, but
1925 <pre>
1926 (?&#60;!dogs?|cats?)
1927 </pre>
1928 causes an error at compile time. Branches that match different length strings
1929 are permitted only at the top level of a lookbehind assertion. This is an
1930 extension compared with Perl, which requires all branches to match the same
1931 length of string. An assertion such as
1932 <pre>
1933 (?&#60;=ab(c|de))
1934 </pre>
1935 is not permitted, because its single top-level branch can match two different
1936 lengths, but it is acceptable to PCRE if rewritten to use two top-level
1937 branches:
1938 <pre>
1939 (?&#60;=abc|abde)
1940 </pre>
1941 In some cases, the escape sequence \K
1942 <a href="#resetmatchstart">(see above)</a>
1943 can be used instead of a lookbehind assertion to get round the fixed-length
1944 restriction.
1945 </P>
1946 <P>
1947 The implementation of lookbehind assertions is, for each alternative, to
1948 temporarily move the current position back by the fixed length and then try to
1949 match. If there are insufficient characters before the current position, the
1950 assertion fails.
1951 </P>
1952 <P>
1953 In UTF-8 mode, PCRE does not allow the \C escape (which matches a single byte,
1954 even in UTF-8 mode) to appear in lookbehind assertions, because it makes it
1955 impossible to calculate the length of the lookbehind. The \X and \R escapes,
1956 which can match different numbers of bytes, are also not permitted.
1957 </P>
1958 <P>
1959 <a href="#subpatternsassubroutines">"Subroutine"</a>
1960 calls (see below) such as (?2) or (?&X) are permitted in lookbehinds, as long
1961 as the subpattern matches a fixed-length string.
1962 <a href="#recursion">Recursion,</a>
1963 however, is not supported.
1964 </P>
1965 <P>
1966 Possessive quantifiers can be used in conjunction with lookbehind assertions to
1967 specify efficient matching of fixed-length strings at the end of subject
1968 strings. Consider a simple pattern such as
1969 <pre>
1970 abcd$
1971 </pre>
1972 when applied to a long string that does not match. Because matching proceeds
1973 from left to right, PCRE will look for each "a" in the subject and then see if
1974 what follows matches the rest of the pattern. If the pattern is specified as
1975 <pre>
1976 ^.*abcd$
1977 </pre>
1978 the initial .* matches the entire string at first, but when this fails (because
1979 there is no following "a"), it backtracks to match all but the last character,
1980 then all but the last two characters, and so on. Once again the search for "a"
1981 covers the entire string, from right to left, so we are no better off. However,
1982 if the pattern is written as
1983 <pre>
1984 ^.*+(?&#60;=abcd)
1985 </pre>
1986 there can be no backtracking for the .*+ item; it can match only the entire
1987 string. The subsequent lookbehind assertion does a single test on the last four
1988 characters. If it fails, the match fails immediately. For long strings, this
1989 approach makes a significant difference to the processing time.
1990 </P>
1991 <br><b>
1992 Using multiple assertions
1993 </b><br>
1994 <P>
1995 Several assertions (of any sort) may occur in succession. For example,
1996 <pre>
1997 (?&#60;=\d{3})(?&#60;!999)foo
1998 </pre>
1999 matches "foo" preceded by three digits that are not "999". Notice that each of
2000 the assertions is applied independently at the same point in the subject
2001 string. First there is a check that the previous three characters are all
2002 digits, and then there is a check that the same three characters are not "999".
2003 This pattern does <i>not</i> match "foo" preceded by six characters, the first
2004 of which are digits and the last three of which are not "999". For example, it
2005 doesn't match "123abcfoo". A pattern to do that is
2006 <pre>
2007 (?&#60;=\d{3}...)(?&#60;!999)foo
2008 </pre>
2009 This time the first assertion looks at the preceding six characters, checking
2010 that the first three are digits, and then the second assertion checks that the
2011 preceding three characters are not "999".
2012 </P>
2013 <P>
2014 Assertions can be nested in any combination. For example,
2015 <pre>
2016 (?&#60;=(?&#60;!foo)bar)baz
2017 </pre>
2018 matches an occurrence of "baz" that is preceded by "bar" which in turn is not
2019 preceded by "foo", while
2020 <pre>
2021 (?&#60;=\d{3}(?!999)...)foo
2022 </pre>
2023 is another pattern that matches "foo" preceded by three digits and any three
2024 characters that are not "999".
2025 <a name="conditions"></a></P>
2026 <br><a name="SEC19" href="#TOC1">CONDITIONAL SUBPATTERNS</a><br>
2027 <P>
2028 It is possible to cause the matching process to obey a subpattern
2029 conditionally or to choose between two alternative subpatterns, depending on
2030 the result of an assertion, or whether a specific capturing subpattern has
2031 already been matched. The two possible forms of conditional subpattern are:
2032 <pre>
2033 (?(condition)yes-pattern)
2034 (?(condition)yes-pattern|no-pattern)
2035 </pre>
2036 If the condition is satisfied, the yes-pattern is used; otherwise the
2037 no-pattern (if present) is used. If there are more than two alternatives in the
2038 subpattern, a compile-time error occurs. Each of the two alternatives may
2039 itself contain nested subpatterns of any form, including conditional
2040 subpatterns; the restriction to two alternatives applies only at the level of
2041 the condition. This pattern fragment is an example where the alternatives are
2042 complex:
2043 <pre>
2044 (?(1) (A|B|C) | (D | (?(2)E|F) | E) )
2046 </PRE>
2047 </P>
2048 <P>
2049 There are four kinds of condition: references to subpatterns, references to
2050 recursion, a pseudo-condition called DEFINE, and assertions.
2051 </P>
2052 <br><b>
2053 Checking for a used subpattern by number
2054 </b><br>
2055 <P>
2056 If the text between the parentheses consists of a sequence of digits, the
2057 condition is true if a capturing subpattern of that number has previously
2058 matched. If there is more than one capturing subpattern with the same number
2059 (see the earlier
2060 <a href="#recursion">section about duplicate subpattern numbers),</a>
2061 the condition is true if any of them have matched. An alternative notation is
2062 to precede the digits with a plus or minus sign. In this case, the subpattern
2063 number is relative rather than absolute. The most recently opened parentheses
2064 can be referenced by (?(-1), the next most recent by (?(-2), and so on. Inside
2065 loops it can also make sense to refer to subsequent groups. The next
2066 parentheses to be opened can be referenced as (?(+1), and so on. (The value
2067 zero in any of these forms is not used; it provokes a compile-time error.)
2068 </P>
2069 <P>
2070 Consider the following pattern, which contains non-significant white space to
2071 make it more readable (assume the PCRE_EXTENDED option) and to divide it into
2072 three parts for ease of discussion:
2073 <pre>
2074 ( \( )? [^()]+ (?(1) \) )
2075 </pre>
2076 The first part matches an optional opening parenthesis, and if that
2077 character is present, sets it as the first captured substring. The second part
2078 matches one or more characters that are not parentheses. The third part is a
2079 conditional subpattern that tests whether or not the first set of parentheses
2080 matched. If they did, that is, if subject started with an opening parenthesis,
2081 the condition is true, and so the yes-pattern is executed and a closing
2082 parenthesis is required. Otherwise, since no-pattern is not present, the
2083 subpattern matches nothing. In other words, this pattern matches a sequence of
2084 non-parentheses, optionally enclosed in parentheses.
2085 </P>
2086 <P>
2087 If you were embedding this pattern in a larger one, you could use a relative
2088 reference:
2089 <pre>
2090 ...other stuff... ( \( )? [^()]+ (?(-1) \) ) ...
2091 </pre>
2092 This makes the fragment independent of the parentheses in the larger pattern.
2093 </P>
2094 <br><b>
2095 Checking for a used subpattern by name
2096 </b><br>
2097 <P>
2098 Perl uses the syntax (?(&#60;name&#62;)...) or (?('name')...) to test for a used
2099 subpattern by name. For compatibility with earlier versions of PCRE, which had
2100 this facility before Perl, the syntax (?(name)...) is also recognized. However,
2101 there is a possible ambiguity with this syntax, because subpattern names may
2102 consist entirely of digits. PCRE looks first for a named subpattern; if it
2103 cannot find one and the name consists entirely of digits, PCRE looks for a
2104 subpattern of that number, which must be greater than zero. Using subpattern
2105 names that consist entirely of digits is not recommended.
2106 </P>
2107 <P>
2108 Rewriting the above example to use a named subpattern gives this:
2109 <pre>
2110 (?&#60;OPEN&#62; \( )? [^()]+ (?(&#60;OPEN&#62;) \) )
2111 </pre>
2112 If the name used in a condition of this kind is a duplicate, the test is
2113 applied to all subpatterns of the same name, and is true if any one of them has
2114 matched.
2115 </P>
2116 <br><b>
2117 Checking for pattern recursion
2118 </b><br>
2119 <P>
2120 If the condition is the string (R), and there is no subpattern with the name R,
2121 the condition is true if a recursive call to the whole pattern or any
2122 subpattern has been made. If digits or a name preceded by ampersand follow the
2123 letter R, for example:
2124 <pre>
2125 (?(R3)...) or (?(R&name)...)
2126 </pre>
2127 the condition is true if the most recent recursion is into a subpattern whose
2128 number or name is given. This condition does not check the entire recursion
2129 stack. If the name used in a condition of this kind is a duplicate, the test is
2130 applied to all subpatterns of the same name, and is true if any one of them is
2131 the most recent recursion.
2132 </P>
2133 <P>
2134 At "top level", all these recursion test conditions are false.
2135 <a href="#recursion">The syntax for recursive patterns</a>
2136 is described below.
2137 <a name="subdefine"></a></P>
2138 <br><b>
2139 Defining subpatterns for use by reference only
2140 </b><br>
2141 <P>
2142 If the condition is the string (DEFINE), and there is no subpattern with the
2143 name DEFINE, the condition is always false. In this case, there may be only one
2144 alternative in the subpattern. It is always skipped if control reaches this
2145 point in the pattern; the idea of DEFINE is that it can be used to define
2146 subroutines that can be referenced from elsewhere. (The use of
2147 <a href="#subpatternsassubroutines">subroutines</a>
2148 is described below.) For example, a pattern to match an IPv4 address such as
2149 "" could be written like this (ignore whitespace and line
2150 breaks):
2151 <pre>
2152 (?(DEFINE) (?&#60;byte&#62; 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
2153 \b (?&byte) (\.(?&byte)){3} \b
2154 </pre>
2155 The first part of the pattern is a DEFINE group inside which a another group
2156 named "byte" is defined. This matches an individual component of an IPv4
2157 address (a number less than 256). When matching takes place, this part of the
2158 pattern is skipped because DEFINE acts like a false condition. The rest of the
2159 pattern uses references to the named group to match the four dot-separated
2160 components of an IPv4 address, insisting on a word boundary at each end.
2161 </P>
2162 <br><b>
2163 Assertion conditions
2164 </b><br>
2165 <P>
2166 If the condition is not in any of the above formats, it must be an assertion.
2167 This may be a positive or negative lookahead or lookbehind assertion. Consider
2168 this pattern, again containing non-significant white space, and with the two
2169 alternatives on the second line:
2170 <pre>
2171 (?(?=[^a-z]*[a-z])
2172 \d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} )
2173 </pre>
2174 The condition is a positive lookahead assertion that matches an optional
2175 sequence of non-letters followed by a letter. In other words, it tests for the
2176 presence of at least one letter in the subject. If a letter is found, the
2177 subject is matched against the first alternative; otherwise it is matched
2178 against the second. This pattern matches strings in one of the two forms
2179 dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
2180 <a name="comments"></a></P>
2181 <br><a name="SEC20" href="#TOC1">COMMENTS</a><br>
2182 <P>
2183 There are two ways of including comments in patterns that are processed by
2184 PCRE. In both cases, the start of the comment must not be in a character class,
2185 nor in the middle of any other sequence of related characters such as (?: or a
2186 subpattern name or number. The characters that make up a comment play no part
2187 in the pattern matching.
2188 </P>
2189 <P>
2190 The sequence (?# marks the start of a comment that continues up to the next
2191 closing parenthesis. Nested parentheses are not permitted. If the PCRE_EXTENDED
2192 option is set, an unescaped # character also introduces a comment, which in
2193 this case continues to immediately after the next newline character or
2194 character sequence in the pattern. Which characters are interpreted as newlines
2195 is controlled by the options passed to <b>pcre_compile()</b> or by a special
2196 sequence at the start of the pattern, as described in the section entitled
2197 <a href="#newlines">"Newline conventions"</a>
2198 above. Note that the end of this type of comment is a literal newline sequence
2199 in the pattern; escape sequences that happen to represent a newline do not
2200 count. For example, consider this pattern when PCRE_EXTENDED is set, and the
2201 default newline convention is in force:
2202 <pre>
2203 abc #comment \n still comment
2204 </pre>
2205 On encountering the # character, <b>pcre_compile()</b> skips along, looking for
2206 a newline in the pattern. The sequence \n is still literal at this stage, so
2207 it does not terminate the comment. Only an actual character with the code value
2208 0x0a (the default newline) does so.
2209 <a name="recursion"></a></P>
2210 <br><a name="SEC21" href="#TOC1">RECURSIVE PATTERNS</a><br>
2211 <P>
2212 Consider the problem of matching a string in parentheses, allowing for
2213 unlimited nested parentheses. Without the use of recursion, the best that can
2214 be done is to use a pattern that matches up to some fixed depth of nesting. It
2215 is not possible to handle an arbitrary nesting depth.
2216 </P>
2217 <P>
2218 For some time, Perl has provided a facility that allows regular expressions to
2219 recurse (amongst other things). It does this by interpolating Perl code in the
2220 expression at run time, and the code can refer to the expression itself. A Perl
2221 pattern using code interpolation to solve the parentheses problem can be
2222 created like this:
2223 <pre>
2224 $re = qr{\( (?: (?&#62;[^()]+) | (?p{$re}) )* \)}x;
2225 </pre>
2226 The (?p{...}) item interpolates Perl code at run time, and in this case refers
2227 recursively to the pattern in which it appears.
2228 </P>
2229 <P>
2230 Obviously, PCRE cannot support the interpolation of Perl code. Instead, it
2231 supports special syntax for recursion of the entire pattern, and also for
2232 individual subpattern recursion. After its introduction in PCRE and Python,
2233 this kind of recursion was subsequently introduced into Perl at release 5.10.
2234 </P>
2235 <P>
2236 A special item that consists of (? followed by a number greater than zero and a
2237 closing parenthesis is a recursive subroutine call of the subpattern of the
2238 given number, provided that it occurs inside that subpattern. (If not, it is a
2239 <a href="#subpatternsassubroutines">non-recursive subroutine</a>
2240 call, which is described in the next section.) The special item (?R) or (?0) is
2241 a recursive call of the entire regular expression.
2242 </P>
2243 <P>
2244 This PCRE pattern solves the nested parentheses problem (assume the
2245 PCRE_EXTENDED option is set so that white space is ignored):
2246 <pre>
2247 \( ( [^()]++ | (?R) )* \)
2248 </pre>
2249 First it matches an opening parenthesis. Then it matches any number of
2250 substrings which can either be a sequence of non-parentheses, or a recursive
2251 match of the pattern itself (that is, a correctly parenthesized substring).
2252 Finally there is a closing parenthesis. Note the use of a possessive quantifier
2253 to avoid backtracking into sequences of non-parentheses.
2254 </P>
2255 <P>
2256 If this were part of a larger pattern, you would not want to recurse the entire
2257 pattern, so instead you could use this:
2258 <pre>
2259 ( \( ( [^()]++ | (?1) )* \) )
2260 </pre>
2261 We have put the pattern into parentheses, and caused the recursion to refer to
2262 them instead of the whole pattern.
2263 </P>
2264 <P>
2265 In a larger pattern, keeping track of parenthesis numbers can be tricky. This
2266 is made easier by the use of relative references. Instead of (?1) in the
2267 pattern above you can write (?-2) to refer to the second most recently opened
2268 parentheses preceding the recursion. In other words, a negative number counts
2269 capturing parentheses leftwards from the point at which it is encountered.
2270 </P>
2271 <P>
2272 It is also possible to refer to subsequently opened parentheses, by writing
2273 references such as (?+2). However, these cannot be recursive because the
2274 reference is not inside the parentheses that are referenced. They are always
2275 <a href="#subpatternsassubroutines">non-recursive subroutine</a>
2276 calls, as described in the next section.
2277 </P>
2278 <P>
2279 An alternative approach is to use named parentheses instead. The Perl syntax
2280 for this is (?&name); PCRE's earlier syntax (?P&#62;name) is also supported. We
2281 could rewrite the above example as follows:
2282 <pre>
2283 (?&#60;pn&#62; \( ( [^()]++ | (?&pn) )* \) )
2284 </pre>
2285 If there is more than one subpattern with the same name, the earliest one is
2286 used.
2287 </P>
2288 <P>
2289 This particular example pattern that we have been looking at contains nested
2290 unlimited repeats, and so the use of a possessive quantifier for matching
2291 strings of non-parentheses is important when applying the pattern to strings
2292 that do not match. For example, when this pattern is applied to
2293 <pre>
2294 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
2295 </pre>
2296 it yields "no match" quickly. However, if a possessive quantifier is not used,
2297 the match runs for a very long time indeed because there are so many different
2298 ways the + and * repeats can carve up the subject, and all have to be tested
2299 before failure can be reported.
2300 </P>
2301 <P>
2302 At the end of a match, the values of capturing parentheses are those from
2303 the outermost level. If you want to obtain intermediate values, a callout
2304 function can be used (see below and the
2305 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2306 documentation). If the pattern above is matched against
2307 <pre>
2308 (ab(cd)ef)
2309 </pre>
2310 the value for the inner capturing parentheses (numbered 2) is "ef", which is
2311 the last value taken on at the top level. If a capturing subpattern is not
2312 matched at the top level, its final captured value is unset, even if it was
2313 (temporarily) set at a deeper level during the matching process.
2314 </P>
2315 <P>
2316 If there are more than 15 capturing parentheses in a pattern, PCRE has to
2317 obtain extra memory to store data during a recursion, which it does by using
2318 <b>pcre_malloc</b>, freeing it via <b>pcre_free</b> afterwards. If no memory can
2319 be obtained, the match fails with the PCRE_ERROR_NOMEMORY error.
2320 </P>
2321 <P>
2322 Do not confuse the (?R) item with the condition (R), which tests for recursion.
2323 Consider this pattern, which matches text in angle brackets, allowing for
2324 arbitrary nesting. Only digits are allowed in nested brackets (that is, when
2325 recursing), whereas any characters are permitted at the outer level.
2326 <pre>
2327 &#60; (?: (?(R) \d++ | [^&#60;&#62;]*+) | (?R)) * &#62;
2328 </pre>
2329 In this pattern, (?(R) is the start of a conditional subpattern, with two
2330 different alternatives for the recursive and non-recursive cases. The (?R) item
2331 is the actual recursive call.
2332 <a name="recursiondifference"></a></P>
2333 <br><b>
2334 Differences in recursion processing between PCRE and Perl
2335 </b><br>
2336 <P>
2337 Recursion processing in PCRE differs from Perl in two important ways. In PCRE
2338 (like Python, but unlike Perl), a recursive subpattern call is always treated
2339 as an atomic group. That is, once it has matched some of the subject string, it
2340 is never re-entered, even if it contains untried alternatives and there is a
2341 subsequent matching failure. This can be illustrated by the following pattern,
2342 which purports to match a palindromic string that contains an odd number of
2343 characters (for example, "a", "aba", "abcba", "abcdcba"):
2344 <pre>
2345 ^(.|(.)(?1)\2)$
2346 </pre>
2347 The idea is that it either matches a single character, or two identical
2348 characters surrounding a sub-palindrome. In Perl, this pattern works; in PCRE
2349 it does not if the pattern is longer than three characters. Consider the
2350 subject string "abcba":
2351 </P>
2352 <P>
2353 At the top level, the first character is matched, but as it is not at the end
2354 of the string, the first alternative fails; the second alternative is taken
2355 and the recursion kicks in. The recursive call to subpattern 1 successfully
2356 matches the next character ("b"). (Note that the beginning and end of line
2357 tests are not part of the recursion).
2358 </P>
2359 <P>
2360 Back at the top level, the next character ("c") is compared with what
2361 subpattern 2 matched, which was "a". This fails. Because the recursion is
2362 treated as an atomic group, there are now no backtracking points, and so the
2363 entire match fails. (Perl is able, at this point, to re-enter the recursion and
2364 try the second alternative.) However, if the pattern is written with the
2365 alternatives in the other order, things are different:
2366 <pre>
2367 ^((.)(?1)\2|.)$
2368 </pre>
2369 This time, the recursing alternative is tried first, and continues to recurse
2370 until it runs out of characters, at which point the recursion fails. But this
2371 time we do have another alternative to try at the higher level. That is the big
2372 difference: in the previous case the remaining alternative is at a deeper
2373 recursion level, which PCRE cannot use.
2374 </P>
2375 <P>
2376 To change the pattern so that it matches all palindromic strings, not just
2377 those with an odd number of characters, it is tempting to change the pattern to
2378 this:
2379 <pre>
2380 ^((.)(?1)\2|.?)$
2381 </pre>
2382 Again, this works in Perl, but not in PCRE, and for the same reason. When a
2383 deeper recursion has matched a single character, it cannot be entered again in
2384 order to match an empty string. The solution is to separate the two cases, and
2385 write out the odd and even cases as alternatives at the higher level:
2386 <pre>
2387 ^(?:((.)(?1)\2|)|((.)(?3)\4|.))
2388 </pre>
2389 If you want to match typical palindromic phrases, the pattern has to ignore all
2390 non-word characters, which can be done like this:
2391 <pre>
2392 ^\W*+(?:((.)\W*+(?1)\W*+\2|)|((.)\W*+(?3)\W*+\4|\W*+.\W*+))\W*+$
2393 </pre>
2394 If run with the PCRE_CASELESS option, this pattern matches phrases such as "A
2395 man, a plan, a canal: Panama!" and it works well in both PCRE and Perl. Note
2396 the use of the possessive quantifier *+ to avoid backtracking into sequences of
2397 non-word characters. Without this, PCRE takes a great deal longer (ten times or
2398 more) to match typical phrases, and Perl takes so long that you think it has
2399 gone into a loop.
2400 </P>
2401 <P>
2402 <b>WARNING</b>: The palindrome-matching patterns above work only if the subject
2403 string does not start with a palindrome that is shorter than the entire string.
2404 For example, although "abcba" is correctly matched, if the subject is "ababa",
2405 PCRE finds the palindrome "aba" at the start, then fails at top level because
2406 the end of the string does not follow. Once again, it cannot jump back into the
2407 recursion to try other alternatives, so the entire match fails.
2408 </P>
2409 <P>
2410 The second way in which PCRE and Perl differ in their recursion processing is
2411 in the handling of captured values. In Perl, when a subpattern is called
2412 recursively or as a subpattern (see the next section), it has no access to any
2413 values that were captured outside the recursion, whereas in PCRE these values
2414 can be referenced. Consider this pattern:
2415 <pre>
2416 ^(.)(\1|a(?2))
2417 </pre>
2418 In PCRE, this pattern matches "bab". The first capturing parentheses match "b",
2419 then in the second group, when the back reference \1 fails to match "b", the
2420 second alternative matches "a" and then recurses. In the recursion, \1 does
2421 now match "b" and so the whole match succeeds. In Perl, the pattern fails to
2422 match because inside the recursive call \1 cannot access the externally set
2423 value.
2424 <a name="subpatternsassubroutines"></a></P>
2425 <br><a name="SEC22" href="#TOC1">SUBPATTERNS AS SUBROUTINES</a><br>
2426 <P>
2427 If the syntax for a recursive subpattern call (either by number or by
2428 name) is used outside the parentheses to which it refers, it operates like a
2429 subroutine in a programming language. The called subpattern may be defined
2430 before or after the reference. A numbered reference can be absolute or
2431 relative, as in these examples:
2432 <pre>
2433 (...(absolute)...)...(?2)...
2434 (...(relative)...)...(?-1)...
2435 (...(?+1)...(relative)...
2436 </pre>
2437 An earlier example pointed out that the pattern
2438 <pre>
2439 (sens|respons)e and \1ibility
2440 </pre>
2441 matches "sense and sensibility" and "response and responsibility", but not
2442 "sense and responsibility". If instead the pattern
2443 <pre>
2444 (sens|respons)e and (?1)ibility
2445 </pre>
2446 is used, it does match "sense and responsibility" as well as the other two
2447 strings. Another example is given in the discussion of DEFINE above.
2448 </P>
2449 <P>
2450 All subroutine calls, whether recursive or not, are always treated as atomic
2451 groups. That is, once a subroutine has matched some of the subject string, it
2452 is never re-entered, even if it contains untried alternatives and there is a
2453 subsequent matching failure. Any capturing parentheses that are set during the
2454 subroutine call revert to their previous values afterwards.
2455 </P>
2456 <P>
2457 Processing options such as case-independence are fixed when a subpattern is
2458 defined, so if it is used as a subroutine, such options cannot be changed for
2459 different calls. For example, consider this pattern:
2460 <pre>
2461 (abc)(?i:(?-1))
2462 </pre>
2463 It matches "abcabc". It does not match "abcABC" because the change of
2464 processing option does not affect the called subpattern.
2465 <a name="onigurumasubroutines"></a></P>
2466 <br><a name="SEC23" href="#TOC1">ONIGURUMA SUBROUTINE SYNTAX</a><br>
2467 <P>
2468 For compatibility with Oniguruma, the non-Perl syntax \g followed by a name or
2469 a number enclosed either in angle brackets or single quotes, is an alternative
2470 syntax for referencing a subpattern as a subroutine, possibly recursively. Here
2471 are two of the examples used above, rewritten using this syntax:
2472 <pre>
2473 (?&#60;pn&#62; \( ( (?&#62;[^()]+) | \g&#60;pn&#62; )* \) )
2474 (sens|respons)e and \g'1'ibility
2475 </pre>
2476 PCRE supports an extension to Oniguruma: if a number is preceded by a
2477 plus or a minus sign it is taken as a relative reference. For example:
2478 <pre>
2479 (abc)(?i:\g&#60;-1&#62;)
2480 </pre>
2481 Note that \g{...} (Perl syntax) and \g&#60;...&#62; (Oniguruma syntax) are <i>not</i>
2482 synonymous. The former is a back reference; the latter is a subroutine call.
2483 </P>
2484 <br><a name="SEC24" href="#TOC1">CALLOUTS</a><br>
2485 <P>
2486 Perl has a feature whereby using the sequence (?{...}) causes arbitrary Perl
2487 code to be obeyed in the middle of matching a regular expression. This makes it
2488 possible, amongst other things, to extract different substrings that match the
2489 same pair of parentheses when there is a repetition.
2490 </P>
2491 <P>
2492 PCRE provides a similar feature, but of course it cannot obey arbitrary Perl
2493 code. The feature is called "callout". The caller of PCRE provides an external
2494 function by putting its entry point in the global variable <i>pcre_callout</i>.
2495 By default, this variable contains NULL, which disables all calling out.
2496 </P>
2497 <P>
2498 Within a regular expression, (?C) indicates the points at which the external
2499 function is to be called. If you want to identify different callout points, you
2500 can put a number less than 256 after the letter C. The default value is zero.
2501 For example, this pattern has two callout points:
2502 <pre>
2503 (?C1)abc(?C2)def
2504 </pre>
2505 If the PCRE_AUTO_CALLOUT flag is passed to <b>pcre_compile()</b>, callouts are
2506 automatically installed before each item in the pattern. They are all numbered
2507 255.
2508 </P>
2509 <P>
2510 During matching, when PCRE reaches a callout point (and <i>pcre_callout</i> is
2511 set), the external function is called. It is provided with the number of the
2512 callout, the position in the pattern, and, optionally, one item of data
2513 originally supplied by the caller of <b>pcre_exec()</b>. The callout function
2514 may cause matching to proceed, to backtrack, or to fail altogether. A complete
2515 description of the interface to the callout function is given in the
2516 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2517 documentation.
2518 <a name="backtrackcontrol"></a></P>
2519 <br><a name="SEC25" href="#TOC1">BACKTRACKING CONTROL</a><br>
2520 <P>
2521 Perl 5.10 introduced a number of "Special Backtracking Control Verbs", which
2522 are described in the Perl documentation as "experimental and subject to change
2523 or removal in a future version of Perl". It goes on to say: "Their usage in
2524 production code should be noted to avoid problems during upgrades." The same
2525 remarks apply to the PCRE features described in this section.
2526 </P>
2527 <P>
2528 Since these verbs are specifically related to backtracking, most of them can be
2529 used only when the pattern is to be matched using <b>pcre_exec()</b>, which uses
2530 a backtracking algorithm. With the exception of (*FAIL), which behaves like a
2531 failing negative assertion, they cause an error if encountered by
2532 <b>pcre_dfa_exec()</b>.
2533 </P>
2534 <P>
2535 If any of these verbs are used in an assertion or in a subpattern that is
2536 called as a subroutine (whether or not recursively), their effect is confined
2537 to that subpattern; it does not extend to the surrounding pattern, with one
2538 exception: the name from a *(MARK), (*PRUNE), or (*THEN) that is encountered in
2539 a successful positive assertion <i>is</i> passed back when a match succeeds
2540 (compare capturing parentheses in assertions). Note that such subpatterns are
2541 processed as anchored at the point where they are tested. Note also that Perl's
2542 treatment of subroutines is different in some cases.
2543 </P>
2544 <P>
2545 The new verbs make use of what was previously invalid syntax: an opening
2546 parenthesis followed by an asterisk. They are generally of the form
2547 (*VERB) or (*VERB:NAME). Some may take either form, with differing behaviour,
2548 depending on whether or not an argument is present. A name is any sequence of
2549 characters that does not include a closing parenthesis. If the name is empty,
2550 that is, if the closing parenthesis immediately follows the colon, the effect
2551 is as if the colon were not there. Any number of these verbs may occur in a
2552 pattern.
2553 </P>
2554 <P>
2555 PCRE contains some optimizations that are used to speed up matching by running
2556 some checks at the start of each match attempt. For example, it may know the
2557 minimum length of matching subject, or that a particular character must be
2558 present. When one of these optimizations suppresses the running of a match, any
2559 included backtracking verbs will not, of course, be processed. You can suppress
2560 the start-of-match optimizations by setting the PCRE_NO_START_OPTIMIZE option
2561 when calling <b>pcre_compile()</b> or <b>pcre_exec()</b>, or by starting the
2562 pattern with (*NO_START_OPT).
2563 </P>
2564 <P>
2565 Experiments with Perl suggest that it too has similar optimizations, sometimes
2566 leading to anomalous results.
2567 </P>
2568 <br><b>
2569 Verbs that act immediately
2570 </b><br>
2571 <P>
2572 The following verbs act as soon as they are encountered. They may not be
2573 followed by a name.
2574 <pre>
2575 (*ACCEPT)
2576 </pre>
2577 This verb causes the match to end successfully, skipping the remainder of the
2578 pattern. However, when it is inside a subpattern that is called as a
2579 subroutine, only that subpattern is ended successfully. Matching then continues
2580 at the outer level. If (*ACCEPT) is inside capturing parentheses, the data so
2581 far is captured. For example:
2582 <pre>
2583 A((?:A|B(*ACCEPT)|C)D)
2584 </pre>
2585 This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is captured by
2586 the outer parentheses.
2587 <pre>
2588 (*FAIL) or (*F)
2589 </pre>
2590 This verb causes a matching failure, forcing backtracking to occur. It is
2591 equivalent to (?!) but easier to read. The Perl documentation notes that it is
2592 probably useful only when combined with (?{}) or (??{}). Those are, of course,
2593 Perl features that are not present in PCRE. The nearest equivalent is the
2594 callout feature, as for example in this pattern:
2595 <pre>
2596 a+(?C)(*FAIL)
2597 </pre>
2598 A match with the string "aaaa" always fails, but the callout is taken before
2599 each backtrack happens (in this example, 10 times).
2600 </P>
2601 <br><b>
2602 Recording which path was taken
2603 </b><br>
2604 <P>
2605 There is one verb whose main purpose is to track how a match was arrived at,
2606 though it also has a secondary use in conjunction with advancing the match
2607 starting point (see (*SKIP) below).
2608 <pre>
2609 (*MARK:NAME) or (*:NAME)
2610 </pre>
2611 A name is always required with this verb. There may be as many instances of
2612 (*MARK) as you like in a pattern, and their names do not have to be unique.
2613 </P>
2614 <P>
2615 When a match succeeds, the name of the last-encountered (*MARK) on the matching
2616 path is passed back to the caller via the <i>pcre_extra</i> data structure, as
2617 described in the
2618 <a href="pcreapi.html#extradata">section on <i>pcre_extra</i></a>
2619 in the
2620 <a href="pcreapi.html"><b>pcreapi</b></a>
2621 documentation. Here is an example of <b>pcretest</b> output, where the /K
2622 modifier requests the retrieval and outputting of (*MARK) data:
2623 <pre>
2624 re&#62; /X(*MARK:A)Y|X(*MARK:B)Z/K
2625 data&#62; XY
2626 0: XY
2627 MK: A
2628 XZ
2629 0: XZ
2630 MK: B
2631 </pre>
2632 The (*MARK) name is tagged with "MK:" in this output, and in this example it
2633 indicates which of the two alternatives matched. This is a more efficient way
2634 of obtaining this information than putting each alternative in its own
2635 capturing parentheses.
2636 </P>
2637 <P>
2638 If (*MARK) is encountered in a positive assertion, its name is recorded and
2639 passed back if it is the last-encountered. This does not happen for negative
2640 assertions.
2641 </P>
2642 <P>
2643 After a partial match or a failed match, the name of the last encountered
2644 (*MARK) in the entire match process is returned. For example:
2645 <pre>
2646 re&#62; /X(*MARK:A)Y|X(*MARK:B)Z/K
2647 data&#62; XP
2648 No match, mark = B
2649 </pre>
2650 Note that in this unanchored example the mark is retained from the match
2651 attempt that started at the letter "X". Subsequent match attempts starting at
2652 "P" and then with an empty string do not get as far as the (*MARK) item, but
2653 nevertheless do not reset it.
2654 </P>
2655 <br><b>
2656 Verbs that act after backtracking
2657 </b><br>
2658 <P>
2659 The following verbs do nothing when they are encountered. Matching continues
2660 with what follows, but if there is no subsequent match, causing a backtrack to
2661 the verb, a failure is forced. That is, backtracking cannot pass to the left of
2662 the verb. However, when one of these verbs appears inside an atomic group, its
2663 effect is confined to that group, because once the group has been matched,
2664 there is never any backtracking into it. In this situation, backtracking can
2665 "jump back" to the left of the entire atomic group. (Remember also, as stated
2666 above, that this localization also applies in subroutine calls and assertions.)
2667 </P>
2668 <P>
2669 These verbs differ in exactly what kind of failure occurs when backtracking
2670 reaches them.
2671 <pre>
2672 (*COMMIT)
2673 </pre>
2674 This verb, which may not be followed by a name, causes the whole match to fail
2675 outright if the rest of the pattern does not match. Even if the pattern is
2676 unanchored, no further attempts to find a match by advancing the starting point
2677 take place. Once (*COMMIT) has been passed, <b>pcre_exec()</b> is committed to
2678 finding a match at the current starting point, or not at all. For example:
2679 <pre>
2680 a+(*COMMIT)b
2681 </pre>
2682 This matches "xxaab" but not "aacaab". It can be thought of as a kind of
2683 dynamic anchor, or "I've started, so I must finish." The name of the most
2684 recently passed (*MARK) in the path is passed back when (*COMMIT) forces a
2685 match failure.
2686 </P>
2687 <P>
2688 Note that (*COMMIT) at the start of a pattern is not the same as an anchor,
2689 unless PCRE's start-of-match optimizations are turned off, as shown in this
2690 <b>pcretest</b> example:
2691 <pre>
2692 re&#62; /(*COMMIT)abc/
2693 data&#62; xyzabc
2694 0: abc
2695 xyzabc\Y
2696 No match
2697 </pre>
2698 PCRE knows that any match must start with "a", so the optimization skips along
2699 the subject to "a" before running the first match attempt, which succeeds. When
2700 the optimization is disabled by the \Y escape in the second subject, the match
2701 starts at "x" and so the (*COMMIT) causes it to fail without trying any other
2702 starting points.
2703 <pre>
2704 (*PRUNE) or (*PRUNE:NAME)
2705 </pre>
2706 This verb causes the match to fail at the current starting position in the
2707 subject if the rest of the pattern does not match. If the pattern is
2708 unanchored, the normal "bumpalong" advance to the next starting character then
2709 happens. Backtracking can occur as usual to the left of (*PRUNE), before it is
2710 reached, or when matching to the right of (*PRUNE), but if there is no match to
2711 the right, backtracking cannot cross (*PRUNE). In simple cases, the use of
2712 (*PRUNE) is just an alternative to an atomic group or possessive quantifier,
2713 but there are some uses of (*PRUNE) that cannot be expressed in any other way.
2714 The behaviour of (*PRUNE:NAME) is the same as (*MARK:NAME)(*PRUNE). In an
2715 anchored pattern (*PRUNE) has the same effect as (*COMMIT).
2716 <pre>
2717 (*SKIP)
2718 </pre>
2719 This verb, when given without a name, is like (*PRUNE), except that if the
2720 pattern is unanchored, the "bumpalong" advance is not to the next character,
2721 but to the position in the subject where (*SKIP) was encountered. (*SKIP)
2722 signifies that whatever text was matched leading up to it cannot be part of a
2723 successful match. Consider:
2724 <pre>
2725 a+(*SKIP)b
2726 </pre>
2727 If the subject is "aaaac...", after the first match attempt fails (starting at
2728 the first character in the string), the starting point skips on to start the
2729 next attempt at "c". Note that a possessive quantifer does not have the same
2730 effect as this example; although it would suppress backtracking during the
2731 first match attempt, the second attempt would start at the second character
2732 instead of skipping on to "c".
2733 <pre>
2734 (*SKIP:NAME)
2735 </pre>
2736 When (*SKIP) has an associated name, its behaviour is modified. If the
2737 following pattern fails to match, the previous path through the pattern is
2738 searched for the most recent (*MARK) that has the same name. If one is found,
2739 the "bumpalong" advance is to the subject position that corresponds to that
2740 (*MARK) instead of to where (*SKIP) was encountered. If no (*MARK) with a
2741 matching name is found, the (*SKIP) is ignored.
2742 <pre>
2743 (*THEN) or (*THEN:NAME)
2744 </pre>
2745 This verb causes a skip to the next innermost alternative if the rest of the
2746 pattern does not match. That is, it cancels pending backtracking, but only
2747 within the current alternative. Its name comes from the observation that it can
2748 be used for a pattern-based if-then-else block:
2749 <pre>
2750 ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
2751 </pre>
2752 If the COND1 pattern matches, FOO is tried (and possibly further items after
2753 the end of the group if FOO succeeds); on failure, the matcher skips to the
2754 second alternative and tries COND2, without backtracking into COND1. The
2755 behaviour of (*THEN:NAME) is exactly the same as (*MARK:NAME)(*THEN).
2756 If (*THEN) is not inside an alternation, it acts like (*PRUNE).
2757 </P>
2758 <P>
2759 Note that a subpattern that does not contain a | character is just a part of
2760 the enclosing alternative; it is not a nested alternation with only one
2761 alternative. The effect of (*THEN) extends beyond such a subpattern to the
2762 enclosing alternative. Consider this pattern, where A, B, etc. are complex
2763 pattern fragments that do not contain any | characters at this level:
2764 <pre>
2765 A (B(*THEN)C) | D
2766 </pre>
2767 If A and B are matched, but there is a failure in C, matching does not
2768 backtrack into A; instead it moves to the next alternative, that is, D.
2769 However, if the subpattern containing (*THEN) is given an alternative, it
2770 behaves differently:
2771 <pre>
2772 A (B(*THEN)C | (*FAIL)) | D
2773 </pre>
2774 The effect of (*THEN) is now confined to the inner subpattern. After a failure
2775 in C, matching moves to (*FAIL), which causes the whole subpattern to fail
2776 because there are no more alternatives to try. In this case, matching does now
2777 backtrack into A.
2778 </P>
2779 <P>
2780 Note also that a conditional subpattern is not considered as having two
2781 alternatives, because only one is ever used. In other words, the | character in
2782 a conditional subpattern has a different meaning. Ignoring white space,
2783 consider:
2784 <pre>
2785 ^.*? (?(?=a) a | b(*THEN)c )
2786 </pre>
2787 If the subject is "ba", this pattern does not match. Because .*? is ungreedy,
2788 it initially matches zero characters. The condition (?=a) then fails, the
2789 character "b" is matched, but "c" is not. At this point, matching does not
2790 backtrack to .*? as might perhaps be expected from the presence of the |
2791 character. The conditional subpattern is part of the single alternative that
2792 comprises the whole pattern, and so the match fails. (If there was a backtrack
2793 into .*?, allowing it to match "b", the match would succeed.)
2794 </P>
2795 <P>
2796 The verbs just described provide four different "strengths" of control when
2797 subsequent matching fails. (*THEN) is the weakest, carrying on the match at the
2798 next alternative. (*PRUNE) comes next, failing the match at the current
2799 starting position, but allowing an advance to the next character (for an
2800 unanchored pattern). (*SKIP) is similar, except that the advance may be more
2801 than one character. (*COMMIT) is the strongest, causing the entire match to
2802 fail.
2803 </P>
2804 <P>
2805 If more than one such verb is present in a pattern, the "strongest" one wins.
2806 For example, consider this pattern, where A, B, etc. are complex pattern
2807 fragments:
2808 <pre>
2809 (A(*COMMIT)B(*THEN)C|D)
2810 </pre>
2811 Once A has matched, PCRE is committed to this match, at the current starting
2812 position. If subsequently B matches, but C does not, the normal (*THEN) action
2813 of trying the next alternative (that is, D) does not happen because (*COMMIT)
2814 overrides.
2815 </P>
2816 <br><a name="SEC26" href="#TOC1">SEE ALSO</a><br>
2817 <P>
2818 <b>pcreapi</b>(3), <b>pcrecallout</b>(3), <b>pcrematching</b>(3),
2819 <b>pcresyntax</b>(3), <b>pcre</b>(3).
2820 </P>
2821 <br><a name="SEC27" href="#TOC1">AUTHOR</a><br>
2822 <P>
2823 Philip Hazel
2824 <br>
2825 University Computing Service
2826 <br>
2827 Cambridge CB2 3QH, England.
2828 <br>
2829 </P>
2830 <br><a name="SEC28" href="#TOC1">REVISION</a><br>
2831 <P>
2832 Last updated: 29 November 2011
2833 <br>
2834 Copyright &copy; 1997-2011 University of Cambridge.
2835 <br>
2836 <p>
2837 Return to the <a href="index.html">PCRE index page</a>.
2838 </p>


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

  ViewVC Help
Powered by ViewVC 1.1.5