ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 453 - (hide annotations) (download) (as text)
Fri Sep 18 19:12:35 2009 UTC (5 years, 7 months ago) by ph10
File MIME type: text/html
File size: 99112 byte(s)
Add more explanation about recursive subpatterns, and make it possible to 
process the documenation without building a whole release.

1 nigel 63 <html>
2     <head>
3     <title>pcrepattern specification</title>
4     </head>
5     <body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
6 nigel 75 <h1>pcrepattern man page</h1>
7     <p>
8     Return to the <a href="index.html">PCRE index page</a>.
9     </p>
10 ph10 111 <p>
11 nigel 75 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 ph10 111 <br>
15 nigel 63 <ul>
16     <li><a name="TOC1" href="#SEC1">PCRE REGULAR EXPRESSION DETAILS</a>
17 ph10 227 <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)</a>
22     <li><a name="TOC7" href="#SEC7">MATCHING A SINGLE BYTE</a>
23     <li><a name="TOC8" href="#SEC8">SQUARE BRACKETS AND CHARACTER CLASSES</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 ph10 345 <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 nigel 63 </ul>
45     <br><a name="SEC1" href="#TOC1">PCRE REGULAR EXPRESSION DETAILS</a><br>
46     <P>
47 ph10 208 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 ph10 345 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 ph10 208 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 nigel 63 </P>
62     <P>
63 nigel 75 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, you must
65     build PCRE to include UTF-8 support, and then call <b>pcre_compile()</b> with
66 ph10 416 the PCRE_UTF8 option. There is also a special sequence that can be given at the
67     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 nigel 63 <a href="pcre.html#utf8support">section on UTF-8 support</a>
76     in the main
77     <a href="pcre.html"><b>pcre</b></a>
78     page.
79     </P>
80     <P>
81 nigel 77 The remainder of this document discusses the patterns that are supported by
82     PCRE when its main matching function, <b>pcre_exec()</b>, is used.
83     From release 6.0, PCRE offers a second matching function,
84     <b>pcre_dfa_exec()</b>, which matches using a different algorithm that is not
85 ph10 172 Perl-compatible. Some of the features discussed below are not available when
86     <b>pcre_dfa_exec()</b> is used. The advantages and disadvantages of the
87     alternative function, and how it differs from the normal function, are
88     discussed in the
89 nigel 77 <a href="pcrematching.html"><b>pcrematching</b></a>
90     page.
91     </P>
92 ph10 227 <br><a name="SEC2" href="#TOC1">NEWLINE CONVENTIONS</a><br>
93 nigel 77 <P>
94 ph10 227 PCRE supports five different conventions for indicating line breaks in
95     strings: a single CR (carriage return) character, a single LF (linefeed)
96     character, the two-character sequence CRLF, any of the three preceding, or any
97     Unicode newline sequence. The
98     <a href="pcreapi.html"><b>pcreapi</b></a>
99     page has
100     <a href="pcreapi.html#newlines">further discussion</a>
101     about newlines, and shows how to set the newline convention in the
102     <i>options</i> arguments for the compiling and matching functions.
103     </P>
104     <P>
105     It is also possible to specify a newline convention by starting a pattern
106     string with one of the following five sequences:
107     <pre>
108     (*CR) carriage return
109     (*LF) linefeed
110     (*CRLF) carriage return, followed by linefeed
111     (*ANYCRLF) any of the three above
112     (*ANY) all Unicode newline sequences
113     </pre>
114     These override the default and the options given to <b>pcre_compile()</b>. For
115     example, on a Unix system where LF is the default newline sequence, the pattern
116     <pre>
117     (*CR)a.b
118     </pre>
119     changes the convention to CR. That pattern matches "a\nb" because LF is no
120     longer a newline. Note that these special settings, which are not
121     Perl-compatible, are recognized only at the very start of a pattern, and that
122 ph10 231 they must be in upper case. If more than one of them is present, the last one
123     is used.
124 ph10 227 </P>
125 ph10 231 <P>
126     The newline convention does not affect what the \R escape sequence matches. By
127     default, this is any Unicode newline sequence, for Perl compatibility. However,
128     this can be changed; see the description of \R in the section entitled
129     <a href="#newlineseq">"Newline sequences"</a>
130 ph10 247 below. A change of \R setting can be combined with a change of newline
131     convention.
132 ph10 231 </P>
133 ph10 227 <br><a name="SEC3" href="#TOC1">CHARACTERS AND METACHARACTERS</a><br>
134     <P>
135 nigel 63 A regular expression is a pattern that is matched against a subject string from
136     left to right. Most characters stand for themselves in a pattern, and match the
137     corresponding characters in the subject. As a trivial example, the pattern
138     <pre>
139     The quick brown fox
140 nigel 75 </pre>
141 nigel 77 matches a portion of a subject string that is identical to itself. When
142     caseless matching is specified (the PCRE_CASELESS option), letters are matched
143     independently of case. In UTF-8 mode, PCRE always understands the concept of
144     case for characters whose values are less than 128, so caseless matching is
145     always possible. For characters with higher values, the concept of case is
146     supported if PCRE is compiled with Unicode property support, but not otherwise.
147     If you want to use caseless matching for characters 128 and above, you must
148     ensure that PCRE is compiled with Unicode property support as well as with
149     UTF-8 support.
150     </P>
151     <P>
152     The power of regular expressions comes from the ability to include alternatives
153     and repetitions in the pattern. These are encoded in the pattern by the use of
154 nigel 75 <i>metacharacters</i>, which do not stand for themselves but instead are
155 nigel 63 interpreted in some special way.
156     </P>
157     <P>
158 nigel 75 There are two different sets of metacharacters: those that are recognized
159 nigel 63 anywhere in the pattern except within square brackets, and those that are
160 nigel 93 recognized within square brackets. Outside square brackets, the metacharacters
161     are as follows:
162 nigel 63 <pre>
163     \ general escape character with several uses
164     ^ assert start of string (or line, in multiline mode)
165     $ assert end of string (or line, in multiline mode)
166     . match any character except newline (by default)
167     [ start character class definition
168     | start of alternative branch
169     ( start subpattern
170     ) end subpattern
171     ? extends the meaning of (
172     also 0 or 1 quantifier
173     also quantifier minimizer
174     * 0 or more quantifier
175     + 1 or more quantifier
176     also "possessive quantifier"
177     { start min/max quantifier
178 nigel 75 </pre>
179 nigel 63 Part of a pattern that is in square brackets is called a "character class". In
180 nigel 75 a character class the only metacharacters are:
181 nigel 63 <pre>
182     \ general escape character
183     ^ negate the class, but only if the first character
184     - indicates character range
185 nigel 75 [ POSIX character class (only if followed by POSIX syntax)
186 nigel 63 ] terminates the character class
187 nigel 75 </pre>
188     The following sections describe the use of each of the metacharacters.
189 nigel 63 </P>
190 ph10 227 <br><a name="SEC4" href="#TOC1">BACKSLASH</a><br>
191 nigel 63 <P>
192     The backslash character has several uses. Firstly, if it is followed by a
193 nigel 91 non-alphanumeric character, it takes away any special meaning that character
194     may have. This use of backslash as an escape character applies both inside and
195 nigel 63 outside character classes.
196     </P>
197     <P>
198     For example, if you want to match a * character, you write \* in the pattern.
199     This escaping action applies whether or not the following character would
200 nigel 75 otherwise be interpreted as a metacharacter, so it is always safe to precede a
201     non-alphanumeric with backslash to specify that it stands for itself. In
202 nigel 63 particular, if you want to match a backslash, you write \\.
203     </P>
204     <P>
205     If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the
206     pattern (other than in a character class) and characters between a # outside
207 nigel 91 a character class and the next newline are ignored. An escaping backslash can
208     be used to include a whitespace or # character as part of the pattern.
209 nigel 63 </P>
210     <P>
211     If you want to remove the special meaning from a sequence of characters, you
212     can do so by putting them between \Q and \E. This is different from Perl in
213     that $ and @ are handled as literals in \Q...\E sequences in PCRE, whereas in
214     Perl, $ and @ cause variable interpolation. Note the following examples:
215     <pre>
216     Pattern PCRE matches Perl matches
217 nigel 75
218     \Qabc$xyz\E abc$xyz abc followed by the contents of $xyz
219 nigel 63 \Qabc\$xyz\E abc\$xyz abc\$xyz
220     \Qabc\E\$\Qxyz\E abc$xyz abc$xyz
221 nigel 75 </pre>
222 nigel 63 The \Q...\E sequence is recognized both inside and outside character classes.
223 nigel 75 <a name="digitsafterbackslash"></a></P>
224     <br><b>
225     Non-printing characters
226     </b><br>
227 nigel 63 <P>
228     A second use of backslash provides a way of encoding non-printing characters
229     in patterns in a visible manner. There is no restriction on the appearance of
230     non-printing characters, apart from the binary zero that terminates a pattern,
231     but when a pattern is being prepared by text editing, it is usually easier to
232     use one of the following escape sequences than the binary character it
233     represents:
234     <pre>
235     \a alarm, that is, the BEL character (hex 07)
236     \cx "control-x", where x is any character
237     \e escape (hex 1B)
238     \f formfeed (hex 0C)
239 ph10 227 \n linefeed (hex 0A)
240 nigel 63 \r carriage return (hex 0D)
241     \t tab (hex 09)
242     \ddd character with octal code ddd, or backreference
243     \xhh character with hex code hh
244 nigel 87 \x{hhh..} character with hex code hhh..
245 nigel 75 </pre>
246 nigel 63 The precise effect of \cx is as follows: if x is a lower case letter, it
247     is converted to upper case. Then bit 6 of the character (hex 40) is inverted.
248     Thus \cz becomes hex 1A, but \c{ becomes hex 3B, while \c; becomes hex
249     7B.
250     </P>
251     <P>
252     After \x, from zero to two hexadecimal digits are read (letters can be in
253 nigel 87 upper or lower case). Any number of hexadecimal digits may appear between \x{
254     and }, but the value of the character code must be less than 256 in non-UTF-8
255 ph10 211 mode, and less than 2**31 in UTF-8 mode. That is, the maximum value in
256     hexadecimal is 7FFFFFFF. Note that this is bigger than the largest Unicode code
257     point, which is 10FFFF.
258 nigel 63 </P>
259     <P>
260 ph10 211 If characters other than hexadecimal digits appear between \x{ and }, or if
261     there is no terminating }, this form of escape is not recognized. Instead, the
262     initial \x will be interpreted as a basic hexadecimal escape, with no
263     following digits, giving a character whose value is zero.
264     </P>
265     <P>
266 nigel 63 Characters whose value is less than 256 can be defined by either of the two
267 nigel 87 syntaxes for \x. There is no difference in the way they are handled. For
268     example, \xdc is exactly the same as \x{dc}.
269 nigel 63 </P>
270     <P>
271 nigel 91 After \0 up to two further octal digits are read. If there are fewer than two
272     digits, just those that are present are used. Thus the sequence \0\x\07
273     specifies two binary zeros followed by a BEL character (code value 7). Make
274     sure you supply two digits after the initial zero if the pattern character that
275     follows is itself an octal digit.
276 nigel 63 </P>
277     <P>
278     The handling of a backslash followed by a digit other than 0 is complicated.
279     Outside a character class, PCRE reads it and any following digits as a decimal
280     number. If the number is less than 10, or if there have been at least that many
281     previous capturing left parentheses in the expression, the entire sequence is
282     taken as a <i>back reference</i>. A description of how this works is given
283 nigel 75 <a href="#backreferences">later,</a>
284     following the discussion of
285     <a href="#subpattern">parenthesized subpatterns.</a>
286 nigel 63 </P>
287     <P>
288     Inside a character class, or if the decimal number is greater than 9 and there
289     have not been that many capturing subpatterns, PCRE re-reads up to three octal
290 nigel 93 digits following the backslash, and uses them to generate a data character. Any
291 nigel 91 subsequent digits stand for themselves. In non-UTF-8 mode, the value of a
292     character specified in octal must be less than \400. In UTF-8 mode, values up
293     to \777 are permitted. For example:
294 nigel 63 <pre>
295     \040 is another way of writing a space
296 nigel 75 \40 is the same, provided there are fewer than 40 previous capturing subpatterns
297 nigel 63 \7 is always a back reference
298 nigel 75 \11 might be a back reference, or another way of writing a tab
299 nigel 63 \011 is always a tab
300     \0113 is a tab followed by the character "3"
301 nigel 75 \113 might be a back reference, otherwise the character with octal code 113
302     \377 might be a back reference, otherwise the byte consisting entirely of 1 bits
303     \81 is either a back reference, or a binary zero followed by the two characters "8" and "1"
304     </pre>
305 nigel 63 Note that octal values of 100 or greater must not be introduced by a leading
306     zero, because no more than three octal digits are ever read.
307     </P>
308     <P>
309 nigel 91 All the sequences that define a single character value can be used both inside
310     and outside character classes. In addition, inside a character class, the
311     sequence \b is interpreted as the backspace character (hex 08), and the
312 nigel 93 sequences \R and \X are interpreted as the characters "R" and "X",
313     respectively. Outside a character class, these sequences have different
314     meanings
315 nigel 75 <a href="#uniextseq">(see below).</a>
316 nigel 63 </P>
317 nigel 75 <br><b>
318 nigel 93 Absolute and relative back references
319     </b><br>
320     <P>
321 ph10 208 The sequence \g followed by an unsigned or a negative number, optionally
322     enclosed in braces, is an absolute or relative back reference. A named back
323     reference can be coded as \g{name}. Back references are discussed
324 nigel 93 <a href="#backreferences">later,</a>
325     following the discussion of
326     <a href="#subpattern">parenthesized subpatterns.</a>
327     </P>
328     <br><b>
329 ph10 345 Absolute and relative subroutine calls
330     </b><br>
331     <P>
332     For compatibility with Oniguruma, the non-Perl syntax \g followed by a name or
333     a number enclosed either in angle brackets or single quotes, is an alternative
334     syntax for referencing a subpattern as a "subroutine". Details are discussed
335     <a href="#onigurumasubroutines">later.</a>
336     Note that \g{...} (Perl syntax) and \g&#60;...&#62; (Oniguruma syntax) are <i>not</i>
337     synonymous. The former is a back reference; the latter is a subroutine call.
338     </P>
339     <br><b>
340 nigel 75 Generic character types
341     </b><br>
342 nigel 63 <P>
343 nigel 93 Another use of backslash is for specifying generic character types. The
344 nigel 75 following are always recognized:
345 nigel 63 <pre>
346     \d any decimal digit
347     \D any character that is not a decimal digit
348 ph10 182 \h any horizontal whitespace character
349     \H any character that is not a horizontal whitespace character
350 nigel 63 \s any whitespace character
351     \S any character that is not a whitespace character
352 ph10 182 \v any vertical whitespace character
353     \V any character that is not a vertical whitespace character
354 nigel 63 \w any "word" character
355     \W any "non-word" character
356 nigel 75 </pre>
357 nigel 63 Each pair of escape sequences partitions the complete set of characters into
358     two disjoint sets. Any given character matches one, and only one, of each pair.
359     </P>
360     <P>
361 nigel 75 These character type sequences can appear both inside and outside character
362     classes. They each match one character of the appropriate type. If the current
363     matching point is at the end of the subject string, all of them fail, since
364     there is no character to match.
365 nigel 63 </P>
366     <P>
367     For compatibility with Perl, \s does not match the VT character (code 11).
368     This makes it different from the the POSIX "space" class. The \s characters
369 ph10 182 are HT (9), LF (10), FF (12), CR (13), and space (32). If "use locale;" is
370 nigel 91 included in a Perl script, \s may match the VT character. In PCRE, it never
371 ph10 182 does.
372 nigel 63 </P>
373     <P>
374 ph10 182 In UTF-8 mode, characters with values greater than 128 never match \d, \s, or
375     \w, and always match \D, \S, and \W. This is true even when Unicode
376     character property support is available. These sequences retain their original
377     meanings from before UTF-8 support was available, mainly for efficiency
378 ph10 406 reasons. Note that this also affects \b, because it is defined in terms of \w
379     and \W.
380 ph10 182 </P>
381     <P>
382     The sequences \h, \H, \v, and \V are Perl 5.10 features. In contrast to the
383     other sequences, these do match certain high-valued codepoints in UTF-8 mode.
384     The horizontal space characters are:
385     <pre>
386     U+0009 Horizontal tab
387     U+0020 Space
388     U+00A0 Non-break space
389     U+1680 Ogham space mark
390     U+180E Mongolian vowel separator
391     U+2000 En quad
392     U+2001 Em quad
393     U+2002 En space
394     U+2003 Em space
395     U+2004 Three-per-em space
396     U+2005 Four-per-em space
397     U+2006 Six-per-em space
398     U+2007 Figure space
399     U+2008 Punctuation space
400     U+2009 Thin space
401     U+200A Hair space
402     U+202F Narrow no-break space
403     U+205F Medium mathematical space
404     U+3000 Ideographic space
405     </pre>
406     The vertical space characters are:
407     <pre>
408     U+000A Linefeed
409     U+000B Vertical tab
410     U+000C Formfeed
411     U+000D Carriage return
412     U+0085 Next line
413     U+2028 Line separator
414     U+2029 Paragraph separator
415     </PRE>
416     </P>
417     <P>
418 nigel 75 A "word" character is an underscore or any character less than 256 that is a
419     letter or digit. The definition of letters and digits is controlled by PCRE's
420     low-valued character tables, and may vary if locale-specific matching is taking
421     place (see
422 nigel 63 <a href="pcreapi.html#localesupport">"Locale support"</a>
423     in the
424     <a href="pcreapi.html"><b>pcreapi</b></a>
425 ph10 142 page). For example, in a French locale such as "fr_FR" in Unix-like systems,
426     or "french" in Windows, some character codes greater than 128 are used for
427 ph10 182 accented letters, and these are matched by \w. The use of locales with Unicode
428     is discouraged.
429 ph10 231 <a name="newlineseq"></a></P>
430 nigel 93 <br><b>
431     Newline sequences
432     </b><br>
433     <P>
434 ph10 231 Outside a character class, by default, the escape sequence \R matches any
435     Unicode newline sequence. This is a Perl 5.10 feature. In non-UTF-8 mode \R is
436     equivalent to the following:
437 nigel 93 <pre>
438     (?&#62;\r\n|\n|\x0b|\f|\r|\x85)
439     </pre>
440     This is an example of an "atomic group", details of which are given
441     <a href="#atomicgroup">below.</a>
442     This particular group matches either the two-character sequence CR followed by
443     LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab,
444     U+000B), FF (formfeed, U+000C), CR (carriage return, U+000D), or NEL (next
445     line, U+0085). The two-character sequence is treated as a single unit that
446     cannot be split.
447     </P>
448     <P>
449     In UTF-8 mode, two additional characters whose codepoints are greater than 255
450     are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029).
451     Unicode character property support is not needed for these characters to be
452     recognized.
453     </P>
454     <P>
455 ph10 231 It is possible to restrict \R to match only CR, LF, or CRLF (instead of the
456     complete set of Unicode line endings) by setting the option PCRE_BSR_ANYCRLF
457 ph10 247 either at compile time or when the pattern is matched. (BSR is an abbrevation
458     for "backslash R".) This can be made the default when PCRE is built; if this is
459     the case, the other behaviour can be requested via the PCRE_BSR_UNICODE option.
460     It is also possible to specify these settings by starting a pattern string with
461     one of the following sequences:
462 ph10 231 <pre>
463     (*BSR_ANYCRLF) CR, LF, or CRLF only
464     (*BSR_UNICODE) any Unicode newline sequence
465     </pre>
466     These override the default and the options given to <b>pcre_compile()</b>, but
467     they can be overridden by options given to <b>pcre_exec()</b>. Note that these
468     special settings, which are not Perl-compatible, are recognized only at the
469     very start of a pattern, and that they must be in upper case. If more than one
470 ph10 247 of them is present, the last one is used. They can be combined with a change of
471     newline convention, for example, a pattern can start with:
472     <pre>
473     (*ANY)(*BSR_ANYCRLF)
474     </pre>
475 nigel 93 Inside a character class, \R matches the letter "R".
476 nigel 75 <a name="uniextseq"></a></P>
477     <br><b>
478     Unicode character properties
479     </b><br>
480     <P>
481     When PCRE is built with Unicode character property support, three additional
482 ph10 185 escape sequences that match characters with specific properties are available.
483     When not in UTF-8 mode, these sequences are of course limited to testing
484     characters whose codepoints are less than 256, but they do work in this mode.
485     The extra escape sequences are:
486 nigel 75 <pre>
487 nigel 87 \p{<i>xx</i>} a character with the <i>xx</i> property
488     \P{<i>xx</i>} a character without the <i>xx</i> property
489     \X an extended Unicode sequence
490 nigel 75 </pre>
491 nigel 87 The property names represented by <i>xx</i> above are limited to the Unicode
492     script names, the general category properties, and "Any", which matches any
493     character (including newline). Other properties such as "InMusicalSymbols" are
494     not currently supported by PCRE. Note that \P{Any} does not match any
495     characters, so always causes a match failure.
496 nigel 63 </P>
497     <P>
498 nigel 87 Sets of Unicode characters are defined as belonging to certain scripts. A
499     character from one of these sets can be matched using a script name. For
500     example:
501 nigel 75 <pre>
502 nigel 87 \p{Greek}
503     \P{Han}
504     </pre>
505     Those that are not part of an identified script are lumped together as
506     "Common". The current list of scripts is:
507     </P>
508     <P>
509     Arabic,
510     Armenian,
511 nigel 93 Balinese,
512 nigel 87 Bengali,
513     Bopomofo,
514     Braille,
515     Buginese,
516     Buhid,
517     Canadian_Aboriginal,
518     Cherokee,
519     Common,
520     Coptic,
521 nigel 93 Cuneiform,
522 nigel 87 Cypriot,
523     Cyrillic,
524     Deseret,
525     Devanagari,
526     Ethiopic,
527     Georgian,
528     Glagolitic,
529     Gothic,
530     Greek,
531     Gujarati,
532     Gurmukhi,
533     Han,
534     Hangul,
535     Hanunoo,
536     Hebrew,
537     Hiragana,
538     Inherited,
539     Kannada,
540     Katakana,
541     Kharoshthi,
542     Khmer,
543     Lao,
544     Latin,
545     Limbu,
546     Linear_B,
547     Malayalam,
548     Mongolian,
549     Myanmar,
550     New_Tai_Lue,
551 nigel 93 Nko,
552 nigel 87 Ogham,
553     Old_Italic,
554     Old_Persian,
555     Oriya,
556     Osmanya,
557 nigel 93 Phags_Pa,
558     Phoenician,
559 nigel 87 Runic,
560     Shavian,
561     Sinhala,
562     Syloti_Nagri,
563     Syriac,
564     Tagalog,
565     Tagbanwa,
566     Tai_Le,
567     Tamil,
568     Telugu,
569     Thaana,
570     Thai,
571     Tibetan,
572     Tifinagh,
573     Ugaritic,
574     Yi.
575     </P>
576     <P>
577     Each character has exactly one general category property, specified by a
578     two-letter abbreviation. For compatibility with Perl, negation can be specified
579     by including a circumflex between the opening brace and the property name. For
580     example, \p{^Lu} is the same as \P{Lu}.
581     </P>
582     <P>
583     If only one letter is specified with \p or \P, it includes all the general
584     category properties that start with that letter. In this case, in the absence
585     of negation, the curly brackets in the escape sequence are optional; these two
586     examples have the same effect:
587     <pre>
588 nigel 75 \p{L}
589     \pL
590     </pre>
591 nigel 87 The following general category property codes are supported:
592 nigel 75 <pre>
593     C Other
594     Cc Control
595     Cf Format
596     Cn Unassigned
597     Co Private use
598     Cs Surrogate
600     L Letter
601     Ll Lower case letter
602     Lm Modifier letter
603     Lo Other letter
604     Lt Title case letter
605     Lu Upper case letter
607     M Mark
608     Mc Spacing mark
609     Me Enclosing mark
610     Mn Non-spacing mark
612     N Number
613     Nd Decimal number
614     Nl Letter number
615     No Other number
617     P Punctuation
618     Pc Connector punctuation
619     Pd Dash punctuation
620     Pe Close punctuation
621     Pf Final punctuation
622     Pi Initial punctuation
623     Po Other punctuation
624     Ps Open punctuation
626     S Symbol
627     Sc Currency symbol
628     Sk Modifier symbol
629     Sm Mathematical symbol
630     So Other symbol
632     Z Separator
633     Zl Line separator
634     Zp Paragraph separator
635     Zs Space separator
636     </pre>
637 nigel 87 The special property L& is also supported: it matches a character that has
638     the Lu, Ll, or Lt property, in other words, a letter that is not classified as
639     a modifier or "other".
640 nigel 75 </P>
641     <P>
642 ph10 211 The Cs (Surrogate) property applies only to characters in the range U+D800 to
643     U+DFFF. Such characters are not valid in UTF-8 strings (see RFC 3629) and so
644     cannot be tested by PCRE, unless UTF-8 validity checking has been turned off
645     (see the discussion of PCRE_NO_UTF8_CHECK in the
646     <a href="pcreapi.html"><b>pcreapi</b></a>
647 ph10 453 page). Perl does not support the Cs property.
648 ph10 211 </P>
649     <P>
650 ph10 453 The long synonyms for property names that Perl supports (such as \p{Letter})
651 nigel 91 are not supported by PCRE, nor is it permitted to prefix any of these
652 nigel 87 properties with "Is".
653     </P>
654     <P>
655     No character that is in the Unicode table has the Cn (unassigned) property.
656     Instead, this property is assumed for any code point that is not in the
657     Unicode table.
658     </P>
659     <P>
660 nigel 75 Specifying caseless matching does not affect these escape sequences. For
661     example, \p{Lu} always matches only upper case letters.
662     </P>
663     <P>
664     The \X escape matches any number of Unicode characters that form an extended
665     Unicode sequence. \X is equivalent to
666     <pre>
667     (?&#62;\PM\pM*)
668     </pre>
669     That is, it matches a character without the "mark" property, followed by zero
670     or more characters with the "mark" property, and treats the sequence as an
671     atomic group
672     <a href="#atomicgroup">(see below).</a>
673     Characters with the "mark" property are typically accents that affect the
674 ph10 185 preceding character. None of them have codepoints less than 256, so in
675     non-UTF-8 mode \X matches any one character.
676 nigel 75 </P>
677     <P>
678     Matching characters by Unicode property is not fast, because PCRE has to search
679     a structure that contains data for over fifteen thousand characters. That is
680     why the traditional escape sequences such as \d and \w do not use Unicode
681     properties in PCRE.
682 ph10 172 <a name="resetmatchstart"></a></P>
683     <br><b>
684     Resetting the match start
685     </b><br>
686     <P>
687     The escape sequence \K, which is a Perl 5.10 feature, causes any previously
688     matched characters not to be included in the final matched sequence. For
689     example, the pattern:
690     <pre>
691     foo\Kbar
692     </pre>
693     matches "foobar", but reports that it has matched "bar". This feature is
694     similar to a lookbehind assertion
695     <a href="#lookbehind">(described below).</a>
696     However, in this case, the part of the subject before the real match does not
697     have to be of fixed length, as lookbehind assertions do. The use of \K does
698     not interfere with the setting of
699     <a href="#subpattern">captured substrings.</a>
700     For example, when the pattern
701     <pre>
702     (foo)\Kbar
703     </pre>
704     matches "foobar", the first substring is still set to "foo".
705 nigel 75 <a name="smallassertions"></a></P>
706     <br><b>
707     Simple assertions
708     </b><br>
709     <P>
710 nigel 93 The final use of backslash is for certain simple assertions. An assertion
711 nigel 63 specifies a condition that has to be met at a particular point in a match,
712     without consuming any characters from the subject string. The use of
713 nigel 75 subpatterns for more complicated assertions is described
714     <a href="#bigassertions">below.</a>
715 nigel 91 The backslashed assertions are:
716 nigel 63 <pre>
717     \b matches at a word boundary
718     \B matches when not at a word boundary
719 nigel 93 \A matches at the start of the subject
720     \Z matches at the end of the subject
721     also matches before a newline at the end of the subject
722     \z matches only at the end of the subject
723     \G matches at the first matching position in the subject
724 nigel 75 </pre>
725 nigel 63 These assertions may not appear in character classes (but note that \b has a
726     different meaning, namely the backspace character, inside a character class).
727     </P>
728     <P>
729     A word boundary is a position in the subject string where the current character
730     and the previous character do not both match \w or \W (i.e. one matches
731     \w and the other matches \W), or the start or end of the string if the
732     first or last character matches \w, respectively.
733     </P>
734     <P>
735     The \A, \Z, and \z assertions differ from the traditional circumflex and
736 nigel 75 dollar (described in the next section) in that they only ever match at the very
737     start and end of the subject string, whatever options are set. Thus, they are
738     independent of multiline mode. These three assertions are not affected by the
739     PCRE_NOTBOL or PCRE_NOTEOL options, which affect only the behaviour of the
740     circumflex and dollar metacharacters. However, if the <i>startoffset</i>
741     argument of <b>pcre_exec()</b> is non-zero, indicating that matching is to start
742     at a point other than the beginning of the subject, \A can never match. The
743 nigel 91 difference between \Z and \z is that \Z matches before a newline at the end
744     of the string as well as at the very end, whereas \z matches only at the end.
745 nigel 63 </P>
746     <P>
747     The \G assertion is true only when the current matching position is at the
748     start point of the match, as specified by the <i>startoffset</i> argument of
749     <b>pcre_exec()</b>. It differs from \A when the value of <i>startoffset</i> is
750     non-zero. By calling <b>pcre_exec()</b> multiple times with appropriate
751     arguments, you can mimic Perl's /g option, and it is in this kind of
752     implementation where \G can be useful.
753     </P>
754     <P>
755     Note, however, that PCRE's interpretation of \G, as the start of the current
756     match, is subtly different from Perl's, which defines it as the end of the
757     previous match. In Perl, these can be different when the previously matched
758     string was empty. Because PCRE does just one match at a time, it cannot
759     reproduce this behaviour.
760     </P>
761     <P>
762     If all the alternatives of a pattern begin with \G, the expression is anchored
763     to the starting match position, and the "anchored" flag is set in the compiled
764     regular expression.
765     </P>
766 ph10 227 <br><a name="SEC5" href="#TOC1">CIRCUMFLEX AND DOLLAR</a><br>
767 nigel 63 <P>
768     Outside a character class, in the default matching mode, the circumflex
769 nigel 75 character is an assertion that is true only if the current matching point is
770 nigel 63 at the start of the subject string. If the <i>startoffset</i> argument of
771     <b>pcre_exec()</b> is non-zero, circumflex can never match if the PCRE_MULTILINE
772     option is unset. Inside a character class, circumflex has an entirely different
773 nigel 75 meaning
774     <a href="#characterclass">(see below).</a>
775 nigel 63 </P>
776     <P>
777     Circumflex need not be the first character of the pattern if a number of
778     alternatives are involved, but it should be the first thing in each alternative
779     in which it appears if the pattern is ever to match that branch. If all
780     possible alternatives start with a circumflex, that is, if the pattern is
781     constrained to match only at the start of the subject, it is said to be an
782     "anchored" pattern. (There are also other constructs that can cause a pattern
783     to be anchored.)
784     </P>
785     <P>
786 nigel 75 A dollar character is an assertion that is true only if the current matching
787 nigel 63 point is at the end of the subject string, or immediately before a newline
788 nigel 91 at the end of the string (by default). Dollar need not be the last character of
789     the pattern if a number of alternatives are involved, but it should be the last
790     item in any branch in which it appears. Dollar has no special meaning in a
791     character class.
792 nigel 63 </P>
793     <P>
794     The meaning of dollar can be changed so that it matches only at the very end of
795     the string, by setting the PCRE_DOLLAR_ENDONLY option at compile time. This
796     does not affect the \Z assertion.
797     </P>
798     <P>
799     The meanings of the circumflex and dollar characters are changed if the
800 nigel 91 PCRE_MULTILINE option is set. When this is the case, a circumflex matches
801     immediately after internal newlines as well as at the start of the subject
802     string. It does not match after a newline that ends the string. A dollar
803     matches before any newlines in the string, as well as at the very end, when
804     PCRE_MULTILINE is set. When newline is specified as the two-character
805     sequence CRLF, isolated CR and LF characters do not indicate newlines.
806 nigel 63 </P>
807     <P>
808 nigel 91 For example, the pattern /^abc$/ matches the subject string "def\nabc" (where
809     \n represents a newline) in multiline mode, but not otherwise. Consequently,
810     patterns that are anchored in single line mode because all branches start with
811     ^ are not anchored in multiline mode, and a match for circumflex is possible
812     when the <i>startoffset</i> argument of <b>pcre_exec()</b> is non-zero. The
813     PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
814     </P>
815     <P>
816 nigel 63 Note that the sequences \A, \Z, and \z can be used to match the start and
817     end of the subject in both modes, and if all branches of a pattern start with
818 nigel 91 \A it is always anchored, whether or not PCRE_MULTILINE is set.
819 nigel 63 </P>
820 ph10 227 <br><a name="SEC6" href="#TOC1">FULL STOP (PERIOD, DOT)</a><br>
821 nigel 63 <P>
822     Outside a character class, a dot in the pattern matches any one character in
823 nigel 91 the subject string except (by default) a character that signifies the end of a
824 nigel 93 line. In UTF-8 mode, the matched character may be more than one byte long.
825 nigel 63 </P>
826 nigel 91 <P>
827 nigel 93 When a line ending is defined as a single character, dot never matches that
828     character; when the two-character sequence CRLF is used, dot does not match CR
829     if it is immediately followed by LF, but otherwise it matches all characters
830     (including isolated CRs and LFs). When any Unicode line endings are being
831     recognized, dot does not match CR or LF or any of the other line ending
832     characters.
833     </P>
834     <P>
835 nigel 91 The behaviour of dot with regard to newlines can be changed. If the PCRE_DOTALL
836 nigel 93 option is set, a dot matches any one character, without exception. If the
837     two-character sequence CRLF is present in the subject string, it takes two dots
838     to match it.
839 nigel 91 </P>
840     <P>
841     The handling of dot is entirely independent of the handling of circumflex and
842     dollar, the only relationship being that they both involve newlines. Dot has no
843     special meaning in a character class.
844     </P>
845 ph10 227 <br><a name="SEC7" href="#TOC1">MATCHING A SINGLE BYTE</a><br>
846 nigel 63 <P>
847     Outside a character class, the escape sequence \C matches any one byte, both
848 nigel 93 in and out of UTF-8 mode. Unlike a dot, it always matches any line-ending
849     characters. The feature is provided in Perl in order to match individual bytes
850     in UTF-8 mode. Because it breaks up UTF-8 characters into individual bytes,
851     what remains in the string may be a malformed UTF-8 string. For this reason,
852     the \C escape sequence is best avoided.
853 nigel 63 </P>
854     <P>
855 nigel 75 PCRE does not allow \C to appear in lookbehind assertions
856     <a href="#lookbehind">(described below),</a>
857     because in UTF-8 mode this would make it impossible to calculate the length of
858     the lookbehind.
859     <a name="characterclass"></a></P>
860 ph10 227 <br><a name="SEC8" href="#TOC1">SQUARE BRACKETS AND CHARACTER CLASSES</a><br>
861 nigel 63 <P>
862     An opening square bracket introduces a character class, terminated by a closing
863     square bracket. A closing square bracket on its own is not special. If a
864     closing square bracket is required as a member of the class, it should be the
865     first data character in the class (after an initial circumflex, if present) or
866     escaped with a backslash.
867     </P>
868     <P>
869     A character class matches a single character in the subject. In UTF-8 mode, the
870     character may occupy more than one byte. A matched character must be in the set
871     of characters defined by the class, unless the first character in the class
872     definition is a circumflex, in which case the subject character must not be in
873     the set defined by the class. If a circumflex is actually required as a member
874     of the class, ensure it is not the first character, or escape it with a
875     backslash.
876     </P>
877     <P>
878     For example, the character class [aeiou] matches any lower case vowel, while
879     [^aeiou] matches any character that is not a lower case vowel. Note that a
880 nigel 75 circumflex is just a convenient notation for specifying the characters that
881     are in the class by enumerating those that are not. A class that starts with a
882     circumflex is not an assertion: it still consumes a character from the subject
883     string, and therefore it fails if the current pointer is at the end of the
884     string.
885 nigel 63 </P>
886     <P>
887     In UTF-8 mode, characters with values greater than 255 can be included in a
888     class as a literal string of bytes, or by using the \x{ escaping mechanism.
889     </P>
890     <P>
891     When caseless matching is set, any letters in a class represent both their
892     upper case and lower case versions, so for example, a caseless [aeiou] matches
893     "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
894 nigel 77 caseful version would. In UTF-8 mode, PCRE always understands the concept of
895     case for characters whose values are less than 128, so caseless matching is
896     always possible. For characters with higher values, the concept of case is
897     supported if PCRE is compiled with Unicode property support, but not otherwise.
898     If you want to use caseless matching for characters 128 and above, you must
899     ensure that PCRE is compiled with Unicode property support as well as with
900     UTF-8 support.
901 nigel 63 </P>
902     <P>
903 nigel 93 Characters that might indicate line breaks are never treated in any special way
904     when matching character classes, whatever line-ending sequence is in use, and
905     whatever setting of the PCRE_DOTALL and PCRE_MULTILINE options is used. A class
906     such as [^a] always matches one of these characters.
907 nigel 63 </P>
908     <P>
909     The minus (hyphen) character can be used to specify a range of characters in a
910     character class. For example, [d-m] matches any letter between d and m,
911     inclusive. If a minus character is required in a class, it must be escaped with
912     a backslash or appear in a position where it cannot be interpreted as
913     indicating a range, typically as the first or last character in the class.
914     </P>
915     <P>
916     It is not possible to have the literal character "]" as the end character of a
917     range. A pattern such as [W-]46] is interpreted as a class of two characters
918     ("W" and "-") followed by a literal string "46]", so it would match "W46]" or
919     "-46]". However, if the "]" is escaped with a backslash it is interpreted as
920 nigel 75 the end of range, so [W-\]46] is interpreted as a class containing a range
921     followed by two other characters. The octal or hexadecimal representation of
922     "]" can also be used to end a range.
923 nigel 63 </P>
924     <P>
925     Ranges operate in the collating sequence of character values. They can also be
926     used for characters specified numerically, for example [\000-\037]. In UTF-8
927     mode, ranges can include characters whose values are greater than 255, for
928     example [\x{100}-\x{2ff}].
929     </P>
930     <P>
931     If a range that includes letters is used when caseless matching is set, it
932     matches the letters in either case. For example, [W-c] is equivalent to
933 nigel 75 [][\\^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if character
934 ph10 142 tables for a French locale are in use, [\xc8-\xcb] matches accented E
935 nigel 75 characters in both cases. In UTF-8 mode, PCRE supports the concept of case for
936     characters with values greater than 128 only when it is compiled with Unicode
937     property support.
938 nigel 63 </P>
939     <P>
940 nigel 75 The character types \d, \D, \p, \P, \s, \S, \w, and \W may also appear
941     in a character class, and add the characters that they match to the class. For
942 nigel 63 example, [\dABCDEF] matches any hexadecimal digit. A circumflex can
943     conveniently be used with the upper case character types to specify a more
944     restricted set of characters than the matching lower case type. For example,
945     the class [^\W_] matches any letter or digit, but not underscore.
946     </P>
947     <P>
948 nigel 75 The only metacharacters that are recognized in character classes are backslash,
949     hyphen (only where it can be interpreted as specifying a range), circumflex
950     (only at the start), opening square bracket (only when it can be interpreted as
951     introducing a POSIX class name - see the next section), and the terminating
952     closing square bracket. However, escaping other non-alphanumeric characters
953     does no harm.
954 nigel 63 </P>
955 ph10 227 <br><a name="SEC9" href="#TOC1">POSIX CHARACTER CLASSES</a><br>
956 nigel 63 <P>
957 nigel 75 Perl supports the POSIX notation for character classes. This uses names
958 nigel 63 enclosed by [: and :] within the enclosing square brackets. PCRE also supports
959     this notation. For example,
960     <pre>
961     [01[:alpha:]%]
962 nigel 75 </pre>
963 nigel 63 matches "0", "1", any alphabetic character, or "%". The supported class names
964     are
965     <pre>
966     alnum letters and digits
967     alpha letters
968     ascii character codes 0 - 127
969     blank space or tab only
970     cntrl control characters
971     digit decimal digits (same as \d)
972     graph printing characters, excluding space
973     lower lower case letters
974     print printing characters, including space
975     punct printing characters, excluding letters and digits
976     space white space (not quite the same as \s)
977     upper upper case letters
978     word "word" characters (same as \w)
979     xdigit hexadecimal digits
980 nigel 75 </pre>
981 nigel 63 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13), and
982     space (32). Notice that this list includes the VT character (code 11). This
983     makes "space" different to \s, which does not include VT (for Perl
984     compatibility).
985     </P>
986     <P>
987     The name "word" is a Perl extension, and "blank" is a GNU extension from Perl
988     5.8. Another Perl extension is negation, which is indicated by a ^ character
989     after the colon. For example,
990     <pre>
991     [12[:^digit:]]
992 nigel 75 </pre>
993 nigel 63 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX
994     syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
995     supported, and an error is given if they are encountered.
996     </P>
997     <P>
998 nigel 75 In UTF-8 mode, characters with values greater than 128 do not match any of
999 nigel 63 the POSIX character classes.
1000     </P>
1001 ph10 227 <br><a name="SEC10" href="#TOC1">VERTICAL BAR</a><br>
1002 nigel 63 <P>
1003     Vertical bar characters are used to separate alternative patterns. For example,
1004     the pattern
1005     <pre>
1006     gilbert|sullivan
1007 nigel 75 </pre>
1008 nigel 63 matches either "gilbert" or "sullivan". Any number of alternatives may appear,
1009 nigel 91 and an empty alternative is permitted (matching the empty string). The matching
1010     process tries each alternative in turn, from left to right, and the first one
1011     that succeeds is used. If the alternatives are within a subpattern
1012 nigel 75 <a href="#subpattern">(defined below),</a>
1013     "succeeds" means matching the rest of the main pattern as well as the
1014     alternative in the subpattern.
1015 nigel 63 </P>
1016 ph10 227 <br><a name="SEC11" href="#TOC1">INTERNAL OPTION SETTING</a><br>
1017 nigel 63 <P>
1018     The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
1019 ph10 243 PCRE_EXTENDED options (which are Perl-compatible) can be changed from within
1020     the pattern by a sequence of Perl option letters enclosed between "(?" and ")".
1021     The option letters are
1022 nigel 63 <pre>
1023     i for PCRE_CASELESS
1024     m for PCRE_MULTILINE
1025     s for PCRE_DOTALL
1026     x for PCRE_EXTENDED
1027 nigel 75 </pre>
1028 nigel 63 For example, (?im) sets caseless, multiline matching. It is also possible to
1029     unset these options by preceding the letter with a hyphen, and a combined
1030     setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
1031     PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
1032     permitted. If a letter appears both before and after the hyphen, the option is
1033     unset.
1034     </P>
1035     <P>
1036 ph10 243 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA can be
1037     changed in the same way as the Perl-compatible options by using the characters
1038     J, U and X respectively.
1039     </P>
1040     <P>
1041 ph10 416 When one of these option changes occurs at top level (that is, not inside
1042     subpattern parentheses), the change applies to the remainder of the pattern
1043     that follows. If the change is placed right at the start of a pattern, PCRE
1044     extracts it into the global options (and it will therefore show up in data
1045     extracted by the <b>pcre_fullinfo()</b> function).
1046 nigel 63 </P>
1047     <P>
1048 nigel 93 An option change within a subpattern (see below for a description of
1049     subpatterns) affects only that part of the current pattern that follows it, so
1050 nigel 63 <pre>
1051     (a(?i)b)c
1052 nigel 75 </pre>
1053 nigel 63 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
1054     By this means, options can be made to have different settings in different
1055     parts of the pattern. Any changes made in one alternative do carry on
1056     into subsequent branches within the same subpattern. For example,
1057     <pre>
1058     (a(?i)b|c)
1059 nigel 75 </pre>
1060 nigel 63 matches "ab", "aB", "c", and "C", even though when matching "C" the first
1061     branch is abandoned before the option setting. This is because the effects of
1062     option settings happen at compile time. There would be some very weird
1063     behaviour otherwise.
1064 ph10 254 </P>
1065     <P>
1066     <b>Note:</b> There are other PCRE-specific options that can be set by the
1067     application when the compile or match functions are called. In some cases the
1068 ph10 416 pattern can contain special leading sequences such as (*CRLF) to override what
1069     the application has set or what has been defaulted. Details are given in the
1070     section entitled
1071 ph10 254 <a href="#newlineseq">"Newline sequences"</a>
1072 ph10 416 above. There is also the (*UTF8) leading sequence that can be used to set UTF-8
1073     mode; this is equivalent to setting the PCRE_UTF8 option.
1074 nigel 75 <a name="subpattern"></a></P>
1075 ph10 227 <br><a name="SEC12" href="#TOC1">SUBPATTERNS</a><br>
1076 nigel 63 <P>
1077     Subpatterns are delimited by parentheses (round brackets), which can be nested.
1078 nigel 75 Turning part of a pattern into a subpattern does two things:
1079     <br>
1080     <br>
1081 nigel 63 1. It localizes a set of alternatives. For example, the pattern
1082     <pre>
1083     cat(aract|erpillar|)
1084 nigel 75 </pre>
1085 nigel 63 matches one of the words "cat", "cataract", or "caterpillar". Without the
1086 nigel 93 parentheses, it would match "cataract", "erpillar" or an empty string.
1087 nigel 75 <br>
1088     <br>
1089     2. It sets up the subpattern as a capturing subpattern. This means that, when
1090     the whole pattern matches, that portion of the subject string that matched the
1091     subpattern is passed back to the caller via the <i>ovector</i> argument of
1092 nigel 63 <b>pcre_exec()</b>. Opening parentheses are counted from left to right (starting
1093 nigel 75 from 1) to obtain numbers for the capturing subpatterns.
1094 nigel 63 </P>
1095     <P>
1096     For example, if the string "the red king" is matched against the pattern
1097     <pre>
1098     the ((red|white) (king|queen))
1099 nigel 75 </pre>
1100 nigel 63 the captured substrings are "red king", "red", and "king", and are numbered 1,
1101     2, and 3, respectively.
1102     </P>
1103     <P>
1104     The fact that plain parentheses fulfil two functions is not always helpful.
1105     There are often times when a grouping subpattern is required without a
1106     capturing requirement. If an opening parenthesis is followed by a question mark
1107     and a colon, the subpattern does not do any capturing, and is not counted when
1108     computing the number of any subsequent capturing subpatterns. For example, if
1109     the string "the white queen" is matched against the pattern
1110     <pre>
1111     the ((?:red|white) (king|queen))
1112 nigel 75 </pre>
1113 nigel 63 the captured substrings are "white queen" and "queen", and are numbered 1 and
1114 nigel 93 2. The maximum number of capturing subpatterns is 65535.
1115 nigel 63 </P>
1116     <P>
1117     As a convenient shorthand, if any option settings are required at the start of
1118     a non-capturing subpattern, the option letters may appear between the "?" and
1119     the ":". Thus the two patterns
1120     <pre>
1121     (?i:saturday|sunday)
1122     (?:(?i)saturday|sunday)
1123 nigel 75 </pre>
1124 nigel 63 match exactly the same set of strings. Because alternative branches are tried
1125     from left to right, and options are not reset until the end of the subpattern
1126     is reached, an option setting in one branch does affect subsequent branches, so
1127     the above patterns match "SUNDAY" as well as "Saturday".
1128     </P>
1129 ph10 227 <br><a name="SEC13" href="#TOC1">DUPLICATE SUBPATTERN NUMBERS</a><br>
1130 nigel 63 <P>
1131 ph10 182 Perl 5.10 introduced a feature whereby each alternative in a subpattern uses
1132     the same numbers for its capturing parentheses. Such a subpattern starts with
1133     (?| and is itself a non-capturing subpattern. For example, consider this
1134     pattern:
1135     <pre>
1136     (?|(Sat)ur|(Sun))day
1137     </pre>
1138     Because the two alternatives are inside a (?| group, both sets of capturing
1139     parentheses are numbered one. Thus, when the pattern matches, you can look
1140     at captured substring number one, whichever alternative matched. This construct
1141     is useful when you want to capture part, but not all, of one of a number of
1142     alternatives. Inside a (?| group, parentheses are numbered as usual, but the
1143     number is reset at the start of each branch. The numbers of any capturing
1144     buffers that follow the subpattern start after the highest number used in any
1145     branch. The following example is taken from the Perl documentation.
1146     The numbers underneath show in which buffer the captured content will be
1147     stored.
1148     <pre>
1149     # before ---------------branch-reset----------- after
1150     / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
1151     # 1 2 2 3 2 3 4
1152     </pre>
1153     A backreference or a recursive call to a numbered subpattern always refers to
1154     the first one in the pattern with the given number.
1155     </P>
1156     <P>
1157     An alternative approach to using this "branch reset" feature is to use
1158     duplicate named subpatterns, as described in the next section.
1159     </P>
1160 ph10 227 <br><a name="SEC14" href="#TOC1">NAMED SUBPATTERNS</a><br>
1161 ph10 182 <P>
1162 nigel 63 Identifying capturing parentheses by number is simple, but it can be very hard
1163     to keep track of the numbers in complicated regular expressions. Furthermore,
1164 nigel 75 if an expression is modified, the numbers may change. To help with this
1165 nigel 93 difficulty, PCRE supports the naming of subpatterns. This feature was not
1166     added to Perl until release 5.10. Python had the feature earlier, and PCRE
1167     introduced it at release 4.0, using the Python syntax. PCRE now supports both
1168     the Perl and the Python syntax.
1169     </P>
1170     <P>
1171     In PCRE, a subpattern can be named in one of three ways: (?&#60;name&#62;...) or
1172     (?'name'...) as in Perl, or (?P&#60;name&#62;...) as in Python. References to capturing
1173 nigel 91 parentheses from other parts of the pattern, such as
1174     <a href="#backreferences">backreferences,</a>
1175     <a href="#recursion">recursion,</a>
1176     and
1177     <a href="#conditions">conditions,</a>
1178     can be made by name as well as by number.
1179 nigel 63 </P>
1180     <P>
1181 nigel 91 Names consist of up to 32 alphanumeric characters and underscores. Named
1182 nigel 93 capturing parentheses are still allocated numbers as well as names, exactly as
1183     if the names were not present. The PCRE API provides function calls for
1184     extracting the name-to-number translation table from a compiled pattern. There
1185     is also a convenience function for extracting a captured substring by name.
1186 nigel 91 </P>
1187     <P>
1188     By default, a name must be unique within a pattern, but it is possible to relax
1189     this constraint by setting the PCRE_DUPNAMES option at compile time. This can
1190     be useful for patterns where only one instance of the named parentheses can
1191     match. Suppose you want to match the name of a weekday, either as a 3-letter
1192     abbreviation or as the full name, and in both cases you want to extract the
1193     abbreviation. This pattern (ignoring the line breaks) does the job:
1194     <pre>
1195 nigel 93 (?&#60;DN&#62;Mon|Fri|Sun)(?:day)?|
1196     (?&#60;DN&#62;Tue)(?:sday)?|
1197     (?&#60;DN&#62;Wed)(?:nesday)?|
1198     (?&#60;DN&#62;Thu)(?:rsday)?|
1199     (?&#60;DN&#62;Sat)(?:urday)?
1200 nigel 91 </pre>
1201     There are five capturing substrings, but only one is ever set after a match.
1202 ph10 182 (An alternative way of solving this problem is to use a "branch reset"
1203     subpattern, as described in the previous section.)
1204     </P>
1205     <P>
1206 nigel 91 The convenience function for extracting the data by name returns the substring
1207 nigel 93 for the first (and in this example, the only) subpattern of that name that
1208 nigel 91 matched. This saves searching to find which numbered subpattern it was. If you
1209     make a reference to a non-unique named subpattern from elsewhere in the
1210     pattern, the one that corresponds to the lowest number is used. For further
1211     details of the interfaces for handling named subpatterns, see the
1212 nigel 63 <a href="pcreapi.html"><b>pcreapi</b></a>
1213     documentation.
1214     </P>
1215 ph10 392 <P>
1216     <b>Warning:</b> You cannot use different names to distinguish between two
1217     subpatterns with the same number (see the previous section) because PCRE uses
1218     only the numbers when matching.
1219     </P>
1220 ph10 227 <br><a name="SEC15" href="#TOC1">REPETITION</a><br>
1221 nigel 63 <P>
1222     Repetition is specified by quantifiers, which can follow any of the following
1223     items:
1224     <pre>
1225     a literal data character
1226 nigel 93 the dot metacharacter
1227 nigel 63 the \C escape sequence
1228 nigel 75 the \X escape sequence (in UTF-8 mode with Unicode properties)
1229 nigel 93 the \R escape sequence
1230 nigel 75 an escape such as \d that matches a single character
1231 nigel 63 a character class
1232     a back reference (see next section)
1233     a parenthesized subpattern (unless it is an assertion)
1234 nigel 75 </pre>
1235 nigel 63 The general repetition quantifier specifies a minimum and maximum number of
1236     permitted matches, by giving the two numbers in curly brackets (braces),
1237     separated by a comma. The numbers must be less than 65536, and the first must
1238     be less than or equal to the second. For example:
1239     <pre>
1240     z{2,4}
1241 nigel 75 </pre>
1242 nigel 63 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
1243     character. If the second number is omitted, but the comma is present, there is
1244     no upper limit; if the second number and the comma are both omitted, the
1245     quantifier specifies an exact number of required matches. Thus
1246     <pre>
1247     [aeiou]{3,}
1248 nigel 75 </pre>
1249 nigel 63 matches at least 3 successive vowels, but may match many more, while
1250     <pre>
1251     \d{8}
1252 nigel 75 </pre>
1253 nigel 63 matches exactly 8 digits. An opening curly bracket that appears in a position
1254     where a quantifier is not allowed, or one that does not match the syntax of a
1255     quantifier, is taken as a literal character. For example, {,6} is not a
1256     quantifier, but a literal string of four characters.
1257     </P>
1258     <P>
1259     In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to individual
1260     bytes. Thus, for example, \x{100}{2} matches two UTF-8 characters, each of
1261 nigel 75 which is represented by a two-byte sequence. Similarly, when Unicode property
1262     support is available, \X{3} matches three Unicode extended sequences, each of
1263     which may be several bytes long (and they may be of different lengths).
1264 nigel 63 </P>
1265     <P>
1266     The quantifier {0} is permitted, causing the expression to behave as if the
1267 ph10 345 previous item and the quantifier were not present. This may be useful for
1268     subpatterns that are referenced as
1269     <a href="#subpatternsassubroutines">subroutines</a>
1270     from elsewhere in the pattern. Items other than subpatterns that have a {0}
1271     quantifier are omitted from the compiled pattern.
1272 nigel 63 </P>
1273     <P>
1274 nigel 93 For convenience, the three most common quantifiers have single-character
1275     abbreviations:
1276 nigel 63 <pre>
1277     * is equivalent to {0,}
1278     + is equivalent to {1,}
1279     ? is equivalent to {0,1}
1280 nigel 75 </pre>
1281 nigel 63 It is possible to construct infinite loops by following a subpattern that can
1282     match no characters with a quantifier that has no upper limit, for example:
1283     <pre>
1284     (a?)*
1285 nigel 75 </pre>
1286 nigel 63 Earlier versions of Perl and PCRE used to give an error at compile time for
1287     such patterns. However, because there are cases where this can be useful, such
1288     patterns are now accepted, but if any repetition of the subpattern does in fact
1289     match no characters, the loop is forcibly broken.
1290     </P>
1291     <P>
1292     By default, the quantifiers are "greedy", that is, they match as much as
1293     possible (up to the maximum number of permitted times), without causing the
1294     rest of the pattern to fail. The classic example of where this gives problems
1295 nigel 75 is in trying to match comments in C programs. These appear between /* and */
1296     and within the comment, individual * and / characters may appear. An attempt to
1297     match C comments by applying the pattern
1298 nigel 63 <pre>
1299     /\*.*\*/
1300 nigel 75 </pre>
1301 nigel 63 to the string
1302     <pre>
1303 nigel 75 /* first comment */ not comment /* second comment */
1304     </pre>
1305 nigel 63 fails, because it matches the entire string owing to the greediness of the .*
1306     item.
1307     </P>
1308     <P>
1309     However, if a quantifier is followed by a question mark, it ceases to be
1310     greedy, and instead matches the minimum number of times possible, so the
1311     pattern
1312     <pre>
1313     /\*.*?\*/
1314 nigel 75 </pre>
1315 nigel 63 does the right thing with the C comments. The meaning of the various
1316     quantifiers is not otherwise changed, just the preferred number of matches.
1317     Do not confuse this use of question mark with its use as a quantifier in its
1318     own right. Because it has two uses, it can sometimes appear doubled, as in
1319     <pre>
1320     \d??\d
1321 nigel 75 </pre>
1322 nigel 63 which matches one digit by preference, but can match two if that is the only
1323     way the rest of the pattern matches.
1324     </P>
1325     <P>
1326 nigel 93 If the PCRE_UNGREEDY option is set (an option that is not available in Perl),
1327 nigel 63 the quantifiers are not greedy by default, but individual ones can be made
1328     greedy by following them with a question mark. In other words, it inverts the
1329     default behaviour.
1330     </P>
1331     <P>
1332     When a parenthesized subpattern is quantified with a minimum repeat count that
1333 nigel 75 is greater than 1 or with a limited maximum, more memory is required for the
1334 nigel 63 compiled pattern, in proportion to the size of the minimum or maximum.
1335     </P>
1336     <P>
1337     If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent
1338 nigel 93 to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is
1339 nigel 63 implicitly anchored, because whatever follows will be tried against every
1340     character position in the subject string, so there is no point in retrying the
1341     overall match at any position after the first. PCRE normally treats such a
1342     pattern as though it were preceded by \A.
1343     </P>
1344     <P>
1345     In cases where it is known that the subject string contains no newlines, it is
1346     worth setting PCRE_DOTALL in order to obtain this optimization, or
1347     alternatively using ^ to indicate anchoring explicitly.
1348     </P>
1349     <P>
1350     However, there is one situation where the optimization cannot be used. When .*
1351     is inside capturing parentheses that are the subject of a backreference
1352 nigel 93 elsewhere in the pattern, a match at the start may fail where a later one
1353     succeeds. Consider, for example:
1354 nigel 63 <pre>
1355     (.*)abc\1
1356 nigel 75 </pre>
1357 nigel 63 If the subject is "xyz123abc123" the match point is the fourth character. For
1358     this reason, such a pattern is not implicitly anchored.
1359     </P>
1360     <P>
1361     When a capturing subpattern is repeated, the value captured is the substring
1362     that matched the final iteration. For example, after
1363     <pre>
1364     (tweedle[dume]{3}\s*)+
1365 nigel 75 </pre>
1366 nigel 63 has matched "tweedledum tweedledee" the value of the captured substring is
1367     "tweedledee". However, if there are nested capturing subpatterns, the
1368     corresponding captured values may have been set in previous iterations. For
1369     example, after
1370     <pre>
1371     /(a|(b))+/
1372 nigel 75 </pre>
1373 nigel 63 matches "aba" the value of the second captured substring is "b".
1374 nigel 75 <a name="atomicgroup"></a></P>
1375 ph10 227 <br><a name="SEC16" href="#TOC1">ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS</a><br>
1376 nigel 63 <P>
1377 nigel 93 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
1378     repetition, failure of what follows normally causes the repeated item to be
1379     re-evaluated to see if a different number of repeats allows the rest of the
1380     pattern to match. Sometimes it is useful to prevent this, either to change the
1381     nature of the match, or to cause it fail earlier than it otherwise might, when
1382     the author of the pattern knows there is no point in carrying on.
1383 nigel 63 </P>
1384     <P>
1385     Consider, for example, the pattern \d+foo when applied to the subject line
1386     <pre>
1387     123456bar
1388 nigel 75 </pre>
1389 nigel 63 After matching all 6 digits and then failing to match "foo", the normal
1390     action of the matcher is to try again with only 5 digits matching the \d+
1391     item, and then with 4, and so on, before ultimately failing. "Atomic grouping"
1392     (a term taken from Jeffrey Friedl's book) provides the means for specifying
1393     that once a subpattern has matched, it is not to be re-evaluated in this way.
1394     </P>
1395     <P>
1396 nigel 93 If we use atomic grouping for the previous example, the matcher gives up
1397 nigel 63 immediately on failing to match "foo" the first time. The notation is a kind of
1398     special parenthesis, starting with (?&#62; as in this example:
1399     <pre>
1400 nigel 73 (?&#62;\d+)foo
1401 nigel 75 </pre>
1402 nigel 63 This kind of parenthesis "locks up" the part of the pattern it contains once
1403     it has matched, and a failure further into the pattern is prevented from
1404     backtracking into it. Backtracking past it to previous items, however, works as
1405     normal.
1406     </P>
1407     <P>
1408     An alternative description is that a subpattern of this type matches the string
1409     of characters that an identical standalone pattern would match, if anchored at
1410     the current point in the subject string.
1411     </P>
1412     <P>
1413     Atomic grouping subpatterns are not capturing subpatterns. Simple cases such as
1414     the above example can be thought of as a maximizing repeat that must swallow
1415     everything it can. So, while both \d+ and \d+? are prepared to adjust the
1416     number of digits they match in order to make the rest of the pattern match,
1417     (?&#62;\d+) can only match an entire sequence of digits.
1418     </P>
1419     <P>
1420     Atomic groups in general can of course contain arbitrarily complicated
1421     subpatterns, and can be nested. However, when the subpattern for an atomic
1422     group is just a single repeated item, as in the example above, a simpler
1423     notation, called a "possessive quantifier" can be used. This consists of an
1424     additional + character following a quantifier. Using this notation, the
1425     previous example can be rewritten as
1426     <pre>
1427 nigel 75 \d++foo
1428     </pre>
1429 ph10 208 Note that a possessive quantifier can be used with an entire group, for
1430     example:
1431     <pre>
1432     (abc|xyz){2,3}+
1433     </pre>
1434 nigel 63 Possessive quantifiers are always greedy; the setting of the PCRE_UNGREEDY
1435     option is ignored. They are a convenient notation for the simpler forms of
1436 nigel 93 atomic group. However, there is no difference in the meaning of a possessive
1437     quantifier and the equivalent atomic group, though there may be a performance
1438     difference; possessive quantifiers should be slightly faster.
1439 nigel 63 </P>
1440     <P>
1441 nigel 93 The possessive quantifier syntax is an extension to the Perl 5.8 syntax.
1442     Jeffrey Friedl originated the idea (and the name) in the first edition of his
1443     book. Mike McCloskey liked it, so implemented it when he built Sun's Java
1444     package, and PCRE copied it from there. It ultimately found its way into Perl
1445     at release 5.10.
1446 nigel 63 </P>
1447     <P>
1448 nigel 93 PCRE has an optimization that automatically "possessifies" certain simple
1449     pattern constructs. For example, the sequence A+B is treated as A++B because
1450     there is no point in backtracking into a sequence of A's when B must follow.
1451     </P>
1452     <P>
1453 nigel 63 When a pattern contains an unlimited repeat inside a subpattern that can itself
1454     be repeated an unlimited number of times, the use of an atomic group is the
1455     only way to avoid some failing matches taking a very long time indeed. The
1456     pattern
1457     <pre>
1458     (\D+|&#60;\d+&#62;)*[!?]
1459 nigel 75 </pre>
1460 nigel 63 matches an unlimited number of substrings that either consist of non-digits, or
1461     digits enclosed in &#60;&#62;, followed by either ! or ?. When it matches, it runs
1462     quickly. However, if it is applied to
1463     <pre>
1464     aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
1465 nigel 75 </pre>
1466 nigel 63 it takes a long time before reporting failure. This is because the string can
1467 nigel 75 be divided between the internal \D+ repeat and the external * repeat in a
1468     large number of ways, and all have to be tried. (The example uses [!?] rather
1469     than a single character at the end, because both PCRE and Perl have an
1470     optimization that allows for fast failure when a single character is used. They
1471     remember the last single character that is required for a match, and fail early
1472     if it is not present in the string.) If the pattern is changed so that it uses
1473     an atomic group, like this:
1474 nigel 63 <pre>
1475     ((?&#62;\D+)|&#60;\d+&#62;)*[!?]
1476 nigel 75 </pre>
1477 nigel 63 sequences of non-digits cannot be broken, and failure happens quickly.
1478 nigel 75 <a name="backreferences"></a></P>
1479 ph10 227 <br><a name="SEC17" href="#TOC1">BACK REFERENCES</a><br>
1480 nigel 63 <P>
1481     Outside a character class, a backslash followed by a digit greater than 0 (and
1482     possibly further digits) is a back reference to a capturing subpattern earlier
1483     (that is, to its left) in the pattern, provided there have been that many
1484     previous capturing left parentheses.
1485     </P>
1486     <P>
1487     However, if the decimal number following the backslash is less than 10, it is
1488     always taken as a back reference, and causes an error only if there are not
1489     that many capturing left parentheses in the entire pattern. In other words, the
1490     parentheses that are referenced need not be to the left of the reference for
1491 nigel 91 numbers less than 10. A "forward back reference" of this type can make sense
1492     when a repetition is involved and the subpattern to the right has participated
1493     in an earlier iteration.
1494     </P>
1495     <P>
1496 nigel 93 It is not possible to have a numerical "forward back reference" to a subpattern
1497     whose number is 10 or more using this syntax because a sequence such as \50 is
1498     interpreted as a character defined in octal. See the subsection entitled
1499 nigel 91 "Non-printing characters"
1500 nigel 75 <a href="#digitsafterbackslash">above</a>
1501 nigel 93 for further details of the handling of digits following a backslash. There is
1502     no such problem when named parentheses are used. A back reference to any
1503     subpattern is possible using named parentheses (see below).
1504 nigel 63 </P>
1505     <P>
1506 nigel 93 Another way of avoiding the ambiguity inherent in the use of digits following a
1507     backslash is to use the \g escape sequence, which is a feature introduced in
1508 ph10 208 Perl 5.10. This escape must be followed by an unsigned number or a negative
1509     number, optionally enclosed in braces. These examples are all identical:
1510 nigel 93 <pre>
1511     (ring), \1
1512     (ring), \g1
1513     (ring), \g{1}
1514     </pre>
1515 ph10 208 An unsigned number specifies an absolute reference without the ambiguity that
1516     is present in the older syntax. It is also useful when literal digits follow
1517     the reference. A negative number is a relative reference. Consider this
1518     example:
1519 nigel 93 <pre>
1520     (abc(def)ghi)\g{-1}
1521     </pre>
1522     The sequence \g{-1} is a reference to the most recently started capturing
1523     subpattern before \g, that is, is it equivalent to \2. Similarly, \g{-2}
1524     would be equivalent to \1. The use of relative references can be helpful in
1525     long patterns, and also in patterns that are created by joining together
1526     fragments that contain references within themselves.
1527     </P>
1528     <P>
1529 nigel 63 A back reference matches whatever actually matched the capturing subpattern in
1530     the current subject string, rather than anything matching the subpattern
1531     itself (see
1532     <a href="#subpatternsassubroutines">"Subpatterns as subroutines"</a>
1533     below for a way of doing that). So the pattern
1534     <pre>
1535     (sens|respons)e and \1ibility
1536 nigel 75 </pre>
1537 nigel 63 matches "sense and sensibility" and "response and responsibility", but not
1538     "sense and responsibility". If caseful matching is in force at the time of the
1539     back reference, the case of letters is relevant. For example,
1540     <pre>
1541     ((?i)rah)\s+\1
1542 nigel 75 </pre>
1543 nigel 63 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
1544     capturing subpattern is matched caselessly.
1545     </P>
1546     <P>
1547 ph10 172 There are several different ways of writing back references to named
1548     subpatterns. The .NET syntax \k{name} and the Perl syntax \k&#60;name&#62; or
1549     \k'name' are supported, as is the Python syntax (?P=name). Perl 5.10's unified
1550     back reference syntax, in which \g can be used for both numeric and named
1551     references, is also supported. We could rewrite the above example in any of
1552 nigel 93 the following ways:
1553 nigel 63 <pre>
1554 nigel 93 (?&#60;p1&#62;(?i)rah)\s+\k&#60;p1&#62;
1555 ph10 172 (?'p1'(?i)rah)\s+\k{p1}
1556 nigel 91 (?P&#60;p1&#62;(?i)rah)\s+(?P=p1)
1557 ph10 172 (?&#60;p1&#62;(?i)rah)\s+\g{p1}
1558 nigel 75 </pre>
1559 nigel 91 A subpattern that is referenced by name may appear in the pattern before or
1560     after the reference.
1561     </P>
1562     <P>
1563 nigel 63 There may be more than one back reference to the same subpattern. If a
1564     subpattern has not actually been used in a particular match, any back
1565     references to it always fail. For example, the pattern
1566     <pre>
1567     (a|(bc))\2
1568 nigel 75 </pre>
1569 nigel 63 always fails if it starts to match "a" rather than "bc". Because there may be
1570     many capturing parentheses in a pattern, all digits following the backslash are
1571     taken as part of a potential back reference number. If the pattern continues
1572     with a digit character, some delimiter must be used to terminate the back
1573     reference. If the PCRE_EXTENDED option is set, this can be whitespace.
1574 nigel 75 Otherwise an empty comment (see
1575     <a href="#comments">"Comments"</a>
1576     below) can be used.
1577 nigel 63 </P>
1578     <P>
1579     A back reference that occurs inside the parentheses to which it refers fails
1580     when the subpattern is first used, so, for example, (a\1) never matches.
1581     However, such references can be useful inside repeated subpatterns. For
1582     example, the pattern
1583     <pre>
1584     (a|b\1)+
1585 nigel 75 </pre>
1586 nigel 63 matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
1587     the subpattern, the back reference matches the character string corresponding
1588     to the previous iteration. In order for this to work, the pattern must be such
1589     that the first iteration does not need to match the back reference. This can be
1590     done using alternation, as in the example above, or by a quantifier with a
1591     minimum of zero.
1592 nigel 75 <a name="bigassertions"></a></P>
1593 ph10 227 <br><a name="SEC18" href="#TOC1">ASSERTIONS</a><br>
1594 nigel 63 <P>
1595     An assertion is a test on the characters following or preceding the current
1596     matching point that does not actually consume any characters. The simple
1597 nigel 75 assertions coded as \b, \B, \A, \G, \Z, \z, ^ and $ are described
1598     <a href="#smallassertions">above.</a>
1599     </P>
1600     <P>
1601 nigel 63 More complicated assertions are coded as subpatterns. There are two kinds:
1602     those that look ahead of the current position in the subject string, and those
1603 nigel 75 that look behind it. An assertion subpattern is matched in the normal way,
1604     except that it does not cause the current matching position to be changed.
1605 nigel 63 </P>
1606     <P>
1607 nigel 75 Assertion subpatterns are not capturing subpatterns, and may not be repeated,
1608     because it makes no sense to assert the same thing several times. If any kind
1609     of assertion contains capturing subpatterns within it, these are counted for
1610     the purposes of numbering the capturing subpatterns in the whole pattern.
1611     However, substring capturing is carried out only for positive assertions,
1612     because it does not make sense for negative assertions.
1613 nigel 63 </P>
1614 nigel 75 <br><b>
1615     Lookahead assertions
1616     </b><br>
1617 nigel 63 <P>
1618 nigel 91 Lookahead assertions start with (?= for positive assertions and (?! for
1619     negative assertions. For example,
1620 nigel 63 <pre>
1621     \w+(?=;)
1622 nigel 75 </pre>
1623 nigel 63 matches a word followed by a semicolon, but does not include the semicolon in
1624     the match, and
1625     <pre>
1626     foo(?!bar)
1627 nigel 75 </pre>
1628 nigel 63 matches any occurrence of "foo" that is not followed by "bar". Note that the
1629     apparently similar pattern
1630     <pre>
1631     (?!foo)bar
1632 nigel 75 </pre>
1633 nigel 63 does not find an occurrence of "bar" that is preceded by something other than
1634     "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
1635     (?!foo) is always true when the next three characters are "bar". A
1636 nigel 75 lookbehind assertion is needed to achieve the other effect.
1637 nigel 63 </P>
1638     <P>
1639     If you want to force a matching failure at some point in a pattern, the most
1640     convenient way to do it is with (?!) because an empty string always matches, so
1641     an assertion that requires there not to be an empty string must always fail.
1642 nigel 75 <a name="lookbehind"></a></P>
1643     <br><b>
1644     Lookbehind assertions
1645     </b><br>
1646 nigel 63 <P>
1647     Lookbehind assertions start with (?&#60;= for positive assertions and (?&#60;! for
1648     negative assertions. For example,
1649     <pre>
1650     (?&#60;!foo)bar
1651 nigel 75 </pre>
1652 nigel 63 does find an occurrence of "bar" that is not preceded by "foo". The contents of
1653     a lookbehind assertion are restricted such that all the strings it matches must
1654 nigel 91 have a fixed length. However, if there are several top-level alternatives, they
1655     do not all have to have the same fixed length. Thus
1656 nigel 63 <pre>
1657     (?&#60;=bullock|donkey)
1658 nigel 75 </pre>
1659 nigel 63 is permitted, but
1660     <pre>
1661     (?&#60;!dogs?|cats?)
1662 nigel 75 </pre>
1663 nigel 63 causes an error at compile time. Branches that match different length strings
1664     are permitted only at the top level of a lookbehind assertion. This is an
1665     extension compared with Perl (at least for 5.8), which requires all branches to
1666     match the same length of string. An assertion such as
1667     <pre>
1668     (?&#60;=ab(c|de))
1669 nigel 75 </pre>
1670 nigel 63 is not permitted, because its single top-level branch can match two different
1671     lengths, but it is acceptable if rewritten to use two top-level branches:
1672     <pre>
1673     (?&#60;=abc|abde)
1674 nigel 75 </pre>
1675 ph10 172 In some cases, the Perl 5.10 escape sequence \K
1676     <a href="#resetmatchstart">(see above)</a>
1677     can be used instead of a lookbehind assertion; this is not restricted to a
1678     fixed-length.
1679     </P>
1680     <P>
1681 nigel 63 The implementation of lookbehind assertions is, for each alternative, to
1682 nigel 93 temporarily move the current position back by the fixed length and then try to
1683 nigel 63 match. If there are insufficient characters before the current position, the
1684 nigel 93 assertion fails.
1685 nigel 63 </P>
1686     <P>
1687     PCRE does not allow the \C escape (which matches a single byte in UTF-8 mode)
1688     to appear in lookbehind assertions, because it makes it impossible to calculate
1689 nigel 93 the length of the lookbehind. The \X and \R escapes, which can match
1690     different numbers of bytes, are also not permitted.
1691 nigel 63 </P>
1692     <P>
1693 nigel 93 Possessive quantifiers can be used in conjunction with lookbehind assertions to
1694     specify efficient matching at the end of the subject string. Consider a simple
1695     pattern such as
1696 nigel 63 <pre>
1697     abcd$
1698 nigel 75 </pre>
1699 nigel 63 when applied to a long string that does not match. Because matching proceeds
1700     from left to right, PCRE will look for each "a" in the subject and then see if
1701     what follows matches the rest of the pattern. If the pattern is specified as
1702     <pre>
1703     ^.*abcd$
1704 nigel 75 </pre>
1705 nigel 63 the initial .* matches the entire string at first, but when this fails (because
1706     there is no following "a"), it backtracks to match all but the last character,
1707     then all but the last two characters, and so on. Once again the search for "a"
1708     covers the entire string, from right to left, so we are no better off. However,
1709     if the pattern is written as
1710     <pre>
1711     ^.*+(?&#60;=abcd)
1712 nigel 75 </pre>
1713 nigel 93 there can be no backtracking for the .*+ item; it can match only the entire
1714 nigel 63 string. The subsequent lookbehind assertion does a single test on the last four
1715     characters. If it fails, the match fails immediately. For long strings, this
1716     approach makes a significant difference to the processing time.
1717     </P>
1718 nigel 75 <br><b>
1719     Using multiple assertions
1720     </b><br>
1721 nigel 63 <P>
1722     Several assertions (of any sort) may occur in succession. For example,
1723     <pre>
1724     (?&#60;=\d{3})(?&#60;!999)foo
1725 nigel 75 </pre>
1726 nigel 63 matches "foo" preceded by three digits that are not "999". Notice that each of
1727     the assertions is applied independently at the same point in the subject
1728     string. First there is a check that the previous three characters are all
1729     digits, and then there is a check that the same three characters are not "999".
1730     This pattern does <i>not</i> match "foo" preceded by six characters, the first
1731     of which are digits and the last three of which are not "999". For example, it
1732     doesn't match "123abcfoo". A pattern to do that is
1733     <pre>
1734     (?&#60;=\d{3}...)(?&#60;!999)foo
1735 nigel 75 </pre>
1736 nigel 63 This time the first assertion looks at the preceding six characters, checking
1737     that the first three are digits, and then the second assertion checks that the
1738     preceding three characters are not "999".
1739     </P>
1740     <P>
1741     Assertions can be nested in any combination. For example,
1742     <pre>
1743     (?&#60;=(?&#60;!foo)bar)baz
1744 nigel 75 </pre>
1745 nigel 63 matches an occurrence of "baz" that is preceded by "bar" which in turn is not
1746     preceded by "foo", while
1747     <pre>
1748     (?&#60;=\d{3}(?!999)...)foo
1749 nigel 75 </pre>
1750     is another pattern that matches "foo" preceded by three digits and any three
1751 nigel 63 characters that are not "999".
1752 nigel 91 <a name="conditions"></a></P>
1753 ph10 227 <br><a name="SEC19" href="#TOC1">CONDITIONAL SUBPATTERNS</a><br>
1754 nigel 63 <P>
1755     It is possible to cause the matching process to obey a subpattern
1756     conditionally or to choose between two alternative subpatterns, depending on
1757     the result of an assertion, or whether a previous capturing subpattern matched
1758     or not. The two possible forms of conditional subpattern are
1759     <pre>
1760     (?(condition)yes-pattern)
1761     (?(condition)yes-pattern|no-pattern)
1762 nigel 75 </pre>
1763 nigel 63 If the condition is satisfied, the yes-pattern is used; otherwise the
1764     no-pattern (if present) is used. If there are more than two alternatives in the
1765     subpattern, a compile-time error occurs.
1766     </P>
1767     <P>
1768 nigel 93 There are four kinds of condition: references to subpatterns, references to
1769     recursion, a pseudo-condition called DEFINE, and assertions.
1770 nigel 91 </P>
1771 nigel 93 <br><b>
1772     Checking for a used subpattern by number
1773     </b><br>
1774 nigel 91 <P>
1775 nigel 93 If the text between the parentheses consists of a sequence of digits, the
1776     condition is true if the capturing subpattern of that number has previously
1777 ph10 172 matched. An alternative notation is to precede the digits with a plus or minus
1778     sign. In this case, the subpattern number is relative rather than absolute.
1779     The most recently opened parentheses can be referenced by (?(-1), the next most
1780     recent by (?(-2), and so on. In looping constructs it can also make sense to
1781     refer to subsequent groups with constructs such as (?(+2).
1782 nigel 93 </P>
1783     <P>
1784 nigel 91 Consider the following pattern, which contains non-significant white space to
1785     make it more readable (assume the PCRE_EXTENDED option) and to divide it into
1786     three parts for ease of discussion:
1787 nigel 63 <pre>
1788     ( \( )? [^()]+ (?(1) \) )
1789 nigel 75 </pre>
1790 nigel 63 The first part matches an optional opening parenthesis, and if that
1791     character is present, sets it as the first captured substring. The second part
1792     matches one or more characters that are not parentheses. The third part is a
1793     conditional subpattern that tests whether the first set of parentheses matched
1794     or not. If they did, that is, if subject started with an opening parenthesis,
1795     the condition is true, and so the yes-pattern is executed and a closing
1796     parenthesis is required. Otherwise, since no-pattern is not present, the
1797     subpattern matches nothing. In other words, this pattern matches a sequence of
1798 nigel 93 non-parentheses, optionally enclosed in parentheses.
1799     </P>
1800 ph10 172 <P>
1801     If you were embedding this pattern in a larger one, you could use a relative
1802     reference:
1803     <pre>
1804     ...other stuff... ( \( )? [^()]+ (?(-1) \) ) ...
1805     </pre>
1806     This makes the fragment independent of the parentheses in the larger pattern.
1807     </P>
1808 nigel 93 <br><b>
1809     Checking for a used subpattern by name
1810     </b><br>
1811     <P>
1812     Perl uses the syntax (?(&#60;name&#62;)...) or (?('name')...) to test for a used
1813     subpattern by name. For compatibility with earlier versions of PCRE, which had
1814     this facility before Perl, the syntax (?(name)...) is also recognized. However,
1815     there is a possible ambiguity with this syntax, because subpattern names may
1816     consist entirely of digits. PCRE looks first for a named subpattern; if it
1817     cannot find one and the name consists entirely of digits, PCRE looks for a
1818     subpattern of that number, which must be greater than zero. Using subpattern
1819     names that consist entirely of digits is not recommended.
1820     </P>
1821     <P>
1822     Rewriting the above example to use a named subpattern gives this:
1823 nigel 91 <pre>
1824 nigel 93 (?&#60;OPEN&#62; \( )? [^()]+ (?(&#60;OPEN&#62;) \) )
1826     </PRE>
1827     </P>
1828     <br><b>
1829     Checking for pattern recursion
1830     </b><br>
1831     <P>
1832     If the condition is the string (R), and there is no subpattern with the name R,
1833     the condition is true if a recursive call to the whole pattern or any
1834     subpattern has been made. If digits or a name preceded by ampersand follow the
1835     letter R, for example:
1836     <pre>
1837     (?(R3)...) or (?(R&name)...)
1838 nigel 91 </pre>
1839 nigel 93 the condition is true if the most recent recursion is into the subpattern whose
1840     number or name is given. This condition does not check the entire recursion
1841     stack.
1842 nigel 63 </P>
1843     <P>
1844 nigel 93 At "top level", all these recursion test conditions are false. Recursive
1845     patterns are described below.
1846     </P>
1847     <br><b>
1848     Defining subpatterns for use by reference only
1849     </b><br>
1850     <P>
1851     If the condition is the string (DEFINE), and there is no subpattern with the
1852     name DEFINE, the condition is always false. In this case, there may be only one
1853     alternative in the subpattern. It is always skipped if control reaches this
1854     point in the pattern; the idea of DEFINE is that it can be used to define
1855     "subroutines" that can be referenced from elsewhere. (The use of "subroutines"
1856     is described below.) For example, a pattern to match an IPv4 address could be
1857     written like this (ignore whitespace and line breaks):
1858     <pre>
1859     (?(DEFINE) (?&#60;byte&#62; 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
1860     \b (?&byte) (\.(?&byte)){3} \b
1861     </pre>
1862     The first part of the pattern is a DEFINE group inside which a another group
1863     named "byte" is defined. This matches an individual component of an IPv4
1864     address (a number less than 256). When matching takes place, this part of the
1865     pattern is skipped because DEFINE acts like a false condition.
1866     </P>
1867     <P>
1868     The rest of the pattern uses references to the named group to match the four
1869     dot-separated components of an IPv4 address, insisting on a word boundary at
1870     each end.
1871     </P>
1872     <br><b>
1873     Assertion conditions
1874     </b><br>
1875     <P>
1876     If the condition is not in any of the above formats, it must be an assertion.
1877 nigel 63 This may be a positive or negative lookahead or lookbehind assertion. Consider
1878     this pattern, again containing non-significant white space, and with the two
1879     alternatives on the second line:
1880     <pre>
1881     (?(?=[^a-z]*[a-z])
1882     \d{2}-[a-z]{3}-\d{2} | \d{2}-\d{2}-\d{2} )
1883 nigel 75 </pre>
1884 nigel 63 The condition is a positive lookahead assertion that matches an optional
1885     sequence of non-letters followed by a letter. In other words, it tests for the
1886     presence of at least one letter in the subject. If a letter is found, the
1887     subject is matched against the first alternative; otherwise it is matched
1888     against the second. This pattern matches strings in one of the two forms
1889     dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
1890 nigel 75 <a name="comments"></a></P>
1891 ph10 227 <br><a name="SEC20" href="#TOC1">COMMENTS</a><br>
1892 nigel 63 <P>
1893 nigel 75 The sequence (?# marks the start of a comment that continues up to the next
1894 nigel 63 closing parenthesis. Nested parentheses are not permitted. The characters
1895     that make up a comment play no part in the pattern matching at all.
1896     </P>
1897     <P>
1898     If the PCRE_EXTENDED option is set, an unescaped # character outside a
1899 nigel 91 character class introduces a comment that continues to immediately after the
1900     next newline in the pattern.
1901     <a name="recursion"></a></P>
1902 ph10 227 <br><a name="SEC21" href="#TOC1">RECURSIVE PATTERNS</a><br>
1903 nigel 63 <P>
1904     Consider the problem of matching a string in parentheses, allowing for
1905     unlimited nested parentheses. Without the use of recursion, the best that can
1906     be done is to use a pattern that matches up to some fixed depth of nesting. It
1907 nigel 93 is not possible to handle an arbitrary nesting depth.
1908     </P>
1909     <P>
1910     For some time, Perl has provided a facility that allows regular expressions to
1911     recurse (amongst other things). It does this by interpolating Perl code in the
1912     expression at run time, and the code can refer to the expression itself. A Perl
1913     pattern using code interpolation to solve the parentheses problem can be
1914     created like this:
1915 nigel 63 <pre>
1916     $re = qr{\( (?: (?&#62;[^()]+) | (?p{$re}) )* \)}x;
1917 nigel 75 </pre>
1918 nigel 63 The (?p{...}) item interpolates Perl code at run time, and in this case refers
1919 nigel 93 recursively to the pattern in which it appears.
1920 nigel 63 </P>
1921     <P>
1922 nigel 93 Obviously, PCRE cannot support the interpolation of Perl code. Instead, it
1923     supports special syntax for recursion of the entire pattern, and also for
1924     individual subpattern recursion. After its introduction in PCRE and Python,
1925 ph10 453 this kind of recursion was subsequently introduced into Perl at release 5.10.
1926 nigel 63 </P>
1927     <P>
1928 nigel 93 A special item that consists of (? followed by a number greater than zero and a
1929     closing parenthesis is a recursive call of the subpattern of the given number,
1930     provided that it occurs inside that subpattern. (If not, it is a "subroutine"
1931     call, which is described in the next section.) The special item (?R) or (?0) is
1932     a recursive call of the entire regular expression.
1933 nigel 87 </P>
1934     <P>
1935     This PCRE pattern solves the nested parentheses problem (assume the
1936     PCRE_EXTENDED option is set so that white space is ignored):
1937 nigel 63 <pre>
1938     \( ( (?&#62;[^()]+) | (?R) )* \)
1939 nigel 75 </pre>
1940 nigel 63 First it matches an opening parenthesis. Then it matches any number of
1941     substrings which can either be a sequence of non-parentheses, or a recursive
1942 nigel 87 match of the pattern itself (that is, a correctly parenthesized substring).
1943 nigel 63 Finally there is a closing parenthesis.
1944     </P>
1945     <P>
1946     If this were part of a larger pattern, you would not want to recurse the entire
1947     pattern, so instead you could use this:
1948     <pre>
1949     ( \( ( (?&#62;[^()]+) | (?1) )* \) )
1950 nigel 75 </pre>
1951 nigel 63 We have put the pattern into parentheses, and caused the recursion to refer to
1952 ph10 172 them instead of the whole pattern.
1953     </P>
1954     <P>
1955     In a larger pattern, keeping track of parenthesis numbers can be tricky. This
1956     is made easier by the use of relative references. (A Perl 5.10 feature.)
1957     Instead of (?1) in the pattern above you can write (?-2) to refer to the second
1958     most recently opened parentheses preceding the recursion. In other words, a
1959     negative number counts capturing parentheses leftwards from the point at which
1960     it is encountered.
1961     </P>
1962     <P>
1963     It is also possible to refer to subsequently opened parentheses, by writing
1964     references such as (?+2). However, these cannot be recursive because the
1965     reference is not inside the parentheses that are referenced. They are always
1966     "subroutine" calls, as described in the next section.
1967     </P>
1968     <P>
1969     An alternative approach is to use named parentheses instead. The Perl syntax
1970     for this is (?&name); PCRE's earlier syntax (?P&#62;name) is also supported. We
1971     could rewrite the above example as follows:
1972 nigel 63 <pre>
1973 nigel 93 (?&#60;pn&#62; \( ( (?&#62;[^()]+) | (?&pn) )* \) )
1974 nigel 75 </pre>
1975 nigel 93 If there is more than one subpattern with the same name, the earliest one is
1976 ph10 172 used.
1977     </P>
1978     <P>
1979     This particular example pattern that we have been looking at contains nested
1980     unlimited repeats, and so the use of atomic grouping for matching strings of
1981     non-parentheses is important when applying the pattern to strings that do not
1982     match. For example, when this pattern is applied to
1983 nigel 63 <pre>
1984     (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
1985 nigel 75 </pre>
1986 nigel 63 it yields "no match" quickly. However, if atomic grouping is not used,
1987     the match runs for a very long time indeed because there are so many different
1988     ways the + and * repeats can carve up the subject, and all have to be tested
1989     before failure can be reported.
1990     </P>
1991     <P>
1992     At the end of a match, the values set for any capturing subpatterns are those
1993     from the outermost level of the recursion at which the subpattern value is set.
1994     If you want to obtain intermediate values, a callout function can be used (see
1995 nigel 93 below and the
1996 nigel 63 <a href="pcrecallout.html"><b>pcrecallout</b></a>
1997     documentation). If the pattern above is matched against
1998     <pre>
1999     (ab(cd)ef)
2000 nigel 75 </pre>
2001 nigel 63 the value for the capturing parentheses is "ef", which is the last value taken
2002     on at the top level. If additional parentheses are added, giving
2003     <pre>
2004     \( ( ( (?&#62;[^()]+) | (?R) )* ) \)
2005     ^ ^
2006     ^ ^
2007 nigel 75 </pre>
2008 nigel 63 the string they capture is "ab(cd)ef", the contents of the top level
2009     parentheses. If there are more than 15 capturing parentheses in a pattern, PCRE
2010     has to obtain extra memory to store data during a recursion, which it does by
2011     using <b>pcre_malloc</b>, freeing it via <b>pcre_free</b> afterwards. If no
2012     memory can be obtained, the match fails with the PCRE_ERROR_NOMEMORY error.
2013     </P>
2014     <P>
2015     Do not confuse the (?R) item with the condition (R), which tests for recursion.
2016     Consider this pattern, which matches text in angle brackets, allowing for
2017     arbitrary nesting. Only digits are allowed in nested brackets (that is, when
2018     recursing), whereas any characters are permitted at the outer level.
2019     <pre>
2020     &#60; (?: (?(R) \d++ | [^&#60;&#62;]*+) | (?R)) * &#62;
2021 nigel 75 </pre>
2022 nigel 63 In this pattern, (?(R) is the start of a conditional subpattern, with two
2023     different alternatives for the recursive and non-recursive cases. The (?R) item
2024     is the actual recursive call.
2025 ph10 453 <a name="recursiondifference"></a></P>
2026     <br><b>
2027     Recursion difference from Perl
2028     </b><br>
2029     <P>
2030     In PCRE (like Python, but unlike Perl), a recursive subpattern call is always
2031     treated as an atomic group. That is, once it has matched some of the subject
2032     string, it is never re-entered, even if it contains untried alternatives and
2033     there is a subsequent matching failure. This can be illustrated by the
2034     following pattern, which purports to match a palindromic string that contains
2035     an odd number of characters (for example, "a", "aba", "abcba", "abcdcba"):
2036     <pre>
2037     ^(.|(.)(?1)\2)$
2038     </pre>
2039     The idea is that it either matches a single character, or two identical
2040     characters surrounding a sub-palindrome. In Perl, this pattern works; in PCRE
2041     it does not if the pattern is longer than three characters. Consider the
2042     subject string "abcba":
2043     </P>
2044     <P>
2045     At the top level, the first character is matched, but as it is not at the end
2046     of the string, the first alternative fails; the second alternative is taken
2047     and the recursion kicks in. The recursive call to subpattern 1 successfully
2048     matches the next character ("b"). (Note that the beginning and end of line
2049     tests are not part of the recursion).
2050     </P>
2051     <P>
2052     Back at the top level, the next character ("c") is compared with what
2053     subpattern 2 matched, which was "a". This fails. Because the recursion is
2054     treated as an atomic group, there are now no backtracking points, and so the
2055     entire match fails. (Perl is able, at this point, to re-enter the recursion and
2056     try the second alternative.) However, if the pattern is written with the
2057     alternatives in the other order, things are different:
2058     <pre>
2059     ^((.)(?1)\2|.)$
2060     </pre>
2061     This time, the recursing alternative is tried first, and continues to recurse
2062     until it runs out of characters, at which point the recursion fails. But this
2063     time we do have another alternative to try at the higher level. That is the big
2064     difference: in the previous case the remaining alternative is at a deeper
2065     recursion level, which PCRE cannot use.
2066     </P>
2067     <P>
2068     To change the pattern so that matches all palindromic strings, not just those
2069     with an odd number of characters, it is tempting to change the pattern to this:
2070     <pre>
2071     ^((.)(?1)\2|.?)$
2072     </pre>
2073     Again, this works in Perl, but not in PCRE, and for the same reason. When a
2074     deeper recursion has matched a single character, it cannot be entered again in
2075     order to match an empty string. The solution is to separate the two cases, and
2076     write out the odd and even cases as alternatives at the higher level:
2077     <pre>
2078     ^(?:((.)(?1)\2|)|((.)(?3)\4|.))
2079     </pre>
2080     If you want to match typical palindromic phrases, the pattern has to ignore all
2081     non-word characters, which can be done like this:
2082     <pre>
2083     ^\W*+(?:((.)\W*+(?1)\W*+\2|)|((.)\W*+(?3)\W*+\4|\W*+.\W*+))\W*+$
2084     </pre>
2085     If run with the PCRE_CASELESS option, this pattern matches phrases such as "A
2086     man, a plan, a canal: Panama!" and it works well in both PCRE and Perl. Note
2087     the use of the possessive quantifier *+ to avoid backtracking into sequences of
2088     non-word characters. Without this, PCRE takes a great deal longer (ten times or
2089     more) to match typical phrases, and Perl takes so long that you think it has
2090     gone into a loop.
2091 nigel 75 <a name="subpatternsassubroutines"></a></P>
2092 ph10 227 <br><a name="SEC22" href="#TOC1">SUBPATTERNS AS SUBROUTINES</a><br>
2093 nigel 63 <P>
2094     If the syntax for a recursive subpattern reference (either by number or by
2095     name) is used outside the parentheses to which it refers, it operates like a
2096 nigel 93 subroutine in a programming language. The "called" subpattern may be defined
2097 ph10 172 before or after the reference. A numbered reference can be absolute or
2098     relative, as in these examples:
2099 nigel 63 <pre>
2100 ph10 172 (...(absolute)...)...(?2)...
2101     (...(relative)...)...(?-1)...
2102     (...(?+1)...(relative)...
2103     </pre>
2104     An earlier example pointed out that the pattern
2105     <pre>
2106 nigel 63 (sens|respons)e and \1ibility
2107 nigel 75 </pre>
2108 nigel 63 matches "sense and sensibility" and "response and responsibility", but not
2109     "sense and responsibility". If instead the pattern
2110     <pre>
2111     (sens|respons)e and (?1)ibility
2112 nigel 75 </pre>
2113 nigel 63 is used, it does match "sense and responsibility" as well as the other two
2114 nigel 93 strings. Another example is given in the discussion of DEFINE above.
2115 nigel 63 </P>
2116 nigel 87 <P>
2117     Like recursive subpatterns, a "subroutine" call is always treated as an atomic
2118     group. That is, once it has matched some of the subject string, it is never
2119     re-entered, even if it contains untried alternatives and there is a subsequent
2120     matching failure.
2121     </P>
2122 nigel 63 <P>
2123 nigel 93 When a subpattern is used as a subroutine, processing options such as
2124     case-independence are fixed when the subpattern is defined. They cannot be
2125     changed for different calls. For example, consider this pattern:
2126     <pre>
2127 ph10 172 (abc)(?i:(?-1))
2128 nigel 93 </pre>
2129     It matches "abcabc". It does not match "abcABC" because the change of
2130     processing option does not affect the called subpattern.
2131 ph10 345 <a name="onigurumasubroutines"></a></P>
2132     <br><a name="SEC23" href="#TOC1">ONIGURUMA SUBROUTINE SYNTAX</a><br>
2133     <P>
2134     For compatibility with Oniguruma, the non-Perl syntax \g followed by a name or
2135     a number enclosed either in angle brackets or single quotes, is an alternative
2136     syntax for referencing a subpattern as a subroutine, possibly recursively. Here
2137     are two of the examples used above, rewritten using this syntax:
2138     <pre>
2139     (?&#60;pn&#62; \( ( (?&#62;[^()]+) | \g&#60;pn&#62; )* \) )
2140     (sens|respons)e and \g'1'ibility
2141     </pre>
2142     PCRE supports an extension to Oniguruma: if a number is preceded by a
2143     plus or a minus sign it is taken as a relative reference. For example:
2144     <pre>
2145     (abc)(?i:\g&#60;-1&#62;)
2146     </pre>
2147     Note that \g{...} (Perl syntax) and \g&#60;...&#62; (Oniguruma syntax) are <i>not</i>
2148     synonymous. The former is a back reference; the latter is a subroutine call.
2149 nigel 93 </P>
2150 ph10 345 <br><a name="SEC24" href="#TOC1">CALLOUTS</a><br>
2151 nigel 93 <P>
2152 nigel 63 Perl has a feature whereby using the sequence (?{...}) causes arbitrary Perl
2153     code to be obeyed in the middle of matching a regular expression. This makes it
2154     possible, amongst other things, to extract different substrings that match the
2155     same pair of parentheses when there is a repetition.
2156     </P>
2157     <P>
2158     PCRE provides a similar feature, but of course it cannot obey arbitrary Perl
2159     code. The feature is called "callout". The caller of PCRE provides an external
2160     function by putting its entry point in the global variable <i>pcre_callout</i>.
2161     By default, this variable contains NULL, which disables all calling out.
2162     </P>
2163     <P>
2164     Within a regular expression, (?C) indicates the points at which the external
2165     function is to be called. If you want to identify different callout points, you
2166     can put a number less than 256 after the letter C. The default value is zero.
2167     For example, this pattern has two callout points:
2168     <pre>
2169 ph10 155 (?C1)abc(?C2)def
2170 nigel 75 </pre>
2171     If the PCRE_AUTO_CALLOUT flag is passed to <b>pcre_compile()</b>, callouts are
2172     automatically installed before each item in the pattern. They are all numbered
2173     255.
2174 nigel 63 </P>
2175     <P>
2176     During matching, when PCRE reaches a callout point (and <i>pcre_callout</i> is
2177     set), the external function is called. It is provided with the number of the
2178 nigel 75 callout, the position in the pattern, and, optionally, one item of data
2179     originally supplied by the caller of <b>pcre_exec()</b>. The callout function
2180     may cause matching to proceed, to backtrack, or to fail altogether. A complete
2181     description of the interface to the callout function is given in the
2182 nigel 63 <a href="pcrecallout.html"><b>pcrecallout</b></a>
2183     documentation.
2184     </P>
2185 ph10 345 <br><a name="SEC25" href="#TOC1">BACKTRACKING CONTROL</a><br>
2186 nigel 63 <P>
2187 ph10 211 Perl 5.10 introduced a number of "Special Backtracking Control Verbs", which
2188     are described in the Perl documentation as "experimental and subject to change
2189     or removal in a future version of Perl". It goes on to say: "Their usage in
2190     production code should be noted to avoid problems during upgrades." The same
2191     remarks apply to the PCRE features described in this section.
2192     </P>
2193     <P>
2194 ph10 345 Since these verbs are specifically related to backtracking, most of them can be
2195     used only when the pattern is to be matched using <b>pcre_exec()</b>, which uses
2196     a backtracking algorithm. With the exception of (*FAIL), which behaves like a
2197     failing negative assertion, they cause an error if encountered by
2198 ph10 211 <b>pcre_dfa_exec()</b>.
2199     </P>
2200     <P>
2201 ph10 453 If any of these verbs are used in an assertion subpattern, their effect is
2202     confined to that subpattern; it does not extend to the surrounding pattern.
2203     Note that assertion subpatterns are processed as anchored at the point where
2204     they are tested.
2205     </P>
2206     <P>
2207 ph10 211 The new verbs make use of what was previously invalid syntax: an opening
2208     parenthesis followed by an asterisk. In Perl, they are generally of the form
2209     (*VERB:ARG) but PCRE does not support the use of arguments, so its general
2210     form is just (*VERB). Any number of these verbs may occur in a pattern. There
2211     are two kinds:
2212     </P>
2213     <br><b>
2214     Verbs that act immediately
2215     </b><br>
2216     <P>
2217     The following verbs act as soon as they are encountered:
2218     <pre>
2219     (*ACCEPT)
2220     </pre>
2221     This verb causes the match to end successfully, skipping the remainder of the
2222     pattern. When inside a recursion, only the innermost pattern is ended
2223 ph10 453 immediately. If the (*ACCEPT) is inside capturing parentheses, the data so far
2224     is captured. (This feature was added to PCRE at release 8.00.) For example:
2225 ph10 211 <pre>
2226 ph10 453 A((?:A|B(*ACCEPT)|C)D)
2227 ph10 211 </pre>
2228 ph10 453 This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is captured by
2229     the outer parentheses.
2230 ph10 211 <pre>
2231     (*FAIL) or (*F)
2232     </pre>
2233     This verb causes the match to fail, forcing backtracking to occur. It is
2234     equivalent to (?!) but easier to read. The Perl documentation notes that it is
2235     probably useful only when combined with (?{}) or (??{}). Those are, of course,
2236     Perl features that are not present in PCRE. The nearest equivalent is the
2237     callout feature, as for example in this pattern:
2238     <pre>
2239     a+(?C)(*FAIL)
2240     </pre>
2241     A match with the string "aaaa" always fails, but the callout is taken before
2242     each backtrack happens (in this example, 10 times).
2243     </P>
2244     <br><b>
2245     Verbs that act after backtracking
2246     </b><br>
2247     <P>
2248     The following verbs do nothing when they are encountered. Matching continues
2249     with what follows, but if there is no subsequent match, a failure is forced.
2250     The verbs differ in exactly what kind of failure occurs.
2251     <pre>
2252     (*COMMIT)
2253     </pre>
2254     This verb causes the whole match to fail outright if the rest of the pattern
2255     does not match. Even if the pattern is unanchored, no further attempts to find
2256     a match by advancing the start point take place. Once (*COMMIT) has been
2257     passed, <b>pcre_exec()</b> is committed to finding a match at the current
2258     starting point, or not at all. For example:
2259     <pre>
2260     a+(*COMMIT)b
2261     </pre>
2262     This matches "xxaab" but not "aacaab". It can be thought of as a kind of
2263     dynamic anchor, or "I've started, so I must finish."
2264     <pre>
2265     (*PRUNE)
2266     </pre>
2267     This verb causes the match to fail at the current position if the rest of the
2268     pattern does not match. If the pattern is unanchored, the normal "bumpalong"
2269     advance to the next starting character then happens. Backtracking can occur as
2270     usual to the left of (*PRUNE), or when matching to the right of (*PRUNE), but
2271     if there is no match to the right, backtracking cannot cross (*PRUNE).
2272     In simple cases, the use of (*PRUNE) is just an alternative to an atomic
2273     group or possessive quantifier, but there are some uses of (*PRUNE) that cannot
2274     be expressed in any other way.
2275     <pre>
2276     (*SKIP)
2277     </pre>
2278     This verb is like (*PRUNE), except that if the pattern is unanchored, the
2279     "bumpalong" advance is not to the next character, but to the position in the
2280     subject where (*SKIP) was encountered. (*SKIP) signifies that whatever text
2281     was matched leading up to it cannot be part of a successful match. Consider:
2282     <pre>
2283     a+(*SKIP)b
2284     </pre>
2285     If the subject is "aaaac...", after the first match attempt fails (starting at
2286     the first character in the string), the starting point skips on to start the
2287     next attempt at "c". Note that a possessive quantifer does not have the same
2288     effect in this example; although it would suppress backtracking during the
2289     first match attempt, the second attempt would start at the second character
2290     instead of skipping on to "c".
2291     <pre>
2292     (*THEN)
2293     </pre>
2294     This verb causes a skip to the next alternation if the rest of the pattern does
2295     not match. That is, it cancels pending backtracking, but only within the
2296     current alternation. Its name comes from the observation that it can be used
2297     for a pattern-based if-then-else block:
2298     <pre>
2299     ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
2300     </pre>
2301     If the COND1 pattern matches, FOO is tried (and possibly further items after
2302     the end of the group if FOO succeeds); on failure the matcher skips to the
2303     second alternative and tries COND2, without backtracking into COND1. If (*THEN)
2304     is used outside of any alternation, it acts exactly like (*PRUNE).
2305     </P>
2306 ph10 345 <br><a name="SEC26" href="#TOC1">SEE ALSO</a><br>
2307 ph10 211 <P>
2308 nigel 93 <b>pcreapi</b>(3), <b>pcrecallout</b>(3), <b>pcrematching</b>(3), <b>pcre</b>(3).
2309     </P>
2310 ph10 345 <br><a name="SEC27" href="#TOC1">AUTHOR</a><br>
2311 nigel 93 <P>
2312 ph10 99 Philip Hazel
2313 nigel 63 <br>
2314 ph10 99 University Computing Service
2315     <br>
2316     Cambridge CB2 3QH, England.
2317     <br>
2318     </P>
2319 ph10 345 <br><a name="SEC28" href="#TOC1">REVISION</a><br>
2320 ph10 99 <P>
2321 ph10 453 Last updated: 18 September 2009
2322 ph10 99 <br>
2323 ph10 392 Copyright &copy; 1997-2009 University of Cambridge.
2324 ph10 99 <br>
2325 nigel 75 <p>
2326     Return to the <a href="index.html">PCRE index page</a>.
2327     </p>


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

ViewVC Help
Powered by ViewVC 1.1.12