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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 745 - (show annotations) (download)
Mon Nov 14 11:41:03 2011 UTC (3 years ago) by ph10
File size: 120327 byte(s)
Small tidies, and documentation update for JavaScript \x, \u, \U support.

1 .TH PCREPATTERN 3
2 .SH NAME
3 PCRE - Perl-compatible regular expressions
4 .SH "PCRE REGULAR EXPRESSION DETAILS"
5 .rs
6 .sp
7 The syntax and semantics of the regular expressions that are supported by PCRE
8 are described in detail below. There is a quick-reference syntax summary in the
9 .\" HREF
10 \fBpcresyntax\fP
11 .\"
12 page. PCRE tries to match Perl syntax and semantics as closely as it can. PCRE
13 also supports some alternative regular expression syntax (which does not
14 conflict with the Perl syntax) in order to provide some compatibility with
15 regular expressions in Python, .NET, and Oniguruma.
16 .P
17 Perl's regular expressions are described in its own documentation, and
18 regular expressions in general are covered in a number of books, some of which
19 have copious examples. Jeffrey Friedl's "Mastering Regular Expressions",
20 published by O'Reilly, covers regular expressions in great detail. This
21 description of PCRE's regular expressions is intended as reference material.
22 .P
23 The original operation of PCRE was on strings of one-byte characters. However,
24 there is now also support for UTF-8 character strings. To use this,
25 PCRE must be built to include UTF-8 support, and you must call
26 \fBpcre_compile()\fP or \fBpcre_compile2()\fP with the PCRE_UTF8 option. There
27 is also a special sequence that can be given at the start of a pattern:
28 .sp
29 (*UTF8)
30 .sp
31 Starting a pattern with this sequence is equivalent to setting the PCRE_UTF8
32 option. This feature is not Perl-compatible. How setting UTF-8 mode affects
33 pattern matching is mentioned in several places below. There is also a summary
34 of UTF-8 features in the
35 .\" HREF
36 \fBpcreunicode\fP
37 .\"
38 page.
39 .P
40 Another special sequence that may appear at the start of a pattern or in
41 combination with (*UTF8) is:
42 .sp
43 (*UCP)
44 .sp
45 This has the same effect as setting the PCRE_UCP option: it causes sequences
46 such as \ed and \ew to use Unicode properties to determine character types,
47 instead of recognizing only characters with codes less than 128 via a lookup
48 table.
49 .P
50 If a pattern starts with (*NO_START_OPT), it has the same effect as setting the
51 PCRE_NO_START_OPTIMIZE option either at compile or matching time. There are
52 also some more of these special sequences that are concerned with the handling
53 of newlines; they are described below.
54 .P
55 The remainder of this document discusses the patterns that are supported by
56 PCRE when its main matching function, \fBpcre_exec()\fP, is used.
57 From release 6.0, PCRE offers a second matching function,
58 \fBpcre_dfa_exec()\fP, which matches using a different algorithm that is not
59 Perl-compatible. Some of the features discussed below are not available when
60 \fBpcre_dfa_exec()\fP is used. The advantages and disadvantages of the
61 alternative function, and how it differs from the normal function, are
62 discussed in the
63 .\" HREF
64 \fBpcrematching\fP
65 .\"
66 page.
67 .
68 .
69 .\" HTML <a name="newlines"></a>
70 .SH "NEWLINE CONVENTIONS"
71 .rs
72 .sp
73 PCRE supports five different conventions for indicating line breaks in
74 strings: a single CR (carriage return) character, a single LF (linefeed)
75 character, the two-character sequence CRLF, any of the three preceding, or any
76 Unicode newline sequence. The
77 .\" HREF
78 \fBpcreapi\fP
79 .\"
80 page has
81 .\" HTML <a href="pcreapi.html#newlines">
82 .\" </a>
83 further discussion
84 .\"
85 about newlines, and shows how to set the newline convention in the
86 \fIoptions\fP arguments for the compiling and matching functions.
87 .P
88 It is also possible to specify a newline convention by starting a pattern
89 string with one of the following five sequences:
90 .sp
91 (*CR) carriage return
92 (*LF) linefeed
93 (*CRLF) carriage return, followed by linefeed
94 (*ANYCRLF) any of the three above
95 (*ANY) all Unicode newline sequences
96 .sp
97 These override the default and the options given to \fBpcre_compile()\fP or
98 \fBpcre_compile2()\fP. For example, on a Unix system where LF is the default
99 newline sequence, the pattern
100 .sp
101 (*CR)a.b
102 .sp
103 changes the convention to CR. That pattern matches "a\enb" because LF is no
104 longer a newline. Note that these special settings, which are not
105 Perl-compatible, are recognized only at the very start of a pattern, and that
106 they must be in upper case. If more than one of them is present, the last one
107 is used.
108 .P
109 The newline convention affects the interpretation of the dot metacharacter when
110 PCRE_DOTALL is not set, and also the behaviour of \eN. However, it does not
111 affect what the \eR escape sequence matches. By default, this is any Unicode
112 newline sequence, for Perl compatibility. However, this can be changed; see the
113 description of \eR in the section entitled
114 .\" HTML <a href="#newlineseq">
115 .\" </a>
116 "Newline sequences"
117 .\"
118 below. A change of \eR setting can be combined with a change of newline
119 convention.
120 .
121 .
122 .SH "CHARACTERS AND METACHARACTERS"
123 .rs
124 .sp
125 A regular expression is a pattern that is matched against a subject string from
126 left to right. Most characters stand for themselves in a pattern, and match the
127 corresponding characters in the subject. As a trivial example, the pattern
128 .sp
129 The quick brown fox
130 .sp
131 matches a portion of a subject string that is identical to itself. When
132 caseless matching is specified (the PCRE_CASELESS option), letters are matched
133 independently of case. In UTF-8 mode, PCRE always understands the concept of
134 case for characters whose values are less than 128, so caseless matching is
135 always possible. For characters with higher values, the concept of case is
136 supported if PCRE is compiled with Unicode property support, but not otherwise.
137 If you want to use caseless matching for characters 128 and above, you must
138 ensure that PCRE is compiled with Unicode property support as well as with
139 UTF-8 support.
140 .P
141 The power of regular expressions comes from the ability to include alternatives
142 and repetitions in the pattern. These are encoded in the pattern by the use of
143 \fImetacharacters\fP, which do not stand for themselves but instead are
144 interpreted in some special way.
145 .P
146 There are two different sets of metacharacters: those that are recognized
147 anywhere in the pattern except within square brackets, and those that are
148 recognized within square brackets. Outside square brackets, the metacharacters
149 are as follows:
150 .sp
151 \e general escape character with several uses
152 ^ assert start of string (or line, in multiline mode)
153 $ assert end of string (or line, in multiline mode)
154 . match any character except newline (by default)
155 [ start character class definition
156 | start of alternative branch
157 ( start subpattern
158 ) end subpattern
159 ? extends the meaning of (
160 also 0 or 1 quantifier
161 also quantifier minimizer
162 * 0 or more quantifier
163 + 1 or more quantifier
164 also "possessive quantifier"
165 { start min/max quantifier
166 .sp
167 Part of a pattern that is in square brackets is called a "character class". In
168 a character class the only metacharacters are:
169 .sp
170 \e general escape character
171 ^ negate the class, but only if the first character
172 - indicates character range
173 .\" JOIN
174 [ POSIX character class (only if followed by POSIX
175 syntax)
176 ] terminates the character class
177 .sp
178 The following sections describe the use of each of the metacharacters.
179 .
180 .
181 .SH BACKSLASH
182 .rs
183 .sp
184 The backslash character has several uses. Firstly, if it is followed by a
185 character that is not a number or a letter, it takes away any special meaning
186 that character may have. This use of backslash as an escape character applies
187 both inside and outside character classes.
188 .P
189 For example, if you want to match a * character, you write \e* in the pattern.
190 This escaping action applies whether or not the following character would
191 otherwise be interpreted as a metacharacter, so it is always safe to precede a
192 non-alphanumeric with backslash to specify that it stands for itself. In
193 particular, if you want to match a backslash, you write \e\e.
194 .P
195 In UTF-8 mode, only ASCII numbers and letters have any special meaning after a
196 backslash. All other characters (in particular, those whose codepoints are
197 greater than 127) are treated as literals.
198 .P
199 If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the
200 pattern (other than in a character class) and characters between a # outside
201 a character class and the next newline are ignored. An escaping backslash can
202 be used to include a whitespace or # character as part of the pattern.
203 .P
204 If you want to remove the special meaning from a sequence of characters, you
205 can do so by putting them between \eQ and \eE. This is different from Perl in
206 that $ and @ are handled as literals in \eQ...\eE sequences in PCRE, whereas in
207 Perl, $ and @ cause variable interpolation. Note the following examples:
208 .sp
209 Pattern PCRE matches Perl matches
210 .sp
211 .\" JOIN
212 \eQabc$xyz\eE abc$xyz abc followed by the
213 contents of $xyz
214 \eQabc\e$xyz\eE abc\e$xyz abc\e$xyz
215 \eQabc\eE\e$\eQxyz\eE abc$xyz abc$xyz
216 .sp
217 The \eQ...\eE sequence is recognized both inside and outside character classes.
218 An isolated \eE that is not preceded by \eQ is ignored. If \eQ is not followed
219 by \eE later in the pattern, the literal interpretation continues to the end of
220 the pattern (that is, \eE is assumed at the end). If the isolated \eQ is inside
221 a character class, this causes an error, because the character class is not
222 terminated.
223 .
224 .
225 .\" HTML <a name="digitsafterbackslash"></a>
226 .SS "Non-printing characters"
227 .rs
228 .sp
229 A second use of backslash provides a way of encoding non-printing characters
230 in patterns in a visible manner. There is no restriction on the appearance of
231 non-printing characters, apart from the binary zero that terminates a pattern,
232 but when a pattern is being prepared by text editing, it is often easier to use
233 one of the following escape sequences than the binary character it represents:
234 .sp
235 \ea alarm, that is, the BEL character (hex 07)
236 \ecx "control-x", where x is any ASCII character
237 \ee escape (hex 1B)
238 \ef formfeed (hex 0C)
239 \en linefeed (hex 0A)
240 \er carriage return (hex 0D)
241 \et tab (hex 09)
242 \eddd character with octal code ddd, or back reference
243 \exhh character with hex code hh
244 \ex{hhh..} character with hex code hhh.. (non-JavaScript mode)
245 \euhhhh character with hex code hhhh (JavaScript mode only)
246 .sp
247 The precise effect of \ecx is as follows: if x is a lower case letter, it
248 is converted to upper case. Then bit 6 of the character (hex 40) is inverted.
249 Thus \ecz becomes hex 1A (z is 7A), but \ec{ becomes hex 3B ({ is 7B), while
250 \ec; becomes hex 7B (; is 3B). If the byte following \ec has a value greater
251 than 127, a compile-time error occurs. This locks out non-ASCII characters in
252 both byte mode and UTF-8 mode. (When PCRE is compiled in EBCDIC mode, all byte
253 values are valid. A lower case letter is converted to upper case, and then the
254 0xc0 bits are flipped.)
255 .P
256 By default, after \ex, from zero to two hexadecimal digits are read (letters
257 can be in upper or lower case). Any number of hexadecimal digits may appear
258 between \ex{ and }, but the value of the character code must be less than 256
259 in non-UTF-8 mode, and less than 2**31 in UTF-8 mode. That is, the maximum
260 value in hexadecimal is 7FFFFFFF. Note that this is bigger than the largest
261 Unicode code point, which is 10FFFF.
262 .P
263 If characters other than hexadecimal digits appear between \ex{ and }, or if
264 there is no terminating }, this form of escape is not recognized. Instead, the
265 initial \ex will be interpreted as a basic hexadecimal escape, with no
266 following digits, giving a character whose value is zero.
267 .P
268 If the PCRE_JAVASCRIPT_COMPAT option is set, the interpretation of \ex is
269 as just described only when it is followed by two hexadecimal digits.
270 Otherwise, it matches a literal "x" character. In JavaScript mode, support for
271 code points greater than 256 is provided by \eu, which must be followed by
272 four hexadecimal digits; otherwise it matches a literal "u" character.
273 .P
274 Characters whose value is less than 256 can be defined by either of the two
275 syntaxes for \ex (or by \eu in JavaScript mode). There is no difference in the
276 way they are handled. For example, \exdc is exactly the same as \ex{dc} (or
277 \eu00dc in JavaScript mode).
278 .P
279 After \e0 up to two further octal digits are read. If there are fewer than two
280 digits, just those that are present are used. Thus the sequence \e0\ex\e07
281 specifies two binary zeros followed by a BEL character (code value 7). Make
282 sure you supply two digits after the initial zero if the pattern character that
283 follows is itself an octal digit.
284 .P
285 The handling of a backslash followed by a digit other than 0 is complicated.
286 Outside a character class, PCRE reads it and any following digits as a decimal
287 number. If the number is less than 10, or if there have been at least that many
288 previous capturing left parentheses in the expression, the entire sequence is
289 taken as a \fIback reference\fP. A description of how this works is given
290 .\" HTML <a href="#backreferences">
291 .\" </a>
292 later,
293 .\"
294 following the discussion of
295 .\" HTML <a href="#subpattern">
296 .\" </a>
297 parenthesized subpatterns.
298 .\"
299 .P
300 Inside a character class, or if the decimal number is greater than 9 and there
301 have not been that many capturing subpatterns, PCRE re-reads up to three octal
302 digits following the backslash, and uses them to generate a data character. Any
303 subsequent digits stand for themselves. In non-UTF-8 mode, the value of a
304 character specified in octal must be less than \e400. In UTF-8 mode, values up
305 to \e777 are permitted. For example:
306 .sp
307 \e040 is another way of writing a space
308 .\" JOIN
309 \e40 is the same, provided there are fewer than 40
310 previous capturing subpatterns
311 \e7 is always a back reference
312 .\" JOIN
313 \e11 might be a back reference, or another way of
314 writing a tab
315 \e011 is always a tab
316 \e0113 is a tab followed by the character "3"
317 .\" JOIN
318 \e113 might be a back reference, otherwise the
319 character with octal code 113
320 .\" JOIN
321 \e377 might be a back reference, otherwise
322 the byte consisting entirely of 1 bits
323 .\" JOIN
324 \e81 is either a back reference, or a binary zero
325 followed by the two characters "8" and "1"
326 .sp
327 Note that octal values of 100 or greater must not be introduced by a leading
328 zero, because no more than three octal digits are ever read.
329 .P
330 All the sequences that define a single character value can be used both inside
331 and outside character classes. In addition, inside a character class, the
332 sequence \eb is interpreted as the backspace character (hex 08). The sequences
333 \eB, \eN, \eR, and \eX are not special inside a character class. Like any other
334 unrecognized escape sequences, they are treated as the literal characters "B",
335 "N", "R", and "X" by default, but cause an error if the PCRE_EXTRA option is
336 set. Outside a character class, these sequences have different meanings.
337 .
338 .
339 .SS "Unsupported escape sequences"
340 .rs
341 .sp
342 In Perl, the sequences \el, \eL, \eu, and \eU are recognized by its string
343 handler and used to modify the case of following characters. By default, PCRE
344 does not support these escape sequences. However, if the PCRE_JAVASCRIPT_COMPAT
345 option is set, \eU matches a "U" character, and \eu can be used to define a
346 character by code point, as described in the previous section.
347 .
348 .
349 .SS "Absolute and relative back references"
350 .rs
351 .sp
352 The sequence \eg followed by an unsigned or a negative number, optionally
353 enclosed in braces, is an absolute or relative back reference. A named back
354 reference can be coded as \eg{name}. Back references are discussed
355 .\" HTML <a href="#backreferences">
356 .\" </a>
357 later,
358 .\"
359 following the discussion of
360 .\" HTML <a href="#subpattern">
361 .\" </a>
362 parenthesized subpatterns.
363 .\"
364 .
365 .
366 .SS "Absolute and relative subroutine calls"
367 .rs
368 .sp
369 For compatibility with Oniguruma, the non-Perl syntax \eg followed by a name or
370 a number enclosed either in angle brackets or single quotes, is an alternative
371 syntax for referencing a subpattern as a "subroutine". Details are discussed
372 .\" HTML <a href="#onigurumasubroutines">
373 .\" </a>
374 later.
375 .\"
376 Note that \eg{...} (Perl syntax) and \eg<...> (Oniguruma syntax) are \fInot\fP
377 synonymous. The former is a back reference; the latter is a
378 .\" HTML <a href="#subpatternsassubroutines">
379 .\" </a>
380 subroutine
381 .\"
382 call.
383 .
384 .
385 .\" HTML <a name="genericchartypes"></a>
386 .SS "Generic character types"
387 .rs
388 .sp
389 Another use of backslash is for specifying generic character types:
390 .sp
391 \ed any decimal digit
392 \eD any character that is not a decimal digit
393 \eh any horizontal whitespace character
394 \eH any character that is not a horizontal whitespace character
395 \es any whitespace character
396 \eS any character that is not a whitespace character
397 \ev any vertical whitespace character
398 \eV any character that is not a vertical whitespace character
399 \ew any "word" character
400 \eW any "non-word" character
401 .sp
402 There is also the single sequence \eN, which matches a non-newline character.
403 This is the same as
404 .\" HTML <a href="#fullstopdot">
405 .\" </a>
406 the "." metacharacter
407 .\"
408 when PCRE_DOTALL is not set. Perl also uses \eN to match characters by name;
409 PCRE does not support this.
410 .P
411 Each pair of lower and upper case escape sequences partitions the complete set
412 of characters into two disjoint sets. Any given character matches one, and only
413 one, of each pair. The sequences can appear both inside and outside character
414 classes. They each match one character of the appropriate type. If the current
415 matching point is at the end of the subject string, all of them fail, because
416 there is no character to match.
417 .P
418 For compatibility with Perl, \es does not match the VT character (code 11).
419 This makes it different from the the POSIX "space" class. The \es characters
420 are HT (9), LF (10), FF (12), CR (13), and space (32). If "use locale;" is
421 included in a Perl script, \es may match the VT character. In PCRE, it never
422 does.
423 .P
424 A "word" character is an underscore or any character that is a letter or digit.
425 By default, the definition of letters and digits is controlled by PCRE's
426 low-valued character tables, and may vary if locale-specific matching is taking
427 place (see
428 .\" HTML <a href="pcreapi.html#localesupport">
429 .\" </a>
430 "Locale support"
431 .\"
432 in the
433 .\" HREF
434 \fBpcreapi\fP
435 .\"
436 page). For example, in a French locale such as "fr_FR" in Unix-like systems,
437 or "french" in Windows, some character codes greater than 128 are used for
438 accented letters, and these are then matched by \ew. The use of locales with
439 Unicode is discouraged.
440 .P
441 By default, in UTF-8 mode, characters with values greater than 128 never match
442 \ed, \es, or \ew, and always match \eD, \eS, and \eW. These sequences retain
443 their original meanings from before UTF-8 support was available, mainly for
444 efficiency reasons. However, if PCRE is compiled with Unicode property support,
445 and the PCRE_UCP option is set, the behaviour is changed so that Unicode
446 properties are used to determine character types, as follows:
447 .sp
448 \ed any character that \ep{Nd} matches (decimal digit)
449 \es any character that \ep{Z} matches, plus HT, LF, FF, CR
450 \ew any character that \ep{L} or \ep{N} matches, plus underscore
451 .sp
452 The upper case escapes match the inverse sets of characters. Note that \ed
453 matches only decimal digits, whereas \ew matches any Unicode digit, as well as
454 any Unicode letter, and underscore. Note also that PCRE_UCP affects \eb, and
455 \eB because they are defined in terms of \ew and \eW. Matching these sequences
456 is noticeably slower when PCRE_UCP is set.
457 .P
458 The sequences \eh, \eH, \ev, and \eV are features that were added to Perl at
459 release 5.10. In contrast to the other sequences, which match only ASCII
460 characters by default, these always match certain high-valued codepoints in
461 UTF-8 mode, whether or not PCRE_UCP is set. The horizontal space characters
462 are:
463 .sp
464 U+0009 Horizontal tab
465 U+0020 Space
466 U+00A0 Non-break space
467 U+1680 Ogham space mark
468 U+180E Mongolian vowel separator
469 U+2000 En quad
470 U+2001 Em quad
471 U+2002 En space
472 U+2003 Em space
473 U+2004 Three-per-em space
474 U+2005 Four-per-em space
475 U+2006 Six-per-em space
476 U+2007 Figure space
477 U+2008 Punctuation space
478 U+2009 Thin space
479 U+200A Hair space
480 U+202F Narrow no-break space
481 U+205F Medium mathematical space
482 U+3000 Ideographic space
483 .sp
484 The vertical space characters are:
485 .sp
486 U+000A Linefeed
487 U+000B Vertical tab
488 U+000C Formfeed
489 U+000D Carriage return
490 U+0085 Next line
491 U+2028 Line separator
492 U+2029 Paragraph separator
493 .
494 .
495 .\" HTML <a name="newlineseq"></a>
496 .SS "Newline sequences"
497 .rs
498 .sp
499 Outside a character class, by default, the escape sequence \eR matches any
500 Unicode newline sequence. In non-UTF-8 mode \eR is equivalent to the following:
501 .sp
502 (?>\er\en|\en|\ex0b|\ef|\er|\ex85)
503 .sp
504 This is an example of an "atomic group", details of which are given
505 .\" HTML <a href="#atomicgroup">
506 .\" </a>
507 below.
508 .\"
509 This particular group matches either the two-character sequence CR followed by
510 LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab,
511 U+000B), FF (formfeed, U+000C), CR (carriage return, U+000D), or NEL (next
512 line, U+0085). The two-character sequence is treated as a single unit that
513 cannot be split.
514 .P
515 In UTF-8 mode, two additional characters whose codepoints are greater than 255
516 are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029).
517 Unicode character property support is not needed for these characters to be
518 recognized.
519 .P
520 It is possible to restrict \eR to match only CR, LF, or CRLF (instead of the
521 complete set of Unicode line endings) by setting the option PCRE_BSR_ANYCRLF
522 either at compile time or when the pattern is matched. (BSR is an abbrevation
523 for "backslash R".) This can be made the default when PCRE is built; if this is
524 the case, the other behaviour can be requested via the PCRE_BSR_UNICODE option.
525 It is also possible to specify these settings by starting a pattern string with
526 one of the following sequences:
527 .sp
528 (*BSR_ANYCRLF) CR, LF, or CRLF only
529 (*BSR_UNICODE) any Unicode newline sequence
530 .sp
531 These override the default and the options given to \fBpcre_compile()\fP or
532 \fBpcre_compile2()\fP, but they can be overridden by options given to
533 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. Note that these special settings,
534 which are not Perl-compatible, are recognized only at the very start of a
535 pattern, and that they must be in upper case. If more than one of them is
536 present, the last one is used. They can be combined with a change of newline
537 convention; for example, a pattern can start with:
538 .sp
539 (*ANY)(*BSR_ANYCRLF)
540 .sp
541 They can also be combined with the (*UTF8) or (*UCP) special sequences. Inside
542 a character class, \eR is treated as an unrecognized escape sequence, and so
543 matches the letter "R" by default, but causes an error if PCRE_EXTRA is set.
544 .
545 .
546 .\" HTML <a name="uniextseq"></a>
547 .SS Unicode character properties
548 .rs
549 .sp
550 When PCRE is built with Unicode character property support, three additional
551 escape sequences that match characters with specific properties are available.
552 When not in UTF-8 mode, these sequences are of course limited to testing
553 characters whose codepoints are less than 256, but they do work in this mode.
554 The extra escape sequences are:
555 .sp
556 \ep{\fIxx\fP} a character with the \fIxx\fP property
557 \eP{\fIxx\fP} a character without the \fIxx\fP property
558 \eX an extended Unicode sequence
559 .sp
560 The property names represented by \fIxx\fP above are limited to the Unicode
561 script names, the general category properties, "Any", which matches any
562 character (including newline), and some special PCRE properties (described
563 in the
564 .\" HTML <a href="#extraprops">
565 .\" </a>
566 next section).
567 .\"
568 Other Perl properties such as "InMusicalSymbols" are not currently supported by
569 PCRE. Note that \eP{Any} does not match any characters, so always causes a
570 match failure.
571 .P
572 Sets of Unicode characters are defined as belonging to certain scripts. A
573 character from one of these sets can be matched using a script name. For
574 example:
575 .sp
576 \ep{Greek}
577 \eP{Han}
578 .sp
579 Those that are not part of an identified script are lumped together as
580 "Common". The current list of scripts is:
581 .P
582 Arabic,
583 Armenian,
584 Avestan,
585 Balinese,
586 Bamum,
587 Bengali,
588 Bopomofo,
589 Braille,
590 Buginese,
591 Buhid,
592 Canadian_Aboriginal,
593 Carian,
594 Cham,
595 Cherokee,
596 Common,
597 Coptic,
598 Cuneiform,
599 Cypriot,
600 Cyrillic,
601 Deseret,
602 Devanagari,
603 Egyptian_Hieroglyphs,
604 Ethiopic,
605 Georgian,
606 Glagolitic,
607 Gothic,
608 Greek,
609 Gujarati,
610 Gurmukhi,
611 Han,
612 Hangul,
613 Hanunoo,
614 Hebrew,
615 Hiragana,
616 Imperial_Aramaic,
617 Inherited,
618 Inscriptional_Pahlavi,
619 Inscriptional_Parthian,
620 Javanese,
621 Kaithi,
622 Kannada,
623 Katakana,
624 Kayah_Li,
625 Kharoshthi,
626 Khmer,
627 Lao,
628 Latin,
629 Lepcha,
630 Limbu,
631 Linear_B,
632 Lisu,
633 Lycian,
634 Lydian,
635 Malayalam,
636 Meetei_Mayek,
637 Mongolian,
638 Myanmar,
639 New_Tai_Lue,
640 Nko,
641 Ogham,
642 Old_Italic,
643 Old_Persian,
644 Old_South_Arabian,
645 Old_Turkic,
646 Ol_Chiki,
647 Oriya,
648 Osmanya,
649 Phags_Pa,
650 Phoenician,
651 Rejang,
652 Runic,
653 Samaritan,
654 Saurashtra,
655 Shavian,
656 Sinhala,
657 Sundanese,
658 Syloti_Nagri,
659 Syriac,
660 Tagalog,
661 Tagbanwa,
662 Tai_Le,
663 Tai_Tham,
664 Tai_Viet,
665 Tamil,
666 Telugu,
667 Thaana,
668 Thai,
669 Tibetan,
670 Tifinagh,
671 Ugaritic,
672 Vai,
673 Yi.
674 .P
675 Each character has exactly one Unicode general category property, specified by
676 a two-letter abbreviation. For compatibility with Perl, negation can be
677 specified by including a circumflex between the opening brace and the property
678 name. For example, \ep{^Lu} is the same as \eP{Lu}.
679 .P
680 If only one letter is specified with \ep or \eP, it includes all the general
681 category properties that start with that letter. In this case, in the absence
682 of negation, the curly brackets in the escape sequence are optional; these two
683 examples have the same effect:
684 .sp
685 \ep{L}
686 \epL
687 .sp
688 The following general category property codes are supported:
689 .sp
690 C Other
691 Cc Control
692 Cf Format
693 Cn Unassigned
694 Co Private use
695 Cs Surrogate
696 .sp
697 L Letter
698 Ll Lower case letter
699 Lm Modifier letter
700 Lo Other letter
701 Lt Title case letter
702 Lu Upper case letter
703 .sp
704 M Mark
705 Mc Spacing mark
706 Me Enclosing mark
707 Mn Non-spacing mark
708 .sp
709 N Number
710 Nd Decimal number
711 Nl Letter number
712 No Other number
713 .sp
714 P Punctuation
715 Pc Connector punctuation
716 Pd Dash punctuation
717 Pe Close punctuation
718 Pf Final punctuation
719 Pi Initial punctuation
720 Po Other punctuation
721 Ps Open punctuation
722 .sp
723 S Symbol
724 Sc Currency symbol
725 Sk Modifier symbol
726 Sm Mathematical symbol
727 So Other symbol
728 .sp
729 Z Separator
730 Zl Line separator
731 Zp Paragraph separator
732 Zs Space separator
733 .sp
734 The special property L& is also supported: it matches a character that has
735 the Lu, Ll, or Lt property, in other words, a letter that is not classified as
736 a modifier or "other".
737 .P
738 The Cs (Surrogate) property applies only to characters in the range U+D800 to
739 U+DFFF. Such characters are not valid in UTF-8 strings (see RFC 3629) and so
740 cannot be tested by PCRE, unless UTF-8 validity checking has been turned off
741 (see the discussion of PCRE_NO_UTF8_CHECK in the
742 .\" HREF
743 \fBpcreapi\fP
744 .\"
745 page). Perl does not support the Cs property.
746 .P
747 The long synonyms for property names that Perl supports (such as \ep{Letter})
748 are not supported by PCRE, nor is it permitted to prefix any of these
749 properties with "Is".
750 .P
751 No character that is in the Unicode table has the Cn (unassigned) property.
752 Instead, this property is assumed for any code point that is not in the
753 Unicode table.
754 .P
755 Specifying caseless matching does not affect these escape sequences. For
756 example, \ep{Lu} always matches only upper case letters.
757 .P
758 The \eX escape matches any number of Unicode characters that form an extended
759 Unicode sequence. \eX is equivalent to
760 .sp
761 (?>\ePM\epM*)
762 .sp
763 That is, it matches a character without the "mark" property, followed by zero
764 or more characters with the "mark" property, and treats the sequence as an
765 atomic group
766 .\" HTML <a href="#atomicgroup">
767 .\" </a>
768 (see below).
769 .\"
770 Characters with the "mark" property are typically accents that affect the
771 preceding character. None of them have codepoints less than 256, so in
772 non-UTF-8 mode \eX matches any one character.
773 .P
774 Note that recent versions of Perl have changed \eX to match what Unicode calls
775 an "extended grapheme cluster", which has a more complicated definition.
776 .P
777 Matching characters by Unicode property is not fast, because PCRE has to search
778 a structure that contains data for over fifteen thousand characters. That is
779 why the traditional escape sequences such as \ed and \ew do not use Unicode
780 properties in PCRE by default, though you can make them do so by setting the
781 PCRE_UCP option for \fBpcre_compile()\fP or by starting the pattern with
782 (*UCP).
783 .
784 .
785 .\" HTML <a name="extraprops"></a>
786 .SS PCRE's additional properties
787 .rs
788 .sp
789 As well as the standard Unicode properties described in the previous
790 section, PCRE supports four more that make it possible to convert traditional
791 escape sequences such as \ew and \es and POSIX character classes to use Unicode
792 properties. PCRE uses these non-standard, non-Perl properties internally when
793 PCRE_UCP is set. They are:
794 .sp
795 Xan Any alphanumeric character
796 Xps Any POSIX space character
797 Xsp Any Perl space character
798 Xwd Any Perl "word" character
799 .sp
800 Xan matches characters that have either the L (letter) or the N (number)
801 property. Xps matches the characters tab, linefeed, vertical tab, formfeed, or
802 carriage return, and any other character that has the Z (separator) property.
803 Xsp is the same as Xps, except that vertical tab is excluded. Xwd matches the
804 same characters as Xan, plus underscore.
805 .
806 .
807 .\" HTML <a name="resetmatchstart"></a>
808 .SS "Resetting the match start"
809 .rs
810 .sp
811 The escape sequence \eK causes any previously matched characters not to be
812 included in the final matched sequence. For example, the pattern:
813 .sp
814 foo\eKbar
815 .sp
816 matches "foobar", but reports that it has matched "bar". This feature is
817 similar to a lookbehind assertion
818 .\" HTML <a href="#lookbehind">
819 .\" </a>
820 (described below).
821 .\"
822 However, in this case, the part of the subject before the real match does not
823 have to be of fixed length, as lookbehind assertions do. The use of \eK does
824 not interfere with the setting of
825 .\" HTML <a href="#subpattern">
826 .\" </a>
827 captured substrings.
828 .\"
829 For example, when the pattern
830 .sp
831 (foo)\eKbar
832 .sp
833 matches "foobar", the first substring is still set to "foo".
834 .P
835 Perl documents that the use of \eK within assertions is "not well defined". In
836 PCRE, \eK is acted upon when it occurs inside positive assertions, but is
837 ignored in negative assertions.
838 .
839 .
840 .\" HTML <a name="smallassertions"></a>
841 .SS "Simple assertions"
842 .rs
843 .sp
844 The final use of backslash is for certain simple assertions. An assertion
845 specifies a condition that has to be met at a particular point in a match,
846 without consuming any characters from the subject string. The use of
847 subpatterns for more complicated assertions is described
848 .\" HTML <a href="#bigassertions">
849 .\" </a>
850 below.
851 .\"
852 The backslashed assertions are:
853 .sp
854 \eb matches at a word boundary
855 \eB matches when not at a word boundary
856 \eA matches at the start of the subject
857 \eZ matches at the end of the subject
858 also matches before a newline at the end of the subject
859 \ez matches only at the end of the subject
860 \eG matches at the first matching position in the subject
861 .sp
862 Inside a character class, \eb has a different meaning; it matches the backspace
863 character. If any other of these assertions appears in a character class, by
864 default it matches the corresponding literal character (for example, \eB
865 matches the letter B). However, if the PCRE_EXTRA option is set, an "invalid
866 escape sequence" error is generated instead.
867 .P
868 A word boundary is a position in the subject string where the current character
869 and the previous character do not both match \ew or \eW (i.e. one matches
870 \ew and the other matches \eW), or the start or end of the string if the
871 first or last character matches \ew, respectively. In UTF-8 mode, the meanings
872 of \ew and \eW can be changed by setting the PCRE_UCP option. When this is
873 done, it also affects \eb and \eB. Neither PCRE nor Perl has a separate "start
874 of word" or "end of word" metasequence. However, whatever follows \eb normally
875 determines which it is. For example, the fragment \eba matches "a" at the start
876 of a word.
877 .P
878 The \eA, \eZ, and \ez assertions differ from the traditional circumflex and
879 dollar (described in the next section) in that they only ever match at the very
880 start and end of the subject string, whatever options are set. Thus, they are
881 independent of multiline mode. These three assertions are not affected by the
882 PCRE_NOTBOL or PCRE_NOTEOL options, which affect only the behaviour of the
883 circumflex and dollar metacharacters. However, if the \fIstartoffset\fP
884 argument of \fBpcre_exec()\fP is non-zero, indicating that matching is to start
885 at a point other than the beginning of the subject, \eA can never match. The
886 difference between \eZ and \ez is that \eZ matches before a newline at the end
887 of the string as well as at the very end, whereas \ez matches only at the end.
888 .P
889 The \eG assertion is true only when the current matching position is at the
890 start point of the match, as specified by the \fIstartoffset\fP argument of
891 \fBpcre_exec()\fP. It differs from \eA when the value of \fIstartoffset\fP is
892 non-zero. By calling \fBpcre_exec()\fP multiple times with appropriate
893 arguments, you can mimic Perl's /g option, and it is in this kind of
894 implementation where \eG can be useful.
895 .P
896 Note, however, that PCRE's interpretation of \eG, as the start of the current
897 match, is subtly different from Perl's, which defines it as the end of the
898 previous match. In Perl, these can be different when the previously matched
899 string was empty. Because PCRE does just one match at a time, it cannot
900 reproduce this behaviour.
901 .P
902 If all the alternatives of a pattern begin with \eG, the expression is anchored
903 to the starting match position, and the "anchored" flag is set in the compiled
904 regular expression.
905 .
906 .
907 .SH "CIRCUMFLEX AND DOLLAR"
908 .rs
909 .sp
910 Outside a character class, in the default matching mode, the circumflex
911 character is an assertion that is true only if the current matching point is
912 at the start of the subject string. If the \fIstartoffset\fP argument of
913 \fBpcre_exec()\fP is non-zero, circumflex can never match if the PCRE_MULTILINE
914 option is unset. Inside a character class, circumflex has an entirely different
915 meaning
916 .\" HTML <a href="#characterclass">
917 .\" </a>
918 (see below).
919 .\"
920 .P
921 Circumflex need not be the first character of the pattern if a number of
922 alternatives are involved, but it should be the first thing in each alternative
923 in which it appears if the pattern is ever to match that branch. If all
924 possible alternatives start with a circumflex, that is, if the pattern is
925 constrained to match only at the start of the subject, it is said to be an
926 "anchored" pattern. (There are also other constructs that can cause a pattern
927 to be anchored.)
928 .P
929 A dollar character is an assertion that is true only if the current matching
930 point is at the end of the subject string, or immediately before a newline
931 at the end of the string (by default). Dollar need not be the last character of
932 the pattern if a number of alternatives are involved, but it should be the last
933 item in any branch in which it appears. Dollar has no special meaning in a
934 character class.
935 .P
936 The meaning of dollar can be changed so that it matches only at the very end of
937 the string, by setting the PCRE_DOLLAR_ENDONLY option at compile time. This
938 does not affect the \eZ assertion.
939 .P
940 The meanings of the circumflex and dollar characters are changed if the
941 PCRE_MULTILINE option is set. When this is the case, a circumflex matches
942 immediately after internal newlines as well as at the start of the subject
943 string. It does not match after a newline that ends the string. A dollar
944 matches before any newlines in the string, as well as at the very end, when
945 PCRE_MULTILINE is set. When newline is specified as the two-character
946 sequence CRLF, isolated CR and LF characters do not indicate newlines.
947 .P
948 For example, the pattern /^abc$/ matches the subject string "def\enabc" (where
949 \en represents a newline) in multiline mode, but not otherwise. Consequently,
950 patterns that are anchored in single line mode because all branches start with
951 ^ are not anchored in multiline mode, and a match for circumflex is possible
952 when the \fIstartoffset\fP argument of \fBpcre_exec()\fP is non-zero. The
953 PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
954 .P
955 Note that the sequences \eA, \eZ, and \ez can be used to match the start and
956 end of the subject in both modes, and if all branches of a pattern start with
957 \eA it is always anchored, whether or not PCRE_MULTILINE is set.
958 .
959 .
960 .\" HTML <a name="fullstopdot"></a>
961 .SH "FULL STOP (PERIOD, DOT) AND \eN"
962 .rs
963 .sp
964 Outside a character class, a dot in the pattern matches any one character in
965 the subject string except (by default) a character that signifies the end of a
966 line. In UTF-8 mode, the matched character may be more than one byte long.
967 .P
968 When a line ending is defined as a single character, dot never matches that
969 character; when the two-character sequence CRLF is used, dot does not match CR
970 if it is immediately followed by LF, but otherwise it matches all characters
971 (including isolated CRs and LFs). When any Unicode line endings are being
972 recognized, dot does not match CR or LF or any of the other line ending
973 characters.
974 .P
975 The behaviour of dot with regard to newlines can be changed. If the PCRE_DOTALL
976 option is set, a dot matches any one character, without exception. If the
977 two-character sequence CRLF is present in the subject string, it takes two dots
978 to match it.
979 .P
980 The handling of dot is entirely independent of the handling of circumflex and
981 dollar, the only relationship being that they both involve newlines. Dot has no
982 special meaning in a character class.
983 .P
984 The escape sequence \eN behaves like a dot, except that it is not affected by
985 the PCRE_DOTALL option. In other words, it matches any character except one
986 that signifies the end of a line. Perl also uses \eN to match characters by
987 name; PCRE does not support this.
988 .
989 .
990 .SH "MATCHING A SINGLE BYTE"
991 .rs
992 .sp
993 Outside a character class, the escape sequence \eC matches any one byte, both
994 in and out of UTF-8 mode. Unlike a dot, it always matches line-ending
995 characters. The feature is provided in Perl in order to match individual bytes
996 in UTF-8 mode, but it is unclear how it can usefully be used. Because \eC
997 breaks up characters into individual bytes, matching one byte with \eC in UTF-8
998 mode means that the rest of the string may start with a malformed UTF-8
999 character. This has undefined results, because PCRE assumes that it is dealing
1000 with valid UTF-8 strings (and by default it checks this at the start of
1001 processing unless the PCRE_NO_UTF8_CHECK option is used).
1002 .P
1003 PCRE does not allow \eC to appear in lookbehind assertions
1004 .\" HTML <a href="#lookbehind">
1005 .\" </a>
1006 (described below),
1007 .\"
1008 because in UTF-8 mode this would make it impossible to calculate the length of
1009 the lookbehind.
1010 .P
1011 In general, the \eC escape sequence is best avoided in UTF-8 mode. However, one
1012 way of using it that avoids the problem of malformed UTF-8 characters is to
1013 use a lookahead to check the length of the next character, as in this pattern
1014 (ignore white space and line breaks):
1015 .sp
1016 (?| (?=[\ex00-\ex7f])(\eC) |
1017 (?=[\ex80-\ex{7ff}])(\eC)(\eC) |
1018 (?=[\ex{800}-\ex{ffff}])(\eC)(\eC)(\eC) |
1019 (?=[\ex{10000}-\ex{1fffff}])(\eC)(\eC)(\eC)(\eC))
1020 .sp
1021 A group that starts with (?| resets the capturing parentheses numbers in each
1022 alternative (see
1023 .\" HTML <a href="#dupsubpatternnumber">
1024 .\" </a>
1025 "Duplicate Subpattern Numbers"
1026 .\"
1027 below). The assertions at the start of each branch check the next UTF-8
1028 character for values whose encoding uses 1, 2, 3, or 4 bytes, respectively. The
1029 character's individual bytes are then captured by the appropriate number of
1030 groups.
1031 .
1032 .
1033 .\" HTML <a name="characterclass"></a>
1034 .SH "SQUARE BRACKETS AND CHARACTER CLASSES"
1035 .rs
1036 .sp
1037 An opening square bracket introduces a character class, terminated by a closing
1038 square bracket. A closing square bracket on its own is not special by default.
1039 However, if the PCRE_JAVASCRIPT_COMPAT option is set, a lone closing square
1040 bracket causes a compile-time error. If a closing square bracket is required as
1041 a member of the class, it should be the first data character in the class
1042 (after an initial circumflex, if present) or escaped with a backslash.
1043 .P
1044 A character class matches a single character in the subject. In UTF-8 mode, the
1045 character may be more than one byte long. A matched character must be in the
1046 set of characters defined by the class, unless the first character in the class
1047 definition is a circumflex, in which case the subject character must not be in
1048 the set defined by the class. If a circumflex is actually required as a member
1049 of the class, ensure it is not the first character, or escape it with a
1050 backslash.
1051 .P
1052 For example, the character class [aeiou] matches any lower case vowel, while
1053 [^aeiou] matches any character that is not a lower case vowel. Note that a
1054 circumflex is just a convenient notation for specifying the characters that
1055 are in the class by enumerating those that are not. A class that starts with a
1056 circumflex is not an assertion; it still consumes a character from the subject
1057 string, and therefore it fails if the current pointer is at the end of the
1058 string.
1059 .P
1060 In UTF-8 mode, characters with values greater than 255 can be included in a
1061 class as a literal string of bytes, or by using the \ex{ escaping mechanism.
1062 .P
1063 When caseless matching is set, any letters in a class represent both their
1064 upper case and lower case versions, so for example, a caseless [aeiou] matches
1065 "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
1066 caseful version would. In UTF-8 mode, PCRE always understands the concept of
1067 case for characters whose values are less than 128, so caseless matching is
1068 always possible. For characters with higher values, the concept of case is
1069 supported if PCRE is compiled with Unicode property support, but not otherwise.
1070 If you want to use caseless matching in UTF8-mode for characters 128 and above,
1071 you must ensure that PCRE is compiled with Unicode property support as well as
1072 with UTF-8 support.
1073 .P
1074 Characters that might indicate line breaks are never treated in any special way
1075 when matching character classes, whatever line-ending sequence is in use, and
1076 whatever setting of the PCRE_DOTALL and PCRE_MULTILINE options is used. A class
1077 such as [^a] always matches one of these characters.
1078 .P
1079 The minus (hyphen) character can be used to specify a range of characters in a
1080 character class. For example, [d-m] matches any letter between d and m,
1081 inclusive. If a minus character is required in a class, it must be escaped with
1082 a backslash or appear in a position where it cannot be interpreted as
1083 indicating a range, typically as the first or last character in the class.
1084 .P
1085 It is not possible to have the literal character "]" as the end character of a
1086 range. A pattern such as [W-]46] is interpreted as a class of two characters
1087 ("W" and "-") followed by a literal string "46]", so it would match "W46]" or
1088 "-46]". However, if the "]" is escaped with a backslash it is interpreted as
1089 the end of range, so [W-\e]46] is interpreted as a class containing a range
1090 followed by two other characters. The octal or hexadecimal representation of
1091 "]" can also be used to end a range.
1092 .P
1093 Ranges operate in the collating sequence of character values. They can also be
1094 used for characters specified numerically, for example [\e000-\e037]. In UTF-8
1095 mode, ranges can include characters whose values are greater than 255, for
1096 example [\ex{100}-\ex{2ff}].
1097 .P
1098 If a range that includes letters is used when caseless matching is set, it
1099 matches the letters in either case. For example, [W-c] is equivalent to
1100 [][\e\e^_`wxyzabc], matched caselessly, and in non-UTF-8 mode, if character
1101 tables for a French locale are in use, [\exc8-\excb] matches accented E
1102 characters in both cases. In UTF-8 mode, PCRE supports the concept of case for
1103 characters with values greater than 128 only when it is compiled with Unicode
1104 property support.
1105 .P
1106 The character escape sequences \ed, \eD, \eh, \eH, \ep, \eP, \es, \eS, \ev,
1107 \eV, \ew, and \eW may appear in a character class, and add the characters that
1108 they match to the class. For example, [\edABCDEF] matches any hexadecimal
1109 digit. In UTF-8 mode, the PCRE_UCP option affects the meanings of \ed, \es, \ew
1110 and their upper case partners, just as it does when they appear outside a
1111 character class, as described in the section entitled
1112 .\" HTML <a href="#genericchartypes">
1113 .\" </a>
1114 "Generic character types"
1115 .\"
1116 above. The escape sequence \eb has a different meaning inside a character
1117 class; it matches the backspace character. The sequences \eB, \eN, \eR, and \eX
1118 are not special inside a character class. Like any other unrecognized escape
1119 sequences, they are treated as the literal characters "B", "N", "R", and "X" by
1120 default, but cause an error if the PCRE_EXTRA option is set.
1121 .P
1122 A circumflex can conveniently be used with the upper case character types to
1123 specify a more restricted set of characters than the matching lower case type.
1124 For example, the class [^\eW_] matches any letter or digit, but not underscore,
1125 whereas [\ew] includes underscore. A positive character class should be read as
1126 "something OR something OR ..." and a negative class as "NOT something AND NOT
1127 something AND NOT ...".
1128 .P
1129 The only metacharacters that are recognized in character classes are backslash,
1130 hyphen (only where it can be interpreted as specifying a range), circumflex
1131 (only at the start), opening square bracket (only when it can be interpreted as
1132 introducing a POSIX class name - see the next section), and the terminating
1133 closing square bracket. However, escaping other non-alphanumeric characters
1134 does no harm.
1135 .
1136 .
1137 .SH "POSIX CHARACTER CLASSES"
1138 .rs
1139 .sp
1140 Perl supports the POSIX notation for character classes. This uses names
1141 enclosed by [: and :] within the enclosing square brackets. PCRE also supports
1142 this notation. For example,
1143 .sp
1144 [01[:alpha:]%]
1145 .sp
1146 matches "0", "1", any alphabetic character, or "%". The supported class names
1147 are:
1148 .sp
1149 alnum letters and digits
1150 alpha letters
1151 ascii character codes 0 - 127
1152 blank space or tab only
1153 cntrl control characters
1154 digit decimal digits (same as \ed)
1155 graph printing characters, excluding space
1156 lower lower case letters
1157 print printing characters, including space
1158 punct printing characters, excluding letters and digits and space
1159 space white space (not quite the same as \es)
1160 upper upper case letters
1161 word "word" characters (same as \ew)
1162 xdigit hexadecimal digits
1163 .sp
1164 The "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13), and
1165 space (32). Notice that this list includes the VT character (code 11). This
1166 makes "space" different to \es, which does not include VT (for Perl
1167 compatibility).
1168 .P
1169 The name "word" is a Perl extension, and "blank" is a GNU extension from Perl
1170 5.8. Another Perl extension is negation, which is indicated by a ^ character
1171 after the colon. For example,
1172 .sp
1173 [12[:^digit:]]
1174 .sp
1175 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX
1176 syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
1177 supported, and an error is given if they are encountered.
1178 .P
1179 By default, in UTF-8 mode, characters with values greater than 128 do not match
1180 any of the POSIX character classes. However, if the PCRE_UCP option is passed
1181 to \fBpcre_compile()\fP, some of the classes are changed so that Unicode
1182 character properties are used. This is achieved by replacing the POSIX classes
1183 by other sequences, as follows:
1184 .sp
1185 [:alnum:] becomes \ep{Xan}
1186 [:alpha:] becomes \ep{L}
1187 [:blank:] becomes \eh
1188 [:digit:] becomes \ep{Nd}
1189 [:lower:] becomes \ep{Ll}
1190 [:space:] becomes \ep{Xps}
1191 [:upper:] becomes \ep{Lu}
1192 [:word:] becomes \ep{Xwd}
1193 .sp
1194 Negated versions, such as [:^alpha:] use \eP instead of \ep. The other POSIX
1195 classes are unchanged, and match only characters with code points less than
1196 128.
1197 .
1198 .
1199 .SH "VERTICAL BAR"
1200 .rs
1201 .sp
1202 Vertical bar characters are used to separate alternative patterns. For example,
1203 the pattern
1204 .sp
1205 gilbert|sullivan
1206 .sp
1207 matches either "gilbert" or "sullivan". Any number of alternatives may appear,
1208 and an empty alternative is permitted (matching the empty string). The matching
1209 process tries each alternative in turn, from left to right, and the first one
1210 that succeeds is used. If the alternatives are within a subpattern
1211 .\" HTML <a href="#subpattern">
1212 .\" </a>
1213 (defined below),
1214 .\"
1215 "succeeds" means matching the rest of the main pattern as well as the
1216 alternative in the subpattern.
1217 .
1218 .
1219 .SH "INTERNAL OPTION SETTING"
1220 .rs
1221 .sp
1222 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
1223 PCRE_EXTENDED options (which are Perl-compatible) can be changed from within
1224 the pattern by a sequence of Perl option letters enclosed between "(?" and ")".
1225 The option letters are
1226 .sp
1227 i for PCRE_CASELESS
1228 m for PCRE_MULTILINE
1229 s for PCRE_DOTALL
1230 x for PCRE_EXTENDED
1231 .sp
1232 For example, (?im) sets caseless, multiline matching. It is also possible to
1233 unset these options by preceding the letter with a hyphen, and a combined
1234 setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
1235 PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
1236 permitted. If a letter appears both before and after the hyphen, the option is
1237 unset.
1238 .P
1239 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA can be
1240 changed in the same way as the Perl-compatible options by using the characters
1241 J, U and X respectively.
1242 .P
1243 When one of these option changes occurs at top level (that is, not inside
1244 subpattern parentheses), the change applies to the remainder of the pattern
1245 that follows. If the change is placed right at the start of a pattern, PCRE
1246 extracts it into the global options (and it will therefore show up in data
1247 extracted by the \fBpcre_fullinfo()\fP function).
1248 .P
1249 An option change within a subpattern (see below for a description of
1250 subpatterns) affects only that part of the subpattern that follows it, so
1251 .sp
1252 (a(?i)b)c
1253 .sp
1254 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
1255 By this means, options can be made to have different settings in different
1256 parts of the pattern. Any changes made in one alternative do carry on
1257 into subsequent branches within the same subpattern. For example,
1258 .sp
1259 (a(?i)b|c)
1260 .sp
1261 matches "ab", "aB", "c", and "C", even though when matching "C" the first
1262 branch is abandoned before the option setting. This is because the effects of
1263 option settings happen at compile time. There would be some very weird
1264 behaviour otherwise.
1265 .P
1266 \fBNote:\fP There are other PCRE-specific options that can be set by the
1267 application when the compile or match functions are called. In some cases the
1268 pattern can contain special leading sequences such as (*CRLF) to override what
1269 the application has set or what has been defaulted. Details are given in the
1270 section entitled
1271 .\" HTML <a href="#newlineseq">
1272 .\" </a>
1273 "Newline sequences"
1274 .\"
1275 above. There are also the (*UTF8) and (*UCP) leading sequences that can be used
1276 to set UTF-8 and Unicode property modes; they are equivalent to setting the
1277 PCRE_UTF8 and the PCRE_UCP options, respectively.
1278 .
1279 .
1280 .\" HTML <a name="subpattern"></a>
1281 .SH SUBPATTERNS
1282 .rs
1283 .sp
1284 Subpatterns are delimited by parentheses (round brackets), which can be nested.
1285 Turning part of a pattern into a subpattern does two things:
1286 .sp
1287 1. It localizes a set of alternatives. For example, the pattern
1288 .sp
1289 cat(aract|erpillar|)
1290 .sp
1291 matches "cataract", "caterpillar", or "cat". Without the parentheses, it would
1292 match "cataract", "erpillar" or an empty string.
1293 .sp
1294 2. It sets up the subpattern as a capturing subpattern. This means that, when
1295 the whole pattern matches, that portion of the subject string that matched the
1296 subpattern is passed back to the caller via the \fIovector\fP argument of
1297 \fBpcre_exec()\fP. Opening parentheses are counted from left to right (starting
1298 from 1) to obtain numbers for the capturing subpatterns. For example, if the
1299 string "the red king" is matched against the pattern
1300 .sp
1301 the ((red|white) (king|queen))
1302 .sp
1303 the captured substrings are "red king", "red", and "king", and are numbered 1,
1304 2, and 3, respectively.
1305 .P
1306 The fact that plain parentheses fulfil two functions is not always helpful.
1307 There are often times when a grouping subpattern is required without a
1308 capturing requirement. If an opening parenthesis is followed by a question mark
1309 and a colon, the subpattern does not do any capturing, and is not counted when
1310 computing the number of any subsequent capturing subpatterns. For example, if
1311 the string "the white queen" is matched against the pattern
1312 .sp
1313 the ((?:red|white) (king|queen))
1314 .sp
1315 the captured substrings are "white queen" and "queen", and are numbered 1 and
1316 2. The maximum number of capturing subpatterns is 65535.
1317 .P
1318 As a convenient shorthand, if any option settings are required at the start of
1319 a non-capturing subpattern, the option letters may appear between the "?" and
1320 the ":". Thus the two patterns
1321 .sp
1322 (?i:saturday|sunday)
1323 (?:(?i)saturday|sunday)
1324 .sp
1325 match exactly the same set of strings. Because alternative branches are tried
1326 from left to right, and options are not reset until the end of the subpattern
1327 is reached, an option setting in one branch does affect subsequent branches, so
1328 the above patterns match "SUNDAY" as well as "Saturday".
1329 .
1330 .
1331 .\" HTML <a name="dupsubpatternnumber"></a>
1332 .SH "DUPLICATE SUBPATTERN NUMBERS"
1333 .rs
1334 .sp
1335 Perl 5.10 introduced a feature whereby each alternative in a subpattern uses
1336 the same numbers for its capturing parentheses. Such a subpattern starts with
1337 (?| and is itself a non-capturing subpattern. For example, consider this
1338 pattern:
1339 .sp
1340 (?|(Sat)ur|(Sun))day
1341 .sp
1342 Because the two alternatives are inside a (?| group, both sets of capturing
1343 parentheses are numbered one. Thus, when the pattern matches, you can look
1344 at captured substring number one, whichever alternative matched. This construct
1345 is useful when you want to capture part, but not all, of one of a number of
1346 alternatives. Inside a (?| group, parentheses are numbered as usual, but the
1347 number is reset at the start of each branch. The numbers of any capturing
1348 parentheses that follow the subpattern start after the highest number used in
1349 any branch. The following example is taken from the Perl documentation. The
1350 numbers underneath show in which buffer the captured content will be stored.
1351 .sp
1352 # before ---------------branch-reset----------- after
1353 / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
1354 # 1 2 2 3 2 3 4
1355 .sp
1356 A back reference to a numbered subpattern uses the most recent value that is
1357 set for that number by any subpattern. The following pattern matches "abcabc"
1358 or "defdef":
1359 .sp
1360 /(?|(abc)|(def))\e1/
1361 .sp
1362 In contrast, a subroutine call to a numbered subpattern always refers to the
1363 first one in the pattern with the given number. The following pattern matches
1364 "abcabc" or "defabc":
1365 .sp
1366 /(?|(abc)|(def))(?1)/
1367 .sp
1368 If a
1369 .\" HTML <a href="#conditions">
1370 .\" </a>
1371 condition test
1372 .\"
1373 for a subpattern's having matched refers to a non-unique number, the test is
1374 true if any of the subpatterns of that number have matched.
1375 .P
1376 An alternative approach to using this "branch reset" feature is to use
1377 duplicate named subpatterns, as described in the next section.
1378 .
1379 .
1380 .SH "NAMED SUBPATTERNS"
1381 .rs
1382 .sp
1383 Identifying capturing parentheses by number is simple, but it can be very hard
1384 to keep track of the numbers in complicated regular expressions. Furthermore,
1385 if an expression is modified, the numbers may change. To help with this
1386 difficulty, PCRE supports the naming of subpatterns. This feature was not
1387 added to Perl until release 5.10. Python had the feature earlier, and PCRE
1388 introduced it at release 4.0, using the Python syntax. PCRE now supports both
1389 the Perl and the Python syntax. Perl allows identically numbered subpatterns to
1390 have different names, but PCRE does not.
1391 .P
1392 In PCRE, a subpattern can be named in one of three ways: (?<name>...) or
1393 (?'name'...) as in Perl, or (?P<name>...) as in Python. References to capturing
1394 parentheses from other parts of the pattern, such as
1395 .\" HTML <a href="#backreferences">
1396 .\" </a>
1397 back references,
1398 .\"
1399 .\" HTML <a href="#recursion">
1400 .\" </a>
1401 recursion,
1402 .\"
1403 and
1404 .\" HTML <a href="#conditions">
1405 .\" </a>
1406 conditions,
1407 .\"
1408 can be made by name as well as by number.
1409 .P
1410 Names consist of up to 32 alphanumeric characters and underscores. Named
1411 capturing parentheses are still allocated numbers as well as names, exactly as
1412 if the names were not present. The PCRE API provides function calls for
1413 extracting the name-to-number translation table from a compiled pattern. There
1414 is also a convenience function for extracting a captured substring by name.
1415 .P
1416 By default, a name must be unique within a pattern, but it is possible to relax
1417 this constraint by setting the PCRE_DUPNAMES option at compile time. (Duplicate
1418 names are also always permitted for subpatterns with the same number, set up as
1419 described in the previous section.) Duplicate names can be useful for patterns
1420 where only one instance of the named parentheses can match. Suppose you want to
1421 match the name of a weekday, either as a 3-letter abbreviation or as the full
1422 name, and in both cases you want to extract the abbreviation. This pattern
1423 (ignoring the line breaks) does the job:
1424 .sp
1425 (?<DN>Mon|Fri|Sun)(?:day)?|
1426 (?<DN>Tue)(?:sday)?|
1427 (?<DN>Wed)(?:nesday)?|
1428 (?<DN>Thu)(?:rsday)?|
1429 (?<DN>Sat)(?:urday)?
1430 .sp
1431 There are five capturing substrings, but only one is ever set after a match.
1432 (An alternative way of solving this problem is to use a "branch reset"
1433 subpattern, as described in the previous section.)
1434 .P
1435 The convenience function for extracting the data by name returns the substring
1436 for the first (and in this example, the only) subpattern of that name that
1437 matched. This saves searching to find which numbered subpattern it was.
1438 .P
1439 If you make a back reference to a non-unique named subpattern from elsewhere in
1440 the pattern, the one that corresponds to the first occurrence of the name is
1441 used. In the absence of duplicate numbers (see the previous section) this is
1442 the one with the lowest number. If you use a named reference in a condition
1443 test (see the
1444 .\"
1445 .\" HTML <a href="#conditions">
1446 .\" </a>
1447 section about conditions
1448 .\"
1449 below), either to check whether a subpattern has matched, or to check for
1450 recursion, all subpatterns with the same name are tested. If the condition is
1451 true for any one of them, the overall condition is true. This is the same
1452 behaviour as testing by number. For further details of the interfaces for
1453 handling named subpatterns, see the
1454 .\" HREF
1455 \fBpcreapi\fP
1456 .\"
1457 documentation.
1458 .P
1459 \fBWarning:\fP You cannot use different names to distinguish between two
1460 subpatterns with the same number because PCRE uses only the numbers when
1461 matching. For this reason, an error is given at compile time if different names
1462 are given to subpatterns with the same number. However, you can give the same
1463 name to subpatterns with the same number, even when PCRE_DUPNAMES is not set.
1464 .
1465 .
1466 .SH REPETITION
1467 .rs
1468 .sp
1469 Repetition is specified by quantifiers, which can follow any of the following
1470 items:
1471 .sp
1472 a literal data character
1473 the dot metacharacter
1474 the \eC escape sequence
1475 the \eX escape sequence (in UTF-8 mode with Unicode properties)
1476 the \eR escape sequence
1477 an escape such as \ed or \epL that matches a single character
1478 a character class
1479 a back reference (see next section)
1480 a parenthesized subpattern (including assertions)
1481 a subroutine call to a subpattern (recursive or otherwise)
1482 .sp
1483 The general repetition quantifier specifies a minimum and maximum number of
1484 permitted matches, by giving the two numbers in curly brackets (braces),
1485 separated by a comma. The numbers must be less than 65536, and the first must
1486 be less than or equal to the second. For example:
1487 .sp
1488 z{2,4}
1489 .sp
1490 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
1491 character. If the second number is omitted, but the comma is present, there is
1492 no upper limit; if the second number and the comma are both omitted, the
1493 quantifier specifies an exact number of required matches. Thus
1494 .sp
1495 [aeiou]{3,}
1496 .sp
1497 matches at least 3 successive vowels, but may match many more, while
1498 .sp
1499 \ed{8}
1500 .sp
1501 matches exactly 8 digits. An opening curly bracket that appears in a position
1502 where a quantifier is not allowed, or one that does not match the syntax of a
1503 quantifier, is taken as a literal character. For example, {,6} is not a
1504 quantifier, but a literal string of four characters.
1505 .P
1506 In UTF-8 mode, quantifiers apply to UTF-8 characters rather than to individual
1507 bytes. Thus, for example, \ex{100}{2} matches two UTF-8 characters, each of
1508 which is represented by a two-byte sequence. Similarly, when Unicode property
1509 support is available, \eX{3} matches three Unicode extended sequences, each of
1510 which may be several bytes long (and they may be of different lengths).
1511 .P
1512 The quantifier {0} is permitted, causing the expression to behave as if the
1513 previous item and the quantifier were not present. This may be useful for
1514 subpatterns that are referenced as
1515 .\" HTML <a href="#subpatternsassubroutines">
1516 .\" </a>
1517 subroutines
1518 .\"
1519 from elsewhere in the pattern (but see also the section entitled
1520 .\" HTML <a href="#subdefine">
1521 .\" </a>
1522 "Defining subpatterns for use by reference only"
1523 .\"
1524 below). Items other than subpatterns that have a {0} quantifier are omitted
1525 from the compiled pattern.
1526 .P
1527 For convenience, the three most common quantifiers have single-character
1528 abbreviations:
1529 .sp
1530 * is equivalent to {0,}
1531 + is equivalent to {1,}
1532 ? is equivalent to {0,1}
1533 .sp
1534 It is possible to construct infinite loops by following a subpattern that can
1535 match no characters with a quantifier that has no upper limit, for example:
1536 .sp
1537 (a?)*
1538 .sp
1539 Earlier versions of Perl and PCRE used to give an error at compile time for
1540 such patterns. However, because there are cases where this can be useful, such
1541 patterns are now accepted, but if any repetition of the subpattern does in fact
1542 match no characters, the loop is forcibly broken.
1543 .P
1544 By default, the quantifiers are "greedy", that is, they match as much as
1545 possible (up to the maximum number of permitted times), without causing the
1546 rest of the pattern to fail. The classic example of where this gives problems
1547 is in trying to match comments in C programs. These appear between /* and */
1548 and within the comment, individual * and / characters may appear. An attempt to
1549 match C comments by applying the pattern
1550 .sp
1551 /\e*.*\e*/
1552 .sp
1553 to the string
1554 .sp
1555 /* first comment */ not comment /* second comment */
1556 .sp
1557 fails, because it matches the entire string owing to the greediness of the .*
1558 item.
1559 .P
1560 However, if a quantifier is followed by a question mark, it ceases to be
1561 greedy, and instead matches the minimum number of times possible, so the
1562 pattern
1563 .sp
1564 /\e*.*?\e*/
1565 .sp
1566 does the right thing with the C comments. The meaning of the various
1567 quantifiers is not otherwise changed, just the preferred number of matches.
1568 Do not confuse this use of question mark with its use as a quantifier in its
1569 own right. Because it has two uses, it can sometimes appear doubled, as in
1570 .sp
1571 \ed??\ed
1572 .sp
1573 which matches one digit by preference, but can match two if that is the only
1574 way the rest of the pattern matches.
1575 .P
1576 If the PCRE_UNGREEDY option is set (an option that is not available in Perl),
1577 the quantifiers are not greedy by default, but individual ones can be made
1578 greedy by following them with a question mark. In other words, it inverts the
1579 default behaviour.
1580 .P
1581 When a parenthesized subpattern is quantified with a minimum repeat count that
1582 is greater than 1 or with a limited maximum, more memory is required for the
1583 compiled pattern, in proportion to the size of the minimum or maximum.
1584 .P
1585 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent
1586 to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is
1587 implicitly anchored, because whatever follows will be tried against every
1588 character position in the subject string, so there is no point in retrying the
1589 overall match at any position after the first. PCRE normally treats such a
1590 pattern as though it were preceded by \eA.
1591 .P
1592 In cases where it is known that the subject string contains no newlines, it is
1593 worth setting PCRE_DOTALL in order to obtain this optimization, or
1594 alternatively using ^ to indicate anchoring explicitly.
1595 .P
1596 However, there is one situation where the optimization cannot be used. When .*
1597 is inside capturing parentheses that are the subject of a back reference
1598 elsewhere in the pattern, a match at the start may fail where a later one
1599 succeeds. Consider, for example:
1600 .sp
1601 (.*)abc\e1
1602 .sp
1603 If the subject is "xyz123abc123" the match point is the fourth character. For
1604 this reason, such a pattern is not implicitly anchored.
1605 .P
1606 When a capturing subpattern is repeated, the value captured is the substring
1607 that matched the final iteration. For example, after
1608 .sp
1609 (tweedle[dume]{3}\es*)+
1610 .sp
1611 has matched "tweedledum tweedledee" the value of the captured substring is
1612 "tweedledee". However, if there are nested capturing subpatterns, the
1613 corresponding captured values may have been set in previous iterations. For
1614 example, after
1615 .sp
1616 /(a|(b))+/
1617 .sp
1618 matches "aba" the value of the second captured substring is "b".
1619 .
1620 .
1621 .\" HTML <a name="atomicgroup"></a>
1622 .SH "ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS"
1623 .rs
1624 .sp
1625 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
1626 repetition, failure of what follows normally causes the repeated item to be
1627 re-evaluated to see if a different number of repeats allows the rest of the
1628 pattern to match. Sometimes it is useful to prevent this, either to change the
1629 nature of the match, or to cause it fail earlier than it otherwise might, when
1630 the author of the pattern knows there is no point in carrying on.
1631 .P
1632 Consider, for example, the pattern \ed+foo when applied to the subject line
1633 .sp
1634 123456bar
1635 .sp
1636 After matching all 6 digits and then failing to match "foo", the normal
1637 action of the matcher is to try again with only 5 digits matching the \ed+
1638 item, and then with 4, and so on, before ultimately failing. "Atomic grouping"
1639 (a term taken from Jeffrey Friedl's book) provides the means for specifying
1640 that once a subpattern has matched, it is not to be re-evaluated in this way.
1641 .P
1642 If we use atomic grouping for the previous example, the matcher gives up
1643 immediately on failing to match "foo" the first time. The notation is a kind of
1644 special parenthesis, starting with (?> as in this example:
1645 .sp
1646 (?>\ed+)foo
1647 .sp
1648 This kind of parenthesis "locks up" the part of the pattern it contains once
1649 it has matched, and a failure further into the pattern is prevented from
1650 backtracking into it. Backtracking past it to previous items, however, works as
1651 normal.
1652 .P
1653 An alternative description is that a subpattern of this type matches the string
1654 of characters that an identical standalone pattern would match, if anchored at
1655 the current point in the subject string.
1656 .P
1657 Atomic grouping subpatterns are not capturing subpatterns. Simple cases such as
1658 the above example can be thought of as a maximizing repeat that must swallow
1659 everything it can. So, while both \ed+ and \ed+? are prepared to adjust the
1660 number of digits they match in order to make the rest of the pattern match,
1661 (?>\ed+) can only match an entire sequence of digits.
1662 .P
1663 Atomic groups in general can of course contain arbitrarily complicated
1664 subpatterns, and can be nested. However, when the subpattern for an atomic
1665 group is just a single repeated item, as in the example above, a simpler
1666 notation, called a "possessive quantifier" can be used. This consists of an
1667 additional + character following a quantifier. Using this notation, the
1668 previous example can be rewritten as
1669 .sp
1670 \ed++foo
1671 .sp
1672 Note that a possessive quantifier can be used with an entire group, for
1673 example:
1674 .sp
1675 (abc|xyz){2,3}+
1676 .sp
1677 Possessive quantifiers are always greedy; the setting of the PCRE_UNGREEDY
1678 option is ignored. They are a convenient notation for the simpler forms of
1679 atomic group. However, there is no difference in the meaning of a possessive
1680 quantifier and the equivalent atomic group, though there may be a performance
1681 difference; possessive quantifiers should be slightly faster.
1682 .P
1683 The possessive quantifier syntax is an extension to the Perl 5.8 syntax.
1684 Jeffrey Friedl originated the idea (and the name) in the first edition of his
1685 book. Mike McCloskey liked it, so implemented it when he built Sun's Java
1686 package, and PCRE copied it from there. It ultimately found its way into Perl
1687 at release 5.10.
1688 .P
1689 PCRE has an optimization that automatically "possessifies" certain simple
1690 pattern constructs. For example, the sequence A+B is treated as A++B because
1691 there is no point in backtracking into a sequence of A's when B must follow.
1692 .P
1693 When a pattern contains an unlimited repeat inside a subpattern that can itself
1694 be repeated an unlimited number of times, the use of an atomic group is the
1695 only way to avoid some failing matches taking a very long time indeed. The
1696 pattern
1697 .sp
1698 (\eD+|<\ed+>)*[!?]
1699 .sp
1700 matches an unlimited number of substrings that either consist of non-digits, or
1701 digits enclosed in <>, followed by either ! or ?. When it matches, it runs
1702 quickly. However, if it is applied to
1703 .sp
1704 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
1705 .sp
1706 it takes a long time before reporting failure. This is because the string can
1707 be divided between the internal \eD+ repeat and the external * repeat in a
1708 large number of ways, and all have to be tried. (The example uses [!?] rather
1709 than a single character at the end, because both PCRE and Perl have an
1710 optimization that allows for fast failure when a single character is used. They
1711 remember the last single character that is required for a match, and fail early
1712 if it is not present in the string.) If the pattern is changed so that it uses
1713 an atomic group, like this:
1714 .sp
1715 ((?>\eD+)|<\ed+>)*[!?]
1716 .sp
1717 sequences of non-digits cannot be broken, and failure happens quickly.
1718 .
1719 .
1720 .\" HTML <a name="backreferences"></a>
1721 .SH "BACK REFERENCES"
1722 .rs
1723 .sp
1724 Outside a character class, a backslash followed by a digit greater than 0 (and
1725 possibly further digits) is a back reference to a capturing subpattern earlier
1726 (that is, to its left) in the pattern, provided there have been that many
1727 previous capturing left parentheses.
1728 .P
1729 However, if the decimal number following the backslash is less than 10, it is
1730 always taken as a back reference, and causes an error only if there are not
1731 that many capturing left parentheses in the entire pattern. In other words, the
1732 parentheses that are referenced need not be to the left of the reference for
1733 numbers less than 10. A "forward back reference" of this type can make sense
1734 when a repetition is involved and the subpattern to the right has participated
1735 in an earlier iteration.
1736 .P
1737 It is not possible to have a numerical "forward back reference" to a subpattern
1738 whose number is 10 or more using this syntax because a sequence such as \e50 is
1739 interpreted as a character defined in octal. See the subsection entitled
1740 "Non-printing characters"
1741 .\" HTML <a href="#digitsafterbackslash">
1742 .\" </a>
1743 above
1744 .\"
1745 for further details of the handling of digits following a backslash. There is
1746 no such problem when named parentheses are used. A back reference to any
1747 subpattern is possible using named parentheses (see below).
1748 .P
1749 Another way of avoiding the ambiguity inherent in the use of digits following a
1750 backslash is to use the \eg escape sequence. This escape must be followed by an
1751 unsigned number or a negative number, optionally enclosed in braces. These
1752 examples are all identical:
1753 .sp
1754 (ring), \e1
1755 (ring), \eg1
1756 (ring), \eg{1}
1757 .sp
1758 An unsigned number specifies an absolute reference without the ambiguity that
1759 is present in the older syntax. It is also useful when literal digits follow
1760 the reference. A negative number is a relative reference. Consider this
1761 example:
1762 .sp
1763 (abc(def)ghi)\eg{-1}
1764 .sp
1765 The sequence \eg{-1} is a reference to the most recently started capturing
1766 subpattern before \eg, that is, is it equivalent to \e2 in this example.
1767 Similarly, \eg{-2} would be equivalent to \e1. The use of relative references
1768 can be helpful in long patterns, and also in patterns that are created by
1769 joining together fragments that contain references within themselves.
1770 .P
1771 A back reference matches whatever actually matched the capturing subpattern in
1772 the current subject string, rather than anything matching the subpattern
1773 itself (see
1774 .\" HTML <a href="#subpatternsassubroutines">
1775 .\" </a>
1776 "Subpatterns as subroutines"
1777 .\"
1778 below for a way of doing that). So the pattern
1779 .sp
1780 (sens|respons)e and \e1ibility
1781 .sp
1782 matches "sense and sensibility" and "response and responsibility", but not
1783 "sense and responsibility". If caseful matching is in force at the time of the
1784 back reference, the case of letters is relevant. For example,
1785 .sp
1786 ((?i)rah)\es+\e1
1787 .sp
1788 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
1789 capturing subpattern is matched caselessly.
1790 .P
1791 There are several different ways of writing back references to named
1792 subpatterns. The .NET syntax \ek{name} and the Perl syntax \ek<name> or
1793 \ek'name' are supported, as is the Python syntax (?P=name). Perl 5.10's unified
1794 back reference syntax, in which \eg can be used for both numeric and named
1795 references, is also supported. We could rewrite the above example in any of
1796 the following ways:
1797 .sp
1798 (?<p1>(?i)rah)\es+\ek<p1>
1799 (?'p1'(?i)rah)\es+\ek{p1}
1800 (?P<p1>(?i)rah)\es+(?P=p1)
1801 (?<p1>(?i)rah)\es+\eg{p1}
1802 .sp
1803 A subpattern that is referenced by name may appear in the pattern before or
1804 after the reference.
1805 .P
1806 There may be more than one back reference to the same subpattern. If a
1807 subpattern has not actually been used in a particular match, any back
1808 references to it always fail by default. For example, the pattern
1809 .sp
1810 (a|(bc))\e2
1811 .sp
1812 always fails if it starts to match "a" rather than "bc". However, if the
1813 PCRE_JAVASCRIPT_COMPAT option is set at compile time, a back reference to an
1814 unset value matches an empty string.
1815 .P
1816 Because there may be many capturing parentheses in a pattern, all digits
1817 following a backslash are taken as part of a potential back reference number.
1818 If the pattern continues with a digit character, some delimiter must be used to
1819 terminate the back reference. If the PCRE_EXTENDED option is set, this can be
1820 whitespace. Otherwise, the \eg{ syntax or an empty comment (see
1821 .\" HTML <a href="#comments">
1822 .\" </a>
1823 "Comments"
1824 .\"
1825 below) can be used.
1826 .
1827 .SS "Recursive back references"
1828 .rs
1829 .sp
1830 A back reference that occurs inside the parentheses to which it refers fails
1831 when the subpattern is first used, so, for example, (a\e1) never matches.
1832 However, such references can be useful inside repeated subpatterns. For
1833 example, the pattern
1834 .sp
1835 (a|b\e1)+
1836 .sp
1837 matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
1838 the subpattern, the back reference matches the character string corresponding
1839 to the previous iteration. In order for this to work, the pattern must be such
1840 that the first iteration does not need to match the back reference. This can be
1841 done using alternation, as in the example above, or by a quantifier with a
1842 minimum of zero.
1843 .P
1844 Back references of this type cause the group that they reference to be treated
1845 as an
1846 .\" HTML <a href="#atomicgroup">
1847 .\" </a>
1848 atomic group.
1849 .\"
1850 Once the whole group has been matched, a subsequent matching failure cannot
1851 cause backtracking into the middle of the group.
1852 .
1853 .
1854 .\" HTML <a name="bigassertions"></a>
1855 .SH ASSERTIONS
1856 .rs
1857 .sp
1858 An assertion is a test on the characters following or preceding the current
1859 matching point that does not actually consume any characters. The simple
1860 assertions coded as \eb, \eB, \eA, \eG, \eZ, \ez, ^ and $ are described
1861 .\" HTML <a href="#smallassertions">
1862 .\" </a>
1863 above.
1864 .\"
1865 .P
1866 More complicated assertions are coded as subpatterns. There are two kinds:
1867 those that look ahead of the current position in the subject string, and those
1868 that look behind it. An assertion subpattern is matched in the normal way,
1869 except that it does not cause the current matching position to be changed.
1870 .P
1871 Assertion subpatterns are not capturing subpatterns. If such an assertion
1872 contains capturing subpatterns within it, these are counted for the purposes of
1873 numbering the capturing subpatterns in the whole pattern. However, substring
1874 capturing is carried out only for positive assertions, because it does not make
1875 sense for negative assertions.
1876 .P
1877 For compatibility with Perl, assertion subpatterns may be repeated; though
1878 it makes no sense to assert the same thing several times, the side effect of
1879 capturing parentheses may occasionally be useful. In practice, there only three
1880 cases:
1881 .sp
1882 (1) If the quantifier is {0}, the assertion is never obeyed during matching.
1883 However, it may contain internal capturing parenthesized groups that are called
1884 from elsewhere via the
1885 .\" HTML <a href="#subpatternsassubroutines">
1886 .\" </a>
1887 subroutine mechanism.
1888 .\"
1889 .sp
1890 (2) If quantifier is {0,n} where n is greater than zero, it is treated as if it
1891 were {0,1}. At run time, the rest of the pattern match is tried with and
1892 without the assertion, the order depending on the greediness of the quantifier.
1893 .sp
1894 (3) If the minimum repetition is greater than zero, the quantifier is ignored.
1895 The assertion is obeyed just once when encountered during matching.
1896 .
1897 .
1898 .SS "Lookahead assertions"
1899 .rs
1900 .sp
1901 Lookahead assertions start with (?= for positive assertions and (?! for
1902 negative assertions. For example,
1903 .sp
1904 \ew+(?=;)
1905 .sp
1906 matches a word followed by a semicolon, but does not include the semicolon in
1907 the match, and
1908 .sp
1909 foo(?!bar)
1910 .sp
1911 matches any occurrence of "foo" that is not followed by "bar". Note that the
1912 apparently similar pattern
1913 .sp
1914 (?!foo)bar
1915 .sp
1916 does not find an occurrence of "bar" that is preceded by something other than
1917 "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
1918 (?!foo) is always true when the next three characters are "bar". A
1919 lookbehind assertion is needed to achieve the other effect.
1920 .P
1921 If you want to force a matching failure at some point in a pattern, the most
1922 convenient way to do it is with (?!) because an empty string always matches, so
1923 an assertion that requires there not to be an empty string must always fail.
1924 The backtracking control verb (*FAIL) or (*F) is a synonym for (?!).
1925 .
1926 .
1927 .\" HTML <a name="lookbehind"></a>
1928 .SS "Lookbehind assertions"
1929 .rs
1930 .sp
1931 Lookbehind assertions start with (?<= for positive assertions and (?<! for
1932 negative assertions. For example,
1933 .sp
1934 (?<!foo)bar
1935 .sp
1936 does find an occurrence of "bar" that is not preceded by "foo". The contents of
1937 a lookbehind assertion are restricted such that all the strings it matches must
1938 have a fixed length. However, if there are several top-level alternatives, they
1939 do not all have to have the same fixed length. Thus
1940 .sp
1941 (?<=bullock|donkey)
1942 .sp
1943 is permitted, but
1944 .sp
1945 (?<!dogs?|cats?)
1946 .sp
1947 causes an error at compile time. Branches that match different length strings
1948 are permitted only at the top level of a lookbehind assertion. This is an
1949 extension compared with Perl, which requires all branches to match the same
1950 length of string. An assertion such as
1951 .sp
1952 (?<=ab(c|de))
1953 .sp
1954 is not permitted, because its single top-level branch can match two different
1955 lengths, but it is acceptable to PCRE if rewritten to use two top-level
1956 branches:
1957 .sp
1958 (?<=abc|abde)
1959 .sp
1960 In some cases, the escape sequence \eK
1961 .\" HTML <a href="#resetmatchstart">
1962 .\" </a>
1963 (see above)
1964 .\"
1965 can be used instead of a lookbehind assertion to get round the fixed-length
1966 restriction.
1967 .P
1968 The implementation of lookbehind assertions is, for each alternative, to
1969 temporarily move the current position back by the fixed length and then try to
1970 match. If there are insufficient characters before the current position, the
1971 assertion fails.
1972 .P
1973 PCRE does not allow the \eC escape (which matches a single byte in UTF-8 mode)
1974 to appear in lookbehind assertions, because it makes it impossible to calculate
1975 the length of the lookbehind. The \eX and \eR escapes, which can match
1976 different numbers of bytes, are also not permitted.
1977 .P
1978 .\" HTML <a href="#subpatternsassubroutines">
1979 .\" </a>
1980 "Subroutine"
1981 .\"
1982 calls (see below) such as (?2) or (?&X) are permitted in lookbehinds, as long
1983 as the subpattern matches a fixed-length string.
1984 .\" HTML <a href="#recursion">
1985 .\" </a>
1986 Recursion,
1987 .\"
1988 however, is not supported.
1989 .P
1990 Possessive quantifiers can be used in conjunction with lookbehind assertions to
1991 specify efficient matching of fixed-length strings at the end of subject
1992 strings. Consider a simple pattern such as
1993 .sp
1994 abcd$
1995 .sp
1996 when applied to a long string that does not match. Because matching proceeds
1997 from left to right, PCRE will look for each "a" in the subject and then see if
1998 what follows matches the rest of the pattern. If the pattern is specified as
1999 .sp
2000 ^.*abcd$
2001 .sp
2002 the initial .* matches the entire string at first, but when this fails (because
2003 there is no following "a"), it backtracks to match all but the last character,
2004 then all but the last two characters, and so on. Once again the search for "a"
2005 covers the entire string, from right to left, so we are no better off. However,
2006 if the pattern is written as
2007 .sp
2008 ^.*+(?<=abcd)
2009 .sp
2010 there can be no backtracking for the .*+ item; it can match only the entire
2011 string. The subsequent lookbehind assertion does a single test on the last four
2012 characters. If it fails, the match fails immediately. For long strings, this
2013 approach makes a significant difference to the processing time.
2014 .
2015 .
2016 .SS "Using multiple assertions"
2017 .rs
2018 .sp
2019 Several assertions (of any sort) may occur in succession. For example,
2020 .sp
2021 (?<=\ed{3})(?<!999)foo
2022 .sp
2023 matches "foo" preceded by three digits that are not "999". Notice that each of
2024 the assertions is applied independently at the same point in the subject
2025 string. First there is a check that the previous three characters are all
2026 digits, and then there is a check that the same three characters are not "999".
2027 This pattern does \fInot\fP match "foo" preceded by six characters, the first
2028 of which are digits and the last three of which are not "999". For example, it
2029 doesn't match "123abcfoo". A pattern to do that is
2030 .sp
2031 (?<=\ed{3}...)(?<!999)foo
2032 .sp
2033 This time the first assertion looks at the preceding six characters, checking
2034 that the first three are digits, and then the second assertion checks that the
2035 preceding three characters are not "999".
2036 .P
2037 Assertions can be nested in any combination. For example,
2038 .sp
2039 (?<=(?<!foo)bar)baz
2040 .sp
2041 matches an occurrence of "baz" that is preceded by "bar" which in turn is not
2042 preceded by "foo", while
2043 .sp
2044 (?<=\ed{3}(?!999)...)foo
2045 .sp
2046 is another pattern that matches "foo" preceded by three digits and any three
2047 characters that are not "999".
2048 .
2049 .
2050 .\" HTML <a name="conditions"></a>
2051 .SH "CONDITIONAL SUBPATTERNS"
2052 .rs
2053 .sp
2054 It is possible to cause the matching process to obey a subpattern
2055 conditionally or to choose between two alternative subpatterns, depending on
2056 the result of an assertion, or whether a specific capturing subpattern has
2057 already been matched. The two possible forms of conditional subpattern are:
2058 .sp
2059 (?(condition)yes-pattern)
2060 (?(condition)yes-pattern|no-pattern)
2061 .sp
2062 If the condition is satisfied, the yes-pattern is used; otherwise the
2063 no-pattern (if present) is used. If there are more than two alternatives in the
2064 subpattern, a compile-time error occurs. Each of the two alternatives may
2065 itself contain nested subpatterns of any form, including conditional
2066 subpatterns; the restriction to two alternatives applies only at the level of
2067 the condition. This pattern fragment is an example where the alternatives are
2068 complex:
2069 .sp
2070 (?(1) (A|B|C) | (D | (?(2)E|F) | E) )
2071 .sp
2072 .P
2073 There are four kinds of condition: references to subpatterns, references to
2074 recursion, a pseudo-condition called DEFINE, and assertions.
2075 .
2076 .SS "Checking for a used subpattern by number"
2077 .rs
2078 .sp
2079 If the text between the parentheses consists of a sequence of digits, the
2080 condition is true if a capturing subpattern of that number has previously
2081 matched. If there is more than one capturing subpattern with the same number
2082 (see the earlier
2083 .\"
2084 .\" HTML <a href="#recursion">
2085 .\" </a>
2086 section about duplicate subpattern numbers),
2087 .\"
2088 the condition is true if any of them have matched. An alternative notation is
2089 to precede the digits with a plus or minus sign. In this case, the subpattern
2090 number is relative rather than absolute. The most recently opened parentheses
2091 can be referenced by (?(-1), the next most recent by (?(-2), and so on. Inside
2092 loops it can also make sense to refer to subsequent groups. The next
2093 parentheses to be opened can be referenced as (?(+1), and so on. (The value
2094 zero in any of these forms is not used; it provokes a compile-time error.)
2095 .P
2096 Consider the following pattern, which contains non-significant white space to
2097 make it more readable (assume the PCRE_EXTENDED option) and to divide it into
2098 three parts for ease of discussion:
2099 .sp
2100 ( \e( )? [^()]+ (?(1) \e) )
2101 .sp
2102 The first part matches an optional opening parenthesis, and if that
2103 character is present, sets it as the first captured substring. The second part
2104 matches one or more characters that are not parentheses. The third part is a
2105 conditional subpattern that tests whether or not the first set of parentheses
2106 matched. If they did, that is, if subject started with an opening parenthesis,
2107 the condition is true, and so the yes-pattern is executed and a closing
2108 parenthesis is required. Otherwise, since no-pattern is not present, the
2109 subpattern matches nothing. In other words, this pattern matches a sequence of
2110 non-parentheses, optionally enclosed in parentheses.
2111 .P
2112 If you were embedding this pattern in a larger one, you could use a relative
2113 reference:
2114 .sp
2115 ...other stuff... ( \e( )? [^()]+ (?(-1) \e) ) ...
2116 .sp
2117 This makes the fragment independent of the parentheses in the larger pattern.
2118 .
2119 .SS "Checking for a used subpattern by name"
2120 .rs
2121 .sp
2122 Perl uses the syntax (?(<name>)...) or (?('name')...) to test for a used
2123 subpattern by name. For compatibility with earlier versions of PCRE, which had
2124 this facility before Perl, the syntax (?(name)...) is also recognized. However,
2125 there is a possible ambiguity with this syntax, because subpattern names may
2126 consist entirely of digits. PCRE looks first for a named subpattern; if it
2127 cannot find one and the name consists entirely of digits, PCRE looks for a
2128 subpattern of that number, which must be greater than zero. Using subpattern
2129 names that consist entirely of digits is not recommended.
2130 .P
2131 Rewriting the above example to use a named subpattern gives this:
2132 .sp
2133 (?<OPEN> \e( )? [^()]+ (?(<OPEN>) \e) )
2134 .sp
2135 If the name used in a condition of this kind is a duplicate, the test is
2136 applied to all subpatterns of the same name, and is true if any one of them has
2137 matched.
2138 .
2139 .SS "Checking for pattern recursion"
2140 .rs
2141 .sp
2142 If the condition is the string (R), and there is no subpattern with the name R,
2143 the condition is true if a recursive call to the whole pattern or any
2144 subpattern has been made. If digits or a name preceded by ampersand follow the
2145 letter R, for example:
2146 .sp
2147 (?(R3)...) or (?(R&name)...)
2148 .sp
2149 the condition is true if the most recent recursion is into a subpattern whose
2150 number or name is given. This condition does not check the entire recursion
2151 stack. If the name used in a condition of this kind is a duplicate, the test is
2152 applied to all subpatterns of the same name, and is true if any one of them is
2153 the most recent recursion.
2154 .P
2155 At "top level", all these recursion test conditions are false.
2156 .\" HTML <a href="#recursion">
2157 .\" </a>
2158 The syntax for recursive patterns
2159 .\"
2160 is described below.
2161 .
2162 .\" HTML <a name="subdefine"></a>
2163 .SS "Defining subpatterns for use by reference only"
2164 .rs
2165 .sp
2166 If the condition is the string (DEFINE), and there is no subpattern with the
2167 name DEFINE, the condition is always false. In this case, there may be only one
2168 alternative in the subpattern. It is always skipped if control reaches this
2169 point in the pattern; the idea of DEFINE is that it can be used to define
2170 subroutines that can be referenced from elsewhere. (The use of
2171 .\" HTML <a href="#subpatternsassubroutines">
2172 .\" </a>
2173 subroutines
2174 .\"
2175 is described below.) For example, a pattern to match an IPv4 address such as
2176 "192.168.23.245" could be written like this (ignore whitespace and line
2177 breaks):
2178 .sp
2179 (?(DEFINE) (?<byte> 2[0-4]\ed | 25[0-5] | 1\ed\ed | [1-9]?\ed) )
2180 \eb (?&byte) (\e.(?&byte)){3} \eb
2181 .sp
2182 The first part of the pattern is a DEFINE group inside which a another group
2183 named "byte" is defined. This matches an individual component of an IPv4
2184 address (a number less than 256). When matching takes place, this part of the
2185 pattern is skipped because DEFINE acts like a false condition. The rest of the
2186 pattern uses references to the named group to match the four dot-separated
2187 components of an IPv4 address, insisting on a word boundary at each end.
2188 .
2189 .SS "Assertion conditions"
2190 .rs
2191 .sp
2192 If the condition is not in any of the above formats, it must be an assertion.
2193 This may be a positive or negative lookahead or lookbehind assertion. Consider
2194 this pattern, again containing non-significant white space, and with the two
2195 alternatives on the second line:
2196 .sp
2197 (?(?=[^a-z]*[a-z])
2198 \ed{2}-[a-z]{3}-\ed{2} | \ed{2}-\ed{2}-\ed{2} )
2199 .sp
2200 The condition is a positive lookahead assertion that matches an optional
2201 sequence of non-letters followed by a letter. In other words, it tests for the
2202 presence of at least one letter in the subject. If a letter is found, the
2203 subject is matched against the first alternative; otherwise it is matched
2204 against the second. This pattern matches strings in one of the two forms
2205 dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
2206 .
2207 .
2208 .\" HTML <a name="comments"></a>
2209 .SH COMMENTS
2210 .rs
2211 .sp
2212 There are two ways of including comments in patterns that are processed by
2213 PCRE. In both cases, the start of the comment must not be in a character class,
2214 nor in the middle of any other sequence of related characters such as (?: or a
2215 subpattern name or number. The characters that make up a comment play no part
2216 in the pattern matching.
2217 .P
2218 The sequence (?# marks the start of a comment that continues up to the next
2219 closing parenthesis. Nested parentheses are not permitted. If the PCRE_EXTENDED
2220 option is set, an unescaped # character also introduces a comment, which in
2221 this case continues to immediately after the next newline character or
2222 character sequence in the pattern. Which characters are interpreted as newlines
2223 is controlled by the options passed to \fBpcre_compile()\fP or by a special
2224 sequence at the start of the pattern, as described in the section entitled
2225 .\" HTML <a href="#newlines">
2226 .\" </a>
2227 "Newline conventions"
2228 .\"
2229 above. Note that the end of this type of comment is a literal newline sequence
2230 in the pattern; escape sequences that happen to represent a newline do not
2231 count. For example, consider this pattern when PCRE_EXTENDED is set, and the
2232 default newline convention is in force:
2233 .sp
2234 abc #comment \en still comment
2235 .sp
2236 On encountering the # character, \fBpcre_compile()\fP skips along, looking for
2237 a newline in the pattern. The sequence \en is still literal at this stage, so
2238 it does not terminate the comment. Only an actual character with the code value
2239 0x0a (the default newline) does so.
2240 .
2241 .
2242 .\" HTML <a name="recursion"></a>
2243 .SH "RECURSIVE PATTERNS"
2244 .rs
2245 .sp
2246 Consider the problem of matching a string in parentheses, allowing for
2247 unlimited nested parentheses. Without the use of recursion, the best that can
2248 be done is to use a pattern that matches up to some fixed depth of nesting. It
2249 is not possible to handle an arbitrary nesting depth.
2250 .P
2251 For some time, Perl has provided a facility that allows regular expressions to
2252 recurse (amongst other things). It does this by interpolating Perl code in the
2253 expression at run time, and the code can refer to the expression itself. A Perl
2254 pattern using code interpolation to solve the parentheses problem can be
2255 created like this:
2256 .sp
2257 $re = qr{\e( (?: (?>[^()]+) | (?p{$re}) )* \e)}x;
2258 .sp
2259 The (?p{...}) item interpolates Perl code at run time, and in this case refers
2260 recursively to the pattern in which it appears.
2261 .P
2262 Obviously, PCRE cannot support the interpolation of Perl code. Instead, it
2263 supports special syntax for recursion of the entire pattern, and also for
2264 individual subpattern recursion. After its introduction in PCRE and Python,
2265 this kind of recursion was subsequently introduced into Perl at release 5.10.
2266 .P
2267 A special item that consists of (? followed by a number greater than zero and a
2268 closing parenthesis is a recursive subroutine call of the subpattern of the
2269 given number, provided that it occurs inside that subpattern. (If not, it is a
2270 .\" HTML <a href="#subpatternsassubroutines">
2271 .\" </a>
2272 non-recursive subroutine
2273 .\"
2274 call, which is described in the next section.) The special item (?R) or (?0) is
2275 a recursive call of the entire regular expression.
2276 .P
2277 This PCRE pattern solves the nested parentheses problem (assume the
2278 PCRE_EXTENDED option is set so that white space is ignored):
2279 .sp
2280 \e( ( [^()]++ | (?R) )* \e)
2281 .sp
2282 First it matches an opening parenthesis. Then it matches any number of
2283 substrings which can either be a sequence of non-parentheses, or a recursive
2284 match of the pattern itself (that is, a correctly parenthesized substring).
2285 Finally there is a closing parenthesis. Note the use of a possessive quantifier
2286 to avoid backtracking into sequences of non-parentheses.
2287 .P
2288 If this were part of a larger pattern, you would not want to recurse the entire
2289 pattern, so instead you could use this:
2290 .sp
2291 ( \e( ( [^()]++ | (?1) )* \e) )
2292 .sp
2293 We have put the pattern into parentheses, and caused the recursion to refer to
2294 them instead of the whole pattern.
2295 .P
2296 In a larger pattern, keeping track of parenthesis numbers can be tricky. This
2297 is made easier by the use of relative references. Instead of (?1) in the
2298 pattern above you can write (?-2) to refer to the second most recently opened
2299 parentheses preceding the recursion. In other words, a negative number counts
2300 capturing parentheses leftwards from the point at which it is encountered.
2301 .P
2302 It is also possible to refer to subsequently opened parentheses, by writing
2303 references such as (?+2). However, these cannot be recursive because the
2304 reference is not inside the parentheses that are referenced. They are always
2305 .\" HTML <a href="#subpatternsassubroutines">
2306 .\" </a>
2307 non-recursive subroutine
2308 .\"
2309 calls, as described in the next section.
2310 .P
2311 An alternative approach is to use named parentheses instead. The Perl syntax
2312 for this is (?&name); PCRE's earlier syntax (?P>name) is also supported. We
2313 could rewrite the above example as follows:
2314 .sp
2315 (?<pn> \e( ( [^()]++ | (?&pn) )* \e) )
2316 .sp
2317 If there is more than one subpattern with the same name, the earliest one is
2318 used.
2319 .P
2320 This particular example pattern that we have been looking at contains nested
2321 unlimited repeats, and so the use of a possessive quantifier for matching
2322 strings of non-parentheses is important when applying the pattern to strings
2323 that do not match. For example, when this pattern is applied to
2324 .sp
2325 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
2326 .sp
2327 it yields "no match" quickly. However, if a possessive quantifier is not used,
2328 the match runs for a very long time indeed because there are so many different
2329 ways the + and * repeats can carve up the subject, and all have to be tested
2330 before failure can be reported.
2331 .P
2332 At the end of a match, the values of capturing parentheses are those from
2333 the outermost level. If you want to obtain intermediate values, a callout
2334 function can be used (see below and the
2335 .\" HREF
2336 \fBpcrecallout\fP
2337 .\"
2338 documentation). If the pattern above is matched against
2339 .sp
2340 (ab(cd)ef)
2341 .sp
2342 the value for the inner capturing parentheses (numbered 2) is "ef", which is
2343 the last value taken on at the top level. If a capturing subpattern is not
2344 matched at the top level, its final captured value is unset, even if it was
2345 (temporarily) set at a deeper level during the matching process.
2346 .P
2347 If there are more than 15 capturing parentheses in a pattern, PCRE has to
2348 obtain extra memory to store data during a recursion, which it does by using
2349 \fBpcre_malloc\fP, freeing it via \fBpcre_free\fP afterwards. If no memory can
2350 be obtained, the match fails with the PCRE_ERROR_NOMEMORY error.
2351 .P
2352 Do not confuse the (?R) item with the condition (R), which tests for recursion.
2353 Consider this pattern, which matches text in angle brackets, allowing for
2354 arbitrary nesting. Only digits are allowed in nested brackets (that is, when
2355 recursing), whereas any characters are permitted at the outer level.
2356 .sp
2357 < (?: (?(R) \ed++ | [^<>]*+) | (?R)) * >
2358 .sp
2359 In this pattern, (?(R) is the start of a conditional subpattern, with two
2360 different alternatives for the recursive and non-recursive cases. The (?R) item
2361 is the actual recursive call.
2362 .
2363 .
2364 .\" HTML <a name="recursiondifference"></a>
2365 .SS "Differences in recursion processing between PCRE and Perl"
2366 .rs
2367 .sp
2368 Recursion processing in PCRE differs from Perl in two important ways. In PCRE
2369 (like Python, but unlike Perl), a recursive subpattern call is always treated
2370 as an atomic group. That is, once it has matched some of the subject string, it
2371 is never re-entered, even if it contains untried alternatives and there is a
2372 subsequent matching failure. This can be illustrated by the following pattern,
2373 which purports to match a palindromic string that contains an odd number of
2374 characters (for example, "a", "aba", "abcba", "abcdcba"):
2375 .sp
2376 ^(.|(.)(?1)\e2)$
2377 .sp
2378 The idea is that it either matches a single character, or two identical
2379 characters surrounding a sub-palindrome. In Perl, this pattern works; in PCRE
2380 it does not if the pattern is longer than three characters. Consider the
2381 subject string "abcba":
2382 .P
2383 At the top level, the first character is matched, but as it is not at the end
2384 of the string, the first alternative fails; the second alternative is taken
2385 and the recursion kicks in. The recursive call to subpattern 1 successfully
2386 matches the next character ("b"). (Note that the beginning and end of line
2387 tests are not part of the recursion).
2388 .P
2389 Back at the top level, the next character ("c") is compared with what
2390 subpattern 2 matched, which was "a". This fails. Because the recursion is
2391 treated as an atomic group, there are now no backtracking points, and so the
2392 entire match fails. (Perl is able, at this point, to re-enter the recursion and
2393 try the second alternative.) However, if the pattern is written with the
2394 alternatives in the other order, things are different:
2395 .sp
2396 ^((.)(?1)\e2|.)$
2397 .sp
2398 This time, the recursing alternative is tried first, and continues to recurse
2399 until it runs out of characters, at which point the recursion fails. But this
2400 time we do have another alternative to try at the higher level. That is the big
2401 difference: in the previous case the remaining alternative is at a deeper
2402 recursion level, which PCRE cannot use.
2403 .P
2404 To change the pattern so that it matches all palindromic strings, not just
2405 those with an odd number of characters, it is tempting to change the pattern to
2406 this:
2407 .sp
2408 ^((.)(?1)\e2|.?)$
2409 .sp
2410 Again, this works in Perl, but not in PCRE, and for the same reason. When a
2411 deeper recursion has matched a single character, it cannot be entered again in
2412 order to match an empty string. The solution is to separate the two cases, and
2413 write out the odd and even cases as alternatives at the higher level:
2414 .sp
2415 ^(?:((.)(?1)\e2|)|((.)(?3)\e4|.))
2416 .sp
2417 If you want to match typical palindromic phrases, the pattern has to ignore all
2418 non-word characters, which can be done like this:
2419 .sp
2420 ^\eW*+(?:((.)\eW*+(?1)\eW*+\e2|)|((.)\eW*+(?3)\eW*+\e4|\eW*+.\eW*+))\eW*+$
2421 .sp
2422 If run with the PCRE_CASELESS option, this pattern matches phrases such as "A
2423 man, a plan, a canal: Panama!" and it works well in both PCRE and Perl. Note
2424 the use of the possessive quantifier *+ to avoid backtracking into sequences of
2425 non-word characters. Without this, PCRE takes a great deal longer (ten times or
2426 more) to match typical phrases, and Perl takes so long that you think it has
2427 gone into a loop.
2428 .P
2429 \fBWARNING\fP: The palindrome-matching patterns above work only if the subject
2430 string does not start with a palindrome that is shorter than the entire string.
2431 For example, although "abcba" is correctly matched, if the subject is "ababa",
2432 PCRE finds the palindrome "aba" at the start, then fails at top level because
2433 the end of the string does not follow. Once again, it cannot jump back into the
2434 recursion to try other alternatives, so the entire match fails.
2435 .P
2436 The second way in which PCRE and Perl differ in their recursion processing is
2437 in the handling of captured values. In Perl, when a subpattern is called
2438 recursively or as a subpattern (see the next section), it has no access to any
2439 values that were captured outside the recursion, whereas in PCRE these values
2440 can be referenced. Consider this pattern:
2441 .sp
2442 ^(.)(\e1|a(?2))
2443 .sp
2444 In PCRE, this pattern matches "bab". The first capturing parentheses match "b",
2445 then in the second group, when the back reference \e1 fails to match "b", the
2446 second alternative matches "a" and then recurses. In the recursion, \e1 does
2447 now match "b" and so the whole match succeeds. In Perl, the pattern fails to
2448 match because inside the recursive call \e1 cannot access the externally set
2449 value.
2450 .
2451 .
2452 .\" HTML <a name="subpatternsassubroutines"></a>
2453 .SH "SUBPATTERNS AS SUBROUTINES"
2454 .rs
2455 .sp
2456 If the syntax for a recursive subpattern call (either by number or by
2457 name) is used outside the parentheses to which it refers, it operates like a
2458 subroutine in a programming language. The called subpattern may be defined
2459 before or after the reference. A numbered reference can be absolute or
2460 relative, as in these examples:
2461 .sp
2462 (...(absolute)...)...(?2)...
2463 (...(relative)...)...(?-1)...
2464 (...(?+1)...(relative)...
2465 .sp
2466 An earlier example pointed out that the pattern
2467 .sp
2468 (sens|respons)e and \e1ibility
2469 .sp
2470 matches "sense and sensibility" and "response and responsibility", but not
2471 "sense and responsibility". If instead the pattern
2472 .sp
2473 (sens|respons)e and (?1)ibility
2474 .sp
2475 is used, it does match "sense and responsibility" as well as the other two
2476 strings. Another example is given in the discussion of DEFINE above.
2477 .P
2478 All subroutine calls, whether recursive or not, are always treated as atomic
2479 groups. That is, once a subroutine has matched some of the subject string, it
2480 is never re-entered, even if it contains untried alternatives and there is a
2481 subsequent matching failure. Any capturing parentheses that are set during the
2482 subroutine call revert to their previous values afterwards.
2483 .P
2484 Processing options such as case-independence are fixed when a subpattern is
2485 defined, so if it is used as a subroutine, such options cannot be changed for
2486 different calls. For example, consider this pattern:
2487 .sp
2488 (abc)(?i:(?-1))
2489 .sp
2490 It matches "abcabc". It does not match "abcABC" because the change of
2491 processing option does not affect the called subpattern.
2492 .
2493 .
2494 .\" HTML <a name="onigurumasubroutines"></a>
2495 .SH "ONIGURUMA SUBROUTINE SYNTAX"
2496 .rs
2497 .sp
2498 For compatibility with Oniguruma, the non-Perl syntax \eg followed by a name or
2499 a number enclosed either in angle brackets or single quotes, is an alternative
2500 syntax for referencing a subpattern as a subroutine, possibly recursively. Here
2501 are two of the examples used above, rewritten using this syntax:
2502 .sp
2503 (?<pn> \e( ( (?>[^()]+) | \eg<pn> )* \e) )
2504 (sens|respons)e and \eg'1'ibility
2505 .sp
2506 PCRE supports an extension to Oniguruma: if a number is preceded by a
2507 plus or a minus sign it is taken as a relative reference. For example:
2508 .sp
2509 (abc)(?i:\eg<-1>)
2510 .sp
2511 Note that \eg{...} (Perl syntax) and \eg<...> (Oniguruma syntax) are \fInot\fP
2512 synonymous. The former is a back reference; the latter is a subroutine call.
2513 .
2514 .
2515 .SH CALLOUTS
2516 .rs
2517 .sp
2518 Perl has a feature whereby using the sequence (?{...}) causes arbitrary Perl
2519 code to be obeyed in the middle of matching a regular expression. This makes it
2520 possible, amongst other things, to extract different substrings that match the
2521 same pair of parentheses when there is a repetition.
2522 .P
2523 PCRE provides a similar feature, but of course it cannot obey arbitrary Perl
2524 code. The feature is called "callout". The caller of PCRE provides an external
2525 function by putting its entry point in the global variable \fIpcre_callout\fP.
2526 By default, this variable contains NULL, which disables all calling out.
2527 .P
2528 Within a regular expression, (?C) indicates the points at which the external
2529 function is to be called. If you want to identify different callout points, you
2530 can put a number less than 256 after the letter C. The default value is zero.
2531 For example, this pattern has two callout points:
2532 .sp
2533 (?C1)abc(?C2)def
2534 .sp
2535 If the PCRE_AUTO_CALLOUT flag is passed to \fBpcre_compile()\fP, callouts are
2536 automatically installed before each item in the pattern. They are all numbered
2537 255.
2538 .P
2539 During matching, when PCRE reaches a callout point (and \fIpcre_callout\fP is
2540 set), the external function is called. It is provided with the number of the
2541 callout, the position in the pattern, and, optionally, one item of data
2542 originally supplied by the caller of \fBpcre_exec()\fP. The callout function
2543 may cause matching to proceed, to backtrack, or to fail altogether. A complete
2544 description of the interface to the callout function is given in the
2545 .\" HREF
2546 \fBpcrecallout\fP
2547 .\"
2548 documentation.
2549 .
2550 .
2551 .\" HTML <a name="backtrackcontrol"></a>
2552 .SH "BACKTRACKING CONTROL"
2553 .rs
2554 .sp
2555 Perl 5.10 introduced a number of "Special Backtracking Control Verbs", which
2556 are described in the Perl documentation as "experimental and subject to change
2557 or removal in a future version of Perl". It goes on to say: "Their usage in
2558 production code should be noted to avoid problems during upgrades." The same
2559 remarks apply to the PCRE features described in this section.
2560 .P
2561 Since these verbs are specifically related to backtracking, most of them can be
2562 used only when the pattern is to be matched using \fBpcre_exec()\fP, which uses
2563 a backtracking algorithm. With the exception of (*FAIL), which behaves like a
2564 failing negative assertion, they cause an error if encountered by
2565 \fBpcre_dfa_exec()\fP.
2566 .P
2567 If any of these verbs are used in an assertion or in a subpattern that is
2568 called as a subroutine (whether or not recursively), their effect is confined
2569 to that subpattern; it does not extend to the surrounding pattern, with one
2570 exception: a *MARK that is encountered in a positive assertion \fIis\fP passed
2571 back (compare capturing parentheses in assertions). Note that such subpatterns
2572 are processed as anchored at the point where they are tested. Note also that
2573 Perl's treatment of subroutines is different in some cases.
2574 .P
2575 The new verbs make use of what was previously invalid syntax: an opening
2576 parenthesis followed by an asterisk. They are generally of the form
2577 (*VERB) or (*VERB:NAME). Some may take either form, with differing behaviour,
2578 depending on whether or not an argument is present. A name is any sequence of
2579 characters that does not include a closing parenthesis. If the name is empty,
2580 that is, if the closing parenthesis immediately follows the colon, the effect
2581 is as if the colon were not there. Any number of these verbs may occur in a
2582 pattern.
2583 .P
2584 PCRE contains some optimizations that are used to speed up matching by running
2585 some checks at the start of each match attempt. For example, it may know the
2586 minimum length of matching subject, or that a particular character must be
2587 present. When one of these optimizations suppresses the running of a match, any
2588 included backtracking verbs will not, of course, be processed. You can suppress
2589 the start-of-match optimizations by setting the PCRE_NO_START_OPTIMIZE option
2590 when calling \fBpcre_compile()\fP or \fBpcre_exec()\fP, or by starting the
2591 pattern with (*NO_START_OPT).
2592 .
2593 .
2594 .SS "Verbs that act immediately"
2595 .rs
2596 .sp
2597 The following verbs act as soon as they are encountered. They may not be
2598 followed by a name.
2599 .sp
2600 (*ACCEPT)
2601 .sp
2602 This verb causes the match to end successfully, skipping the remainder of the
2603 pattern. However, when it is inside a subpattern that is called as a
2604 subroutine, only that subpattern is ended successfully. Matching then continues
2605 at the outer level. If (*ACCEPT) is inside capturing parentheses, the data so
2606 far is captured. For example:
2607 .sp
2608 A((?:A|B(*ACCEPT)|C)D)
2609 .sp
2610 This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is captured by
2611 the outer parentheses.
2612 .sp
2613 (*FAIL) or (*F)
2614 .sp
2615 This verb causes a matching failure, forcing backtracking to occur. It is
2616 equivalent to (?!) but easier to read. The Perl documentation notes that it is
2617 probably useful only when combined with (?{}) or (??{}). Those are, of course,
2618 Perl features that are not present in PCRE. The nearest equivalent is the
2619 callout feature, as for example in this pattern:
2620 .sp
2621 a+(?C)(*FAIL)
2622 .sp
2623 A match with the string "aaaa" always fails, but the callout is taken before
2624 each backtrack happens (in this example, 10 times).
2625 .
2626 .
2627 .SS "Recording which path was taken"
2628 .rs
2629 .sp
2630 There is one verb whose main purpose is to track how a match was arrived at,
2631 though it also has a secondary use in conjunction with advancing the match
2632 starting point (see (*SKIP) below).
2633 .sp
2634 (*MARK:NAME) or (*:NAME)
2635 .sp
2636 A name is always required with this verb. There may be as many instances of
2637 (*MARK) as you like in a pattern, and their names do not have to be unique.
2638 .P
2639 When a match succeeds, the name of the last-encountered (*MARK) is passed back
2640 to the caller via the \fIpcre_extra\fP data structure, as described in the
2641 .\" HTML <a href="pcreapi.html#extradata">
2642 .\" </a>
2643 section on \fIpcre_extra\fP
2644 .\"
2645 in the
2646 .\" HREF
2647 \fBpcreapi\fP
2648 .\"
2649 documentation. No data is returned for a partial match. Here is an example of
2650 \fBpcretest\fP output, where the /K modifier requests the retrieval and
2651 outputting of (*MARK) data:
2652 .sp
2653 /X(*MARK:A)Y|X(*MARK:B)Z/K
2654 XY
2655 0: XY
2656 MK: A
2657 XZ
2658 0: XZ
2659 MK: B
2660 .sp
2661 The (*MARK) name is tagged with "MK:" in this output, and in this example it
2662 indicates which of the two alternatives matched. This is a more efficient way
2663 of obtaining this information than putting each alternative in its own
2664 capturing parentheses.
2665 .P
2666 If (*MARK) is encountered in a positive assertion, its name is recorded and
2667 passed back if it is the last-encountered. This does not happen for negative
2668 assertions.
2669 .P
2670 A name may also be returned after a failed match if the final path through the
2671 pattern involves (*MARK). However, unless (*MARK) used in conjunction with
2672 (*COMMIT), this is unlikely to happen for an unanchored pattern because, as the
2673 starting point for matching is advanced, the final check is often with an empty
2674 string, causing a failure before (*MARK) is reached. For example:
2675 .sp
2676 /X(*MARK:A)Y|X(*MARK:B)Z/K
2677 XP
2678 No match
2679 .sp
2680 There are three potential starting points for this match (starting with X,
2681 starting with P, and with an empty string). If the pattern is anchored, the
2682 result is different:
2683 .sp
2684 /^X(*MARK:A)Y|^X(*MARK:B)Z/K
2685 XP
2686 No match, mark = B
2687 .sp
2688 PCRE's start-of-match optimizations can also interfere with this. For example,
2689 if, as a result of a call to \fBpcre_study()\fP, it knows the minimum
2690 subject length for a match, a shorter subject will not be scanned at all.
2691 .P
2692 Note that similar anomalies (though different in detail) exist in Perl, no
2693 doubt for the same reasons. The use of (*MARK) data after a failed match of an
2694 unanchored pattern is not recommended, unless (*COMMIT) is involved.
2695 .
2696 .
2697 .SS "Verbs that act after backtracking"
2698 .rs
2699 .sp
2700 The following verbs do nothing when they are encountered. Matching continues
2701 with what follows, but if there is no subsequent match, causing a backtrack to
2702 the verb, a failure is forced. That is, backtracking cannot pass to the left of
2703 the verb. However, when one of these verbs appears inside an atomic group, its
2704 effect is confined to that group, because once the group has been matched,
2705 there is never any backtracking into it. In this situation, backtracking can
2706 "jump back" to the left of the entire atomic group. (Remember also, as stated
2707 above, that this localization also applies in subroutine calls and assertions.)
2708 .P
2709 These verbs differ in exactly what kind of failure occurs when backtracking
2710 reaches them.
2711 .sp
2712 (*COMMIT)
2713 .sp
2714 This verb, which may not be followed by a name, causes the whole match to fail
2715 outright if the rest of the pattern does not match. Even if the pattern is
2716 unanchored, no further attempts to find a match by advancing the starting point
2717 take place. Once (*COMMIT) has been passed, \fBpcre_exec()\fP is committed to
2718 finding a match at the current starting point, or not at all. For example:
2719 .sp
2720 a+(*COMMIT)b
2721 .sp
2722 This matches "xxaab" but not "aacaab". It can be thought of as a kind of
2723 dynamic anchor, or "I've started, so I must finish." The name of the most
2724 recently passed (*MARK) in the path is passed back when (*COMMIT) forces a
2725 match failure.
2726 .P
2727 Note that (*COMMIT) at the start of a pattern is not the same as an anchor,
2728 unless PCRE's start-of-match optimizations are turned off, as shown in this
2729 \fBpcretest\fP example:
2730 .sp
2731 /(*COMMIT)abc/
2732 xyzabc
2733 0: abc
2734 xyzabc\eY
2735 No match
2736 .sp
2737 PCRE knows that any match must start with "a", so the optimization skips along
2738 the subject to "a" before running the first match attempt, which succeeds. When
2739 the optimization is disabled by the \eY escape in the second subject, the match
2740 starts at "x" and so the (*COMMIT) causes it to fail without trying any other
2741 starting points.
2742 .sp
2743 (*PRUNE) or (*PRUNE:NAME)
2744 .sp
2745 This verb causes the match to fail at the current starting position in the
2746 subject if the rest of the pattern does not match. If the pattern is
2747 unanchored, the normal "bumpalong" advance to the next starting character then
2748 happens. Backtracking can occur as usual to the left of (*PRUNE), before it is
2749 reached, or when matching to the right of (*PRUNE), but if there is no match to
2750 the right, backtracking cannot cross (*PRUNE). In simple cases, the use of
2751 (*PRUNE) is just an alternative to an atomic group or possessive quantifier,
2752 but there are some uses of (*PRUNE) that cannot be expressed in any other way.
2753 The behaviour of (*PRUNE:NAME) is the same as (*MARK:NAME)(*PRUNE) when the
2754 match fails completely; the name is passed back if this is the final attempt.
2755 (*PRUNE:NAME) does not pass back a name if the match succeeds. In an anchored
2756 pattern (*PRUNE) has the same effect as (*COMMIT).
2757 .sp
2758 (*SKIP)
2759 .sp
2760 This verb, when given without a name, is like (*PRUNE), except that if the
2761 pattern is unanchored, the "bumpalong" advance is not to the next character,
2762 but to the position in the subject where (*SKIP) was encountered. (*SKIP)
2763 signifies that whatever text was matched leading up to it cannot be part of a
2764 successful match. Consider:
2765 .sp
2766 a+(*SKIP)b
2767 .sp
2768 If the subject is "aaaac...", after the first match attempt fails (starting at
2769 the first character in the string), the starting point skips on to start the
2770 next attempt at "c". Note that a possessive quantifer does not have the same
2771 effect as this example; although it would suppress backtracking during the
2772 first match attempt, the second attempt would start at the second character
2773 instead of skipping on to "c".
2774 .sp
2775 (*SKIP:NAME)
2776 .sp
2777 When (*SKIP) has an associated name, its behaviour is modified. If the
2778 following pattern fails to match, the previous path through the pattern is
2779 searched for the most recent (*MARK) that has the same name. If one is found,
2780 the "bumpalong" advance is to the subject position that corresponds to that
2781 (*MARK) instead of to where (*SKIP) was encountered. If no (*MARK) with a
2782 matching name is found, normal "bumpalong" of one character happens (that is,
2783 the (*SKIP) is ignored).
2784 .sp
2785 (*THEN) or (*THEN:NAME)
2786 .sp
2787 This verb causes a skip to the next innermost alternative if the rest of the
2788 pattern does not match. That is, it cancels pending backtracking, but only
2789 within the current alternative. Its name comes from the observation that it can
2790 be used for a pattern-based if-then-else block:
2791 .sp
2792 ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
2793 .sp
2794 If the COND1 pattern matches, FOO is tried (and possibly further items after
2795 the end of the group if FOO succeeds); on failure, the matcher skips to the
2796 second alternative and tries COND2, without backtracking into COND1. The
2797 behaviour of (*THEN:NAME) is exactly the same as (*MARK:NAME)(*THEN) if the
2798 overall match fails. If (*THEN) is not inside an alternation, it acts like
2799 (*PRUNE).
2800 .P
2801 Note that a subpattern that does not contain a | character is just a part of
2802 the enclosing alternative; it is not a nested alternation with only one
2803 alternative. The effect of (*THEN) extends beyond such a subpattern to the
2804 enclosing alternative. Consider this pattern, where A, B, etc. are complex
2805 pattern fragments that do not contain any | characters at this level:
2806 .sp
2807 A (B(*THEN)C) | D
2808 .sp
2809 If A and B are matched, but there is a failure in C, matching does not
2810 backtrack into A; instead it moves to the next alternative, that is, D.
2811 However, if the subpattern containing (*THEN) is given an alternative, it
2812 behaves differently:
2813 .sp
2814 A (B(*THEN)C | (*FAIL)) | D
2815 .sp
2816 The effect of (*THEN) is now confined to the inner subpattern. After a failure
2817 in C, matching moves to (*FAIL), which causes the whole subpattern to fail
2818 because there are no more alternatives to try. In this case, matching does now
2819 backtrack into A.
2820 .P
2821 Note also that a conditional subpattern is not considered as having two
2822 alternatives, because only one is ever used. In other words, the | character in
2823 a conditional subpattern has a different meaning. Ignoring white space,
2824 consider:
2825 .sp
2826 ^.*? (?(?=a) a | b(*THEN)c )
2827 .sp
2828 If the subject is "ba", this pattern does not match. Because .*? is ungreedy,
2829 it initially matches zero characters. The condition (?=a) then fails, the
2830 character "b" is matched, but "c" is not. At this point, matching does not
2831 backtrack to .*? as might perhaps be expected from the presence of the |
2832 character. The conditional subpattern is part of the single alternative that
2833 comprises the whole pattern, and so the match fails. (If there was a backtrack
2834 into .*?, allowing it to match "b", the match would succeed.)
2835 .P
2836 The verbs just described provide four different "strengths" of control when
2837 subsequent matching fails. (*THEN) is the weakest, carrying on the match at the
2838 next alternative. (*PRUNE) comes next, failing the match at the current
2839 starting position, but allowing an advance to the next character (for an
2840 unanchored pattern). (*SKIP) is similar, except that the advance may be more
2841 than one character. (*COMMIT) is the strongest, causing the entire match to
2842 fail.
2843 .P
2844 If more than one such verb is present in a pattern, the "strongest" one wins.
2845 For example, consider this pattern, where A, B, etc. are complex pattern
2846 fragments:
2847 .sp
2848 (A(*COMMIT)B(*THEN)C|D)
2849 .sp
2850 Once A has matched, PCRE is committed to this match, at the current starting
2851 position. If subsequently B matches, but C does not, the normal (*THEN) action
2852 of trying the next alternative (that is, D) does not happen because (*COMMIT)
2853 overrides.
2854 .
2855 .
2856 .SH "SEE ALSO"
2857 .rs
2858 .sp
2859 \fBpcreapi\fP(3), \fBpcrecallout\fP(3), \fBpcrematching\fP(3),
2860 \fBpcresyntax\fP(3), \fBpcre\fP(3).
2861 .
2862 .
2863 .SH AUTHOR
2864 .rs
2865 .sp
2866 .nf
2867 Philip Hazel
2868 University Computing Service
2869 Cambridge CB2 3QH, England.
2870 .fi
2871 .
2872 .
2873 .SH REVISION
2874 .rs
2875 .sp
2876 .nf
2877 Last updated: 14 November 2011
2878 Copyright (c) 1997-2011 University of Cambridge.
2879 .fi

Properties

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

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12