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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1404 - (show annotations)
Tue Nov 19 15:36:57 2013 UTC (5 years, 10 months ago) by ph10
File size: 133778 byte(s)
Source tidies for 8.34-RC1.
1 .TH PCREPATTERN 3 "12 November 2013" "PCRE 8.34"
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 This document discusses the patterns that are supported by PCRE when one its
24 main matching functions, \fBpcre_exec()\fP (8-bit) or \fBpcre[16|32]_exec()\fP
25 (16- or 32-bit), is used. PCRE also has alternative matching functions,
26 \fBpcre_dfa_exec()\fP and \fBpcre[16|32_dfa_exec()\fP, which match using a
27 different algorithm that is not Perl-compatible. Some of the features discussed
28 below are not available when DFA matching is used. The advantages and
29 disadvantages of the alternative functions, and how they differ from the normal
30 functions, are discussed in the
31 .\" HREF
32 \fBpcrematching\fP
33 .\"
34 page.
35 .
36 .
37 .SH "SPECIAL START-OF-PATTERN ITEMS"
38 .rs
39 .sp
40 A number of options that can be passed to \fBpcre_compile()\fP can also be set
41 by special items at the start of a pattern. These are not Perl-compatible, but
42 are provided to make these options accessible to pattern writers who are not
43 able to change the program that processes the pattern. Any number of these
44 items may appear, but they must all be together right at the start of the
45 pattern string, and the letters must be in upper case.
46 .
47 .
48 .SS "UTF support"
49 .rs
50 .sp
51 The original operation of PCRE was on strings of one-byte characters. However,
52 there is now also support for UTF-8 strings in the original library, an
53 extra library that supports 16-bit and UTF-16 character strings, and a
54 third library that supports 32-bit and UTF-32 character strings. To use these
55 features, PCRE must be built to include appropriate support. When using UTF
56 strings you must either call the compiling function with the PCRE_UTF8,
57 PCRE_UTF16, or PCRE_UTF32 option, or the pattern must start with one of
58 these special sequences:
59 .sp
60 (*UTF8)
61 (*UTF16)
62 (*UTF32)
63 (*UTF)
64 .sp
65 (*UTF) is a generic sequence that can be used with any of the libraries.
66 Starting a pattern with such a sequence is equivalent to setting the relevant
67 option. How setting a UTF mode affects pattern matching is mentioned in several
68 places below. There is also a summary of features in the
69 .\" HREF
70 \fBpcreunicode\fP
71 .\"
72 page.
73 .P
74 Some applications that allow their users to supply patterns may wish to
75 restrict them to non-UTF data for security reasons. If the PCRE_NEVER_UTF
76 option is set at compile time, (*UTF) etc. are not allowed, and their
77 appearance causes an error.
78 .
79 .
80 .SS "Unicode property support"
81 .rs
82 .sp
83 Another special sequence that may appear at the start of a pattern is (*UCP).
84 This has the same effect as setting the PCRE_UCP option: it causes sequences
85 such as \ed and \ew to use Unicode properties to determine character types,
86 instead of recognizing only characters with codes less than 128 via a lookup
87 table.
88 .
89 .
90 .SS "Disabling auto-possessification"
91 .rs
92 .sp
93 If a pattern starts with (*NO_AUTO_POSSESS), it has the same effect as setting
94 the PCRE_NO_AUTO_POSSESS option at compile time. This stops PCRE from making
95 quantifiers possessive when what follows cannot match the repeated item. For
96 example, by default a+b is treated as a++b. For more details, see the
97 .\" HREF
98 \fBpcreapi\fP
99 .\"
100 documentation.
101 .
102 .
103 .SS "Disabling start-up optimizations"
104 .rs
105 .sp
106 If a pattern starts with (*NO_START_OPT), it has the same effect as setting the
107 PCRE_NO_START_OPTIMIZE option either at compile or matching time. This disables
108 several optimizations for quickly reaching "no match" results. For more
109 details, see the
110 .\" HREF
111 \fBpcreapi\fP
112 .\"
113 documentation.
114 .
115 .
116 .\" HTML <a name="newlines"></a>
117 .SS "Newline conventions"
118 .rs
119 .sp
120 PCRE supports five different conventions for indicating line breaks in
121 strings: a single CR (carriage return) character, a single LF (linefeed)
122 character, the two-character sequence CRLF, any of the three preceding, or any
123 Unicode newline sequence. The
124 .\" HREF
125 \fBpcreapi\fP
126 .\"
127 page has
128 .\" HTML <a href="pcreapi.html#newlines">
129 .\" </a>
130 further discussion
131 .\"
132 about newlines, and shows how to set the newline convention in the
133 \fIoptions\fP arguments for the compiling and matching functions.
134 .P
135 It is also possible to specify a newline convention by starting a pattern
136 string with one of the following five sequences:
137 .sp
138 (*CR) carriage return
139 (*LF) linefeed
140 (*CRLF) carriage return, followed by linefeed
141 (*ANYCRLF) any of the three above
142 (*ANY) all Unicode newline sequences
143 .sp
144 These override the default and the options given to the compiling function. For
145 example, on a Unix system where LF is the default newline sequence, the pattern
146 .sp
147 (*CR)a.b
148 .sp
149 changes the convention to CR. That pattern matches "a\enb" because LF is no
150 longer a newline. If more than one of these settings is present, the last one
151 is used.
152 .P
153 The newline convention affects where the circumflex and dollar assertions are
154 true. It also affects the interpretation of the dot metacharacter when
155 PCRE_DOTALL is not set, and the behaviour of \eN. However, it does not affect
156 what the \eR escape sequence matches. By default, this is any Unicode newline
157 sequence, for Perl compatibility. However, this can be changed; see the
158 description of \eR in the section entitled
159 .\" HTML <a href="#newlineseq">
160 .\" </a>
161 "Newline sequences"
162 .\"
163 below. A change of \eR setting can be combined with a change of newline
164 convention.
165 .
166 .
167 .SS "Setting match and recursion limits"
168 .rs
169 .sp
170 The caller of \fBpcre_exec()\fP can set a limit on the number of times the
171 internal \fBmatch()\fP function is called and on the maximum depth of
172 recursive calls. These facilities are provided to catch runaway matches that
173 are provoked by patterns with huge matching trees (a typical example is a
174 pattern with nested unlimited repeats) and to avoid running out of system stack
175 by too much recursion. When one of these limits is reached, \fBpcre_exec()\fP
176 gives an error return. The limits can also be set by items at the start of the
177 pattern of the form
178 .sp
179 (*LIMIT_MATCH=d)
180 (*LIMIT_RECURSION=d)
181 .sp
182 where d is any number of decimal digits. However, the value of the setting must
183 be less than the value set (or defaulted) by the caller of \fBpcre_exec()\fP
184 for it to have any effect. In other words, the pattern writer can lower the
185 limits set by the programmer, but not raise them. If there is more than one
186 setting of one of these limits, the lower value is used.
187 .
188 .
189 .SH "EBCDIC CHARACTER CODES"
190 .rs
191 .sp
192 PCRE can be compiled to run in an environment that uses EBCDIC as its character
193 code rather than ASCII or Unicode (typically a mainframe system). In the
194 sections below, character code values are ASCII or Unicode; in an EBCDIC
195 environment these characters may have different code values, and there are no
196 code points greater than 255.
197 .
198 .
199 .SH "CHARACTERS AND METACHARACTERS"
200 .rs
201 .sp
202 A regular expression is a pattern that is matched against a subject string from
203 left to right. Most characters stand for themselves in a pattern, and match the
204 corresponding characters in the subject. As a trivial example, the pattern
205 .sp
206 The quick brown fox
207 .sp
208 matches a portion of a subject string that is identical to itself. When
209 caseless matching is specified (the PCRE_CASELESS option), letters are matched
210 independently of case. In a UTF mode, PCRE always understands the concept of
211 case for characters whose values are less than 128, so caseless matching is
212 always possible. For characters with higher values, the concept of case is
213 supported if PCRE is compiled with Unicode property support, but not otherwise.
214 If you want to use caseless matching for characters 128 and above, you must
215 ensure that PCRE is compiled with Unicode property support as well as with
216 UTF support.
217 .P
218 The power of regular expressions comes from the ability to include alternatives
219 and repetitions in the pattern. These are encoded in the pattern by the use of
220 \fImetacharacters\fP, which do not stand for themselves but instead are
221 interpreted in some special way.
222 .P
223 There are two different sets of metacharacters: those that are recognized
224 anywhere in the pattern except within square brackets, and those that are
225 recognized within square brackets. Outside square brackets, the metacharacters
226 are as follows:
227 .sp
228 \e general escape character with several uses
229 ^ assert start of string (or line, in multiline mode)
230 $ assert end of string (or line, in multiline mode)
231 . match any character except newline (by default)
232 [ start character class definition
233 | start of alternative branch
234 ( start subpattern
235 ) end subpattern
236 ? extends the meaning of (
237 also 0 or 1 quantifier
238 also quantifier minimizer
239 * 0 or more quantifier
240 + 1 or more quantifier
241 also "possessive quantifier"
242 { start min/max quantifier
243 .sp
244 Part of a pattern that is in square brackets is called a "character class". In
245 a character class the only metacharacters are:
246 .sp
247 \e general escape character
248 ^ negate the class, but only if the first character
249 - indicates character range
250 .\" JOIN
251 [ POSIX character class (only if followed by POSIX
252 syntax)
253 ] terminates the character class
254 .sp
255 The following sections describe the use of each of the metacharacters.
256 .
257 .
258 .SH BACKSLASH
259 .rs
260 .sp
261 The backslash character has several uses. Firstly, if it is followed by a
262 character that is not a number or a letter, it takes away any special meaning
263 that character may have. This use of backslash as an escape character applies
264 both inside and outside character classes.
265 .P
266 For example, if you want to match a * character, you write \e* in the pattern.
267 This escaping action applies whether or not the following character would
268 otherwise be interpreted as a metacharacter, so it is always safe to precede a
269 non-alphanumeric with backslash to specify that it stands for itself. In
270 particular, if you want to match a backslash, you write \e\e.
271 .P
272 In a UTF mode, only ASCII numbers and letters have any special meaning after a
273 backslash. All other characters (in particular, those whose codepoints are
274 greater than 127) are treated as literals.
275 .P
276 If a pattern is compiled with the PCRE_EXTENDED option, most white space in the
277 pattern (other than in a character class), and characters between a # outside a
278 character class and the next newline, inclusive, are ignored. An escaping
279 backslash can be used to include a white space or # character as part of the
280 pattern.
281 .P
282 If you want to remove the special meaning from a sequence of characters, you
283 can do so by putting them between \eQ and \eE. This is different from Perl in
284 that $ and @ are handled as literals in \eQ...\eE sequences in PCRE, whereas in
285 Perl, $ and @ cause variable interpolation. Note the following examples:
286 .sp
287 Pattern PCRE matches Perl matches
288 .sp
289 .\" JOIN
290 \eQabc$xyz\eE abc$xyz abc followed by the
291 contents of $xyz
292 \eQabc\e$xyz\eE abc\e$xyz abc\e$xyz
293 \eQabc\eE\e$\eQxyz\eE abc$xyz abc$xyz
294 .sp
295 The \eQ...\eE sequence is recognized both inside and outside character classes.
296 An isolated \eE that is not preceded by \eQ is ignored. If \eQ is not followed
297 by \eE later in the pattern, the literal interpretation continues to the end of
298 the pattern (that is, \eE is assumed at the end). If the isolated \eQ is inside
299 a character class, this causes an error, because the character class is not
300 terminated.
301 .
302 .
303 .\" HTML <a name="digitsafterbackslash"></a>
304 .SS "Non-printing characters"
305 .rs
306 .sp
307 A second use of backslash provides a way of encoding non-printing characters
308 in patterns in a visible manner. There is no restriction on the appearance of
309 non-printing characters, apart from the binary zero that terminates a pattern,
310 but when a pattern is being prepared by text editing, it is often easier to use
311 one of the following escape sequences than the binary character it represents:
312 .sp
313 \ea alarm, that is, the BEL character (hex 07)
314 \ecx "control-x", where x is any ASCII character
315 \ee escape (hex 1B)
316 \ef form feed (hex 0C)
317 \en linefeed (hex 0A)
318 \er carriage return (hex 0D)
319 \et tab (hex 09)
320 \e0dd character with octal code 0dd
321 \eddd character with octal code ddd, or back reference
322 \eo{ddd..} character with octal code ddd..
323 \exhh character with hex code hh
324 \ex{hhh..} character with hex code hhh.. (non-JavaScript mode)
325 \euhhhh character with hex code hhhh (JavaScript mode only)
326 .sp
327 The precise effect of \ecx on ASCII characters is as follows: if x is a lower
328 case letter, it is converted to upper case. Then bit 6 of the character (hex
329 40) is inverted. Thus \ecA to \ecZ become hex 01 to hex 1A (A is 41, Z is 5A),
330 but \ec{ becomes hex 3B ({ is 7B), and \ec; becomes hex 7B (; is 3B). If the
331 data item (byte or 16-bit value) following \ec has a value greater than 127, a
332 compile-time error occurs. This locks out non-ASCII characters in all modes.
333 .P
334 The \ec facility was designed for use with ASCII characters, but with the
335 extension to Unicode it is even less useful than it once was. It is, however,
336 recognized when PCRE is compiled in EBCDIC mode, where data items are always
337 bytes. In this mode, all values are valid after \ec. If the next character is a
338 lower case letter, it is converted to upper case. Then the 0xc0 bits of the
339 byte are inverted. Thus \ecA becomes hex 01, as in ASCII (A is C1), but because
340 the EBCDIC letters are disjoint, \ecZ becomes hex 29 (Z is E9), and other
341 characters also generate different values.
342 .P
343 After \e0 up to two further octal digits are read. If there are fewer than two
344 digits, just those that are present are used. Thus the sequence \e0\ex\e07
345 specifies two binary zeros followed by a BEL character (code value 7). Make
346 sure you supply two digits after the initial zero if the pattern character that
347 follows is itself an octal digit.
348 .P
349 The escape \eo must be followed by a sequence of octal digits, enclosed in
350 braces. An error occurs if this is not the case. This escape is a recent
351 addition to Perl; it provides way of specifying character code points as octal
352 numbers greater than 0777, and it also allows octal numbers and back references
353 to be unambiguously specified.
354 .P
355 For greater clarity and unambiguity, it is best to avoid following \e by a
356 digit greater than zero. Instead, use \eo{} or \ex{} to specify character
357 numbers, and \eg{} to specify back references. The following paragraphs
358 describe the old, ambiguous syntax.
359 .P
360 The handling of a backslash followed by a digit other than 0 is complicated,
361 and Perl has changed in recent releases, causing PCRE also to change. Outside a
362 character class, PCRE reads the digit and any following digits as a decimal
363 number. If the number is less than 8, or if there have been at least that many
364 previous capturing left parentheses in the expression, the entire sequence is
365 taken as a \fIback reference\fP. A description of how this works is given
366 .\" HTML <a href="#backreferences">
367 .\" </a>
368 later,
369 .\"
370 following the discussion of
371 .\" HTML <a href="#subpattern">
372 .\" </a>
373 parenthesized subpatterns.
374 .\"
375 .P
376 Inside a character class, or if the decimal number following \e is greater than
377 7 and there have not been that many capturing subpatterns, PCRE handles \e8 and
378 \e9 as the literal characters "8" and "9", and otherwise re-reads up to three
379 octal digits following the backslash, using them to generate a data character.
380 Any subsequent digits stand for themselves. For example:
381 .sp
382 \e040 is another way of writing an ASCII space
383 .\" JOIN
384 \e40 is the same, provided there are fewer than 40
385 previous capturing subpatterns
386 \e7 is always a back reference
387 .\" JOIN
388 \e11 might be a back reference, or another way of
389 writing a tab
390 \e011 is always a tab
391 \e0113 is a tab followed by the character "3"
392 .\" JOIN
393 \e113 might be a back reference, otherwise the
394 character with octal code 113
395 .\" JOIN
396 \e377 might be a back reference, otherwise
397 the value 255 (decimal)
398 .\" JOIN
399 \e81 is either a back reference, or the two
400 characters "8" and "1"
401 .sp
402 Note that octal values of 100 or greater that are specified using this syntax
403 must not be introduced by a leading zero, because no more than three octal
404 digits are ever read.
405 .P
406 By default, after \ex that is not followed by {, from zero to two hexadecimal
407 digits are read (letters can be in upper or lower case). Any number of
408 hexadecimal digits may appear between \ex{ and }. If a character other than
409 a hexadecimal digit appears between \ex{ and }, or if there is no terminating
410 }, an error occurs.
411 .P
412 If the PCRE_JAVASCRIPT_COMPAT option is set, the interpretation of \ex is
413 as just described only when it is followed by two hexadecimal digits.
414 Otherwise, it matches a literal "x" character. In JavaScript mode, support for
415 code points greater than 256 is provided by \eu, which must be followed by
416 four hexadecimal digits; otherwise it matches a literal "u" character.
417 .P
418 Characters whose value is less than 256 can be defined by either of the two
419 syntaxes for \ex (or by \eu in JavaScript mode). There is no difference in the
420 way they are handled. For example, \exdc is exactly the same as \ex{dc} (or
421 \eu00dc in JavaScript mode).
422 .
423 .
424 .SS "Constraints on character values"
425 .rs
426 .sp
427 Characters that are specified using octal or hexadecimal numbers are
428 limited to certain values, as follows:
429 .sp
430 8-bit non-UTF mode less than 0x100
431 8-bit UTF-8 mode less than 0x10ffff and a valid codepoint
432 16-bit non-UTF mode less than 0x10000
433 16-bit UTF-16 mode less than 0x10ffff and a valid codepoint
434 32-bit non-UTF mode less than 0x100000000
435 32-bit UTF-32 mode less than 0x10ffff and a valid codepoint
436 .sp
437 Invalid Unicode codepoints are the range 0xd800 to 0xdfff (the so-called
438 "surrogate" codepoints), and 0xffef.
439 .
440 .
441 .SS "Escape sequences in character classes"
442 .rs
443 .sp
444 All the sequences that define a single character value can be used both inside
445 and outside character classes. In addition, inside a character class, \eb is
446 interpreted as the backspace character (hex 08).
447 .P
448 \eN is not allowed in a character class. \eB, \eR, and \eX are not special
449 inside a character class. Like other unrecognized escape sequences, they are
450 treated as the literal characters "B", "R", and "X" by default, but cause an
451 error if the PCRE_EXTRA option is set. Outside a character class, these
452 sequences have different meanings.
453 .
454 .
455 .SS "Unsupported escape sequences"
456 .rs
457 .sp
458 In Perl, the sequences \el, \eL, \eu, and \eU are recognized by its string
459 handler and used to modify the case of following characters. By default, PCRE
460 does not support these escape sequences. However, if the PCRE_JAVASCRIPT_COMPAT
461 option is set, \eU matches a "U" character, and \eu can be used to define a
462 character by code point, as described in the previous section.
463 .
464 .
465 .SS "Absolute and relative back references"
466 .rs
467 .sp
468 The sequence \eg followed by an unsigned or a negative number, optionally
469 enclosed in braces, is an absolute or relative back reference. A named back
470 reference can be coded as \eg{name}. Back references are discussed
471 .\" HTML <a href="#backreferences">
472 .\" </a>
473 later,
474 .\"
475 following the discussion of
476 .\" HTML <a href="#subpattern">
477 .\" </a>
478 parenthesized subpatterns.
479 .\"
480 .
481 .
482 .SS "Absolute and relative subroutine calls"
483 .rs
484 .sp
485 For compatibility with Oniguruma, the non-Perl syntax \eg followed by a name or
486 a number enclosed either in angle brackets or single quotes, is an alternative
487 syntax for referencing a subpattern as a "subroutine". Details are discussed
488 .\" HTML <a href="#onigurumasubroutines">
489 .\" </a>
490 later.
491 .\"
492 Note that \eg{...} (Perl syntax) and \eg<...> (Oniguruma syntax) are \fInot\fP
493 synonymous. The former is a back reference; the latter is a
494 .\" HTML <a href="#subpatternsassubroutines">
495 .\" </a>
496 subroutine
497 .\"
498 call.
499 .
500 .
501 .\" HTML <a name="genericchartypes"></a>
502 .SS "Generic character types"
503 .rs
504 .sp
505 Another use of backslash is for specifying generic character types:
506 .sp
507 \ed any decimal digit
508 \eD any character that is not a decimal digit
509 \eh any horizontal white space character
510 \eH any character that is not a horizontal white space character
511 \es any white space character
512 \eS any character that is not a white space character
513 \ev any vertical white space character
514 \eV any character that is not a vertical white space character
515 \ew any "word" character
516 \eW any "non-word" character
517 .sp
518 There is also the single sequence \eN, which matches a non-newline character.
519 This is the same as
520 .\" HTML <a href="#fullstopdot">
521 .\" </a>
522 the "." metacharacter
523 .\"
524 when PCRE_DOTALL is not set. Perl also uses \eN to match characters by name;
525 PCRE does not support this.
526 .P
527 Each pair of lower and upper case escape sequences partitions the complete set
528 of characters into two disjoint sets. Any given character matches one, and only
529 one, of each pair. The sequences can appear both inside and outside character
530 classes. They each match one character of the appropriate type. If the current
531 matching point is at the end of the subject string, all of them fail, because
532 there is no character to match.
533 .P
534 For compatibility with Perl, \es did not used to match the VT character (code
535 11), which made it different from the the POSIX "space" class. However, Perl
536 added VT at release 5.18, and PCRE followed suit at release 8.34. The default
537 \es characters are now HT (9), LF (10), VT (11), FF (12), CR (13), and space
538 (32), which are defined as white space in the "C" locale. This list may vary if
539 locale-specific matching is taking place; in particular, in some locales the
540 "non-breaking space" character (\exA0) is recognized as white space.
541 .P
542 A "word" character is an underscore or any character that is a letter or digit.
543 By default, the definition of letters and digits is controlled by PCRE's
544 low-valued character tables, and may vary if locale-specific matching is taking
545 place (see
546 .\" HTML <a href="pcreapi.html#localesupport">
547 .\" </a>
548 "Locale support"
549 .\"
550 in the
551 .\" HREF
552 \fBpcreapi\fP
553 .\"
554 page). For example, in a French locale such as "fr_FR" in Unix-like systems,
555 or "french" in Windows, some character codes greater than 127 are used for
556 accented letters, and these are then matched by \ew. The use of locales with
557 Unicode is discouraged.
558 .P
559 By default, characters whose code points are greater than 127 never match \ed,
560 \es, or \ew, and always match \eD, \eS, and \eW, although this may vary for
561 characters in the range 128-255 when locale-specific matching is happening.
562 These escape sequences retain their original meanings from before Unicode
563 support was available, mainly for efficiency reasons. If PCRE is compiled with
564 Unicode property support, and the PCRE_UCP option is set, the behaviour is
565 changed so that Unicode properties are used to determine character types, as
566 follows:
567 .sp
568 \ed any character that matches \ep{Nd} (decimal digit)
569 \es any character that matches \ep{Z} or \eh or \ev
570 \ew any character that matches \ep{L} or \ep{N}, plus underscore
571 .sp
572 The upper case escapes match the inverse sets of characters. Note that \ed
573 matches only decimal digits, whereas \ew matches any Unicode digit, as well as
574 any Unicode letter, and underscore. Note also that PCRE_UCP affects \eb, and
575 \eB because they are defined in terms of \ew and \eW. Matching these sequences
576 is noticeably slower when PCRE_UCP is set.
577 .P
578 The sequences \eh, \eH, \ev, and \eV are features that were added to Perl at
579 release 5.10. In contrast to the other sequences, which match only ASCII
580 characters by default, these always match certain high-valued code points,
581 whether or not PCRE_UCP is set. The horizontal space characters are:
582 .sp
583 U+0009 Horizontal tab (HT)
584 U+0020 Space
585 U+00A0 Non-break space
586 U+1680 Ogham space mark
587 U+180E Mongolian vowel separator
588 U+2000 En quad
589 U+2001 Em quad
590 U+2002 En space
591 U+2003 Em space
592 U+2004 Three-per-em space
593 U+2005 Four-per-em space
594 U+2006 Six-per-em space
595 U+2007 Figure space
596 U+2008 Punctuation space
597 U+2009 Thin space
598 U+200A Hair space
599 U+202F Narrow no-break space
600 U+205F Medium mathematical space
601 U+3000 Ideographic space
602 .sp
603 The vertical space characters are:
604 .sp
605 U+000A Linefeed (LF)
606 U+000B Vertical tab (VT)
607 U+000C Form feed (FF)
608 U+000D Carriage return (CR)
609 U+0085 Next line (NEL)
610 U+2028 Line separator
611 U+2029 Paragraph separator
612 .sp
613 In 8-bit, non-UTF-8 mode, only the characters with codepoints less than 256 are
614 relevant.
615 .
616 .
617 .\" HTML <a name="newlineseq"></a>
618 .SS "Newline sequences"
619 .rs
620 .sp
621 Outside a character class, by default, the escape sequence \eR matches any
622 Unicode newline sequence. In 8-bit non-UTF-8 mode \eR is equivalent to the
623 following:
624 .sp
625 (?>\er\en|\en|\ex0b|\ef|\er|\ex85)
626 .sp
627 This is an example of an "atomic group", details of which are given
628 .\" HTML <a href="#atomicgroup">
629 .\" </a>
630 below.
631 .\"
632 This particular group matches either the two-character sequence CR followed by
633 LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab,
634 U+000B), FF (form feed, U+000C), CR (carriage return, U+000D), or NEL (next
635 line, U+0085). The two-character sequence is treated as a single unit that
636 cannot be split.
637 .P
638 In other modes, two additional characters whose codepoints are greater than 255
639 are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029).
640 Unicode character property support is not needed for these characters to be
641 recognized.
642 .P
643 It is possible to restrict \eR to match only CR, LF, or CRLF (instead of the
644 complete set of Unicode line endings) by setting the option PCRE_BSR_ANYCRLF
645 either at compile time or when the pattern is matched. (BSR is an abbrevation
646 for "backslash R".) This can be made the default when PCRE is built; if this is
647 the case, the other behaviour can be requested via the PCRE_BSR_UNICODE option.
648 It is also possible to specify these settings by starting a pattern string with
649 one of the following sequences:
650 .sp
651 (*BSR_ANYCRLF) CR, LF, or CRLF only
652 (*BSR_UNICODE) any Unicode newline sequence
653 .sp
654 These override the default and the options given to the compiling function, but
655 they can themselves be overridden by options given to a matching function. Note
656 that these special settings, which are not Perl-compatible, are recognized only
657 at the very start of a pattern, and that they must be in upper case. If more
658 than one of them is present, the last one is used. They can be combined with a
659 change of newline convention; for example, a pattern can start with:
660 .sp
661 (*ANY)(*BSR_ANYCRLF)
662 .sp
663 They can also be combined with the (*UTF8), (*UTF16), (*UTF32), (*UTF) or
664 (*UCP) special sequences. Inside a character class, \eR is treated as an
665 unrecognized escape sequence, and so matches the letter "R" by default, but
666 causes an error if PCRE_EXTRA is set.
667 .
668 .
669 .\" HTML <a name="uniextseq"></a>
670 .SS Unicode character properties
671 .rs
672 .sp
673 When PCRE is built with Unicode character property support, three additional
674 escape sequences that match characters with specific properties are available.
675 When in 8-bit non-UTF-8 mode, these sequences are of course limited to testing
676 characters whose codepoints are less than 256, but they do work in this mode.
677 The extra escape sequences are:
678 .sp
679 \ep{\fIxx\fP} a character with the \fIxx\fP property
680 \eP{\fIxx\fP} a character without the \fIxx\fP property
681 \eX a Unicode extended grapheme cluster
682 .sp
683 The property names represented by \fIxx\fP above are limited to the Unicode
684 script names, the general category properties, "Any", which matches any
685 character (including newline), and some special PCRE properties (described
686 in the
687 .\" HTML <a href="#extraprops">
688 .\" </a>
689 next section).
690 .\"
691 Other Perl properties such as "InMusicalSymbols" are not currently supported by
692 PCRE. Note that \eP{Any} does not match any characters, so always causes a
693 match failure.
694 .P
695 Sets of Unicode characters are defined as belonging to certain scripts. A
696 character from one of these sets can be matched using a script name. For
697 example:
698 .sp
699 \ep{Greek}
700 \eP{Han}
701 .sp
702 Those that are not part of an identified script are lumped together as
703 "Common". The current list of scripts is:
704 .P
705 Arabic,
706 Armenian,
707 Avestan,
708 Balinese,
709 Bamum,
710 Batak,
711 Bengali,
712 Bopomofo,
713 Brahmi,
714 Braille,
715 Buginese,
716 Buhid,
717 Canadian_Aboriginal,
718 Carian,
719 Chakma,
720 Cham,
721 Cherokee,
722 Common,
723 Coptic,
724 Cuneiform,
725 Cypriot,
726 Cyrillic,
727 Deseret,
728 Devanagari,
729 Egyptian_Hieroglyphs,
730 Ethiopic,
731 Georgian,
732 Glagolitic,
733 Gothic,
734 Greek,
735 Gujarati,
736 Gurmukhi,
737 Han,
738 Hangul,
739 Hanunoo,
740 Hebrew,
741 Hiragana,
742 Imperial_Aramaic,
743 Inherited,
744 Inscriptional_Pahlavi,
745 Inscriptional_Parthian,
746 Javanese,
747 Kaithi,
748 Kannada,
749 Katakana,
750 Kayah_Li,
751 Kharoshthi,
752 Khmer,
753 Lao,
754 Latin,
755 Lepcha,
756 Limbu,
757 Linear_B,
758 Lisu,
759 Lycian,
760 Lydian,
761 Malayalam,
762 Mandaic,
763 Meetei_Mayek,
764 Meroitic_Cursive,
765 Meroitic_Hieroglyphs,
766 Miao,
767 Mongolian,
768 Myanmar,
769 New_Tai_Lue,
770 Nko,
771 Ogham,
772 Old_Italic,
773 Old_Persian,
774 Old_South_Arabian,
775 Old_Turkic,
776 Ol_Chiki,
777 Oriya,
778 Osmanya,
779 Phags_Pa,
780 Phoenician,
781 Rejang,
782 Runic,
783 Samaritan,
784 Saurashtra,
785 Sharada,
786 Shavian,
787 Sinhala,
788 Sora_Sompeng,
789 Sundanese,
790 Syloti_Nagri,
791 Syriac,
792 Tagalog,
793 Tagbanwa,
794 Tai_Le,
795 Tai_Tham,
796 Tai_Viet,
797 Takri,
798 Tamil,
799 Telugu,
800 Thaana,
801 Thai,
802 Tibetan,
803 Tifinagh,
804 Ugaritic,
805 Vai,
806 Yi.
807 .P
808 Each character has exactly one Unicode general category property, specified by
809 a two-letter abbreviation. For compatibility with Perl, negation can be
810 specified by including a circumflex between the opening brace and the property
811 name. For example, \ep{^Lu} is the same as \eP{Lu}.
812 .P
813 If only one letter is specified with \ep or \eP, it includes all the general
814 category properties that start with that letter. In this case, in the absence
815 of negation, the curly brackets in the escape sequence are optional; these two
816 examples have the same effect:
817 .sp
818 \ep{L}
819 \epL
820 .sp
821 The following general category property codes are supported:
822 .sp
823 C Other
824 Cc Control
825 Cf Format
826 Cn Unassigned
827 Co Private use
828 Cs Surrogate
829 .sp
830 L Letter
831 Ll Lower case letter
832 Lm Modifier letter
833 Lo Other letter
834 Lt Title case letter
835 Lu Upper case letter
836 .sp
837 M Mark
838 Mc Spacing mark
839 Me Enclosing mark
840 Mn Non-spacing mark
841 .sp
842 N Number
843 Nd Decimal number
844 Nl Letter number
845 No Other number
846 .sp
847 P Punctuation
848 Pc Connector punctuation
849 Pd Dash punctuation
850 Pe Close punctuation
851 Pf Final punctuation
852 Pi Initial punctuation
853 Po Other punctuation
854 Ps Open punctuation
855 .sp
856 S Symbol
857 Sc Currency symbol
858 Sk Modifier symbol
859 Sm Mathematical symbol
860 So Other symbol
861 .sp
862 Z Separator
863 Zl Line separator
864 Zp Paragraph separator
865 Zs Space separator
866 .sp
867 The special property L& is also supported: it matches a character that has
868 the Lu, Ll, or Lt property, in other words, a letter that is not classified as
869 a modifier or "other".
870 .P
871 The Cs (Surrogate) property applies only to characters in the range U+D800 to
872 U+DFFF. Such characters are not valid in Unicode strings and so
873 cannot be tested by PCRE, unless UTF validity checking has been turned off
874 (see the discussion of PCRE_NO_UTF8_CHECK, PCRE_NO_UTF16_CHECK and
875 PCRE_NO_UTF32_CHECK in the
876 .\" HREF
877 \fBpcreapi\fP
878 .\"
879 page). Perl does not support the Cs property.
880 .P
881 The long synonyms for property names that Perl supports (such as \ep{Letter})
882 are not supported by PCRE, nor is it permitted to prefix any of these
883 properties with "Is".
884 .P
885 No character that is in the Unicode table has the Cn (unassigned) property.
886 Instead, this property is assumed for any code point that is not in the
887 Unicode table.
888 .P
889 Specifying caseless matching does not affect these escape sequences. For
890 example, \ep{Lu} always matches only upper case letters. This is different from
891 the behaviour of current versions of Perl.
892 .P
893 Matching characters by Unicode property is not fast, because PCRE has to do a
894 multistage table lookup in order to find a character's property. That is why
895 the traditional escape sequences such as \ed and \ew do not use Unicode
896 properties in PCRE by default, though you can make them do so by setting the
897 PCRE_UCP option or by starting the pattern with (*UCP).
898 .
899 .
900 .SS Extended grapheme clusters
901 .rs
902 .sp
903 The \eX escape matches any number of Unicode characters that form an "extended
904 grapheme cluster", and treats the sequence as an atomic group
905 .\" HTML <a href="#atomicgroup">
906 .\" </a>
907 (see below).
908 .\"
909 Up to and including release 8.31, PCRE matched an earlier, simpler definition
910 that was equivalent to
911 .sp
912 (?>\ePM\epM*)
913 .sp
914 That is, it matched a character without the "mark" property, followed by zero
915 or more characters with the "mark" property. Characters with the "mark"
916 property are typically non-spacing accents that affect the preceding character.
917 .P
918 This simple definition was extended in Unicode to include more complicated
919 kinds of composite character by giving each character a grapheme breaking
920 property, and creating rules that use these properties to define the boundaries
921 of extended grapheme clusters. In releases of PCRE later than 8.31, \eX matches
922 one of these clusters.
923 .P
924 \eX always matches at least one character. Then it decides whether to add
925 additional characters according to the following rules for ending a cluster:
926 .P
927 1. End at the end of the subject string.
928 .P
929 2. Do not end between CR and LF; otherwise end after any control character.
930 .P
931 3. Do not break Hangul (a Korean script) syllable sequences. Hangul characters
932 are of five types: L, V, T, LV, and LVT. An L character may be followed by an
933 L, V, LV, or LVT character; an LV or V character may be followed by a V or T
934 character; an LVT or T character may be follwed only by a T character.
935 .P
936 4. Do not end before extending characters or spacing marks. Characters with
937 the "mark" property always have the "extend" grapheme breaking property.
938 .P
939 5. Do not end after prepend characters.
940 .P
941 6. Otherwise, end the cluster.
942 .
943 .
944 .\" HTML <a name="extraprops"></a>
945 .SS PCRE's additional properties
946 .rs
947 .sp
948 As well as the standard Unicode properties described above, PCRE supports four
949 more that make it possible to convert traditional escape sequences such as \ew
950 and \es to use Unicode properties. PCRE uses these non-standard, non-Perl
951 properties internally when PCRE_UCP is set. However, they may also be used
952 explicitly. These properties are:
953 .sp
954 Xan Any alphanumeric character
955 Xps Any POSIX space character
956 Xsp Any Perl space character
957 Xwd Any Perl "word" character
958 .sp
959 Xan matches characters that have either the L (letter) or the N (number)
960 property. Xps matches the characters tab, linefeed, vertical tab, form feed, or
961 carriage return, and any other character that has the Z (separator) property.
962 Xsp is the same as Xps; it used to exclude vertical tab, for Perl
963 compatibility, but Perl changed, and so PCRE followed at release 8.34. Xwd
964 matches the same characters as Xan, plus underscore.
965 .P
966 There is another non-standard property, Xuc, which matches any character that
967 can be represented by a Universal Character Name in C++ and other programming
968 languages. These are the characters $, @, ` (grave accent), and all characters
969 with Unicode code points greater than or equal to U+00A0, except for the
970 surrogates U+D800 to U+DFFF. Note that most base (ASCII) characters are
971 excluded. (Universal Character Names are of the form \euHHHH or \eUHHHHHHHH
972 where H is a hexadecimal digit. Note that the Xuc property does not match these
973 sequences but the characters that they represent.)
974 .
975 .
976 .\" HTML <a name="resetmatchstart"></a>
977 .SS "Resetting the match start"
978 .rs
979 .sp
980 The escape sequence \eK causes any previously matched characters not to be
981 included in the final matched sequence. For example, the pattern:
982 .sp
983 foo\eKbar
984 .sp
985 matches "foobar", but reports that it has matched "bar". This feature is
986 similar to a lookbehind assertion
987 .\" HTML <a href="#lookbehind">
988 .\" </a>
989 (described below).
990 .\"
991 However, in this case, the part of the subject before the real match does not
992 have to be of fixed length, as lookbehind assertions do. The use of \eK does
993 not interfere with the setting of
994 .\" HTML <a href="#subpattern">
995 .\" </a>
996 captured substrings.
997 .\"
998 For example, when the pattern
999 .sp
1000 (foo)\eKbar
1001 .sp
1002 matches "foobar", the first substring is still set to "foo".
1003 .P
1004 Perl documents that the use of \eK within assertions is "not well defined". In
1005 PCRE, \eK is acted upon when it occurs inside positive assertions, but is
1006 ignored in negative assertions.
1007 .
1008 .
1009 .\" HTML <a name="smallassertions"></a>
1010 .SS "Simple assertions"
1011 .rs
1012 .sp
1013 The final use of backslash is for certain simple assertions. An assertion
1014 specifies a condition that has to be met at a particular point in a match,
1015 without consuming any characters from the subject string. The use of
1016 subpatterns for more complicated assertions is described
1017 .\" HTML <a href="#bigassertions">
1018 .\" </a>
1019 below.
1020 .\"
1021 The backslashed assertions are:
1022 .sp
1023 \eb matches at a word boundary
1024 \eB matches when not at a word boundary
1025 \eA matches at the start of the subject
1026 \eZ matches at the end of the subject
1027 also matches before a newline at the end of the subject
1028 \ez matches only at the end of the subject
1029 \eG matches at the first matching position in the subject
1030 .sp
1031 Inside a character class, \eb has a different meaning; it matches the backspace
1032 character. If any other of these assertions appears in a character class, by
1033 default it matches the corresponding literal character (for example, \eB
1034 matches the letter B). However, if the PCRE_EXTRA option is set, an "invalid
1035 escape sequence" error is generated instead.
1036 .P
1037 A word boundary is a position in the subject string where the current character
1038 and the previous character do not both match \ew or \eW (i.e. one matches
1039 \ew and the other matches \eW), or the start or end of the string if the
1040 first or last character matches \ew, respectively. In a UTF mode, the meanings
1041 of \ew and \eW can be changed by setting the PCRE_UCP option. When this is
1042 done, it also affects \eb and \eB. Neither PCRE nor Perl has a separate "start
1043 of word" or "end of word" metasequence. However, whatever follows \eb normally
1044 determines which it is. For example, the fragment \eba matches "a" at the start
1045 of a word.
1046 .P
1047 The \eA, \eZ, and \ez assertions differ from the traditional circumflex and
1048 dollar (described in the next section) in that they only ever match at the very
1049 start and end of the subject string, whatever options are set. Thus, they are
1050 independent of multiline mode. These three assertions are not affected by the
1051 PCRE_NOTBOL or PCRE_NOTEOL options, which affect only the behaviour of the
1052 circumflex and dollar metacharacters. However, if the \fIstartoffset\fP
1053 argument of \fBpcre_exec()\fP is non-zero, indicating that matching is to start
1054 at a point other than the beginning of the subject, \eA can never match. The
1055 difference between \eZ and \ez is that \eZ matches before a newline at the end
1056 of the string as well as at the very end, whereas \ez matches only at the end.
1057 .P
1058 The \eG assertion is true only when the current matching position is at the
1059 start point of the match, as specified by the \fIstartoffset\fP argument of
1060 \fBpcre_exec()\fP. It differs from \eA when the value of \fIstartoffset\fP is
1061 non-zero. By calling \fBpcre_exec()\fP multiple times with appropriate
1062 arguments, you can mimic Perl's /g option, and it is in this kind of
1063 implementation where \eG can be useful.
1064 .P
1065 Note, however, that PCRE's interpretation of \eG, as the start of the current
1066 match, is subtly different from Perl's, which defines it as the end of the
1067 previous match. In Perl, these can be different when the previously matched
1068 string was empty. Because PCRE does just one match at a time, it cannot
1069 reproduce this behaviour.
1070 .P
1071 If all the alternatives of a pattern begin with \eG, the expression is anchored
1072 to the starting match position, and the "anchored" flag is set in the compiled
1073 regular expression.
1074 .
1075 .
1076 .SH "CIRCUMFLEX AND DOLLAR"
1077 .rs
1078 .sp
1079 The circumflex and dollar metacharacters are zero-width assertions. That is,
1080 they test for a particular condition being true without consuming any
1081 characters from the subject string.
1082 .P
1083 Outside a character class, in the default matching mode, the circumflex
1084 character is an assertion that is true only if the current matching point is at
1085 the start of the subject string. If the \fIstartoffset\fP argument of
1086 \fBpcre_exec()\fP is non-zero, circumflex can never match if the PCRE_MULTILINE
1087 option is unset. Inside a character class, circumflex has an entirely different
1088 meaning
1089 .\" HTML <a href="#characterclass">
1090 .\" </a>
1091 (see below).
1092 .\"
1093 .P
1094 Circumflex need not be the first character of the pattern if a number of
1095 alternatives are involved, but it should be the first thing in each alternative
1096 in which it appears if the pattern is ever to match that branch. If all
1097 possible alternatives start with a circumflex, that is, if the pattern is
1098 constrained to match only at the start of the subject, it is said to be an
1099 "anchored" pattern. (There are also other constructs that can cause a pattern
1100 to be anchored.)
1101 .P
1102 The dollar character is an assertion that is true only if the current matching
1103 point is at the end of the subject string, or immediately before a newline at
1104 the end of the string (by default). Note, however, that it does not actually
1105 match the newline. Dollar need not be the last character of the pattern if a
1106 number of alternatives are involved, but it should be the last item in any
1107 branch in which it appears. Dollar has no special meaning in a character class.
1108 .P
1109 The meaning of dollar can be changed so that it matches only at the very end of
1110 the string, by setting the PCRE_DOLLAR_ENDONLY option at compile time. This
1111 does not affect the \eZ assertion.
1112 .P
1113 The meanings of the circumflex and dollar characters are changed if the
1114 PCRE_MULTILINE option is set. When this is the case, a circumflex matches
1115 immediately after internal newlines as well as at the start of the subject
1116 string. It does not match after a newline that ends the string. A dollar
1117 matches before any newlines in the string, as well as at the very end, when
1118 PCRE_MULTILINE is set. When newline is specified as the two-character
1119 sequence CRLF, isolated CR and LF characters do not indicate newlines.
1120 .P
1121 For example, the pattern /^abc$/ matches the subject string "def\enabc" (where
1122 \en represents a newline) in multiline mode, but not otherwise. Consequently,
1123 patterns that are anchored in single line mode because all branches start with
1124 ^ are not anchored in multiline mode, and a match for circumflex is possible
1125 when the \fIstartoffset\fP argument of \fBpcre_exec()\fP is non-zero. The
1126 PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
1127 .P
1128 Note that the sequences \eA, \eZ, and \ez can be used to match the start and
1129 end of the subject in both modes, and if all branches of a pattern start with
1130 \eA it is always anchored, whether or not PCRE_MULTILINE is set.
1131 .
1132 .
1133 .\" HTML <a name="fullstopdot"></a>
1134 .SH "FULL STOP (PERIOD, DOT) AND \eN"
1135 .rs
1136 .sp
1137 Outside a character class, a dot in the pattern matches any one character in
1138 the subject string except (by default) a character that signifies the end of a
1139 line.
1140 .P
1141 When a line ending is defined as a single character, dot never matches that
1142 character; when the two-character sequence CRLF is used, dot does not match CR
1143 if it is immediately followed by LF, but otherwise it matches all characters
1144 (including isolated CRs and LFs). When any Unicode line endings are being
1145 recognized, dot does not match CR or LF or any of the other line ending
1146 characters.
1147 .P
1148 The behaviour of dot with regard to newlines can be changed. If the PCRE_DOTALL
1149 option is set, a dot matches any one character, without exception. If the
1150 two-character sequence CRLF is present in the subject string, it takes two dots
1151 to match it.
1152 .P
1153 The handling of dot is entirely independent of the handling of circumflex and
1154 dollar, the only relationship being that they both involve newlines. Dot has no
1155 special meaning in a character class.
1156 .P
1157 The escape sequence \eN behaves like a dot, except that it is not affected by
1158 the PCRE_DOTALL option. In other words, it matches any character except one
1159 that signifies the end of a line. Perl also uses \eN to match characters by
1160 name; PCRE does not support this.
1161 .
1162 .
1163 .SH "MATCHING A SINGLE DATA UNIT"
1164 .rs
1165 .sp
1166 Outside a character class, the escape sequence \eC matches any one data unit,
1167 whether or not a UTF mode is set. In the 8-bit library, one data unit is one
1168 byte; in the 16-bit library it is a 16-bit unit; in the 32-bit library it is
1169 a 32-bit unit. Unlike a dot, \eC always
1170 matches line-ending characters. The feature is provided in Perl in order to
1171 match individual bytes in UTF-8 mode, but it is unclear how it can usefully be
1172 used. Because \eC breaks up characters into individual data units, matching one
1173 unit with \eC in a UTF mode means that the rest of the string may start with a
1174 malformed UTF character. This has undefined results, because PCRE assumes that
1175 it is dealing with valid UTF strings (and by default it checks this at the
1176 start of processing unless the PCRE_NO_UTF8_CHECK, PCRE_NO_UTF16_CHECK or
1177 PCRE_NO_UTF32_CHECK option is used).
1178 .P
1179 PCRE does not allow \eC to appear in lookbehind assertions
1180 .\" HTML <a href="#lookbehind">
1181 .\" </a>
1182 (described below)
1183 .\"
1184 in a UTF mode, because this would make it impossible to calculate the length of
1185 the lookbehind.
1186 .P
1187 In general, the \eC escape sequence is best avoided. However, one
1188 way of using it that avoids the problem of malformed UTF characters is to use a
1189 lookahead to check the length of the next character, as in this pattern, which
1190 could be used with a UTF-8 string (ignore white space and line breaks):
1191 .sp
1192 (?| (?=[\ex00-\ex7f])(\eC) |
1193 (?=[\ex80-\ex{7ff}])(\eC)(\eC) |
1194 (?=[\ex{800}-\ex{ffff}])(\eC)(\eC)(\eC) |
1195 (?=[\ex{10000}-\ex{1fffff}])(\eC)(\eC)(\eC)(\eC))
1196 .sp
1197 A group that starts with (?| resets the capturing parentheses numbers in each
1198 alternative (see
1199 .\" HTML <a href="#dupsubpatternnumber">
1200 .\" </a>
1201 "Duplicate Subpattern Numbers"
1202 .\"
1203 below). The assertions at the start of each branch check the next UTF-8
1204 character for values whose encoding uses 1, 2, 3, or 4 bytes, respectively. The
1205 character's individual bytes are then captured by the appropriate number of
1206 groups.
1207 .
1208 .
1209 .\" HTML <a name="characterclass"></a>
1210 .SH "SQUARE BRACKETS AND CHARACTER CLASSES"
1211 .rs
1212 .sp
1213 An opening square bracket introduces a character class, terminated by a closing
1214 square bracket. A closing square bracket on its own is not special by default.
1215 However, if the PCRE_JAVASCRIPT_COMPAT option is set, a lone closing square
1216 bracket causes a compile-time error. If a closing square bracket is required as
1217 a member of the class, it should be the first data character in the class
1218 (after an initial circumflex, if present) or escaped with a backslash.
1219 .P
1220 A character class matches a single character in the subject. In a UTF mode, the
1221 character may be more than one data unit long. A matched character must be in
1222 the set of characters defined by the class, unless the first character in the
1223 class definition is a circumflex, in which case the subject character must not
1224 be in the set defined by the class. If a circumflex is actually required as a
1225 member of the class, ensure it is not the first character, or escape it with a
1226 backslash.
1227 .P
1228 For example, the character class [aeiou] matches any lower case vowel, while
1229 [^aeiou] matches any character that is not a lower case vowel. Note that a
1230 circumflex is just a convenient notation for specifying the characters that
1231 are in the class by enumerating those that are not. A class that starts with a
1232 circumflex is not an assertion; it still consumes a character from the subject
1233 string, and therefore it fails if the current pointer is at the end of the
1234 string.
1235 .P
1236 In UTF-8 (UTF-16, UTF-32) mode, characters with values greater than 255 (0xffff)
1237 can be included in a class as a literal string of data units, or by using the
1238 \ex{ escaping mechanism.
1239 .P
1240 When caseless matching is set, any letters in a class represent both their
1241 upper case and lower case versions, so for example, a caseless [aeiou] matches
1242 "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
1243 caseful version would. In a UTF mode, PCRE always understands the concept of
1244 case for characters whose values are less than 128, so caseless matching is
1245 always possible. For characters with higher values, the concept of case is
1246 supported if PCRE is compiled with Unicode property support, but not otherwise.
1247 If you want to use caseless matching in a UTF mode for characters 128 and
1248 above, you must ensure that PCRE is compiled with Unicode property support as
1249 well as with UTF support.
1250 .P
1251 Characters that might indicate line breaks are never treated in any special way
1252 when matching character classes, whatever line-ending sequence is in use, and
1253 whatever setting of the PCRE_DOTALL and PCRE_MULTILINE options is used. A class
1254 such as [^a] always matches one of these characters.
1255 .P
1256 The minus (hyphen) character can be used to specify a range of characters in a
1257 character class. For example, [d-m] matches any letter between d and m,
1258 inclusive. If a minus character is required in a class, it must be escaped with
1259 a backslash or appear in a position where it cannot be interpreted as
1260 indicating a range, typically as the first or last character in the class, or
1261 immediately after a range. For example, [b-d-z] matches letters in the range b
1262 to d, a hyphen character, or z.
1263 .P
1264 It is not possible to have the literal character "]" as the end character of a
1265 range. A pattern such as [W-]46] is interpreted as a class of two characters
1266 ("W" and "-") followed by a literal string "46]", so it would match "W46]" or
1267 "-46]". However, if the "]" is escaped with a backslash it is interpreted as
1268 the end of range, so [W-\e]46] is interpreted as a class containing a range
1269 followed by two other characters. The octal or hexadecimal representation of
1270 "]" can also be used to end a range.
1271 .P
1272 An error is generated if a POSIX character class (see below) or an escape
1273 sequence other than one that defines a single character appears at a point
1274 where a range ending character is expected. For example, [z-\exff] is valid,
1275 but [A-\ed] and [A-[:digit:]] are not.
1276 .P
1277 Ranges operate in the collating sequence of character values. They can also be
1278 used for characters specified numerically, for example [\e000-\e037]. Ranges
1279 can include any characters that are valid for the current mode.
1280 .P
1281 If a range that includes letters is used when caseless matching is set, it
1282 matches the letters in either case. For example, [W-c] is equivalent to
1283 [][\e\e^_`wxyzabc], matched caselessly, and in a non-UTF mode, if character
1284 tables for a French locale are in use, [\exc8-\excb] matches accented E
1285 characters in both cases. In UTF modes, PCRE supports the concept of case for
1286 characters with values greater than 128 only when it is compiled with Unicode
1287 property support.
1288 .P
1289 The character escape sequences \ed, \eD, \eh, \eH, \ep, \eP, \es, \eS, \ev,
1290 \eV, \ew, and \eW may appear in a character class, and add the characters that
1291 they match to the class. For example, [\edABCDEF] matches any hexadecimal
1292 digit. In UTF modes, the PCRE_UCP option affects the meanings of \ed, \es, \ew
1293 and their upper case partners, just as it does when they appear outside a
1294 character class, as described in the section entitled
1295 .\" HTML <a href="#genericchartypes">
1296 .\" </a>
1297 "Generic character types"
1298 .\"
1299 above. The escape sequence \eb has a different meaning inside a character
1300 class; it matches the backspace character. The sequences \eB, \eN, \eR, and \eX
1301 are not special inside a character class. Like any other unrecognized escape
1302 sequences, they are treated as the literal characters "B", "N", "R", and "X" by
1303 default, but cause an error if the PCRE_EXTRA option is set.
1304 .P
1305 A circumflex can conveniently be used with the upper case character types to
1306 specify a more restricted set of characters than the matching lower case type.
1307 For example, the class [^\eW_] matches any letter or digit, but not underscore,
1308 whereas [\ew] includes underscore. A positive character class should be read as
1309 "something OR something OR ..." and a negative class as "NOT something AND NOT
1310 something AND NOT ...".
1311 .P
1312 The only metacharacters that are recognized in character classes are backslash,
1313 hyphen (only where it can be interpreted as specifying a range), circumflex
1314 (only at the start), opening square bracket (only when it can be interpreted as
1315 introducing a POSIX class name - see the next section), and the terminating
1316 closing square bracket. However, escaping other non-alphanumeric characters
1317 does no harm.
1318 .
1319 .
1320 .SH "POSIX CHARACTER CLASSES"
1321 .rs
1322 .sp
1323 Perl supports the POSIX notation for character classes. This uses names
1324 enclosed by [: and :] within the enclosing square brackets. PCRE also supports
1325 this notation. For example,
1326 .sp
1327 [01[:alpha:]%]
1328 .sp
1329 matches "0", "1", any alphabetic character, or "%". The supported class names
1330 are:
1331 .sp
1332 alnum letters and digits
1333 alpha letters
1334 ascii character codes 0 - 127
1335 blank space or tab only
1336 cntrl control characters
1337 digit decimal digits (same as \ed)
1338 graph printing characters, excluding space
1339 lower lower case letters
1340 print printing characters, including space
1341 punct printing characters, excluding letters and digits and space
1342 space white space (the same as \es from PCRE 8.34)
1343 upper upper case letters
1344 word "word" characters (same as \ew)
1345 xdigit hexadecimal digits
1346 .sp
1347 The default "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13),
1348 and space (32). If locale-specific matching is taking place, there may be
1349 additional space characters. "Space" used to be different to \es, which did not
1350 include VT, for Perl compatibility. However, Perl changed at release 5.18, and
1351 PCRE followed at release 8.34. "Space" and \es now match the same set of
1352 characters.
1353 .P
1354 The name "word" is a Perl extension, and "blank" is a GNU extension from Perl
1355 5.8. Another Perl extension is negation, which is indicated by a ^ character
1356 after the colon. For example,
1357 .sp
1358 [12[:^digit:]]
1359 .sp
1360 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX
1361 syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
1362 supported, and an error is given if they are encountered.
1363 .P
1364 By default, characters with values greater than 128 do not match any of the
1365 POSIX character classes. However, if the PCRE_UCP option is passed to
1366 \fBpcre_compile()\fP, some of the classes are changed so that Unicode character
1367 properties are used. This is achieved by replacing certain POSIX classes by
1368 other sequences, as follows:
1369 .sp
1370 [:alnum:] becomes \ep{Xan}
1371 [:alpha:] becomes \ep{L}
1372 [:blank:] becomes \eh
1373 [:digit:] becomes \ep{Nd}
1374 [:lower:] becomes \ep{Ll}
1375 [:space:] becomes \ep{Xps}
1376 [:upper:] becomes \ep{Lu}
1377 [:word:] becomes \ep{Xwd}
1378 .sp
1379 Negated versions, such as [:^alpha:] use \eP instead of \ep. Three other POSIX
1380 classes are handled specially in UCP mode:
1381 .TP 10
1382 [:graph:]
1383 This matches characters that have glyphs that mark the page when printed. In
1384 Unicode property terms, it matches all characters with the L, M, N, P, S, or Cf
1385 properties, except for:
1386 .sp
1387 U+061C Arabic Letter Mark
1388 U+180E Mongolian Vowel Separator
1389 U+2066 - U+2069 Various "isolate"s
1390 .sp
1391 .TP 10
1392 [:print:]
1393 This matches the same characters as [:graph:] plus space characters that are
1394 not controls, that is, characters with the Zs property.
1395 .TP 10
1396 [:punct:]
1397 This matches all characters that have the Unicode P (punctuation) property,
1398 plus those characters whose code points are less than 128 that have the S
1399 (Symbol) property.
1400 .P
1401 The other POSIX classes are unchanged, and match only characters with code
1402 points less than 128.
1403 .
1404 .
1405 .SH "VERTICAL BAR"
1406 .rs
1407 .sp
1408 Vertical bar characters are used to separate alternative patterns. For example,
1409 the pattern
1410 .sp
1411 gilbert|sullivan
1412 .sp
1413 matches either "gilbert" or "sullivan". Any number of alternatives may appear,
1414 and an empty alternative is permitted (matching the empty string). The matching
1415 process tries each alternative in turn, from left to right, and the first one
1416 that succeeds is used. If the alternatives are within a subpattern
1417 .\" HTML <a href="#subpattern">
1418 .\" </a>
1419 (defined below),
1420 .\"
1421 "succeeds" means matching the rest of the main pattern as well as the
1422 alternative in the subpattern.
1423 .
1424 .
1425 .SH "INTERNAL OPTION SETTING"
1426 .rs
1427 .sp
1428 The settings of the PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and
1429 PCRE_EXTENDED options (which are Perl-compatible) can be changed from within
1430 the pattern by a sequence of Perl option letters enclosed between "(?" and ")".
1431 The option letters are
1432 .sp
1433 i for PCRE_CASELESS
1434 m for PCRE_MULTILINE
1435 s for PCRE_DOTALL
1436 x for PCRE_EXTENDED
1437 .sp
1438 For example, (?im) sets caseless, multiline matching. It is also possible to
1439 unset these options by preceding the letter with a hyphen, and a combined
1440 setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
1441 PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
1442 permitted. If a letter appears both before and after the hyphen, the option is
1443 unset.
1444 .P
1445 The PCRE-specific options PCRE_DUPNAMES, PCRE_UNGREEDY, and PCRE_EXTRA can be
1446 changed in the same way as the Perl-compatible options by using the characters
1447 J, U and X respectively.
1448 .P
1449 When one of these option changes occurs at top level (that is, not inside
1450 subpattern parentheses), the change applies to the remainder of the pattern
1451 that follows. If the change is placed right at the start of a pattern, PCRE
1452 extracts it into the global options (and it will therefore show up in data
1453 extracted by the \fBpcre_fullinfo()\fP function).
1454 .P
1455 An option change within a subpattern (see below for a description of
1456 subpatterns) affects only that part of the subpattern that follows it, so
1457 .sp
1458 (a(?i)b)c
1459 .sp
1460 matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
1461 By this means, options can be made to have different settings in different
1462 parts of the pattern. Any changes made in one alternative do carry on
1463 into subsequent branches within the same subpattern. For example,
1464 .sp
1465 (a(?i)b|c)
1466 .sp
1467 matches "ab", "aB", "c", and "C", even though when matching "C" the first
1468 branch is abandoned before the option setting. This is because the effects of
1469 option settings happen at compile time. There would be some very weird
1470 behaviour otherwise.
1471 .P
1472 \fBNote:\fP There are other PCRE-specific options that can be set by the
1473 application when the compiling or matching functions are called. In some cases
1474 the pattern can contain special leading sequences such as (*CRLF) to override
1475 what the application has set or what has been defaulted. Details are given in
1476 the section entitled
1477 .\" HTML <a href="#newlineseq">
1478 .\" </a>
1479 "Newline sequences"
1480 .\"
1481 above. There are also the (*UTF8), (*UTF16),(*UTF32), and (*UCP) leading
1482 sequences that can be used to set UTF and Unicode property modes; they are
1483 equivalent to setting the PCRE_UTF8, PCRE_UTF16, PCRE_UTF32 and the PCRE_UCP
1484 options, respectively. The (*UTF) sequence is a generic version that can be
1485 used with any of the libraries. However, the application can set the
1486 PCRE_NEVER_UTF option, which locks out the use of the (*UTF) sequences.
1487 .
1488 .
1489 .\" HTML <a name="subpattern"></a>
1490 .SH SUBPATTERNS
1491 .rs
1492 .sp
1493 Subpatterns are delimited by parentheses (round brackets), which can be nested.
1494 Turning part of a pattern into a subpattern does two things:
1495 .sp
1496 1. It localizes a set of alternatives. For example, the pattern
1497 .sp
1498 cat(aract|erpillar|)
1499 .sp
1500 matches "cataract", "caterpillar", or "cat". Without the parentheses, it would
1501 match "cataract", "erpillar" or an empty string.
1502 .sp
1503 2. It sets up the subpattern as a capturing subpattern. This means that, when
1504 the whole pattern matches, that portion of the subject string that matched the
1505 subpattern is passed back to the caller via the \fIovector\fP argument of the
1506 matching function. (This applies only to the traditional matching functions;
1507 the DFA matching functions do not support capturing.)
1508 .P
1509 Opening parentheses are counted from left to right (starting from 1) to obtain
1510 numbers for the capturing subpatterns. For example, if the string "the red
1511 king" is matched against the pattern
1512 .sp
1513 the ((red|white) (king|queen))
1514 .sp
1515 the captured substrings are "red king", "red", and "king", and are numbered 1,
1516 2, and 3, respectively.
1517 .P
1518 The fact that plain parentheses fulfil two functions is not always helpful.
1519 There are often times when a grouping subpattern is required without a
1520 capturing requirement. If an opening parenthesis is followed by a question mark
1521 and a colon, the subpattern does not do any capturing, and is not counted when
1522 computing the number of any subsequent capturing subpatterns. For example, if
1523 the string "the white queen" is matched against the pattern
1524 .sp
1525 the ((?:red|white) (king|queen))
1526 .sp
1527 the captured substrings are "white queen" and "queen", and are numbered 1 and
1528 2. The maximum number of capturing subpatterns is 65535.
1529 .P
1530 As a convenient shorthand, if any option settings are required at the start of
1531 a non-capturing subpattern, the option letters may appear between the "?" and
1532 the ":". Thus the two patterns
1533 .sp
1534 (?i:saturday|sunday)
1535 (?:(?i)saturday|sunday)
1536 .sp
1537 match exactly the same set of strings. Because alternative branches are tried
1538 from left to right, and options are not reset until the end of the subpattern
1539 is reached, an option setting in one branch does affect subsequent branches, so
1540 the above patterns match "SUNDAY" as well as "Saturday".
1541 .
1542 .
1543 .\" HTML <a name="dupsubpatternnumber"></a>
1544 .SH "DUPLICATE SUBPATTERN NUMBERS"
1545 .rs
1546 .sp
1547 Perl 5.10 introduced a feature whereby each alternative in a subpattern uses
1548 the same numbers for its capturing parentheses. Such a subpattern starts with
1549 (?| and is itself a non-capturing subpattern. For example, consider this
1550 pattern:
1551 .sp
1552 (?|(Sat)ur|(Sun))day
1553 .sp
1554 Because the two alternatives are inside a (?| group, both sets of capturing
1555 parentheses are numbered one. Thus, when the pattern matches, you can look
1556 at captured substring number one, whichever alternative matched. This construct
1557 is useful when you want to capture part, but not all, of one of a number of
1558 alternatives. Inside a (?| group, parentheses are numbered as usual, but the
1559 number is reset at the start of each branch. The numbers of any capturing
1560 parentheses that follow the subpattern start after the highest number used in
1561 any branch. The following example is taken from the Perl documentation. The
1562 numbers underneath show in which buffer the captured content will be stored.
1563 .sp
1564 # before ---------------branch-reset----------- after
1565 / ( a ) (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
1566 # 1 2 2 3 2 3 4
1567 .sp
1568 A back reference to a numbered subpattern uses the most recent value that is
1569 set for that number by any subpattern. The following pattern matches "abcabc"
1570 or "defdef":
1571 .sp
1572 /(?|(abc)|(def))\e1/
1573 .sp
1574 In contrast, a subroutine call to a numbered subpattern always refers to the
1575 first one in the pattern with the given number. The following pattern matches
1576 "abcabc" or "defabc":
1577 .sp
1578 /(?|(abc)|(def))(?1)/
1579 .sp
1580 If a
1581 .\" HTML <a href="#conditions">
1582 .\" </a>
1583 condition test
1584 .\"
1585 for a subpattern's having matched refers to a non-unique number, the test is
1586 true if any of the subpatterns of that number have matched.
1587 .P
1588 An alternative approach to using this "branch reset" feature is to use
1589 duplicate named subpatterns, as described in the next section.
1590 .
1591 .
1592 .SH "NAMED SUBPATTERNS"
1593 .rs
1594 .sp
1595 Identifying capturing parentheses by number is simple, but it can be very hard
1596 to keep track of the numbers in complicated regular expressions. Furthermore,
1597 if an expression is modified, the numbers may change. To help with this
1598 difficulty, PCRE supports the naming of subpatterns. This feature was not
1599 added to Perl until release 5.10. Python had the feature earlier, and PCRE
1600 introduced it at release 4.0, using the Python syntax. PCRE now supports both
1601 the Perl and the Python syntax. Perl allows identically numbered subpatterns to
1602 have different names, but PCRE does not.
1603 .P
1604 In PCRE, a subpattern can be named in one of three ways: (?<name>...) or
1605 (?'name'...) as in Perl, or (?P<name>...) as in Python. References to capturing
1606 parentheses from other parts of the pattern, such as
1607 .\" HTML <a href="#backreferences">
1608 .\" </a>
1609 back references,
1610 .\"
1611 .\" HTML <a href="#recursion">
1612 .\" </a>
1613 recursion,
1614 .\"
1615 and
1616 .\" HTML <a href="#conditions">
1617 .\" </a>
1618 conditions,
1619 .\"
1620 can be made by name as well as by number.
1621 .P
1622 Names consist of up to 32 alphanumeric characters and underscores, but must
1623 start with a non-digit. Named capturing parentheses are still allocated numbers
1624 as well as names, exactly as if the names were not present. The PCRE API
1625 provides function calls for extracting the name-to-number translation table
1626 from a compiled pattern. There is also a convenience function for extracting a
1627 captured substring by name.
1628 .P
1629 By default, a name must be unique within a pattern, but it is possible to relax
1630 this constraint by setting the PCRE_DUPNAMES option at compile time. (Duplicate
1631 names are also always permitted for subpatterns with the same number, set up as
1632 described in the previous section.) Duplicate names can be useful for patterns
1633 where only one instance of the named parentheses can match. Suppose you want to
1634 match the name of a weekday, either as a 3-letter abbreviation or as the full
1635 name, and in both cases you want to extract the abbreviation. This pattern
1636 (ignoring the line breaks) does the job:
1637 .sp
1638 (?<DN>Mon|Fri|Sun)(?:day)?|
1639 (?<DN>Tue)(?:sday)?|
1640 (?<DN>Wed)(?:nesday)?|
1641 (?<DN>Thu)(?:rsday)?|
1642 (?<DN>Sat)(?:urday)?
1643 .sp
1644 There are five capturing substrings, but only one is ever set after a match.
1645 (An alternative way of solving this problem is to use a "branch reset"
1646 subpattern, as described in the previous section.)
1647 .P
1648 The convenience function for extracting the data by name returns the substring
1649 for the first (and in this example, the only) subpattern of that name that
1650 matched. This saves searching to find which numbered subpattern it was.
1651 .P
1652 If you make a back reference to a non-unique named subpattern from elsewhere in
1653 the pattern, the subpatterns to which the name refers are checked in the order
1654 in which they appear in the overall pattern. The first one that is set is used
1655 for the reference. For example, this pattern matches both "foofoo" and
1656 "barbar" but not "foobar" or "barfoo":
1657 .sp
1658 (?:(?<n>foo)|(?<n>bar))\ek<n>
1659 .sp
1660 .P
1661 If you make a subroutine call to a non-unique named subpattern, the one that
1662 corresponds to the first occurrence of the name is used. In the absence of
1663 duplicate numbers (see the previous section) this is the one with the lowest
1664 number.
1665 .P
1666 If you use a named reference in a condition
1667 test (see the
1668 .\"
1669 .\" HTML <a href="#conditions">
1670 .\" </a>
1671 section about conditions
1672 .\"
1673 below), either to check whether a subpattern has matched, or to check for
1674 recursion, all subpatterns with the same name are tested. If the condition is
1675 true for any one of them, the overall condition is true. This is the same
1676 behaviour as testing by number. For further details of the interfaces for
1677 handling named subpatterns, see the
1678 .\" HREF
1679 \fBpcreapi\fP
1680 .\"
1681 documentation.
1682 .P
1683 \fBWarning:\fP You cannot use different names to distinguish between two
1684 subpatterns with the same number because PCRE uses only the numbers when
1685 matching. For this reason, an error is given at compile time if different names
1686 are given to subpatterns with the same number. However, you can always give the
1687 same name to subpatterns with the same number, even when PCRE_DUPNAMES is not
1688 set.
1689 .
1690 .
1691 .SH REPETITION
1692 .rs
1693 .sp
1694 Repetition is specified by quantifiers, which can follow any of the following
1695 items:
1696 .sp
1697 a literal data character
1698 the dot metacharacter
1699 the \eC escape sequence
1700 the \eX escape sequence
1701 the \eR escape sequence
1702 an escape such as \ed or \epL that matches a single character
1703 a character class
1704 a back reference (see next section)
1705 a parenthesized subpattern (including assertions)
1706 a subroutine call to a subpattern (recursive or otherwise)
1707 .sp
1708 The general repetition quantifier specifies a minimum and maximum number of
1709 permitted matches, by giving the two numbers in curly brackets (braces),
1710 separated by a comma. The numbers must be less than 65536, and the first must
1711 be less than or equal to the second. For example:
1712 .sp
1713 z{2,4}
1714 .sp
1715 matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
1716 character. If the second number is omitted, but the comma is present, there is
1717 no upper limit; if the second number and the comma are both omitted, the
1718 quantifier specifies an exact number of required matches. Thus
1719 .sp
1720 [aeiou]{3,}
1721 .sp
1722 matches at least 3 successive vowels, but may match many more, while
1723 .sp
1724 \ed{8}
1725 .sp
1726 matches exactly 8 digits. An opening curly bracket that appears in a position
1727 where a quantifier is not allowed, or one that does not match the syntax of a
1728 quantifier, is taken as a literal character. For example, {,6} is not a
1729 quantifier, but a literal string of four characters.
1730 .P
1731 In UTF modes, quantifiers apply to characters rather than to individual data
1732 units. Thus, for example, \ex{100}{2} matches two characters, each of
1733 which is represented by a two-byte sequence in a UTF-8 string. Similarly,
1734 \eX{3} matches three Unicode extended grapheme clusters, each of which may be
1735 several data units long (and they may be of different lengths).
1736 .P
1737 The quantifier {0} is permitted, causing the expression to behave as if the
1738 previous item and the quantifier were not present. This may be useful for
1739 subpatterns that are referenced as
1740 .\" HTML <a href="#subpatternsassubroutines">
1741 .\" </a>
1742 subroutines
1743 .\"
1744 from elsewhere in the pattern (but see also the section entitled
1745 .\" HTML <a href="#subdefine">
1746 .\" </a>
1747 "Defining subpatterns for use by reference only"
1748 .\"
1749 below). Items other than subpatterns that have a {0} quantifier are omitted
1750 from the compiled pattern.
1751 .P
1752 For convenience, the three most common quantifiers have single-character
1753 abbreviations:
1754 .sp
1755 * is equivalent to {0,}
1756 + is equivalent to {1,}
1757 ? is equivalent to {0,1}
1758 .sp
1759 It is possible to construct infinite loops by following a subpattern that can
1760 match no characters with a quantifier that has no upper limit, for example:
1761 .sp
1762 (a?)*
1763 .sp
1764 Earlier versions of Perl and PCRE used to give an error at compile time for
1765 such patterns. However, because there are cases where this can be useful, such
1766 patterns are now accepted, but if any repetition of the subpattern does in fact
1767 match no characters, the loop is forcibly broken.
1768 .P
1769 By default, the quantifiers are "greedy", that is, they match as much as
1770 possible (up to the maximum number of permitted times), without causing the
1771 rest of the pattern to fail. The classic example of where this gives problems
1772 is in trying to match comments in C programs. These appear between /* and */
1773 and within the comment, individual * and / characters may appear. An attempt to
1774 match C comments by applying the pattern
1775 .sp
1776 /\e*.*\e*/
1777 .sp
1778 to the string
1779 .sp
1780 /* first comment */ not comment /* second comment */
1781 .sp
1782 fails, because it matches the entire string owing to the greediness of the .*
1783 item.
1784 .P
1785 However, if a quantifier is followed by a question mark, it ceases to be
1786 greedy, and instead matches the minimum number of times possible, so the
1787 pattern
1788 .sp
1789 /\e*.*?\e*/
1790 .sp
1791 does the right thing with the C comments. The meaning of the various
1792 quantifiers is not otherwise changed, just the preferred number of matches.
1793 Do not confuse this use of question mark with its use as a quantifier in its
1794 own right. Because it has two uses, it can sometimes appear doubled, as in
1795 .sp
1796 \ed??\ed
1797 .sp
1798 which matches one digit by preference, but can match two if that is the only
1799 way the rest of the pattern matches.
1800 .P
1801 If the PCRE_UNGREEDY option is set (an option that is not available in Perl),
1802 the quantifiers are not greedy by default, but individual ones can be made
1803 greedy by following them with a question mark. In other words, it inverts the
1804 default behaviour.
1805 .P
1806 When a parenthesized subpattern is quantified with a minimum repeat count that
1807 is greater than 1 or with a limited maximum, more memory is required for the
1808 compiled pattern, in proportion to the size of the minimum or maximum.
1809 .P
1810 If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent
1811 to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is
1812 implicitly anchored, because whatever follows will be tried against every
1813 character position in the subject string, so there is no point in retrying the
1814 overall match at any position after the first. PCRE normally treats such a
1815 pattern as though it were preceded by \eA.
1816 .P
1817 In cases where it is known that the subject string contains no newlines, it is
1818 worth setting PCRE_DOTALL in order to obtain this optimization, or
1819 alternatively using ^ to indicate anchoring explicitly.
1820 .P
1821 However, there are some cases where the optimization cannot be used. When .*
1822 is inside capturing parentheses that are the subject of a back reference
1823 elsewhere in the pattern, a match at the start may fail where a later one
1824 succeeds. Consider, for example:
1825 .sp
1826 (.*)abc\e1
1827 .sp
1828 If the subject is "xyz123abc123" the match point is the fourth character. For
1829 this reason, such a pattern is not implicitly anchored.
1830 .P
1831 Another case where implicit anchoring is not applied is when the leading .* is
1832 inside an atomic group. Once again, a match at the start may fail where a later
1833 one succeeds. Consider this pattern:
1834 .sp
1835 (?>.*?a)b
1836 .sp
1837 It matches "ab" in the subject "aab". The use of the backtracking control verbs
1838 (*PRUNE) and (*SKIP) also disable this optimization.
1839 .P
1840 When a capturing subpattern is repeated, the value captured is the substring
1841 that matched the final iteration. For example, after
1842 .sp
1843 (tweedle[dume]{3}\es*)+
1844 .sp
1845 has matched "tweedledum tweedledee" the value of the captured substring is
1846 "tweedledee". However, if there are nested capturing subpatterns, the
1847 corresponding captured values may have been set in previous iterations. For
1848 example, after
1849 .sp
1850 /(a|(b))+/
1851 .sp
1852 matches "aba" the value of the second captured substring is "b".
1853 .
1854 .
1855 .\" HTML <a name="atomicgroup"></a>
1856 .SH "ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS"
1857 .rs
1858 .sp
1859 With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
1860 repetition, failure of what follows normally causes the repeated item to be
1861 re-evaluated to see if a different number of repeats allows the rest of the
1862 pattern to match. Sometimes it is useful to prevent this, either to change the
1863 nature of the match, or to cause it fail earlier than it otherwise might, when
1864 the author of the pattern knows there is no point in carrying on.
1865 .P
1866 Consider, for example, the pattern \ed+foo when applied to the subject line
1867 .sp
1868 123456bar
1869 .sp
1870 After matching all 6 digits and then failing to match "foo", the normal
1871 action of the matcher is to try again with only 5 digits matching the \ed+
1872 item, and then with 4, and so on, before ultimately failing. "Atomic grouping"
1873 (a term taken from Jeffrey Friedl's book) provides the means for specifying
1874 that once a subpattern has matched, it is not to be re-evaluated in this way.
1875 .P
1876 If we use atomic grouping for the previous example, the matcher gives up
1877 immediately on failing to match "foo" the first time. The notation is a kind of
1878 special parenthesis, starting with (?> as in this example:
1879 .sp
1880 (?>\ed+)foo
1881 .sp
1882 This kind of parenthesis "locks up" the part of the pattern it contains once
1883 it has matched, and a failure further into the pattern is prevented from
1884 backtracking into it. Backtracking past it to previous items, however, works as
1885 normal.
1886 .P
1887 An alternative description is that a subpattern of this type matches the string
1888 of characters that an identical standalone pattern would match, if anchored at
1889 the current point in the subject string.
1890 .P
1891 Atomic grouping subpatterns are not capturing subpatterns. Simple cases such as
1892 the above example can be thought of as a maximizing repeat that must swallow
1893 everything it can. So, while both \ed+ and \ed+? are prepared to adjust the
1894 number of digits they match in order to make the rest of the pattern match,
1895 (?>\ed+) can only match an entire sequence of digits.
1896 .P
1897 Atomic groups in general can of course contain arbitrarily complicated
1898 subpatterns, and can be nested. However, when the subpattern for an atomic
1899 group is just a single repeated item, as in the example above, a simpler
1900 notation, called a "possessive quantifier" can be used. This consists of an
1901 additional + character following a quantifier. Using this notation, the
1902 previous example can be rewritten as
1903 .sp
1904 \ed++foo
1905 .sp
1906 Note that a possessive quantifier can be used with an entire group, for
1907 example:
1908 .sp
1909 (abc|xyz){2,3}+
1910 .sp
1911 Possessive quantifiers are always greedy; the setting of the PCRE_UNGREEDY
1912 option is ignored. They are a convenient notation for the simpler forms of
1913 atomic group. However, there is no difference in the meaning of a possessive
1914 quantifier and the equivalent atomic group, though there may be a performance
1915 difference; possessive quantifiers should be slightly faster.
1916 .P
1917 The possessive quantifier syntax is an extension to the Perl 5.8 syntax.
1918 Jeffrey Friedl originated the idea (and the name) in the first edition of his
1919 book. Mike McCloskey liked it, so implemented it when he built Sun's Java
1920 package, and PCRE copied it from there. It ultimately found its way into Perl
1921 at release 5.10.
1922 .P
1923 PCRE has an optimization that automatically "possessifies" certain simple
1924 pattern constructs. For example, the sequence A+B is treated as A++B because
1925 there is no point in backtracking into a sequence of A's when B must follow.
1926 .P
1927 When a pattern contains an unlimited repeat inside a subpattern that can itself
1928 be repeated an unlimited number of times, the use of an atomic group is the
1929 only way to avoid some failing matches taking a very long time indeed. The
1930 pattern
1931 .sp
1932 (\eD+|<\ed+>)*[!?]
1933 .sp
1934 matches an unlimited number of substrings that either consist of non-digits, or
1935 digits enclosed in <>, followed by either ! or ?. When it matches, it runs
1936 quickly. However, if it is applied to
1937 .sp
1938 aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
1939 .sp
1940 it takes a long time before reporting failure. This is because the string can
1941 be divided between the internal \eD+ repeat and the external * repeat in a
1942 large number of ways, and all have to be tried. (The example uses [!?] rather
1943 than a single character at the end, because both PCRE and Perl have an
1944 optimization that allows for fast failure when a single character is used. They
1945 remember the last single character that is required for a match, and fail early
1946 if it is not present in the string.) If the pattern is changed so that it uses
1947 an atomic group, like this:
1948 .sp
1949 ((?>\eD+)|<\ed+>)*[!?]
1950 .sp
1951 sequences of non-digits cannot be broken, and failure happens quickly.
1952 .
1953 .
1954 .\" HTML <a name="backreferences"></a>
1955 .SH "BACK REFERENCES"
1956 .rs
1957 .sp
1958 Outside a character class, a backslash followed by a digit greater than 0 (and
1959 possibly further digits) is a back reference to a capturing subpattern earlier
1960 (that is, to its left) in the pattern, provided there have been that many
1961 previous capturing left parentheses.
1962 .P
1963 However, if the decimal number following the backslash is less than 10, it is
1964 always taken as a back reference, and causes an error only if there are not
1965 that many capturing left parentheses in the entire pattern. In other words, the
1966 parentheses that are referenced need not be to the left of the reference for
1967 numbers less than 10. A "forward back reference" of this type can make sense
1968 when a repetition is involved and the subpattern to the right has participated
1969 in an earlier iteration.
1970 .P
1971 It is not possible to have a numerical "forward back reference" to a subpattern
1972 whose number is 10 or more using this syntax because a sequence such as \e50 is
1973 interpreted as a character defined in octal. See the subsection entitled
1974 "Non-printing characters"
1975 .\" HTML <a href="#digitsafterbackslash">
1976 .\" </a>
1977 above
1978 .\"
1979 for further details of the handling of digits following a backslash. There is
1980 no such problem when named parentheses are used. A back reference to any
1981 subpattern is possible using named parentheses (see below).
1982 .P
1983 Another way of avoiding the ambiguity inherent in the use of digits following a
1984 backslash is to use the \eg escape sequence. This escape must be followed by an
1985 unsigned number or a negative number, optionally enclosed in braces. These
1986 examples are all identical:
1987 .sp
1988 (ring), \e1
1989 (ring), \eg1
1990 (ring), \eg{1}
1991 .sp
1992 An unsigned number specifies an absolute reference without the ambiguity that
1993 is present in the older syntax. It is also useful when literal digits follow
1994 the reference. A negative number is a relative reference. Consider this
1995 example:
1996 .sp
1997 (abc(def)ghi)\eg{-1}
1998 .sp
1999 The sequence \eg{-1} is a reference to the most recently started capturing
2000 subpattern before \eg, that is, is it equivalent to \e2 in this example.
2001 Similarly, \eg{-2} would be equivalent to \e1. The use of relative references
2002 can be helpful in long patterns, and also in patterns that are created by
2003 joining together fragments that contain references within themselves.
2004 .P
2005 A back reference matches whatever actually matched the capturing subpattern in
2006 the current subject string, rather than anything matching the subpattern
2007 itself (see
2008 .\" HTML <a href="#subpatternsassubroutines">
2009 .\" </a>
2010 "Subpatterns as subroutines"
2011 .\"
2012 below for a way of doing that). So the pattern
2013 .sp
2014 (sens|respons)e and \e1ibility
2015 .sp
2016 matches "sense and sensibility" and "response and responsibility", but not
2017 "sense and responsibility". If caseful matching is in force at the time of the
2018 back reference, the case of letters is relevant. For example,
2019 .sp
2020 ((?i)rah)\es+\e1
2021 .sp
2022 matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
2023 capturing subpattern is matched caselessly.
2024 .P
2025 There are several different ways of writing back references to named
2026 subpatterns. The .NET syntax \ek{name} and the Perl syntax \ek<name> or
2027 \ek'name' are supported, as is the Python syntax (?P=name). Perl 5.10's unified
2028 back reference syntax, in which \eg can be used for both numeric and named
2029 references, is also supported. We could rewrite the above example in any of
2030 the following ways:
2031 .sp
2032 (?<p1>(?i)rah)\es+\ek<p1>
2033 (?'p1'(?i)rah)\es+\ek{p1}
2034 (?P<p1>(?i)rah)\es+(?P=p1)
2035 (?<p1>(?i)rah)\es+\eg{p1}
2036 .sp
2037 A subpattern that is referenced by name may appear in the pattern before or
2038 after the reference.
2039 .P
2040 There may be more than one back reference to the same subpattern. If a
2041 subpattern has not actually been used in a particular match, any back
2042 references to it always fail by default. For example, the pattern
2043 .sp
2044 (a|(bc))\e2
2045 .sp
2046 always fails if it starts to match "a" rather than "bc". However, if the
2047 PCRE_JAVASCRIPT_COMPAT option is set at compile time, a back reference to an
2048 unset value matches an empty string.
2049 .P
2050 Because there may be many capturing parentheses in a pattern, all digits
2051 following a backslash are taken as part of a potential back reference number.
2052 If the pattern continues with a digit character, some delimiter must be used to
2053 terminate the back reference. If the PCRE_EXTENDED option is set, this can be
2054 white space. Otherwise, the \eg{ syntax or an empty comment (see
2055 .\" HTML <a href="#comments">
2056 .\" </a>
2057 "Comments"
2058 .\"
2059 below) can be used.
2060 .
2061 .SS "Recursive back references"
2062 .rs
2063 .sp
2064 A back reference that occurs inside the parentheses to which it refers fails
2065 when the subpattern is first used, so, for example, (a\e1) never matches.
2066 However, such references can be useful inside repeated subpatterns. For
2067 example, the pattern
2068 .sp
2069 (a|b\e1)+
2070 .sp
2071 matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
2072 the subpattern, the back reference matches the character string corresponding
2073 to the previous iteration. In order for this to work, the pattern must be such
2074 that the first iteration does not need to match the back reference. This can be
2075 done using alternation, as in the example above, or by a quantifier with a
2076 minimum of zero.
2077 .P
2078 Back references of this type cause the group that they reference to be treated
2079 as an
2080 .\" HTML <a href="#atomicgroup">
2081 .\" </a>
2082 atomic group.
2083 .\"
2084 Once the whole group has been matched, a subsequent matching failure cannot
2085 cause backtracking into the middle of the group.
2086 .
2087 .
2088 .\" HTML <a name="bigassertions"></a>
2089 .SH ASSERTIONS
2090 .rs
2091 .sp
2092 An assertion is a test on the characters following or preceding the current
2093 matching point that does not actually consume any characters. The simple
2094 assertions coded as \eb, \eB, \eA, \eG, \eZ, \ez, ^ and $ are described
2095 .\" HTML <a href="#smallassertions">
2096 .\" </a>
2097 above.
2098 .\"
2099 .P
2100 More complicated assertions are coded as subpatterns. There are two kinds:
2101 those that look ahead of the current position in the subject string, and those
2102 that look behind it. An assertion subpattern is matched in the normal way,
2103 except that it does not cause the current matching position to be changed.
2104 .P
2105 Assertion subpatterns are not capturing subpatterns. If such an assertion
2106 contains capturing subpatterns within it, these are counted for the purposes of
2107 numbering the capturing subpatterns in the whole pattern. However, substring
2108 capturing is carried out only for positive assertions. (Perl sometimes, but not
2109 always, does do capturing in negative assertions.)
2110 .P
2111 For compatibility with Perl, assertion subpatterns may be repeated; though
2112 it makes no sense to assert the same thing several times, the side effect of
2113 capturing parentheses may occasionally be useful. In practice, there only three
2114 cases:
2115 .sp
2116 (1) If the quantifier is {0}, the assertion is never obeyed during matching.
2117 However, it may contain internal capturing parenthesized groups that are called
2118 from elsewhere via the
2119 .\" HTML <a href="#subpatternsassubroutines">
2120 .\" </a>
2121 subroutine mechanism.
2122 .\"
2123 .sp
2124 (2) If quantifier is {0,n} where n is greater than zero, it is treated as if it
2125 were {0,1}. At run time, the rest of the pattern match is tried with and
2126 without the assertion, the order depending on the greediness of the quantifier.
2127 .sp
2128 (3) If the minimum repetition is greater than zero, the quantifier is ignored.
2129 The assertion is obeyed just once when encountered during matching.
2130 .
2131 .
2132 .SS "Lookahead assertions"
2133 .rs
2134 .sp
2135 Lookahead assertions start with (?= for positive assertions and (?! for
2136 negative assertions. For example,
2137 .sp
2138 \ew+(?=;)
2139 .sp
2140 matches a word followed by a semicolon, but does not include the semicolon in
2141 the match, and
2142 .sp
2143 foo(?!bar)
2144 .sp
2145 matches any occurrence of "foo" that is not followed by "bar". Note that the
2146 apparently similar pattern
2147 .sp
2148 (?!foo)bar
2149 .sp
2150 does not find an occurrence of "bar" that is preceded by something other than
2151 "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
2152 (?!foo) is always true when the next three characters are "bar". A
2153 lookbehind assertion is needed to achieve the other effect.
2154 .P
2155 If you want to force a matching failure at some point in a pattern, the most
2156 convenient way to do it is with (?!) because an empty string always matches, so
2157 an assertion that requires there not to be an empty string must always fail.
2158 The backtracking control verb (*FAIL) or (*F) is a synonym for (?!).
2159 .
2160 .
2161 .\" HTML <a name="lookbehind"></a>
2162 .SS "Lookbehind assertions"
2163 .rs
2164 .sp
2165 Lookbehind assertions start with (?<= for positive assertions and (?<! for
2166 negative assertions. For example,
2167 .sp
2168 (?<!foo)bar
2169 .sp
2170 does find an occurrence of "bar" that is not preceded by "foo". The contents of
2171 a lookbehind assertion are restricted such that all the strings it matches must
2172 have a fixed length. However, if there are several top-level alternatives, they
2173 do not all have to have the same fixed length. Thus
2174 .sp
2175 (?<=bullock|donkey)
2176 .sp
2177 is permitted, but
2178 .sp
2179 (?<!dogs?|cats?)
2180 .sp
2181 causes an error at compile time. Branches that match different length strings
2182 are permitted only at the top level of a lookbehind assertion. This is an
2183 extension compared with Perl, which requires all branches to match the same
2184 length of string. An assertion such as
2185 .sp
2186 (?<=ab(c|de))
2187 .sp
2188 is not permitted, because its single top-level branch can match two different
2189 lengths, but it is acceptable to PCRE if rewritten to use two top-level
2190 branches:
2191 .sp
2192 (?<=abc|abde)
2193 .sp
2194 In some cases, the escape sequence \eK
2195 .\" HTML <a href="#resetmatchstart">
2196 .\" </a>
2197 (see above)
2198 .\"
2199 can be used instead of a lookbehind assertion to get round the fixed-length
2200 restriction.
2201 .P
2202 The implementation of lookbehind assertions is, for each alternative, to
2203 temporarily move the current position back by the fixed length and then try to
2204 match. If there are insufficient characters before the current position, the
2205 assertion fails.
2206 .P
2207 In a UTF mode, PCRE does not allow the \eC escape (which matches a single data
2208 unit even in a UTF mode) to appear in lookbehind assertions, because it makes
2209 it impossible to calculate the length of the lookbehind. The \eX and \eR
2210 escapes, which can match different numbers of data units, are also not
2211 permitted.
2212 .P
2213 .\" HTML <a href="#subpatternsassubroutines">
2214 .\" </a>
2215 "Subroutine"
2216 .\"
2217 calls (see below) such as (?2) or (?&X) are permitted in lookbehinds, as long
2218 as the subpattern matches a fixed-length string.
2219 .\" HTML <a href="#recursion">
2220 .\" </a>
2221 Recursion,
2222 .\"
2223 however, is not supported.
2224 .P
2225 Possessive quantifiers can be used in conjunction with lookbehind assertions to
2226 specify efficient matching of fixed-length strings at the end of subject
2227 strings. Consider a simple pattern such as
2228 .sp
2229 abcd$
2230 .sp
2231 when applied to a long string that does not match. Because matching proceeds
2232 from left to right, PCRE will look for each "a" in the subject and then see if
2233 what follows matches the rest of the pattern. If the pattern is specified as
2234 .sp
2235 ^.*abcd$
2236 .sp
2237 the initial .* matches the entire string at first, but when this fails (because
2238 there is no following "a"), it backtracks to match all but the last character,
2239 then all but the last two characters, and so on. Once again the search for "a"
2240 covers the entire string, from right to left, so we are no better off. However,
2241 if the pattern is written as
2242 .sp
2243 ^.*+(?<=abcd)
2244 .sp
2245 there can be no backtracking for the .*+ item; it can match only the entire
2246 string. The subsequent lookbehind assertion does a single test on the last four
2247 characters. If it fails, the match fails immediately. For long strings, this
2248 approach makes a significant difference to the processing time.
2249 .
2250 .
2251 .SS "Using multiple assertions"
2252 .rs
2253 .sp
2254 Several assertions (of any sort) may occur in succession. For example,
2255 .sp
2256 (?<=\ed{3})(?<!999)foo
2257 .sp
2258 matches "foo" preceded by three digits that are not "999". Notice that each of
2259 the assertions is applied independently at the same point in the subject
2260 string. First there is a check that the previous three characters are all
2261 digits, and then there is a check that the same three characters are not "999".
2262 This pattern does \fInot\fP match "foo" preceded by six characters, the first
2263 of which are digits and the last three of which are not "999". For example, it
2264 doesn't match "123abcfoo". A pattern to do that is
2265 .sp
2266 (?<=\ed{3}...)(?<!999)foo
2267 .sp
2268 This time the first assertion looks at the preceding six characters, checking
2269 that the first three are digits, and then the second assertion checks that the
2270 preceding three characters are not "999".
2271 .P
2272 Assertions can be nested in any combination. For example,
2273 .sp
2274 (?<=(?<!foo)bar)baz
2275 .sp
2276 matches an occurrence of "baz" that is preceded by "bar" which in turn is not
2277 preceded by "foo", while
2278 .sp
2279 (?<=\ed{3}(?!999)...)foo
2280 .sp
2281 is another pattern that matches "foo" preceded by three digits and any three
2282 characters that are not "999".
2283 .
2284 .
2285 .\" HTML <a name="conditions"></a>
2286 .SH "CONDITIONAL SUBPATTERNS"
2287 .rs
2288 .sp
2289 It is possible to cause the matching process to obey a subpattern
2290 conditionally or to choose between two alternative subpatterns, depending on
2291 the result of an assertion, or whether a specific capturing subpattern has
2292 already been matched. The two possible forms of conditional subpattern are:
2293 .sp
2294 (?(condition)yes-pattern)
2295 (?(condition)yes-pattern|no-pattern)
2296 .sp
2297 If the condition is satisfied, the yes-pattern is used; otherwise the
2298 no-pattern (if present) is used. If there are more than two alternatives in the
2299 subpattern, a compile-time error occurs. Each of the two alternatives may
2300 itself contain nested subpatterns of any form, including conditional
2301 subpatterns; the restriction to two alternatives applies only at the level of
2302 the condition. This pattern fragment is an example where the alternatives are
2303 complex:
2304 .sp
2305 (?(1) (A|B|C) | (D | (?(2)E|F) | E) )
2306 .sp
2307 .P
2308 There are four kinds of condition: references to subpatterns, references to
2309 recursion, a pseudo-condition called DEFINE, and assertions.
2310 .
2311 .SS "Checking for a used subpattern by number"
2312 .rs
2313 .sp
2314 If the text between the parentheses consists of a sequence of digits, the
2315 condition is true if a capturing subpattern of that number has previously
2316 matched. If there is more than one capturing subpattern with the same number
2317 (see the earlier
2318 .\"
2319 .\" HTML <a href="#recursion">
2320 .\" </a>
2321 section about duplicate subpattern numbers),
2322 .\"
2323 the condition is true if any of them have matched. An alternative notation is
2324 to precede the digits with a plus or minus sign. In this case, the subpattern
2325 number is relative rather than absolute. The most recently opened parentheses
2326 can be referenced by (?(-1), the next most recent by (?(-2), and so on. Inside
2327 loops it can also make sense to refer to subsequent groups. The next
2328 parentheses to be opened can be referenced as (?(+1), and so on. (The value
2329 zero in any of these forms is not used; it provokes a compile-time error.)
2330 .P
2331 Consider the following pattern, which contains non-significant white space to
2332 make it more readable (assume the PCRE_EXTENDED option) and to divide it into
2333 three parts for ease of discussion:
2334 .sp
2335 ( \e( )? [^()]+ (?(1) \e) )
2336 .sp
2337 The first part matches an optional opening parenthesis, and if that
2338 character is present, sets it as the first captured substring. The second part
2339 matches one or more characters that are not parentheses. The third part is a
2340 conditional subpattern that tests whether or not the first set of parentheses
2341 matched. If they did, that is, if subject started with an opening parenthesis,
2342 the condition is true, and so the yes-pattern is executed and a closing
2343 parenthesis is required. Otherwise, since no-pattern is not present, the
2344 subpattern matches nothing. In other words, this pattern matches a sequence of
2345 non-parentheses, optionally enclosed in parentheses.
2346 .P
2347 If you were embedding this pattern in a larger one, you could use a relative
2348 reference:
2349 .sp
2350 ...other stuff... ( \e( )? [^()]+ (?(-1) \e) ) ...
2351 .sp
2352 This makes the fragment independent of the parentheses in the larger pattern.
2353 .
2354 .SS "Checking for a used subpattern by name"
2355 .rs
2356 .sp
2357 Perl uses the syntax (?(<name>)...) or (?('name')...) to test for a used
2358 subpattern by name. For compatibility with earlier versions of PCRE, which had
2359 this facility before Perl, the syntax (?(name)...) is also recognized.
2360 .P
2361 Rewriting the above example to use a named subpattern gives this:
2362 .sp
2363 (?<OPEN> \e( )? [^()]+ (?(<OPEN>) \e) )
2364 .sp
2365 If the name used in a condition of this kind is a duplicate, the test is
2366 applied to all subpatterns of the same name, and is true if any one of them has
2367 matched.
2368 .
2369 .SS "Checking for pattern recursion"
2370 .rs
2371 .sp
2372 If the condition is the string (R), and there is no subpattern with the name R,
2373 the condition is true if a recursive call to the whole pattern or any
2374 subpattern has been made. If digits or a name preceded by ampersand follow the
2375 letter R, for example:
2376 .sp
2377 (?(R3)...) or (?(R&name)...)
2378 .sp
2379 the condition is true if the most recent recursion is into a subpattern whose
2380 number or name is given. This condition does not check the entire recursion
2381 stack. If the name used in a condition of this kind is a duplicate, the test is
2382 applied to all subpatterns of the same name, and is true if any one of them is
2383 the most recent recursion.
2384 .P
2385 At "top level", all these recursion test conditions are false.
2386 .\" HTML <a href="#recursion">
2387 .\" </a>
2388 The syntax for recursive patterns
2389 .\"
2390 is described below.
2391 .
2392 .\" HTML <a name="subdefine"></a>
2393 .SS "Defining subpatterns for use by reference only"
2394 .rs
2395 .sp
2396 If the condition is the string (DEFINE), and there is no subpattern with the
2397 name DEFINE, the condition is always false. In this case, there may be only one
2398 alternative in the subpattern. It is always skipped if control reaches this
2399 point in the pattern; the idea of DEFINE is that it can be used to define
2400 subroutines that can be referenced from elsewhere. (The use of
2401 .\" HTML <a href="#subpatternsassubroutines">
2402 .\" </a>
2403 subroutines
2404 .\"
2405 is described below.) For example, a pattern to match an IPv4 address such as
2406 "192.168.23.245" could be written like this (ignore white space and line
2407 breaks):
2408 .sp
2409 (?(DEFINE) (?<byte> 2[0-4]\ed | 25[0-5] | 1\ed\ed | [1-9]?\ed) )
2410 \eb (?&byte) (\e.(?&byte)){3} \eb
2411 .sp
2412 The first part of the pattern is a DEFINE group inside which a another group
2413 named "byte" is defined. This matches an individual component of an IPv4
2414 address (a number less than 256). When matching takes place, this part of the
2415 pattern is skipped because DEFINE acts like a false condition. The rest of the
2416 pattern uses references to the named group to match the four dot-separated
2417 components of an IPv4 address, insisting on a word boundary at each end.
2418 .
2419 .SS "Assertion conditions"
2420 .rs
2421 .sp
2422 If the condition is not in any of the above formats, it must be an assertion.
2423 This may be a positive or negative lookahead or lookbehind assertion. Consider
2424 this pattern, again containing non-significant white space, and with the two
2425 alternatives on the second line:
2426 .sp
2427 (?(?=[^a-z]*[a-z])
2428 \ed{2}-[a-z]{3}-\ed{2} | \ed{2}-\ed{2}-\ed{2} )
2429 .sp
2430 The condition is a positive lookahead assertion that matches an optional
2431 sequence of non-letters followed by a letter. In other words, it tests for the
2432 presence of at least one letter in the subject. If a letter is found, the
2433 subject is matched against the first alternative; otherwise it is matched
2434 against the second. This pattern matches strings in one of the two forms
2435 dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
2436 .
2437 .
2438 .\" HTML <a name="comments"></a>
2439 .SH COMMENTS
2440 .rs
2441 .sp
2442 There are two ways of including comments in patterns that are processed by
2443 PCRE. In both cases, the start of the comment must not be in a character class,
2444 nor in the middle of any other sequence of related characters such as (?: or a
2445 subpattern name or number. The characters that make up a comment play no part
2446 in the pattern matching.
2447 .P
2448 The sequence (?# marks the start of a comment that continues up to the next
2449 closing parenthesis. Nested parentheses are not permitted. If the PCRE_EXTENDED
2450 option is set, an unescaped # character also introduces a comment, which in
2451 this case continues to immediately after the next newline character or
2452 character sequence in the pattern. Which characters are interpreted as newlines
2453 is controlled by the options passed to a compiling function or by a special
2454 sequence at the start of the pattern, as described in the section entitled
2455 .\" HTML <a href="#newlines">
2456 .\" </a>
2457 "Newline conventions"
2458 .\"
2459 above. Note that the end of this type of comment is a literal newline sequence
2460 in the pattern; escape sequences that happen to represent a newline do not
2461 count. For example, consider this pattern when PCRE_EXTENDED is set, and the
2462 default newline convention is in force:
2463 .sp
2464 abc #comment \en still comment
2465 .sp
2466 On encountering the # character, \fBpcre_compile()\fP skips along, looking for
2467 a newline in the pattern. The sequence \en is still literal at this stage, so
2468 it does not terminate the comment. Only an actual character with the code value
2469 0x0a (the default newline) does so.
2470 .
2471 .
2472 .\" HTML <a name="recursion"></a>
2473 .SH "RECURSIVE PATTERNS"
2474 .rs
2475 .sp
2476 Consider the problem of matching a string in parentheses, allowing for
2477 unlimited nested parentheses. Without the use of recursion, the best that can
2478 be done is to use a pattern that matches up to some fixed depth of nesting. It
2479 is not possible to handle an arbitrary nesting depth.
2480 .P
2481 For some time, Perl has provided a facility that allows regular expressions to
2482 recurse (amongst other things). It does this by interpolating Perl code in the
2483 expression at run time, and the code can refer to the expression itself. A Perl
2484 pattern using code interpolation to solve the parentheses problem can be
2485 created like this:
2486 .sp
2487 $re = qr{\e( (?: (?>[^()]+) | (?p{$re}) )* \e)}x;
2488 .sp
2489 The (?p{...}) item interpolates Perl code at run time, and in this case refers
2490 recursively to the pattern in which it appears.
2491 .P
2492 Obviously, PCRE cannot support the interpolation of Perl code. Instead, it
2493 supports special syntax for recursion of the entire pattern, and also for
2494 individual subpattern recursion. After its introduction in PCRE and Python,
2495 this kind of recursion was subsequently introduced into Perl at release 5.10.
2496 .P
2497 A special item that consists of (? followed by a number greater than zero and a
2498 closing parenthesis is a recursive subroutine call of the subpattern of the
2499 given number, provided that it occurs inside that subpattern. (If not, it is a
2500 .\" HTML <a href="#subpatternsassubroutines">
2501 .\" </a>
2502 non-recursive subroutine
2503 .\"
2504 call, which is described in the next section.) The special item (?R) or (?0) is
2505 a recursive call of the entire regular expression.
2506 .P
2507 This PCRE pattern solves the nested parentheses problem (assume the
2508 PCRE_EXTENDED option is set so that white space is ignored):
2509 .sp
2510 \e( ( [^()]++ | (?R) )* \e)
2511 .sp
2512 First it matches an opening parenthesis. Then it matches any number of
2513 substrings which can either be a sequence of non-parentheses, or a recursive
2514 match of the pattern itself (that is, a correctly parenthesized substring).
2515 Finally there is a closing parenthesis. Note the use of a possessive quantifier
2516 to avoid backtracking into sequences of non-parentheses.
2517 .P
2518 If this were part of a larger pattern, you would not want to recurse the entire
2519 pattern, so instead you could use this:
2520 .sp
2521 ( \e( ( [^()]++ | (?1) )* \e) )
2522 .sp
2523 We have put the pattern into parentheses, and caused the recursion to refer to
2524 them instead of the whole pattern.
2525 .P
2526 In a larger pattern, keeping track of parenthesis numbers can be tricky. This
2527 is made easier by the use of relative references. Instead of (?1) in the
2528 pattern above you can write (?-2) to refer to the second most recently opened
2529 parentheses preceding the recursion. In other words, a negative number counts
2530 capturing parentheses leftwards from the point at which it is encountered.
2531 .P
2532 It is also possible to refer to subsequently opened parentheses, by writing
2533 references such as (?+2). However, these cannot be recursive because the
2534 reference is not inside the parentheses that are referenced. They are always
2535 .\" HTML <a href="#subpatternsassubroutines">
2536 .\" </a>
2537 non-recursive subroutine
2538 .\"
2539 calls, as described in the next section.
2540 .P
2541 An alternative approach is to use named parentheses instead. The Perl syntax
2542 for this is (?&name); PCRE's earlier syntax (?P>name) is also supported. We
2543 could rewrite the above example as follows:
2544 .sp
2545 (?<pn> \e( ( [^()]++ | (?&pn) )* \e) )
2546 .sp
2547 If there is more than one subpattern with the same name, the earliest one is
2548 used.
2549 .P
2550 This particular example pattern that we have been looking at contains nested
2551 unlimited repeats, and so the use of a possessive quantifier for matching
2552 strings of non-parentheses is important when applying the pattern to strings
2553 that do not match. For example, when this pattern is applied to
2554 .sp
2555 (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
2556 .sp
2557 it yields "no match" quickly. However, if a possessive quantifier is not used,
2558 the match runs for a very long time indeed because there are so many different
2559 ways the + and * repeats can carve up the subject, and all have to be tested
2560 before failure can be reported.
2561 .P
2562 At the end of a match, the values of capturing parentheses are those from
2563 the outermost level. If you want to obtain intermediate values, a callout
2564 function can be used (see below and the
2565 .\" HREF
2566 \fBpcrecallout\fP
2567 .\"
2568 documentation). If the pattern above is matched against
2569 .sp
2570 (ab(cd)ef)
2571 .sp
2572 the value for the inner capturing parentheses (numbered 2) is "ef", which is
2573 the last value taken on at the top level. If a capturing subpattern is not
2574 matched at the top level, its final captured value is unset, even if it was
2575 (temporarily) set at a deeper level during the matching process.
2576 .P
2577 If there are more than 15 capturing parentheses in a pattern, PCRE has to
2578 obtain extra memory to store data during a recursion, which it does by using
2579 \fBpcre_malloc\fP, freeing it via \fBpcre_free\fP afterwards. If no memory can
2580 be obtained, the match fails with the PCRE_ERROR_NOMEMORY error.
2581 .P
2582 Do not confuse the (?R) item with the condition (R), which tests for recursion.
2583 Consider this pattern, which matches text in angle brackets, allowing for
2584 arbitrary nesting. Only digits are allowed in nested brackets (that is, when
2585 recursing), whereas any characters are permitted at the outer level.
2586 .sp
2587 < (?: (?(R) \ed++ | [^<>]*+) | (?R)) * >
2588 .sp
2589 In this pattern, (?(R) is the start of a conditional subpattern, with two
2590 different alternatives for the recursive and non-recursive cases. The (?R) item
2591 is the actual recursive call.
2592 .
2593 .
2594 .\" HTML <a name="recursiondifference"></a>
2595 .SS "Differences in recursion processing between PCRE and Perl"
2596 .rs
2597 .sp
2598 Recursion processing in PCRE differs from Perl in two important ways. In PCRE
2599 (like Python, but unlike Perl), a recursive subpattern call is always treated
2600 as an atomic group. That is, once it has matched some of the subject string, it
2601 is never re-entered, even if it contains untried alternatives and there is a
2602 subsequent matching failure. This can be illustrated by the following pattern,
2603 which purports to match a palindromic string that contains an odd number of
2604 characters (for example, "a", "aba", "abcba", "abcdcba"):
2605 .sp
2606 ^(.|(.)(?1)\e2)$
2607 .sp
2608 The idea is that it either matches a single character, or two identical
2609 characters surrounding a sub-palindrome. In Perl, this pattern works; in PCRE
2610 it does not if the pattern is longer than three characters. Consider the
2611 subject string "abcba":
2612 .P
2613 At the top level, the first character is matched, but as it is not at the end
2614 of the string, the first alternative fails; the second alternative is taken
2615 and the recursion kicks in. The recursive call to subpattern 1 successfully
2616 matches the next character ("b"). (Note that the beginning and end of line
2617 tests are not part of the recursion).
2618 .P
2619 Back at the top level, the next character ("c") is compared with what
2620 subpattern 2 matched, which was "a". This fails. Because the recursion is
2621 treated as an atomic group, there are now no backtracking points, and so the
2622 entire match fails. (Perl is able, at this point, to re-enter the recursion and
2623 try the second alternative.) However, if the pattern is written with the
2624 alternatives in the other order, things are different:
2625 .sp
2626 ^((.)(?1)\e2|.)$
2627 .sp
2628 This time, the recursing alternative is tried first, and continues to recurse
2629 until it runs out of characters, at which point the recursion fails. But this
2630 time we do have another alternative to try at the higher level. That is the big
2631 difference: in the previous case the remaining alternative is at a deeper
2632 recursion level, which PCRE cannot use.
2633 .P
2634 To change the pattern so that it matches all palindromic strings, not just
2635 those with an odd number of characters, it is tempting to change the pattern to
2636 this:
2637 .sp
2638 ^((.)(?1)\e2|.?)$
2639 .sp
2640 Again, this works in Perl, but not in PCRE, and for the same reason. When a
2641 deeper recursion has matched a single character, it cannot be entered again in
2642 order to match an empty string. The solution is to separate the two cases, and
2643 write out the odd and even cases as alternatives at the higher level:
2644 .sp
2645 ^(?:((.)(?1)\e2|)|((.)(?3)\e4|.))
2646 .sp
2647 If you want to match typical palindromic phrases, the pattern has to ignore all
2648 non-word characters, which can be done like this:
2649 .sp
2650 ^\eW*+(?:((.)\eW*+(?1)\eW*+\e2|)|((.)\eW*+(?3)\eW*+\e4|\eW*+.\eW*+))\eW*+$
2651 .sp
2652 If run with the PCRE_CASELESS option, this pattern matches phrases such as "A
2653 man, a plan, a canal: Panama!" and it works well in both PCRE and Perl. Note
2654 the use of the possessive quantifier *+ to avoid backtracking into sequences of
2655 non-word characters. Without this, PCRE takes a great deal longer (ten times or
2656 more) to match typical phrases, and Perl takes so long that you think it has
2657 gone into a loop.
2658 .P
2659 \fBWARNING\fP: The palindrome-matching patterns above work only if the subject
2660 string does not start with a palindrome that is shorter than the entire string.
2661 For example, although "abcba" is correctly matched, if the subject is "ababa",
2662 PCRE finds the palindrome "aba" at the start, then fails at top level because
2663 the end of the string does not follow. Once again, it cannot jump back into the
2664 recursion to try other alternatives, so the entire match fails.
2665 .P
2666 The second way in which PCRE and Perl differ in their recursion processing is
2667 in the handling of captured values. In Perl, when a subpattern is called
2668 recursively or as a subpattern (see the next section), it has no access to any
2669 values that were captured outside the recursion, whereas in PCRE these values
2670 can be referenced. Consider this pattern:
2671 .sp
2672 ^(.)(\e1|a(?2))
2673 .sp
2674 In PCRE, this pattern matches "bab". The first capturing parentheses match "b",
2675 then in the second group, when the back reference \e1 fails to match "b", the
2676 second alternative matches "a" and then recurses. In the recursion, \e1 does
2677 now match "b" and so the whole match succeeds. In Perl, the pattern fails to
2678 match because inside the recursive call \e1 cannot access the externally set
2679 value.
2680 .
2681 .
2682 .\" HTML <a name="subpatternsassubroutines"></a>
2683 .SH "SUBPATTERNS AS SUBROUTINES"
2684 .rs
2685 .sp
2686 If the syntax for a recursive subpattern call (either by number or by
2687 name) is used outside the parentheses to which it refers, it operates like a
2688 subroutine in a programming language. The called subpattern may be defined
2689 before or after the reference. A numbered reference can be absolute or
2690 relative, as in these examples:
2691 .sp
2692 (...(absolute)...)...(?2)...
2693 (...(relative)...)...(?-1)...
2694 (...(?+1)...(relative)...
2695 .sp
2696 An earlier example pointed out that the pattern
2697 .sp
2698 (sens|respons)e and \e1ibility
2699 .sp
2700 matches "sense and sensibility" and "response and responsibility", but not
2701 "sense and responsibility". If instead the pattern
2702 .sp
2703 (sens|respons)e and (?1)ibility
2704 .sp
2705 is used, it does match "sense and responsibility" as well as the other two
2706 strings. Another example is given in the discussion of DEFINE above.
2707 .P
2708 All subroutine calls, whether recursive or not, are always treated as atomic
2709 groups. That is, once a subroutine has matched some of the subject string, it
2710 is never re-entered, even if it contains untried alternatives and there is a
2711 subsequent matching failure. Any capturing parentheses that are set during the
2712 subroutine call revert to their previous values afterwards.
2713 .P
2714 Processing options such as case-independence are fixed when a subpattern is
2715 defined, so if it is used as a subroutine, such options cannot be changed for
2716 different calls. For example, consider this pattern:
2717 .sp
2718 (abc)(?i:(?-1))
2719 .sp
2720 It matches "abcabc". It does not match "abcABC" because the change of
2721 processing option does not affect the called subpattern.
2722 .
2723 .
2724 .\" HTML <a name="onigurumasubroutines"></a>
2725 .SH "ONIGURUMA SUBROUTINE SYNTAX"
2726 .rs
2727 .sp
2728 For compatibility with Oniguruma, the non-Perl syntax \eg followed by a name or
2729 a number enclosed either in angle brackets or single quotes, is an alternative
2730 syntax for referencing a subpattern as a subroutine, possibly recursively. Here
2731 are two of the examples used above, rewritten using this syntax:
2732 .sp
2733 (?<pn> \e( ( (?>[^()]+) | \eg<pn> )* \e) )
2734 (sens|respons)e and \eg'1'ibility
2735 .sp
2736 PCRE supports an extension to Oniguruma: if a number is preceded by a
2737 plus or a minus sign it is taken as a relative reference. For example:
2738 .sp
2739 (abc)(?i:\eg<-1>)
2740 .sp
2741 Note that \eg{...} (Perl syntax) and \eg<...> (Oniguruma syntax) are \fInot\fP
2742 synonymous. The former is a back reference; the latter is a subroutine call.
2743 .
2744 .
2745 .SH CALLOUTS
2746 .rs
2747 .sp
2748 Perl has a feature whereby using the sequence (?{...}) causes arbitrary Perl
2749 code to be obeyed in the middle of matching a regular expression. This makes it
2750 possible, amongst other things, to extract different substrings that match the
2751 same pair of parentheses when there is a repetition.
2752 .P
2753 PCRE provides a similar feature, but of course it cannot obey arbitrary Perl
2754 code. The feature is called "callout". The caller of PCRE provides an external
2755 function by putting its entry point in the global variable \fIpcre_callout\fP
2756 (8-bit library) or \fIpcre[16|32]_callout\fP (16-bit or 32-bit library).
2757 By default, this variable contains NULL, which disables all calling out.
2758 .P
2759 Within a regular expression, (?C) indicates the points at which the external
2760 function is to be called. If you want to identify different callout points, you
2761 can put a number less than 256 after the letter C. The default value is zero.
2762 For example, this pattern has two callout points:
2763 .sp
2764 (?C1)abc(?C2)def
2765 .sp
2766 If the PCRE_AUTO_CALLOUT flag is passed to a compiling function, callouts are
2767 automatically installed before each item in the pattern. They are all numbered
2768 255. If there is a conditional group in the pattern whose condition is an
2769 assertion, an additional callout is inserted just before the condition. An
2770 explicit callout may also be set at this position, as in this example:
2771 .sp
2772 (?(?C9)(?=a)abc|def)
2773 .sp
2774 Note that this applies only to assertion conditions, not to other types of
2775 condition.
2776 .P
2777 During matching, when PCRE reaches a callout point, the external function is
2778 called. It is provided with the number of the callout, the position in the
2779 pattern, and, optionally, one item of data originally supplied by the caller of
2780 the matching function. The callout function may cause matching to proceed, to
2781 backtrack, or to fail altogether.
2782 .P
2783 By default, PCRE implements a number of optimizations at compile time and
2784 matching time, and one side-effect is that sometimes callouts are skipped. If
2785 you need all possible callouts to happen, you need to set options that disable
2786 the relevant optimizations. More details, and a complete description of the
2787 interface to the callout function, are given in the
2788 .\" HREF
2789 \fBpcrecallout\fP
2790 .\"
2791 documentation.
2792 .
2793 .
2794 .\" HTML <a name="backtrackcontrol"></a>
2795 .SH "BACKTRACKING CONTROL"
2796 .rs
2797 .sp
2798 Perl 5.10 introduced a number of "Special Backtracking Control Verbs", which
2799 are still described in the Perl documentation as "experimental and subject to
2800 change or removal in a future version of Perl". It goes on to say: "Their usage
2801 in production code should be noted to avoid problems during upgrades." The same
2802 remarks apply to the PCRE features described in this section.
2803 .P
2804 The new verbs make use of what was previously invalid syntax: an opening
2805 parenthesis followed by an asterisk. They are generally of the form
2806 (*VERB) or (*VERB:NAME). Some may take either form, possibly behaving
2807 differently depending on whether or not a name is present. A name is any
2808 sequence of characters that does not include a closing parenthesis. The maximum
2809 length of name is 255 in the 8-bit library and 65535 in the 16-bit and 32-bit
2810 libraries. If the name is empty, that is, if the closing parenthesis
2811 immediately follows the colon, the effect is as if the colon were not there.
2812 Any number of these verbs may occur in a pattern.
2813 .P
2814 Since these verbs are specifically related to backtracking, most of them can be
2815 used only when the pattern is to be matched using one of the traditional
2816 matching functions, because these use a backtracking algorithm. With the
2817 exception of (*FAIL), which behaves like a failing negative assertion, the
2818 backtracking control verbs cause an error if encountered by a DFA matching
2819 function.
2820 .P
2821 The behaviour of these verbs in
2822 .\" HTML <a href="#btrepeat">
2823 .\" </a>
2824 repeated groups,
2825 .\"
2826 .\" HTML <a href="#btassert">
2827 .\" </a>
2828 assertions,
2829 .\"
2830 and in
2831 .\" HTML <a href="#btsub">
2832 .\" </a>
2833 subpatterns called as subroutines
2834 .\"
2835 (whether or not recursively) is documented below.
2836 .
2837 .
2838 .\" HTML <a name="nooptimize"></a>
2839 .SS "Optimizations that affect backtracking verbs"
2840 .rs
2841 .sp
2842 PCRE contains some optimizations that are used to speed up matching by running
2843 some checks at the start of each match attempt. For example, it may know the
2844 minimum length of matching subject, or that a particular character must be
2845 present. When one of these optimizations bypasses the running of a match, any
2846 included backtracking verbs will not, of course, be processed. You can suppress
2847 the start-of-match optimizations by setting the PCRE_NO_START_OPTIMIZE option
2848 when calling \fBpcre_compile()\fP or \fBpcre_exec()\fP, or by starting the
2849 pattern with (*NO_START_OPT). There is more discussion of this option in the
2850 section entitled
2851 .\" HTML <a href="pcreapi.html#execoptions">
2852 .\" </a>
2853 "Option bits for \fBpcre_exec()\fP"
2854 .\"
2855 in the
2856 .\" HREF
2857 \fBpcreapi\fP
2858 .\"
2859 documentation.
2860 .P
2861 Experiments with Perl suggest that it too has similar optimizations, sometimes
2862 leading to anomalous results.
2863 .
2864 .
2865 .SS "Verbs that act immediately"
2866 .rs
2867 .sp
2868 The following verbs act as soon as they are encountered. They may not be
2869 followed by a name.
2870 .sp
2871 (*ACCEPT)
2872 .sp
2873 This verb causes the match to end successfully, skipping the remainder of the
2874 pattern. However, when it is inside a subpattern that is called as a
2875 subroutine, only that subpattern is ended successfully. Matching then continues
2876 at the outer level. If (*ACCEPT) in triggered in a positive assertion, the
2877 assertion succeeds; in a negative assertion, the assertion fails.
2878 .P
2879 If (*ACCEPT) is inside capturing parentheses, the data so far is captured. For
2880 example:
2881 .sp
2882 A((?:A|B(*ACCEPT)|C)D)
2883 .sp
2884 This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is captured by
2885 the outer parentheses.
2886 .sp
2887 (*FAIL) or (*F)
2888 .sp
2889 This verb causes a matching failure, forcing backtracking to occur. It is
2890 equivalent to (?!) but easier to read. The Perl documentation notes that it is
2891 probably useful only when combined with (?{}) or (??{}). Those are, of course,
2892 Perl features that are not present in PCRE. The nearest equivalent is the
2893 callout feature, as for example in this pattern:
2894 .sp
2895 a+(?C)(*FAIL)
2896 .sp
2897 A match with the string "aaaa" always fails, but the callout is taken before
2898 each backtrack happens (in this example, 10 times).
2899 .
2900 .
2901 .SS "Recording which path was taken"
2902 .rs
2903 .sp
2904 There is one verb whose main purpose is to track how a match was arrived at,
2905 though it also has a secondary use in conjunction with advancing the match
2906 starting point (see (*SKIP) below).
2907 .sp
2908 (*MARK:NAME) or (*:NAME)
2909 .sp
2910 A name is always required with this verb. There may be as many instances of
2911 (*MARK) as you like in a pattern, and their names do not have to be unique.
2912 .P
2913 When a match succeeds, the name of the last-encountered (*MARK:NAME),
2914 (*PRUNE:NAME), or (*THEN:NAME) on the matching path is passed back to the
2915 caller as described in the section entitled
2916 .\" HTML <a href="pcreapi.html#extradata">
2917 .\" </a>
2918 "Extra data for \fBpcre_exec()\fP"
2919 .\"
2920 in the
2921 .\" HREF
2922 \fBpcreapi\fP
2923 .\"
2924 documentation. Here is an example of \fBpcretest\fP output, where the /K
2925 modifier requests the retrieval and outputting of (*MARK) data:
2926 .sp
2927 re> /X(*MARK:A)Y|X(*MARK:B)Z/K
2928 data> XY
2929 0: XY
2930 MK: A
2931 XZ
2932 0: XZ
2933 MK: B
2934 .sp
2935 The (*MARK) name is tagged with "MK:" in this output, and in this example it
2936 indicates which of the two alternatives matched. This is a more efficient way
2937 of obtaining this information than putting each alternative in its own
2938 capturing parentheses.
2939 .P
2940 If a verb with a name is encountered in a positive assertion that is true, the
2941 name is recorded and passed back if it is the last-encountered. This does not
2942 happen for negative assertions or failing positive assertions.
2943 .P
2944 After a partial match or a failed match, the last encountered name in the
2945 entire match process is returned. For example:
2946 .sp
2947 re> /X(*MARK:A)Y|X(*MARK:B)Z/K
2948 data> XP
2949 No match, mark = B
2950 .sp
2951 Note that in this unanchored example the mark is retained from the match
2952 attempt that started at the letter "X" in the subject. Subsequent match
2953 attempts starting at "P" and then with an empty string do not get as far as the
2954 (*MARK) item, but nevertheless do not reset it.
2955 .P
2956 If you are interested in (*MARK) values after failed matches, you should
2957 probably set the PCRE_NO_START_OPTIMIZE option
2958 .\" HTML <a href="#nooptimize">
2959 .\" </a>
2960 (see above)
2961 .\"
2962 to ensure that the match is always attempted.
2963 .
2964 .
2965 .SS "Verbs that act after backtracking"
2966 .rs
2967 .sp
2968 The following verbs do nothing when they are encountered. Matching continues
2969 with what follows, but if there is no subsequent match, causing a backtrack to
2970 the verb, a failure is forced. That is, backtracking cannot pass to the left of
2971 the verb. However, when one of these verbs appears inside an atomic group or an
2972 assertion that is true, its effect is confined to that group, because once the
2973 group has been matched, there is never any backtracking into it. In this
2974 situation, backtracking can "jump back" to the left of the entire atomic group
2975 or assertion. (Remember also, as stated above, that this localization also
2976 applies in subroutine calls.)
2977 .P
2978 These verbs differ in exactly what kind of failure occurs when backtracking
2979 reaches them. The behaviour described below is what happens when the verb is
2980 not in a subroutine or an assertion. Subsequent sections cover these special
2981 cases.
2982 .sp
2983 (*COMMIT)
2984 .sp
2985 This verb, which may not be followed by a name, causes the whole match to fail
2986 outright if there is a later matching failure that causes backtracking to reach
2987 it. Even if the pattern is unanchored, no further attempts to find a match by
2988 advancing the starting point take place. If (*COMMIT) is the only backtracking
2989 verb that is encountered, once it has been passed \fBpcre_exec()\fP is
2990 committed to finding a match at the current starting point, or not at all. For
2991 example:
2992 .sp
2993 a+(*COMMIT)b
2994 .sp
2995 This matches "xxaab" but not "aacaab". It can be thought of as a kind of
2996 dynamic anchor, or "I've started, so I must finish." The name of the most
2997 recently passed (*MARK) in the path is passed back when (*COMMIT) forces a
2998 match failure.
2999 .P
3000 If there is more than one backtracking verb in a pattern, a different one that
3001 follows (*COMMIT) may be triggered first, so merely passing (*COMMIT) during a
3002 match does not always guarantee that a match must be at this starting point.
3003 .P
3004 Note that (*COMMIT) at the start of a pattern is not the same as an anchor,
3005 unless PCRE's start-of-match optimizations are turned off, as shown in this
3006 \fBpcretest\fP example:
3007 .sp
3008 re> /(*COMMIT)abc/
3009 data> xyzabc
3010 0: abc
3011 xyzabc\eY
3012 No match
3013 .sp
3014 PCRE knows that any match must start with "a", so the optimization skips along
3015 the subject to "a" before running the first match attempt, which succeeds. When
3016 the optimization is disabled by the \eY escape in the second subject, the match
3017 starts at "x" and so the (*COMMIT) causes it to fail without trying any other
3018 starting points.
3019 .sp
3020 (*PRUNE) or (*PRUNE:NAME)
3021 .sp
3022 This verb causes the match to fail at the current starting position in the
3023 subject if there is a later matching failure that causes backtracking to reach
3024 it. If the pattern is unanchored, the normal "bumpalong" advance to the next
3025 starting character then happens. Backtracking can occur as usual to the left of
3026 (*PRUNE), before it is reached, or when matching to the right of (*PRUNE), but
3027 if there is no match to the right, backtracking cannot cross (*PRUNE). In
3028 simple cases, the use of (*PRUNE) is just an alternative to an atomic group or
3029 possessive quantifier, but there are some uses of (*PRUNE) that cannot be
3030 expressed in any other way. In an anchored pattern (*PRUNE) has the same effect
3031 as (*COMMIT).
3032 .P
3033 The behaviour of (*PRUNE:NAME) is the not the same as (*MARK:NAME)(*PRUNE).
3034 It is like (*MARK:NAME) in that the name is remembered for passing back to the
3035 caller. However, (*SKIP:NAME) searches only for names set with (*MARK).
3036 .sp
3037 (*SKIP)
3038 .sp
3039 This verb, when given without a name, is like (*PRUNE), except that if the
3040 pattern is unanchored, the "bumpalong" advance is not to the next character,
3041 but to the position in the subject where (*SKIP) was encountered. (*SKIP)
3042 signifies that whatever text was matched leading up to it cannot be part of a
3043 successful match. Consider:
3044 .sp
3045 a+(*SKIP)b
3046 .sp
3047 If the subject is "aaaac...", after the first match attempt fails (starting at
3048 the first character in the string), the starting point skips on to start the
3049 next attempt at "c". Note that a possessive quantifer does not have the same
3050 effect as this example; although it would suppress backtracking during the
3051 first match attempt, the second attempt would start at the second character
3052 instead of skipping on to "c".
3053 .sp
3054 (*SKIP:NAME)
3055 .sp
3056 When (*SKIP) has an associated name, its behaviour is modified. When it is
3057 triggered, the previous path through the pattern is searched for the most
3058 recent (*MARK) that has the same name. If one is found, the "bumpalong" advance
3059 is to the subject position that corresponds to that (*MARK) instead of to where
3060 (*SKIP) was encountered. If no (*MARK) with a matching name is found, the
3061 (*SKIP) is ignored.
3062 .P
3063 Note that (*SKIP:NAME) searches only for names set by (*MARK:NAME). It ignores
3064 names that are set by (*PRUNE:NAME) or (*THEN:NAME).
3065 .sp
3066 (*THEN) or (*THEN:NAME)
3067 .sp
3068 This verb causes a skip to the next innermost alternative when backtracking
3069 reaches it. That is, it cancels any further backtracking within the current
3070 alternative. Its name comes from the observation that it can be used for a
3071 pattern-based if-then-else block:
3072 .sp
3073 ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
3074 .sp
3075 If the COND1 pattern matches, FOO is tried (and possibly further items after
3076 the end of the group if FOO succeeds); on failure, the matcher skips to the
3077 second alternative and tries COND2, without backtracking into COND1. If that
3078 succeeds and BAR fails, COND3 is tried. If subsequently BAZ fails, there are no
3079 more alternatives, so there is a backtrack to whatever came before the entire
3080 group. If (*THEN) is not inside an alternation, it acts like (*PRUNE).
3081 .P
3082 The behaviour of (*THEN:NAME) is the not the same as (*MARK:NAME)(*THEN).
3083 It is like (*MARK:NAME) in that the name is remembered for passing back to the
3084 caller. However, (*SKIP:NAME) searches only for names set with (*MARK).
3085 .P
3086 A subpattern that does not contain a | character is just a part of the
3087 enclosing alternative; it is not a nested alternation with only one
3088 alternative. The effect of (*THEN) extends beyond such a subpattern to the
3089 enclosing alternative. Consider this pattern, where A, B, etc. are complex
3090 pattern fragments that do not contain any | characters at this level:
3091 .sp
3092 A (B(*THEN)C) | D
3093 .sp
3094 If A and B are matched, but there is a failure in C, matching does not
3095 backtrack into A; instead it moves to the next alternative, that is, D.
3096 However, if the subpattern containing (*THEN) is given an alternative, it
3097 behaves differently:
3098 .sp
3099 A (B(*THEN)C | (*FAIL)) | D
3100 .sp
3101 The effect of (*THEN) is now confined to the inner subpattern. After a failure
3102 in C, matching moves to (*FAIL), which causes the whole subpattern to fail
3103 because there are no more alternatives to try. In this case, matching does now
3104 backtrack into A.
3105 .P
3106 Note that a conditional subpattern is not considered as having two
3107 alternatives, because only one is ever used. In other words, the | character in
3108 a conditional subpattern has a different meaning. Ignoring white space,
3109 consider:
3110 .sp
3111 ^.*? (?(?=a) a | b(*THEN)c )
3112 .sp
3113 If the subject is "ba", this pattern does not match. Because .*? is ungreedy,
3114 it initially matches zero characters. The condition (?=a) then fails, the
3115 character "b" is matched, but "c" is not. At this point, matching does not
3116 backtrack to .*? as might perhaps be expected from the presence of the |
3117 character. The conditional subpattern is part of the single alternative that
3118 comprises the whole pattern, and so the match fails. (If there was a backtrack
3119 into .*?, allowing it to match "b", the match would succeed.)
3120 .P
3121 The verbs just described provide four different "strengths" of control when
3122 subsequent matching fails. (*THEN) is the weakest, carrying on the match at the
3123 next alternative. (*PRUNE) comes next, failing the match at the current
3124 starting position, but allowing an advance to the next character (for an
3125 unanchored pattern). (*SKIP) is similar, except that the advance may be more
3126 than one character. (*COMMIT) is the strongest, causing the entire match to
3127 fail.
3128 .
3129 .
3130 .SS "More than one backtracking verb"
3131 .rs
3132 .sp
3133 If more than one backtracking verb is present in a pattern, the one that is
3134 backtracked onto first acts. For example, consider this pattern, where A, B,
3135 etc. are complex pattern fragments:
3136 .sp
3137 (A(*COMMIT)B(*THEN)C|ABD)
3138 .sp
3139 If A matches but B fails, the backtrack to (*COMMIT) causes the entire match to
3140 fail. However, if A and B match, but C fails, the backtrack to (*THEN) causes
3141 the next alternative (ABD) to be tried. This behaviour is consistent, but is
3142 not always the same as Perl's. It means that if two or more backtracking verbs
3143 appear in succession, all the the last of them has no effect. Consider this
3144 example:
3145 .sp
3146 ...(*COMMIT)(*PRUNE)...
3147 .sp
3148 If there is a matching failure to the right, backtracking onto (*PRUNE) causes
3149 it to be triggered, and its action is taken. There can never be a backtrack
3150 onto (*COMMIT).
3151 .
3152 .
3153 .\" HTML <a name="btrepeat"></a>
3154 .SS "Backtracking verbs in repeated groups"
3155 .rs
3156 .sp
3157 PCRE differs from Perl in its handling of backtracking verbs in repeated
3158 groups. For example, consider:
3159 .sp
3160 /(a(*COMMIT)b)+ac/
3161 .sp
3162 If the subject is "abac", Perl matches, but PCRE fails because the (*COMMIT) in
3163 the second repeat of the group acts.
3164 .
3165 .
3166 .\" HTML <a name="btassert"></a>
3167 .SS "Backtracking verbs in assertions"
3168 .rs
3169 .sp
3170 (*FAIL) in an assertion has its normal effect: it forces an immediate backtrack.
3171 .P
3172 (*ACCEPT) in a positive assertion causes the assertion to succeed without any
3173 further processing. In a negative assertion, (*ACCEPT) causes the assertion to
3174 fail without any further processing.
3175 .P
3176 The other backtracking verbs are not treated specially if they appear in a
3177 positive assertion. In particular, (*THEN) skips to the next alternative in the
3178 innermost enclosing group that has alternations, whether or not this is within
3179 the assertion.
3180 .P
3181 Negative assertions are, however, different, in order to ensure that changing a
3182 positive assertion into a negative assertion changes its result. Backtracking
3183 into (*COMMIT), (*SKIP), or (*PRUNE) causes a negative assertion to be true,
3184 without considering any further alternative branches in the assertion.
3185 Backtracking into (*THEN) causes it to skip to the next enclosing alternative
3186 within the assertion (the normal behaviour), but if the assertion does not have
3187 such an alternative, (*THEN) behaves like (*PRUNE).
3188 .
3189 .
3190 .\" HTML <a name="btsub"></a>
3191 .SS "Backtracking verbs in subroutines"
3192 .rs
3193 .sp
3194 These behaviours occur whether or not the subpattern is called recursively.
3195 Perl's treatment of subroutines is different in some cases.
3196 .P
3197 (*FAIL) in a subpattern called as a subroutine has its normal effect: it forces
3198 an immediate backtrack.
3199 .P
3200 (*ACCEPT) in a subpattern called as a subroutine causes the subroutine match to
3201 succeed without any further processing. Matching then continues after the
3202 subroutine call.
3203 .P
3204 (*COMMIT), (*SKIP), and (*PRUNE) in a subpattern called as a subroutine cause
3205 the subroutine match to fail.
3206 .P
3207 (*THEN) skips to the next alternative in the innermost enclosing group within
3208 the subpattern that has alternatives. If there is no such group within the
3209 subpattern, (*THEN) causes the subroutine match to fail.
3210 .
3211 .
3212 .SH "SEE ALSO"
3213 .rs
3214 .sp
3215 \fBpcreapi\fP(3), \fBpcrecallout\fP(3), \fBpcrematching\fP(3),
3216 \fBpcresyntax\fP(3), \fBpcre\fP(3), \fBpcre16(3)\fP, \fBpcre32(3)\fP.
3217 .
3218 .
3219 .SH AUTHOR
3220 .rs
3221 .sp
3222 .nf
3223 Philip Hazel
3224 University Computing Service
3225 Cambridge CB2 3QH, England.
3226 .fi
3227 .
3228 .
3229 .SH REVISION
3230 .rs
3231 .sp
3232 .nf
3233 Last updated: 12 November 2013
3234 Copyright (c) 1997-2013 University of Cambridge.
3235 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5