1<html>
2<head>
3<title>pcre2pattern specification</title>
4</head>
5<body bgcolor="#FFFFFF" text="#00005A" link="#0066FF" alink="#3399FF" vlink="#2222BB">
6<h1>pcre2pattern man page</h1>
7<p>
8Return to the <a href="index.html">PCRE2 index page</a>.
9</p>
10<p>
11This page is part of the PCRE2 HTML documentation. It was generated
12automatically from the original man page. If there is any nonsense in it,
13please consult the man page, in case the conversion went wrong.
14<br>
15<ul>
16<li><a name="TOC1" href="#SEC1">PCRE2 REGULAR EXPRESSION DETAILS</a>
17<li><a name="TOC2" href="#SEC2">SPECIAL START-OF-PATTERN ITEMS</a>
18<li><a name="TOC3" href="#SEC3">EBCDIC CHARACTER CODES</a>
19<li><a name="TOC4" href="#SEC4">CHARACTERS AND METACHARACTERS</a>
20<li><a name="TOC5" href="#SEC5">BACKSLASH</a>
21<li><a name="TOC6" href="#SEC6">CIRCUMFLEX AND DOLLAR</a>
22<li><a name="TOC7" href="#SEC7">FULL STOP (PERIOD, DOT) AND \N</a>
23<li><a name="TOC8" href="#SEC8">MATCHING A SINGLE CODE UNIT</a>
24<li><a name="TOC9" href="#SEC9">SQUARE BRACKETS AND CHARACTER CLASSES</a>
25<li><a name="TOC10" href="#SEC10">POSIX CHARACTER CLASSES</a>
26<li><a name="TOC11" href="#SEC11">COMPATIBILITY FEATURE FOR WORD BOUNDARIES</a>
27<li><a name="TOC12" href="#SEC12">VERTICAL BAR</a>
28<li><a name="TOC13" href="#SEC13">INTERNAL OPTION SETTING</a>
29<li><a name="TOC14" href="#SEC14">GROUPS</a>
30<li><a name="TOC15" href="#SEC15">DUPLICATE GROUP NUMBERS</a>
31<li><a name="TOC16" href="#SEC16">NAMED CAPTURE GROUPS</a>
32<li><a name="TOC17" href="#SEC17">REPETITION</a>
33<li><a name="TOC18" href="#SEC18">ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS</a>
34<li><a name="TOC19" href="#SEC19">BACKREFERENCES</a>
35<li><a name="TOC20" href="#SEC20">ASSERTIONS</a>
36<li><a name="TOC21" href="#SEC21">NON-ATOMIC ASSERTIONS</a>
37<li><a name="TOC22" href="#SEC22">SCRIPT RUNS</a>
38<li><a name="TOC23" href="#SEC23">CONDITIONAL GROUPS</a>
39<li><a name="TOC24" href="#SEC24">COMMENTS</a>
40<li><a name="TOC25" href="#SEC25">RECURSIVE PATTERNS</a>
41<li><a name="TOC26" href="#SEC26">GROUPS AS SUBROUTINES</a>
42<li><a name="TOC27" href="#SEC27">ONIGURUMA SUBROUTINE SYNTAX</a>
43<li><a name="TOC28" href="#SEC28">CALLOUTS</a>
44<li><a name="TOC29" href="#SEC29">BACKTRACKING CONTROL</a>
45<li><a name="TOC30" href="#SEC30">SEE ALSO</a>
46<li><a name="TOC31" href="#SEC31">AUTHOR</a>
47<li><a name="TOC32" href="#SEC32">REVISION</a>
48</ul>
49<br><a name="SEC1" href="#TOC1">PCRE2 REGULAR EXPRESSION DETAILS</a><br>
50<P>
51The syntax and semantics of the regular expressions that are supported by PCRE2
52are described in detail below. There is a quick-reference syntax summary in the
53<a href="pcre2syntax.html"><b>pcre2syntax</b></a>
54page. PCRE2 tries to match Perl syntax and semantics as closely as it can.
55PCRE2 also supports some alternative regular expression syntax (which does not
56conflict with the Perl syntax) in order to provide some compatibility with
57regular expressions in Python, .NET, and Oniguruma.
58</P>
59<P>
60Perl's regular expressions are described in its own documentation, and regular
61expressions in general are covered in a number of books, some of which have
62copious examples. Jeffrey Friedl's "Mastering Regular Expressions", published
63by O'Reilly, covers regular expressions in great detail. This description of
64PCRE2's regular expressions is intended as reference material.
65</P>
66<P>
67This document discusses the regular expression patterns that are supported by
68PCRE2 when its main matching function, <b>pcre2_match()</b>, is used. PCRE2 also
69has an alternative matching function, <b>pcre2_dfa_match()</b>, which matches
70using a different algorithm that is not Perl-compatible. Some of the features
71discussed below are not available when DFA matching is used. The advantages and
72disadvantages of the alternative function, and how it differs from the normal
73function, are discussed in the
74<a href="pcre2matching.html"><b>pcre2matching</b></a>
75page.
76</P>
77<br><a name="SEC2" href="#TOC1">SPECIAL START-OF-PATTERN ITEMS</a><br>
78<P>
79A number of options that can be passed to <b>pcre2_compile()</b> can also be set
80by special items at the start of a pattern. These are not Perl-compatible, but
81are provided to make these options accessible to pattern writers who are not
82able to change the program that processes the pattern. Any number of these
83items may appear, but they must all be together right at the start of the
84pattern string, and the letters must be in upper case.
85</P>
86<br><b>
87UTF support
88</b><br>
89<P>
90In the 8-bit and 16-bit PCRE2 libraries, characters may be coded either as
91single code units, or as multiple UTF-8 or UTF-16 code units. UTF-32 can be
92specified for the 32-bit library, in which case it constrains the character
93values to valid Unicode code points. To process UTF strings, PCRE2 must be
94built to include Unicode support (which is the default). When using UTF strings
95you must either call the compiling function with one or both of the PCRE2_UTF
96or PCRE2_MATCH_INVALID_UTF options, or the pattern must start with the special
97sequence (*UTF), which is equivalent to setting the relevant PCRE2_UTF. How
98setting a UTF mode affects pattern matching is mentioned in several places
99below. There is also a summary of features in the
100<a href="pcre2unicode.html"><b>pcre2unicode</b></a>
101page.
102</P>
103<P>
104Some applications that allow their users to supply patterns may wish to
105restrict them to non-UTF data for security reasons. If the PCRE2_NEVER_UTF
106option is passed to <b>pcre2_compile()</b>, (*UTF) is not allowed, and its
107appearance in a pattern causes an error.
108</P>
109<br><b>
110Unicode property support
111</b><br>
112<P>
113Another special sequence that may appear at the start of a pattern is (*UCP).
114This has the same effect as setting the PCRE2_UCP option: it causes sequences
115such as \d and \w to use Unicode properties to determine character types,
116instead of recognizing only characters with codes less than 256 via a lookup
117table. If also causes upper/lower casing operations to use Unicode properties
118for characters with code points greater than 127, even when UTF is not set.
119</P>
120<P>
121Some applications that allow their users to supply patterns may wish to
122restrict them for security reasons. If the PCRE2_NEVER_UCP option is passed to
123<b>pcre2_compile()</b>, (*UCP) is not allowed, and its appearance in a pattern
124causes an error.
125</P>
126<br><b>
127Locking out empty string matching
128</b><br>
129<P>
130Starting a pattern with (*NOTEMPTY) or (*NOTEMPTY_ATSTART) has the same effect
131as passing the PCRE2_NOTEMPTY or PCRE2_NOTEMPTY_ATSTART option to whichever
132matching function is subsequently called to match the pattern. These options
133lock out the matching of empty strings, either entirely, or only at the start
134of the subject.
135</P>
136<br><b>
137Disabling auto-possessification
138</b><br>
139<P>
140If a pattern starts with (*NO_AUTO_POSSESS), it has the same effect as setting
141the PCRE2_NO_AUTO_POSSESS option. This stops PCRE2 from making quantifiers
142possessive when what follows cannot match the repeated item. For example, by
143default a+b is treated as a++b. For more details, see the
144<a href="pcre2api.html"><b>pcre2api</b></a>
145documentation.
146</P>
147<br><b>
148Disabling start-up optimizations
149</b><br>
150<P>
151If a pattern starts with (*NO_START_OPT), it has the same effect as setting the
152PCRE2_NO_START_OPTIMIZE option. This disables several optimizations for quickly
153reaching "no match" results. For more details, see the
154<a href="pcre2api.html"><b>pcre2api</b></a>
155documentation.
156</P>
157<br><b>
158Disabling automatic anchoring
159</b><br>
160<P>
161If a pattern starts with (*NO_DOTSTAR_ANCHOR), it has the same effect as
162setting the PCRE2_NO_DOTSTAR_ANCHOR option. This disables optimizations that
163apply to patterns whose top-level branches all start with .* (match any number
164of arbitrary characters). For more details, see the
165<a href="pcre2api.html"><b>pcre2api</b></a>
166documentation.
167</P>
168<br><b>
169Disabling JIT compilation
170</b><br>
171<P>
172If a pattern that starts with (*NO_JIT) is successfully compiled, an attempt by
173the application to apply the JIT optimization by calling
174<b>pcre2_jit_compile()</b> is ignored.
175</P>
176<br><b>
177Setting match resource limits
178</b><br>
179<P>
180The <b>pcre2_match()</b> function contains a counter that is incremented every
181time it goes round its main loop. The caller of <b>pcre2_match()</b> can set a
182limit on this counter, which therefore limits the amount of computing resource
183used for a match. The maximum depth of nested backtracking can also be limited;
184this indirectly restricts the amount of heap memory that is used, but there is
185also an explicit memory limit that can be set.
186</P>
187<P>
188These facilities are provided to catch runaway matches that are provoked by
189patterns with huge matching trees. A common example is a pattern with nested
190unlimited repeats applied to a long string that does not match. When one of
191these limits is reached, <b>pcre2_match()</b> gives an error return. The limits
192can also be set by items at the start of the pattern of the form
193<pre>
194  (*LIMIT_HEAP=d)
195  (*LIMIT_MATCH=d)
196  (*LIMIT_DEPTH=d)
197</pre>
198where d is any number of decimal digits. However, the value of the setting must
199be less than the value set (or defaulted) by the caller of <b>pcre2_match()</b>
200for it to have any effect. In other words, the pattern writer can lower the
201limits set by the programmer, but not raise them. If there is more than one
202setting of one of these limits, the lower value is used. The heap limit is
203specified in kibibytes (units of 1024 bytes).
204</P>
205<P>
206Prior to release 10.30, LIMIT_DEPTH was called LIMIT_RECURSION. This name is
207still recognized for backwards compatibility.
208</P>
209<P>
210The heap limit applies only when the <b>pcre2_match()</b> or
211<b>pcre2_dfa_match()</b> interpreters are used for matching. It does not apply
212to JIT. The match limit is used (but in a different way) when JIT is being
213used, or when <b>pcre2_dfa_match()</b> is called, to limit computing resource
214usage by those matching functions. The depth limit is ignored by JIT but is
215relevant for DFA matching, which uses function recursion for recursions within
216the pattern and for lookaround assertions and atomic groups. In this case, the
217depth limit controls the depth of such recursion.
218<a name="newlines"></a></P>
219<br><b>
220Newline conventions
221</b><br>
222<P>
223PCRE2 supports six different conventions for indicating line breaks in
224strings: a single CR (carriage return) character, a single LF (linefeed)
225character, the two-character sequence CRLF, any of the three preceding, any
226Unicode newline sequence, or the NUL character (binary zero). The
227<a href="pcre2api.html"><b>pcre2api</b></a>
228page has
229<a href="pcre2api.html#newlines">further discussion</a>
230about newlines, and shows how to set the newline convention when calling
231<b>pcre2_compile()</b>.
232</P>
233<P>
234It is also possible to specify a newline convention by starting a pattern
235string with one of the following sequences:
236<pre>
237  (*CR)        carriage return
238  (*LF)        linefeed
239  (*CRLF)      carriage return, followed by linefeed
240  (*ANYCRLF)   any of the three above
241  (*ANY)       all Unicode newline sequences
242  (*NUL)       the NUL character (binary zero)
243</pre>
244These override the default and the options given to the compiling function. For
245example, on a Unix system where LF is the default newline sequence, the pattern
246<pre>
247  (*CR)a.b
248</pre>
249changes the convention to CR. That pattern matches "a\nb" because LF is no
250longer a newline. If more than one of these settings is present, the last one
251is used.
252</P>
253<P>
254The newline convention affects where the circumflex and dollar assertions are
255true. It also affects the interpretation of the dot metacharacter when
256PCRE2_DOTALL is not set, and the behaviour of \N when not followed by an
257opening brace. However, it does not affect what the \R escape sequence
258matches. By default, this is any Unicode newline sequence, for Perl
259compatibility. However, this can be changed; see the next section and the
260description of \R in the section entitled
261<a href="#newlineseq">"Newline sequences"</a>
262below. A change of \R setting can be combined with a change of newline
263convention.
264</P>
265<br><b>
266Specifying what \R matches
267</b><br>
268<P>
269It is possible to restrict \R to match only CR, LF, or CRLF (instead of the
270complete set of Unicode line endings) by setting the option PCRE2_BSR_ANYCRLF
271at compile time. This effect can also be achieved by starting a pattern with
272(*BSR_ANYCRLF). For completeness, (*BSR_UNICODE) is also recognized,
273corresponding to PCRE2_BSR_UNICODE.
274</P>
275<br><a name="SEC3" href="#TOC1">EBCDIC CHARACTER CODES</a><br>
276<P>
277PCRE2 can be compiled to run in an environment that uses EBCDIC as its
278character code instead of ASCII or Unicode (typically a mainframe system). In
279the sections below, character code values are ASCII or Unicode; in an EBCDIC
280environment these characters may have different code values, and there are no
281code points greater than 255.
282</P>
283<br><a name="SEC4" href="#TOC1">CHARACTERS AND METACHARACTERS</a><br>
284<P>
285A regular expression is a pattern that is matched against a subject string from
286left to right. Most characters stand for themselves in a pattern, and match the
287corresponding characters in the subject. As a trivial example, the pattern
288<pre>
289  The quick brown fox
290</pre>
291matches a portion of a subject string that is identical to itself. When
292caseless matching is specified (the PCRE2_CASELESS option or (?i) within the
293pattern), letters are matched independently of case. Note that there are two
294ASCII characters, K and S, that, in addition to their lower case ASCII
295equivalents, are case-equivalent with Unicode U+212A (Kelvin sign) and U+017F
296(long S) respectively when either PCRE2_UTF or PCRE2_UCP is set.
297</P>
298<P>
299The power of regular expressions comes from the ability to include wild cards,
300character classes, alternatives, and repetitions in the pattern. These are
301encoded in the pattern by the use of <i>metacharacters</i>, which do not stand
302for themselves but instead are interpreted in some special way.
303</P>
304<P>
305There are two different sets of metacharacters: those that are recognized
306anywhere in the pattern except within square brackets, and those that are
307recognized within square brackets. Outside square brackets, the metacharacters
308are as follows:
309<pre>
310  \      general escape character with several uses
311  ^      assert start of string (or line, in multiline mode)
312  $      assert end of string (or line, in multiline mode)
313  .      match any character except newline (by default)
314  [      start character class definition
315  |      start of alternative branch
316  (      start group or control verb
317  )      end group or control verb
318  *      0 or more quantifier
319  +      1 or more quantifier; also "possessive quantifier"
320  ?      0 or 1 quantifier; also quantifier minimizer
321  {      start min/max quantifier
322</pre>
323Part of a pattern that is in square brackets is called a "character class". In
324a character class the only metacharacters are:
325<pre>
326  \      general escape character
327  ^      negate the class, but only if the first character
328  -      indicates character range
329  [      POSIX character class (if followed by POSIX syntax)
330  ]      terminates the character class
331</pre>
332If a pattern is compiled with the PCRE2_EXTENDED option, most white space in
333the pattern, other than in a character class, and characters between a #
334outside a character class and the next newline, inclusive, are ignored. An
335escaping backslash can be used to include a white space or a # character as
336part of the pattern. If the PCRE2_EXTENDED_MORE option is set, the same
337applies, but in addition unescaped space and horizontal tab characters are
338ignored inside a character class. Note: only these two characters are ignored,
339not the full set of pattern white space characters that are ignored outside a
340character class. Option settings can be changed within a pattern; see the
341section entitled
342<a href="#internaloptions">"Internal Option Setting"</a>
343below.
344</P>
345<P>
346The following sections describe the use of each of the metacharacters.
347</P>
348<br><a name="SEC5" href="#TOC1">BACKSLASH</a><br>
349<P>
350The backslash character has several uses. Firstly, if it is followed by a
351character that is not a digit or a letter, it takes away any special meaning
352that character may have. This use of backslash as an escape character applies
353both inside and outside character classes.
354</P>
355<P>
356For example, if you want to match a * character, you must write \* in the
357pattern. This escaping action applies whether or not the following character
358would otherwise be interpreted as a metacharacter, so it is always safe to
359precede a non-alphanumeric with backslash to specify that it stands for itself.
360In particular, if you want to match a backslash, you write \\.
361</P>
362<P>
363Only ASCII digits and letters have any special meaning after a backslash. All
364other characters (in particular, those whose code points are greater than 127)
365are treated as literals.
366</P>
367<P>
368If you want to treat all characters in a sequence as literals, you can do so by
369putting them between \Q and \E. This is different from Perl in that $ and @
370are handled as literals in \Q...\E sequences in PCRE2, whereas in Perl, $ and
371@ cause variable interpolation. Also, Perl does "double-quotish backslash
372interpolation" on any backslashes between \Q and \E which, its documentation
373says, "may lead to confusing results". PCRE2 treats a backslash between \Q and
374\E just like any other character. Note the following examples:
375<pre>
376  Pattern            PCRE2 matches   Perl matches
377
378  \Qabc$xyz\E        abc$xyz        abc followed by the contents of $xyz
379  \Qabc\$xyz\E       abc\$xyz       abc\$xyz
380  \Qabc\E\$\Qxyz\E   abc$xyz        abc$xyz
381  \QA\B\E            A\B            A\B
382  \Q\\E              \              \\E
383</pre>
384The \Q...\E sequence is recognized both inside and outside character classes.
385An isolated \E that is not preceded by \Q is ignored. If \Q is not followed
386by \E later in the pattern, the literal interpretation continues to the end of
387the pattern (that is, \E is assumed at the end). If the isolated \Q is inside
388a character class, this causes an error, because the character class is not
389terminated by a closing square bracket.
390<a name="digitsafterbackslash"></a></P>
391<br><b>
392Non-printing characters
393</b><br>
394<P>
395A second use of backslash provides a way of encoding non-printing characters
396in patterns in a visible manner. There is no restriction on the appearance of
397non-printing characters in a pattern, but when a pattern is being prepared by
398text editing, it is often easier to use one of the following escape sequences
399instead of the binary character it represents. In an ASCII or Unicode
400environment, these escapes are as follows:
401<pre>
402  \a          alarm, that is, the BEL character (hex 07)
403  \cx         "control-x", where x is any printable ASCII character
404  \e          escape (hex 1B)
405  \f          form feed (hex 0C)
406  \n          linefeed (hex 0A)
407  \r          carriage return (hex 0D) (but see below)
408  \t          tab (hex 09)
409  \0dd        character with octal code 0dd
410  \ddd        character with octal code ddd, or backreference
411  \o{ddd..}   character with octal code ddd..
412  \xhh        character with hex code hh
413  \x{hhh..}   character with hex code hhh..
414  \N{U+hhh..} character with Unicode hex code point hhh..
415</pre>
416By default, after \x that is not followed by {, from zero to two hexadecimal
417digits are read (letters can be in upper or lower case). Any number of
418hexadecimal digits may appear between \x{ and }. If a character other than a
419hexadecimal digit appears between \x{ and }, or if there is no terminating },
420an error occurs.
421</P>
422<P>
423Characters whose code points are less than 256 can be defined by either of the
424two syntaxes for \x or by an octal sequence. There is no difference in the way
425they are handled. For example, \xdc is exactly the same as \x{dc} or \334.
426However, using the braced versions does make such sequences easier to read.
427</P>
428<P>
429Support is available for some ECMAScript (aka JavaScript) escape sequences via
430two compile-time options. If PCRE2_ALT_BSUX is set, the sequence \x followed
431by { is not recognized. Only if \x is followed by two hexadecimal digits is it
432recognized as a character escape. Otherwise it is interpreted as a literal "x"
433character. In this mode, support for code points greater than 256 is provided
434by \u, which must be followed by four hexadecimal digits; otherwise it is
435interpreted as a literal "u" character.
436</P>
437<P>
438PCRE2_EXTRA_ALT_BSUX has the same effect as PCRE2_ALT_BSUX and, in addition,
439\u{hhh..} is recognized as the character specified by hexadecimal code point.
440There may be any number of hexadecimal digits. This syntax is from ECMAScript
4416.
442</P>
443<P>
444The \N{U+hhh..} escape sequence is recognized only when PCRE2 is operating in
445UTF mode. Perl also uses \N{name} to specify characters by Unicode name; PCRE2
446does not support this. Note that when \N is not followed by an opening brace
447(curly bracket) it has an entirely different meaning, matching any character
448that is not a newline.
449</P>
450<P>
451There are some legacy applications where the escape sequence \r is expected to
452match a newline. If the PCRE2_EXTRA_ESCAPED_CR_IS_LF option is set, \r in a
453pattern is converted to \n so that it matches a LF (linefeed) instead of a CR
454(carriage return) character.
455</P>
456<P>
457The precise effect of \cx on ASCII characters is as follows: if x is a lower
458case letter, it is converted to upper case. Then bit 6 of the character (hex
45940) is inverted. Thus \cA to \cZ become hex 01 to hex 1A (A is 41, Z is 5A),
460but \c{ becomes hex 3B ({ is 7B), and \c; becomes hex 7B (; is 3B). If the
461code unit following \c has a value less than 32 or greater than 126, a
462compile-time error occurs.
463</P>
464<P>
465When PCRE2 is compiled in EBCDIC mode, \N{U+hhh..} is not supported. \a, \e,
466\f, \n, \r, and \t generate the appropriate EBCDIC code values. The \c
467escape is processed as specified for Perl in the <b>perlebcdic</b> document. The
468only characters that are allowed after \c are A-Z, a-z, or one of @, [, \, ],
469^, _, or ?. Any other character provokes a compile-time error. The sequence
470\c@ encodes character code 0; after \c the letters (in either case) encode
471characters 1-26 (hex 01 to hex 1A); [, \, ], ^, and _ encode characters 27-31
472(hex 1B to hex 1F), and \c? becomes either 255 (hex FF) or 95 (hex 5F).
473</P>
474<P>
475Thus, apart from \c?, these escapes generate the same character code values as
476they do in an ASCII environment, though the meanings of the values mostly
477differ. For example, \cG always generates code value 7, which is BEL in ASCII
478but DEL in EBCDIC.
479</P>
480<P>
481The sequence \c? generates DEL (127, hex 7F) in an ASCII environment, but
482because 127 is not a control character in EBCDIC, Perl makes it generate the
483APC character. Unfortunately, there are several variants of EBCDIC. In most of
484them the APC character has the value 255 (hex FF), but in the one Perl calls
485POSIX-BC its value is 95 (hex 5F). If certain other characters have POSIX-BC
486values, PCRE2 makes \c? generate 95; otherwise it generates 255.
487</P>
488<P>
489After \0 up to two further octal digits are read. If there are fewer than two
490digits, just those that are present are used. Thus the sequence \0\x\015
491specifies two binary zeros followed by a CR character (code value 13). Make
492sure you supply two digits after the initial zero if the pattern character that
493follows is itself an octal digit.
494</P>
495<P>
496The escape \o must be followed by a sequence of octal digits, enclosed in
497braces. An error occurs if this is not the case. This escape is a recent
498addition to Perl; it provides way of specifying character code points as octal
499numbers greater than 0777, and it also allows octal numbers and backreferences
500to be unambiguously specified.
501</P>
502<P>
503For greater clarity and unambiguity, it is best to avoid following \ by a
504digit greater than zero. Instead, use \o{} or \x{} to specify numerical
505character code points, and \g{} to specify backreferences. The following
506paragraphs describe the old, ambiguous syntax.
507</P>
508<P>
509The handling of a backslash followed by a digit other than 0 is complicated,
510and Perl has changed over time, causing PCRE2 also to change.
511</P>
512<P>
513Outside a character class, PCRE2 reads the digit and any following digits as a
514decimal number. If the number is less than 10, begins with the digit 8 or 9, or
515if there are at least that many previous capture groups in the expression, the
516entire sequence is taken as a <i>backreference</i>. A description of how this
517works is given
518<a href="#backreferences">later,</a>
519following the discussion of
520<a href="#group">parenthesized groups.</a>
521Otherwise, up to three octal digits are read to form a character code.
522</P>
523<P>
524Inside a character class, PCRE2 handles \8 and \9 as the literal characters
525"8" and "9", and otherwise reads up to three octal digits following the
526backslash, using them to generate a data character. Any subsequent digits stand
527for themselves. For example, outside a character class:
528<pre>
529  \040   is another way of writing an ASCII space
530  \40    is the same, provided there are fewer than 40 previous capture groups
531  \7     is always a backreference
532  \11    might be a backreference, or another way of writing a tab
533  \011   is always a tab
534  \0113  is a tab followed by the character "3"
535  \113   might be a backreference, otherwise the character with octal code 113
536  \377   might be a backreference, otherwise the value 255 (decimal)
537  \81    is always a backreference .sp
538</pre>
539Note that octal values of 100 or greater that are specified using this syntax
540must not be introduced by a leading zero, because no more than three octal
541digits are ever read.
542</P>
543<br><b>
544Constraints on character values
545</b><br>
546<P>
547Characters that are specified using octal or hexadecimal numbers are
548limited to certain values, as follows:
549<pre>
550  8-bit non-UTF mode    no greater than 0xff
551  16-bit non-UTF mode   no greater than 0xffff
552  32-bit non-UTF mode   no greater than 0xffffffff
553  All UTF modes         no greater than 0x10ffff and a valid code point
554</pre>
555Invalid Unicode code points are all those in the range 0xd800 to 0xdfff (the
556so-called "surrogate" code points). The check for these can be disabled by the
557caller of <b>pcre2_compile()</b> by setting the option
558PCRE2_EXTRA_ALLOW_SURROGATE_ESCAPES. However, this is possible only in UTF-8
559and UTF-32 modes, because these values are not representable in UTF-16.
560</P>
561<br><b>
562Escape sequences in character classes
563</b><br>
564<P>
565All the sequences that define a single character value can be used both inside
566and outside character classes. In addition, inside a character class, \b is
567interpreted as the backspace character (hex 08).
568</P>
569<P>
570When not followed by an opening brace, \N is not allowed in a character class.
571\B, \R, and \X are not special inside a character class. Like other
572unrecognized alphabetic escape sequences, they cause an error. Outside a
573character class, these sequences have different meanings.
574</P>
575<br><b>
576Unsupported escape sequences
577</b><br>
578<P>
579In Perl, the sequences \F, \l, \L, \u, and \U are recognized by its string
580handler and used to modify the case of following characters. By default, PCRE2
581does not support these escape sequences in patterns. However, if either of the
582PCRE2_ALT_BSUX or PCRE2_EXTRA_ALT_BSUX options is set, \U matches a "U"
583character, and \u can be used to define a character by code point, as
584described above.
585</P>
586<br><b>
587Absolute and relative backreferences
588</b><br>
589<P>
590The sequence \g followed by a signed or unsigned number, optionally enclosed
591in braces, is an absolute or relative backreference. A named backreference
592can be coded as \g{name}. Backreferences are discussed
593<a href="#backreferences">later,</a>
594following the discussion of
595<a href="#group">parenthesized groups.</a>
596</P>
597<br><b>
598Absolute and relative subroutine calls
599</b><br>
600<P>
601For compatibility with Oniguruma, the non-Perl syntax \g followed by a name or
602a number enclosed either in angle brackets or single quotes, is an alternative
603syntax for referencing a capture group as a subroutine. Details are discussed
604<a href="#onigurumasubroutines">later.</a>
605Note that \g{...} (Perl syntax) and \g&#60;...&#62; (Oniguruma syntax) are <i>not</i>
606synonymous. The former is a backreference; the latter is a
607<a href="#groupsassubroutines">subroutine</a>
608call.
609<a name="genericchartypes"></a></P>
610<br><b>
611Generic character types
612</b><br>
613<P>
614Another use of backslash is for specifying generic character types:
615<pre>
616  \d     any decimal digit
617  \D     any character that is not a decimal digit
618  \h     any horizontal white space character
619  \H     any character that is not a horizontal white space character
620  \N     any character that is not a newline
621  \s     any white space character
622  \S     any character that is not a white space character
623  \v     any vertical white space character
624  \V     any character that is not a vertical white space character
625  \w     any "word" character
626  \W     any "non-word" character
627</pre>
628The \N escape sequence has the same meaning as
629<a href="#fullstopdot">the "." metacharacter</a>
630when PCRE2_DOTALL is not set, but setting PCRE2_DOTALL does not change the
631meaning of \N. Note that when \N is followed by an opening brace it has a
632different meaning. See the section entitled
633<a href="#digitsafterbackslash">"Non-printing characters"</a>
634above for details. Perl also uses \N{name} to specify characters by Unicode
635name; PCRE2 does not support this.
636</P>
637<P>
638Each pair of lower and upper case escape sequences partitions the complete set
639of characters into two disjoint sets. Any given character matches one, and only
640one, of each pair. The sequences can appear both inside and outside character
641classes. They each match one character of the appropriate type. If the current
642matching point is at the end of the subject string, all of them fail, because
643there is no character to match.
644</P>
645<P>
646The default \s characters are HT (9), LF (10), VT (11), FF (12), CR (13), and
647space (32), which are defined as white space in the "C" locale. This list may
648vary if locale-specific matching is taking place. For example, in some locales
649the "non-breaking space" character (\xA0) is recognized as white space, and in
650others the VT character is not.
651</P>
652<P>
653A "word" character is an underscore or any character that is a letter or digit.
654By default, the definition of letters and digits is controlled by PCRE2's
655low-valued character tables, and may vary if locale-specific matching is taking
656place (see
657<a href="pcre2api.html#localesupport">"Locale support"</a>
658in the
659<a href="pcre2api.html"><b>pcre2api</b></a>
660page). For example, in a French locale such as "fr_FR" in Unix-like systems,
661or "french" in Windows, some character codes greater than 127 are used for
662accented letters, and these are then matched by \w. The use of locales with
663Unicode is discouraged.
664</P>
665<P>
666By default, characters whose code points are greater than 127 never match \d,
667\s, or \w, and always match \D, \S, and \W, although this may be different
668for characters in the range 128-255 when locale-specific matching is happening.
669These escape sequences retain their original meanings from before Unicode
670support was available, mainly for efficiency reasons. If the PCRE2_UCP option
671is set, the behaviour is changed so that Unicode properties are used to
672determine character types, as follows:
673<pre>
674  \d  any character that matches \p{Nd} (decimal digit)
675  \s  any character that matches \p{Z} or \h or \v
676  \w  any character that matches \p{L} or \p{N}, plus underscore
677</pre>
678The upper case escapes match the inverse sets of characters. Note that \d
679matches only decimal digits, whereas \w matches any Unicode digit, as well as
680any Unicode letter, and underscore. Note also that PCRE2_UCP affects \b, and
681\B because they are defined in terms of \w and \W. Matching these sequences
682is noticeably slower when PCRE2_UCP is set.
683</P>
684<P>
685The sequences \h, \H, \v, and \V, in contrast to the other sequences, which
686match only ASCII characters by default, always match a specific list of code
687points, whether or not PCRE2_UCP is set. The horizontal space characters are:
688<pre>
689  U+0009     Horizontal tab (HT)
690  U+0020     Space
691  U+00A0     Non-break space
692  U+1680     Ogham space mark
693  U+180E     Mongolian vowel separator
694  U+2000     En quad
695  U+2001     Em quad
696  U+2002     En space
697  U+2003     Em space
698  U+2004     Three-per-em space
699  U+2005     Four-per-em space
700  U+2006     Six-per-em space
701  U+2007     Figure space
702  U+2008     Punctuation space
703  U+2009     Thin space
704  U+200A     Hair space
705  U+202F     Narrow no-break space
706  U+205F     Medium mathematical space
707  U+3000     Ideographic space
708</pre>
709The vertical space characters are:
710<pre>
711  U+000A     Linefeed (LF)
712  U+000B     Vertical tab (VT)
713  U+000C     Form feed (FF)
714  U+000D     Carriage return (CR)
715  U+0085     Next line (NEL)
716  U+2028     Line separator
717  U+2029     Paragraph separator
718</pre>
719In 8-bit, non-UTF-8 mode, only the characters with code points less than 256
720are relevant.
721<a name="newlineseq"></a></P>
722<br><b>
723Newline sequences
724</b><br>
725<P>
726Outside a character class, by default, the escape sequence \R matches any
727Unicode newline sequence. In 8-bit non-UTF-8 mode \R is equivalent to the
728following:
729<pre>
730  (?&#62;\r\n|\n|\x0b|\f|\r|\x85)
731</pre>
732This is an example of an "atomic group", details of which are given
733<a href="#atomicgroup">below.</a>
734This particular group matches either the two-character sequence CR followed by
735LF, or one of the single characters LF (linefeed, U+000A), VT (vertical tab,
736U+000B), FF (form feed, U+000C), CR (carriage return, U+000D), or NEL (next
737line, U+0085). Because this is an atomic group, the two-character sequence is
738treated as a single unit that cannot be split.
739</P>
740<P>
741In other modes, two additional characters whose code points are greater than 255
742are added: LS (line separator, U+2028) and PS (paragraph separator, U+2029).
743Unicode support is not needed for these characters to be recognized.
744</P>
745<P>
746It is possible to restrict \R to match only CR, LF, or CRLF (instead of the
747complete set of Unicode line endings) by setting the option PCRE2_BSR_ANYCRLF
748at compile time. (BSR is an abbrevation for "backslash R".) This can be made
749the default when PCRE2 is built; if this is the case, the other behaviour can
750be requested via the PCRE2_BSR_UNICODE option. It is also possible to specify
751these settings by starting a pattern string with one of the following
752sequences:
753<pre>
754  (*BSR_ANYCRLF)   CR, LF, or CRLF only
755  (*BSR_UNICODE)   any Unicode newline sequence
756</pre>
757These override the default and the options given to the compiling function.
758Note that these special settings, which are not Perl-compatible, are recognized
759only at the very start of a pattern, and that they must be in upper case. If
760more than one of them is present, the last one is used. They can be combined
761with a change of newline convention; for example, a pattern can start with:
762<pre>
763  (*ANY)(*BSR_ANYCRLF)
764</pre>
765They can also be combined with the (*UTF) or (*UCP) special sequences. Inside a
766character class, \R is treated as an unrecognized escape sequence, and causes
767an error.
768<a name="uniextseq"></a></P>
769<br><b>
770Unicode character properties
771</b><br>
772<P>
773When PCRE2 is built with Unicode support (the default), three additional escape
774sequences that match characters with specific properties are available. They
775can be used in any mode, though in 8-bit and 16-bit non-UTF modes these
776sequences are of course limited to testing characters whose code points are
777less than U+0100 and U+10000, respectively. In 32-bit non-UTF mode, code points
778greater than 0x10ffff (the Unicode limit) may be encountered. These are all
779treated as being in the Unknown script and with an unassigned type. The extra
780escape sequences are:
781<pre>
782  \p{<i>xx</i>}   a character with the <i>xx</i> property
783  \P{<i>xx</i>}   a character without the <i>xx</i> property
784  \X       a Unicode extended grapheme cluster
785</pre>
786The property names represented by <i>xx</i> above are case-sensitive. There is
787support for Unicode script names, Unicode general category properties, "Any",
788which matches any character (including newline), and some special PCRE2
789properties (described in the
790<a href="#extraprops">next section).</a>
791Other Perl properties such as "InMusicalSymbols" are not supported by PCRE2.
792Note that \P{Any} does not match any characters, so always causes a match
793failure.
794</P>
795<P>
796Sets of Unicode characters are defined as belonging to certain scripts. A
797character from one of these sets can be matched using a script name. For
798example:
799<pre>
800  \p{Greek}
801  \P{Han}
802</pre>
803Unassigned characters (and in non-UTF 32-bit mode, characters with code points
804greater than 0x10FFFF) are assigned the "Unknown" script. Others that are not
805part of an identified script are lumped together as "Common". The current list
806of scripts is:
807</P>
808<P>
809Adlam,
810Ahom,
811Anatolian_Hieroglyphs,
812Arabic,
813Armenian,
814Avestan,
815Balinese,
816Bamum,
817Bassa_Vah,
818Batak,
819Bengali,
820Bhaiksuki,
821Bopomofo,
822Brahmi,
823Braille,
824Buginese,
825Buhid,
826Canadian_Aboriginal,
827Carian,
828Caucasian_Albanian,
829Chakma,
830Cham,
831Cherokee,
832Chorasmian,
833Common,
834Coptic,
835Cuneiform,
836Cypriot,
837Cyrillic,
838Deseret,
839Devanagari,
840Dives_Akuru,
841Dogra,
842Duployan,
843Egyptian_Hieroglyphs,
844Elbasan,
845Elymaic,
846Ethiopic,
847Georgian,
848Glagolitic,
849Gothic,
850Grantha,
851Greek,
852Gujarati,
853Gunjala_Gondi,
854Gurmukhi,
855Han,
856Hangul,
857Hanifi_Rohingya,
858Hanunoo,
859Hatran,
860Hebrew,
861Hiragana,
862Imperial_Aramaic,
863Inherited,
864Inscriptional_Pahlavi,
865Inscriptional_Parthian,
866Javanese,
867Kaithi,
868Kannada,
869Katakana,
870Kayah_Li,
871Kharoshthi,
872Khitan_Small_Script,
873Khmer,
874Khojki,
875Khudawadi,
876Lao,
877Latin,
878Lepcha,
879Limbu,
880Linear_A,
881Linear_B,
882Lisu,
883Lycian,
884Lydian,
885Mahajani,
886Makasar,
887Malayalam,
888Mandaic,
889Manichaean,
890Marchen,
891Masaram_Gondi,
892Medefaidrin,
893Meetei_Mayek,
894Mende_Kikakui,
895Meroitic_Cursive,
896Meroitic_Hieroglyphs,
897Miao,
898Modi,
899Mongolian,
900Mro,
901Multani,
902Myanmar,
903Nabataean,
904Nandinagari,
905New_Tai_Lue,
906Newa,
907Nko,
908Nushu,
909Nyakeng_Puachue_Hmong,
910Ogham,
911Ol_Chiki,
912Old_Hungarian,
913Old_Italic,
914Old_North_Arabian,
915Old_Permic,
916Old_Persian,
917Old_Sogdian,
918Old_South_Arabian,
919Old_Turkic,
920Oriya,
921Osage,
922Osmanya,
923Pahawh_Hmong,
924Palmyrene,
925Pau_Cin_Hau,
926Phags_Pa,
927Phoenician,
928Psalter_Pahlavi,
929Rejang,
930Runic,
931Samaritan,
932Saurashtra,
933Sharada,
934Shavian,
935Siddham,
936SignWriting,
937Sinhala,
938Sogdian,
939Sora_Sompeng,
940Soyombo,
941Sundanese,
942Syloti_Nagri,
943Syriac,
944Tagalog,
945Tagbanwa,
946Tai_Le,
947Tai_Tham,
948Tai_Viet,
949Takri,
950Tamil,
951Tangut,
952Telugu,
953Thaana,
954Thai,
955Tibetan,
956Tifinagh,
957Tirhuta,
958Ugaritic,
959Unknown,
960Vai,
961Wancho,
962Warang_Citi,
963Yezidi,
964Yi,
965Zanabazar_Square.
966</P>
967<P>
968Each character has exactly one Unicode general category property, specified by
969a two-letter abbreviation. For compatibility with Perl, negation can be
970specified by including a circumflex between the opening brace and the property
971name. For example, \p{^Lu} is the same as \P{Lu}.
972</P>
973<P>
974If only one letter is specified with \p or \P, it includes all the general
975category properties that start with that letter. In this case, in the absence
976of negation, the curly brackets in the escape sequence are optional; these two
977examples have the same effect:
978<pre>
979  \p{L}
980  \pL
981</pre>
982The following general category property codes are supported:
983<pre>
984  C     Other
985  Cc    Control
986  Cf    Format
987  Cn    Unassigned
988  Co    Private use
989  Cs    Surrogate
990
991  L     Letter
992  Ll    Lower case letter
993  Lm    Modifier letter
994  Lo    Other letter
995  Lt    Title case letter
996  Lu    Upper case letter
997
998  M     Mark
999  Mc    Spacing mark
1000  Me    Enclosing mark
1001  Mn    Non-spacing mark
1002
1003  N     Number
1004  Nd    Decimal number
1005  Nl    Letter number
1006  No    Other number
1007
1008  P     Punctuation
1009  Pc    Connector punctuation
1010  Pd    Dash punctuation
1011  Pe    Close punctuation
1012  Pf    Final punctuation
1013  Pi    Initial punctuation
1014  Po    Other punctuation
1015  Ps    Open punctuation
1016
1017  S     Symbol
1018  Sc    Currency symbol
1019  Sk    Modifier symbol
1020  Sm    Mathematical symbol
1021  So    Other symbol
1022
1023  Z     Separator
1024  Zl    Line separator
1025  Zp    Paragraph separator
1026  Zs    Space separator
1027</pre>
1028The special property L& is also supported: it matches a character that has
1029the Lu, Ll, or Lt property, in other words, a letter that is not classified as
1030a modifier or "other".
1031</P>
1032<P>
1033The Cs (Surrogate) property applies only to characters whose code points are in
1034the range U+D800 to U+DFFF. These characters are no different to any other
1035character when PCRE2 is not in UTF mode (using the 16-bit or 32-bit library).
1036However, they are not valid in Unicode strings and so cannot be tested by PCRE2
1037in UTF mode, unless UTF validity checking has been turned off (see the
1038discussion of PCRE2_NO_UTF_CHECK in the
1039<a href="pcre2api.html"><b>pcre2api</b></a>
1040page).
1041</P>
1042<P>
1043The long synonyms for property names that Perl supports (such as \p{Letter})
1044are not supported by PCRE2, nor is it permitted to prefix any of these
1045properties with "Is".
1046</P>
1047<P>
1048No character that is in the Unicode table has the Cn (unassigned) property.
1049Instead, this property is assumed for any code point that is not in the
1050Unicode table.
1051</P>
1052<P>
1053Specifying caseless matching does not affect these escape sequences. For
1054example, \p{Lu} always matches only upper case letters. This is different from
1055the behaviour of current versions of Perl.
1056</P>
1057<P>
1058Matching characters by Unicode property is not fast, because PCRE2 has to do a
1059multistage table lookup in order to find a character's property. That is why
1060the traditional escape sequences such as \d and \w do not use Unicode
1061properties in PCRE2 by default, though you can make them do so by setting the
1062PCRE2_UCP option or by starting the pattern with (*UCP).
1063</P>
1064<br><b>
1065Extended grapheme clusters
1066</b><br>
1067<P>
1068The \X escape matches any number of Unicode characters that form an "extended
1069grapheme cluster", and treats the sequence as an atomic group
1070<a href="#atomicgroup">(see below).</a>
1071Unicode supports various kinds of composite character by giving each character
1072a grapheme breaking property, and having rules that use these properties to
1073define the boundaries of extended grapheme clusters. The rules are defined in
1074Unicode Standard Annex 29, "Unicode Text Segmentation". Unicode 11.0.0
1075abandoned the use of some previous properties that had been used for emojis.
1076Instead it introduced various emoji-specific properties. PCRE2 uses only the
1077Extended Pictographic property.
1078</P>
1079<P>
1080\X always matches at least one character. Then it decides whether to add
1081additional characters according to the following rules for ending a cluster:
1082</P>
1083<P>
10841. End at the end of the subject string.
1085</P>
1086<P>
10872. Do not end between CR and LF; otherwise end after any control character.
1088</P>
1089<P>
10903. Do not break Hangul (a Korean script) syllable sequences. Hangul characters
1091are of five types: L, V, T, LV, and LVT. An L character may be followed by an
1092L, V, LV, or LVT character; an LV or V character may be followed by a V or T
1093character; an LVT or T character may be follwed only by a T character.
1094</P>
1095<P>
10964. Do not end before extending characters or spacing marks or the "zero-width
1097joiner" character. Characters with the "mark" property always have the
1098"extend" grapheme breaking property.
1099</P>
1100<P>
11015. Do not end after prepend characters.
1102</P>
1103<P>
11046. Do not break within emoji modifier sequences or emoji zwj sequences. That
1105is, do not break between characters with the Extended_Pictographic property.
1106Extend and ZWJ characters are allowed between the characters.
1107</P>
1108<P>
11097. Do not break within emoji flag sequences. That is, do not break between
1110regional indicator (RI) characters if there are an odd number of RI characters
1111before the break point.
1112</P>
1113<P>
11148. Otherwise, end the cluster.
1115<a name="extraprops"></a></P>
1116<br><b>
1117PCRE2's additional properties
1118</b><br>
1119<P>
1120As well as the standard Unicode properties described above, PCRE2 supports four
1121more that make it possible to convert traditional escape sequences such as \w
1122and \s to use Unicode properties. PCRE2 uses these non-standard, non-Perl
1123properties internally when PCRE2_UCP is set. However, they may also be used
1124explicitly. These properties are:
1125<pre>
1126  Xan   Any alphanumeric character
1127  Xps   Any POSIX space character
1128  Xsp   Any Perl space character
1129  Xwd   Any Perl "word" character
1130</pre>
1131Xan matches characters that have either the L (letter) or the N (number)
1132property. Xps matches the characters tab, linefeed, vertical tab, form feed, or
1133carriage return, and any other character that has the Z (separator) property.
1134Xsp is the same as Xps; in PCRE1 it used to exclude vertical tab, for Perl
1135compatibility, but Perl changed. Xwd matches the same characters as Xan, plus
1136underscore.
1137</P>
1138<P>
1139There is another non-standard property, Xuc, which matches any character that
1140can be represented by a Universal Character Name in C++ and other programming
1141languages. These are the characters $, @, ` (grave accent), and all characters
1142with Unicode code points greater than or equal to U+00A0, except for the
1143surrogates U+D800 to U+DFFF. Note that most base (ASCII) characters are
1144excluded. (Universal Character Names are of the form \uHHHH or \UHHHHHHHH
1145where H is a hexadecimal digit. Note that the Xuc property does not match these
1146sequences but the characters that they represent.)
1147<a name="resetmatchstart"></a></P>
1148<br><b>
1149Resetting the match start
1150</b><br>
1151<P>
1152In normal use, the escape sequence \K causes any previously matched characters
1153not to be included in the final matched sequence that is returned. For example,
1154the pattern:
1155<pre>
1156  foo\Kbar
1157</pre>
1158matches "foobar", but reports that it has matched "bar". \K does not interact
1159with anchoring in any way. The pattern:
1160<pre>
1161  ^foo\Kbar
1162</pre>
1163matches only when the subject begins with "foobar" (in single line mode),
1164though it again reports the matched string as "bar". This feature is similar to
1165a lookbehind assertion
1166<a href="#lookbehind">(described below).</a>
1167However, in this case, the part of the subject before the real match does not
1168have to be of fixed length, as lookbehind assertions do. The use of \K does
1169not interfere with the setting of
1170<a href="#group">captured substrings.</a>
1171For example, when the pattern
1172<pre>
1173  (foo)\Kbar
1174</pre>
1175matches "foobar", the first substring is still set to "foo".
1176</P>
1177<P>
1178Perl used to document that the use of \K within lookaround assertions is "not
1179well defined", but from version 5.32.0 Perl does not support this usage at all.
1180In PCRE2, \K is acted upon when it occurs inside positive assertions, but is
1181ignored in negative assertions. Note that when a pattern such as (?=ab\K)
1182matches, the reported start of the match can be greater than the end of the
1183match. Using \K in a lookbehind assertion at the start of a pattern can also
1184lead to odd effects. For example, consider this pattern:
1185<pre>
1186  (?&#60;=\Kfoo)bar
1187</pre>
1188If the subject is "foobar", a call to <b>pcre2_match()</b> with a starting
1189offset of 3 succeeds and reports the matching string as "foobar", that is, the
1190start of the reported match is earlier than where the match started.
1191<a name="smallassertions"></a></P>
1192<br><b>
1193Simple assertions
1194</b><br>
1195<P>
1196The final use of backslash is for certain simple assertions. An assertion
1197specifies a condition that has to be met at a particular point in a match,
1198without consuming any characters from the subject string. The use of
1199groups for more complicated assertions is described
1200<a href="#bigassertions">below.</a>
1201The backslashed assertions are:
1202<pre>
1203  \b     matches at a word boundary
1204  \B     matches when not at a word boundary
1205  \A     matches at the start of the subject
1206  \Z     matches at the end of the subject
1207          also matches before a newline at the end of the subject
1208  \z     matches only at the end of the subject
1209  \G     matches at the first matching position in the subject
1210</pre>
1211Inside a character class, \b has a different meaning; it matches the backspace
1212character. If any other of these assertions appears in a character class, an
1213"invalid escape sequence" error is generated.
1214</P>
1215<P>
1216A word boundary is a position in the subject string where the current character
1217and the previous character do not both match \w or \W (i.e. one matches
1218\w and the other matches \W), or the start or end of the string if the
1219first or last character matches \w, respectively. When PCRE2 is built with
1220Unicode support, the meanings of \w and \W can be changed by setting the
1221PCRE2_UCP option. When this is done, it also affects \b and \B. Neither PCRE2
1222nor Perl has a separate "start of word" or "end of word" metasequence. However,
1223whatever follows \b normally determines which it is. For example, the fragment
1224\ba matches "a" at the start of a word.
1225</P>
1226<P>
1227The \A, \Z, and \z assertions differ from the traditional circumflex and
1228dollar (described in the next section) in that they only ever match at the very
1229start and end of the subject string, whatever options are set. Thus, they are
1230independent of multiline mode. These three assertions are not affected by the
1231PCRE2_NOTBOL or PCRE2_NOTEOL options, which affect only the behaviour of the
1232circumflex and dollar metacharacters. However, if the <i>startoffset</i>
1233argument of <b>pcre2_match()</b> is non-zero, indicating that matching is to
1234start at a point other than the beginning of the subject, \A can never match.
1235The difference between \Z and \z is that \Z matches before a newline at the
1236end of the string as well as at the very end, whereas \z matches only at the
1237end.
1238</P>
1239<P>
1240The \G assertion is true only when the current matching position is at the
1241start point of the matching process, as specified by the <i>startoffset</i>
1242argument of <b>pcre2_match()</b>. It differs from \A when the value of
1243<i>startoffset</i> is non-zero. By calling <b>pcre2_match()</b> multiple times
1244with appropriate arguments, you can mimic Perl's /g option, and it is in this
1245kind of implementation where \G can be useful.
1246</P>
1247<P>
1248Note, however, that PCRE2's implementation of \G, being true at the starting
1249character of the matching process, is subtly different from Perl's, which
1250defines it as true at the end of the previous match. In Perl, these can be
1251different when the previously matched string was empty. Because PCRE2 does just
1252one match at a time, it cannot reproduce this behaviour.
1253</P>
1254<P>
1255If all the alternatives of a pattern begin with \G, the expression is anchored
1256to the starting match position, and the "anchored" flag is set in the compiled
1257regular expression.
1258</P>
1259<br><a name="SEC6" href="#TOC1">CIRCUMFLEX AND DOLLAR</a><br>
1260<P>
1261The circumflex and dollar metacharacters are zero-width assertions. That is,
1262they test for a particular condition being true without consuming any
1263characters from the subject string. These two metacharacters are concerned with
1264matching the starts and ends of lines. If the newline convention is set so that
1265only the two-character sequence CRLF is recognized as a newline, isolated CR
1266and LF characters are treated as ordinary data characters, and are not
1267recognized as newlines.
1268</P>
1269<P>
1270Outside a character class, in the default matching mode, the circumflex
1271character is an assertion that is true only if the current matching point is at
1272the start of the subject string. If the <i>startoffset</i> argument of
1273<b>pcre2_match()</b> is non-zero, or if PCRE2_NOTBOL is set, circumflex can
1274never match if the PCRE2_MULTILINE option is unset. Inside a character class,
1275circumflex has an entirely different meaning
1276<a href="#characterclass">(see below).</a>
1277</P>
1278<P>
1279Circumflex need not be the first character of the pattern if a number of
1280alternatives are involved, but it should be the first thing in each alternative
1281in which it appears if the pattern is ever to match that branch. If all
1282possible alternatives start with a circumflex, that is, if the pattern is
1283constrained to match only at the start of the subject, it is said to be an
1284"anchored" pattern. (There are also other constructs that can cause a pattern
1285to be anchored.)
1286</P>
1287<P>
1288The dollar character is an assertion that is true only if the current matching
1289point is at the end of the subject string, or immediately before a newline at
1290the end of the string (by default), unless PCRE2_NOTEOL is set. Note, however,
1291that it does not actually match the newline. Dollar need not be the last
1292character of the pattern if a number of alternatives are involved, but it
1293should be the last item in any branch in which it appears. Dollar has no
1294special meaning in a character class.
1295</P>
1296<P>
1297The meaning of dollar can be changed so that it matches only at the very end of
1298the string, by setting the PCRE2_DOLLAR_ENDONLY option at compile time. This
1299does not affect the \Z assertion.
1300</P>
1301<P>
1302The meanings of the circumflex and dollar metacharacters are changed if the
1303PCRE2_MULTILINE option is set. When this is the case, a dollar character
1304matches before any newlines in the string, as well as at the very end, and a
1305circumflex matches immediately after internal newlines as well as at the start
1306of the subject string. It does not match after a newline that ends the string,
1307for compatibility with Perl. However, this can be changed by setting the
1308PCRE2_ALT_CIRCUMFLEX option.
1309</P>
1310<P>
1311For example, the pattern /^abc$/ matches the subject string "def\nabc" (where
1312\n represents a newline) in multiline mode, but not otherwise. Consequently,
1313patterns that are anchored in single line mode because all branches start with
1314^ are not anchored in multiline mode, and a match for circumflex is possible
1315when the <i>startoffset</i> argument of <b>pcre2_match()</b> is non-zero. The
1316PCRE2_DOLLAR_ENDONLY option is ignored if PCRE2_MULTILINE is set.
1317</P>
1318<P>
1319When the newline convention (see
1320<a href="#newlines">"Newline conventions"</a>
1321below) recognizes the two-character sequence CRLF as a newline, this is
1322preferred, even if the single characters CR and LF are also recognized as
1323newlines. For example, if the newline convention is "any", a multiline mode
1324circumflex matches before "xyz" in the string "abc\r\nxyz" rather than after
1325CR, even though CR on its own is a valid newline. (It also matches at the very
1326start of the string, of course.)
1327</P>
1328<P>
1329Note that the sequences \A, \Z, and \z can be used to match the start and
1330end of the subject in both modes, and if all branches of a pattern start with
1331\A it is always anchored, whether or not PCRE2_MULTILINE is set.
1332<a name="fullstopdot"></a></P>
1333<br><a name="SEC7" href="#TOC1">FULL STOP (PERIOD, DOT) AND \N</a><br>
1334<P>
1335Outside a character class, a dot in the pattern matches any one character in
1336the subject string except (by default) a character that signifies the end of a
1337line.
1338</P>
1339<P>
1340When a line ending is defined as a single character, dot never matches that
1341character; when the two-character sequence CRLF is used, dot does not match CR
1342if it is immediately followed by LF, but otherwise it matches all characters
1343(including isolated CRs and LFs). When any Unicode line endings are being
1344recognized, dot does not match CR or LF or any of the other line ending
1345characters.
1346</P>
1347<P>
1348The behaviour of dot with regard to newlines can be changed. If the
1349PCRE2_DOTALL option is set, a dot matches any one character, without exception.
1350If the two-character sequence CRLF is present in the subject string, it takes
1351two dots to match it.
1352</P>
1353<P>
1354The handling of dot is entirely independent of the handling of circumflex and
1355dollar, the only relationship being that they both involve newlines. Dot has no
1356special meaning in a character class.
1357</P>
1358<P>
1359The escape sequence \N when not followed by an opening brace behaves like a
1360dot, except that it is not affected by the PCRE2_DOTALL option. In other words,
1361it matches any character except one that signifies the end of a line.
1362</P>
1363<P>
1364When \N is followed by an opening brace it has a different meaning. See the
1365section entitled
1366<a href="digitsafterbackslash">"Non-printing characters"</a>
1367above for details. Perl also uses \N{name} to specify characters by Unicode
1368name; PCRE2 does not support this.
1369</P>
1370<br><a name="SEC8" href="#TOC1">MATCHING A SINGLE CODE UNIT</a><br>
1371<P>
1372Outside a character class, the escape sequence \C matches any one code unit,
1373whether or not a UTF mode is set. In the 8-bit library, one code unit is one
1374byte; in the 16-bit library it is a 16-bit unit; in the 32-bit library it is a
137532-bit unit. Unlike a dot, \C always matches line-ending characters. The
1376feature is provided in Perl in order to match individual bytes in UTF-8 mode,
1377but it is unclear how it can usefully be used.
1378</P>
1379<P>
1380Because \C breaks up characters into individual code units, matching one unit
1381with \C in UTF-8 or UTF-16 mode means that the rest of the string may start
1382with a malformed UTF character. This has undefined results, because PCRE2
1383assumes that it is matching character by character in a valid UTF string (by
1384default it checks the subject string's validity at the start of processing
1385unless the PCRE2_NO_UTF_CHECK or PCRE2_MATCH_INVALID_UTF option is used).
1386</P>
1387<P>
1388An application can lock out the use of \C by setting the
1389PCRE2_NEVER_BACKSLASH_C option when compiling a pattern. It is also possible to
1390build PCRE2 with the use of \C permanently disabled.
1391</P>
1392<P>
1393PCRE2 does not allow \C to appear in lookbehind assertions
1394<a href="#lookbehind">(described below)</a>
1395in UTF-8 or UTF-16 modes, because this would make it impossible to calculate
1396the length of the lookbehind. Neither the alternative matching function
1397<b>pcre2_dfa_match()</b> nor the JIT optimizer support \C in these UTF modes.
1398The former gives a match-time error; the latter fails to optimize and so the
1399match is always run using the interpreter.
1400</P>
1401<P>
1402In the 32-bit library, however, \C is always supported (when not explicitly
1403locked out) because it always matches a single code unit, whether or not UTF-32
1404is specified.
1405</P>
1406<P>
1407In general, the \C escape sequence is best avoided. However, one way of using
1408it that avoids the problem of malformed UTF-8 or UTF-16 characters is to use a
1409lookahead to check the length of the next character, as in this pattern, which
1410could be used with a UTF-8 string (ignore white space and line breaks):
1411<pre>
1412  (?| (?=[\x00-\x7f])(\C) |
1413      (?=[\x80-\x{7ff}])(\C)(\C) |
1414      (?=[\x{800}-\x{ffff}])(\C)(\C)(\C) |
1415      (?=[\x{10000}-\x{1fffff}])(\C)(\C)(\C)(\C))
1416</pre>
1417In this example, a group that starts with (?| resets the capturing parentheses
1418numbers in each alternative (see
1419<a href="#dupgroupnumber">"Duplicate Group Numbers"</a>
1420below). The assertions at the start of each branch check the next UTF-8
1421character for values whose encoding uses 1, 2, 3, or 4 bytes, respectively. The
1422character's individual bytes are then captured by the appropriate number of
1423\C groups.
1424<a name="characterclass"></a></P>
1425<br><a name="SEC9" href="#TOC1">SQUARE BRACKETS AND CHARACTER CLASSES</a><br>
1426<P>
1427An opening square bracket introduces a character class, terminated by a closing
1428square bracket. A closing square bracket on its own is not special by default.
1429If a closing square bracket is required as a member of the class, it should be
1430the first data character in the class (after an initial circumflex, if present)
1431or escaped with a backslash. This means that, by default, an empty class cannot
1432be defined. However, if the PCRE2_ALLOW_EMPTY_CLASS option is set, a closing
1433square bracket at the start does end the (empty) class.
1434</P>
1435<P>
1436A character class matches a single character in the subject. A matched
1437character must be in the set of characters defined by the class, unless the
1438first character in the class definition is a circumflex, in which case the
1439subject character must not be in the set defined by the class. If a circumflex
1440is actually required as a member of the class, ensure it is not the first
1441character, or escape it with a backslash.
1442</P>
1443<P>
1444For example, the character class [aeiou] matches any lower case vowel, while
1445[^aeiou] matches any character that is not a lower case vowel. Note that a
1446circumflex is just a convenient notation for specifying the characters that
1447are in the class by enumerating those that are not. A class that starts with a
1448circumflex is not an assertion; it still consumes a character from the subject
1449string, and therefore it fails if the current pointer is at the end of the
1450string.
1451</P>
1452<P>
1453Characters in a class may be specified by their code points using \o, \x, or
1454\N{U+hh..} in the usual way. When caseless matching is set, any letters in a
1455class represent both their upper case and lower case versions, so for example,
1456a caseless [aeiou] matches "A" as well as "a", and a caseless [^aeiou] does not
1457match "A", whereas a caseful version would. Note that there are two ASCII
1458characters, K and S, that, in addition to their lower case ASCII equivalents,
1459are case-equivalent with Unicode U+212A (Kelvin sign) and U+017F (long S)
1460respectively when either PCRE2_UTF or PCRE2_UCP is set.
1461</P>
1462<P>
1463Characters that might indicate line breaks are never treated in any special way
1464when matching character classes, whatever line-ending sequence is in use, and
1465whatever setting of the PCRE2_DOTALL and PCRE2_MULTILINE options is used. A
1466class such as [^a] always matches one of these characters.
1467</P>
1468<P>
1469The generic character type escape sequences \d, \D, \h, \H, \p, \P, \s,
1470\S, \v, \V, \w, and \W may appear in a character class, and add the
1471characters that they match to the class. For example, [\dABCDEF] matches any
1472hexadecimal digit. In UTF modes, the PCRE2_UCP option affects the meanings of
1473\d, \s, \w and their upper case partners, just as it does when they appear
1474outside a character class, as described in the section entitled
1475<a href="#genericchartypes">"Generic character types"</a>
1476above. The escape sequence \b has a different meaning inside a character
1477class; it matches the backspace character. The sequences \B, \R, and \X are
1478not special inside a character class. Like any other unrecognized escape
1479sequences, they cause an error. The same is true for \N when not followed by
1480an opening brace.
1481</P>
1482<P>
1483The minus (hyphen) character can be used to specify a range of characters in a
1484character class. For example, [d-m] matches any letter between d and m,
1485inclusive. If a minus character is required in a class, it must be escaped with
1486a backslash or appear in a position where it cannot be interpreted as
1487indicating a range, typically as the first or last character in the class,
1488or immediately after a range. For example, [b-d-z] matches letters in the range
1489b to d, a hyphen character, or z.
1490</P>
1491<P>
1492Perl treats a hyphen as a literal if it appears before or after a POSIX class
1493(see below) or before or after a character type escape such as as \d or \H.
1494However, unless the hyphen is the last character in the class, Perl outputs a
1495warning in its warning mode, as this is most likely a user error. As PCRE2 has
1496no facility for warning, an error is given in these cases.
1497</P>
1498<P>
1499It is not possible to have the literal character "]" as the end character of a
1500range. A pattern such as [W-]46] is interpreted as a class of two characters
1501("W" and "-") followed by a literal string "46]", so it would match "W46]" or
1502"-46]". However, if the "]" is escaped with a backslash it is interpreted as
1503the end of range, so [W-\]46] is interpreted as a class containing a range
1504followed by two other characters. The octal or hexadecimal representation of
1505"]" can also be used to end a range.
1506</P>
1507<P>
1508Ranges normally include all code points between the start and end characters,
1509inclusive. They can also be used for code points specified numerically, for
1510example [\000-\037]. Ranges can include any characters that are valid for the
1511current mode. In any UTF mode, the so-called "surrogate" characters (those
1512whose code points lie between 0xd800 and 0xdfff inclusive) may not be specified
1513explicitly by default (the PCRE2_EXTRA_ALLOW_SURROGATE_ESCAPES option disables
1514this check). However, ranges such as [\x{d7ff}-\x{e000}], which include the
1515surrogates, are always permitted.
1516</P>
1517<P>
1518There is a special case in EBCDIC environments for ranges whose end points are
1519both specified as literal letters in the same case. For compatibility with
1520Perl, EBCDIC code points within the range that are not letters are omitted. For
1521example, [h-k] matches only four characters, even though the codes for h and k
1522are 0x88 and 0x92, a range of 11 code points. However, if the range is
1523specified numerically, for example, [\x88-\x92] or [h-\x92], all code points
1524are included.
1525</P>
1526<P>
1527If a range that includes letters is used when caseless matching is set, it
1528matches the letters in either case. For example, [W-c] is equivalent to
1529[][\\^_`wxyzabc], matched caselessly, and in a non-UTF mode, if character
1530tables for a French locale are in use, [\xc8-\xcb] matches accented E
1531characters in both cases.
1532</P>
1533<P>
1534A circumflex can conveniently be used with the upper case character types to
1535specify a more restricted set of characters than the matching lower case type.
1536For example, the class [^\W_] matches any letter or digit, but not underscore,
1537whereas [\w] includes underscore. A positive character class should be read as
1538"something OR something OR ..." and a negative class as "NOT something AND NOT
1539something AND NOT ...".
1540</P>
1541<P>
1542The only metacharacters that are recognized in character classes are backslash,
1543hyphen (only where it can be interpreted as specifying a range), circumflex
1544(only at the start), opening square bracket (only when it can be interpreted as
1545introducing a POSIX class name, or for a special compatibility feature - see
1546the next two sections), and the terminating closing square bracket. However,
1547escaping other non-alphanumeric characters does no harm.
1548</P>
1549<br><a name="SEC10" href="#TOC1">POSIX CHARACTER CLASSES</a><br>
1550<P>
1551Perl supports the POSIX notation for character classes. This uses names
1552enclosed by [: and :] within the enclosing square brackets. PCRE2 also supports
1553this notation. For example,
1554<pre>
1555  [01[:alpha:]%]
1556</pre>
1557matches "0", "1", any alphabetic character, or "%". The supported class names
1558are:
1559<pre>
1560  alnum    letters and digits
1561  alpha    letters
1562  ascii    character codes 0 - 127
1563  blank    space or tab only
1564  cntrl    control characters
1565  digit    decimal digits (same as \d)
1566  graph    printing characters, excluding space
1567  lower    lower case letters
1568  print    printing characters, including space
1569  punct    printing characters, excluding letters and digits and space
1570  space    white space (the same as \s from PCRE2 8.34)
1571  upper    upper case letters
1572  word     "word" characters (same as \w)
1573  xdigit   hexadecimal digits
1574</pre>
1575The default "space" characters are HT (9), LF (10), VT (11), FF (12), CR (13),
1576and space (32). If locale-specific matching is taking place, the list of space
1577characters may be different; there may be fewer or more of them. "Space" and
1578\s match the same set of characters.
1579</P>
1580<P>
1581The name "word" is a Perl extension, and "blank" is a GNU extension from Perl
15825.8. Another Perl extension is negation, which is indicated by a ^ character
1583after the colon. For example,
1584<pre>
1585  [12[:^digit:]]
1586</pre>
1587matches "1", "2", or any non-digit. PCRE2 (and Perl) also recognize the POSIX
1588syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
1589supported, and an error is given if they are encountered.
1590</P>
1591<P>
1592By default, characters with values greater than 127 do not match any of the
1593POSIX character classes, although this may be different for characters in the
1594range 128-255 when locale-specific matching is happening. However, if the
1595PCRE2_UCP option is passed to <b>pcre2_compile()</b>, some of the classes are
1596changed so that Unicode character properties are used. This is achieved by
1597replacing certain POSIX classes with other sequences, as follows:
1598<pre>
1599  [:alnum:]  becomes  \p{Xan}
1600  [:alpha:]  becomes  \p{L}
1601  [:blank:]  becomes  \h
1602  [:cntrl:]  becomes  \p{Cc}
1603  [:digit:]  becomes  \p{Nd}
1604  [:lower:]  becomes  \p{Ll}
1605  [:space:]  becomes  \p{Xps}
1606  [:upper:]  becomes  \p{Lu}
1607  [:word:]   becomes  \p{Xwd}
1608</pre>
1609Negated versions, such as [:^alpha:] use \P instead of \p. Three other POSIX
1610classes are handled specially in UCP mode:
1611</P>
1612<P>
1613[:graph:]
1614This matches characters that have glyphs that mark the page when printed. In
1615Unicode property terms, it matches all characters with the L, M, N, P, S, or Cf
1616properties, except for:
1617<pre>
1618  U+061C           Arabic Letter Mark
1619  U+180E           Mongolian Vowel Separator
1620  U+2066 - U+2069  Various "isolate"s
1621
1622</PRE>
1623</P>
1624<P>
1625[:print:]
1626This matches the same characters as [:graph:] plus space characters that are
1627not controls, that is, characters with the Zs property.
1628</P>
1629<P>
1630[:punct:]
1631This matches all characters that have the Unicode P (punctuation) property,
1632plus those characters with code points less than 256 that have the S (Symbol)
1633property.
1634</P>
1635<P>
1636The other POSIX classes are unchanged, and match only characters with code
1637points less than 256.
1638</P>
1639<br><a name="SEC11" href="#TOC1">COMPATIBILITY FEATURE FOR WORD BOUNDARIES</a><br>
1640<P>
1641In the POSIX.2 compliant library that was included in 4.4BSD Unix, the ugly
1642syntax [[:&#60;:]] and [[:&#62;:]] is used for matching "start of word" and "end of
1643word". PCRE2 treats these items as follows:
1644<pre>
1645  [[:&#60;:]]  is converted to  \b(?=\w)
1646  [[:&#62;:]]  is converted to  \b(?&#60;=\w)
1647</pre>
1648Only these exact character sequences are recognized. A sequence such as
1649[a[:&#60;:]b] provokes error for an unrecognized POSIX class name. This support is
1650not compatible with Perl. It is provided to help migrations from other
1651environments, and is best not used in any new patterns. Note that \b matches
1652at the start and the end of a word (see
1653<a href="#smallassertions">"Simple assertions"</a>
1654above), and in a Perl-style pattern the preceding or following character
1655normally shows which is wanted, without the need for the assertions that are
1656used above in order to give exactly the POSIX behaviour.
1657</P>
1658<br><a name="SEC12" href="#TOC1">VERTICAL BAR</a><br>
1659<P>
1660Vertical bar characters are used to separate alternative patterns. For example,
1661the pattern
1662<pre>
1663  gilbert|sullivan
1664</pre>
1665matches either "gilbert" or "sullivan". Any number of alternatives may appear,
1666and an empty alternative is permitted (matching the empty string). The matching
1667process tries each alternative in turn, from left to right, and the first one
1668that succeeds is used. If the alternatives are within a group
1669<a href="#group">(defined below),</a>
1670"succeeds" means matching the rest of the main pattern as well as the
1671alternative in the group.
1672<a name="internaloptions"></a></P>
1673<br><a name="SEC13" href="#TOC1">INTERNAL OPTION SETTING</a><br>
1674<P>
1675The settings of the PCRE2_CASELESS, PCRE2_MULTILINE, PCRE2_DOTALL,
1676PCRE2_EXTENDED, PCRE2_EXTENDED_MORE, and PCRE2_NO_AUTO_CAPTURE options can be
1677changed from within the pattern by a sequence of letters enclosed between "(?"
1678and ")". These options are Perl-compatible, and are described in detail in the
1679<a href="pcre2api.html"><b>pcre2api</b></a>
1680documentation. The option letters are:
1681<pre>
1682  i  for PCRE2_CASELESS
1683  m  for PCRE2_MULTILINE
1684  n  for PCRE2_NO_AUTO_CAPTURE
1685  s  for PCRE2_DOTALL
1686  x  for PCRE2_EXTENDED
1687  xx for PCRE2_EXTENDED_MORE
1688</pre>
1689For example, (?im) sets caseless, multiline matching. It is also possible to
1690unset these options by preceding the relevant letters with a hyphen, for
1691example (?-im). The two "extended" options are not independent; unsetting either
1692one cancels the effects of both of them.
1693</P>
1694<P>
1695A combined setting and unsetting such as (?im-sx), which sets PCRE2_CASELESS
1696and PCRE2_MULTILINE while unsetting PCRE2_DOTALL and PCRE2_EXTENDED, is also
1697permitted. Only one hyphen may appear in the options string. If a letter
1698appears both before and after the hyphen, the option is unset. An empty options
1699setting "(?)" is allowed. Needless to say, it has no effect.
1700</P>
1701<P>
1702If the first character following (? is a circumflex, it causes all of the above
1703options to be unset. Thus, (?^) is equivalent to (?-imnsx). Letters may follow
1704the circumflex to cause some options to be re-instated, but a hyphen may not
1705appear.
1706</P>
1707<P>
1708The PCRE2-specific options PCRE2_DUPNAMES and PCRE2_UNGREEDY can be changed in
1709the same way as the Perl-compatible options by using the characters J and U
1710respectively. However, these are not unset by (?^).
1711</P>
1712<P>
1713When one of these option changes occurs at top level (that is, not inside
1714group parentheses), the change applies to the remainder of the pattern
1715that follows. An option change within a group (see below for a description
1716of groups) affects only that part of the group that follows it, so
1717<pre>
1718  (a(?i)b)c
1719</pre>
1720matches abc and aBc and no other strings (assuming PCRE2_CASELESS is not used).
1721By this means, options can be made to have different settings in different
1722parts of the pattern. Any changes made in one alternative do carry on
1723into subsequent branches within the same group. For example,
1724<pre>
1725  (a(?i)b|c)
1726</pre>
1727matches "ab", "aB", "c", and "C", even though when matching "C" the first
1728branch is abandoned before the option setting. This is because the effects of
1729option settings happen at compile time. There would be some very weird
1730behaviour otherwise.
1731</P>
1732<P>
1733As a convenient shorthand, if any option settings are required at the start of
1734a non-capturing group (see the next section), the option letters may
1735appear between the "?" and the ":". Thus the two patterns
1736<pre>
1737  (?i:saturday|sunday)
1738  (?:(?i)saturday|sunday)
1739</pre>
1740match exactly the same set of strings.
1741</P>
1742<P>
1743<b>Note:</b> There are other PCRE2-specific options, applying to the whole
1744pattern, which can be set by the application when the compiling function is
1745called. In addition, the pattern can contain special leading sequences such as
1746(*CRLF) to override what the application has set or what has been defaulted.
1747Details are given in the section entitled
1748<a href="#newlineseq">"Newline sequences"</a>
1749above. There are also the (*UTF) and (*UCP) leading sequences that can be used
1750to set UTF and Unicode property modes; they are equivalent to setting the
1751PCRE2_UTF and PCRE2_UCP options, respectively. However, the application can set
1752the PCRE2_NEVER_UTF and PCRE2_NEVER_UCP options, which lock out the use of the
1753(*UTF) and (*UCP) sequences.
1754<a name="group"></a></P>
1755<br><a name="SEC14" href="#TOC1">GROUPS</a><br>
1756<P>
1757Groups are delimited by parentheses (round brackets), which can be nested.
1758Turning part of a pattern into a group does two things:
1759<br>
1760<br>
17611. It localizes a set of alternatives. For example, the pattern
1762<pre>
1763  cat(aract|erpillar|)
1764</pre>
1765matches "cataract", "caterpillar", or "cat". Without the parentheses, it would
1766match "cataract", "erpillar" or an empty string.
1767<br>
1768<br>
17692. It creates a "capture group". This means that, when the whole pattern
1770matches, the portion of the subject string that matched the group is passed
1771back to the caller, separately from the portion that matched the whole pattern.
1772(This applies only to the traditional matching function; the DFA matching
1773function does not support capturing.)
1774</P>
1775<P>
1776Opening parentheses are counted from left to right (starting from 1) to obtain
1777numbers for capture groups. For example, if the string "the red king" is
1778matched against the pattern
1779<pre>
1780  the ((red|white) (king|queen))
1781</pre>
1782the captured substrings are "red king", "red", and "king", and are numbered 1,
17832, and 3, respectively.
1784</P>
1785<P>
1786The fact that plain parentheses fulfil two functions is not always helpful.
1787There are often times when grouping is required without capturing. If an
1788opening parenthesis is followed by a question mark and a colon, the group
1789does not do any capturing, and is not counted when computing the number of any
1790subsequent capture groups. For example, if the string "the white queen"
1791is matched against the pattern
1792<pre>
1793  the ((?:red|white) (king|queen))
1794</pre>
1795the captured substrings are "white queen" and "queen", and are numbered 1 and
17962. The maximum number of capture groups is 65535.
1797</P>
1798<P>
1799As a convenient shorthand, if any option settings are required at the start of
1800a non-capturing group, the option letters may appear between the "?" and the
1801":". Thus the two patterns
1802<pre>
1803  (?i:saturday|sunday)
1804  (?:(?i)saturday|sunday)
1805</pre>
1806match exactly the same set of strings. Because alternative branches are tried
1807from left to right, and options are not reset until the end of the group is
1808reached, an option setting in one branch does affect subsequent branches, so
1809the above patterns match "SUNDAY" as well as "Saturday".
1810<a name="dupgroupnumber"></a></P>
1811<br><a name="SEC15" href="#TOC1">DUPLICATE GROUP NUMBERS</a><br>
1812<P>
1813Perl 5.10 introduced a feature whereby each alternative in a group uses the
1814same numbers for its capturing parentheses. Such a group starts with (?| and is
1815itself a non-capturing group. For example, consider this pattern:
1816<pre>
1817  (?|(Sat)ur|(Sun))day
1818</pre>
1819Because the two alternatives are inside a (?| group, both sets of capturing
1820parentheses are numbered one. Thus, when the pattern matches, you can look
1821at captured substring number one, whichever alternative matched. This construct
1822is useful when you want to capture part, but not all, of one of a number of
1823alternatives. Inside a (?| group, parentheses are numbered as usual, but the
1824number is reset at the start of each branch. The numbers of any capturing
1825parentheses that follow the whole group start after the highest number used in
1826any branch. The following example is taken from the Perl documentation. The
1827numbers underneath show in which buffer the captured content will be stored.
1828<pre>
1829  # before  ---------------branch-reset----------- after
1830  / ( a )  (?| x ( y ) z | (p (q) r) | (t) u (v) ) ( z ) /x
1831  # 1            2         2  3        2     3     4
1832</pre>
1833A backreference to a capture group uses the most recent value that is set for
1834the group. The following pattern matches "abcabc" or "defdef":
1835<pre>
1836  /(?|(abc)|(def))\1/
1837</pre>
1838In contrast, a subroutine call to a capture group always refers to the
1839first one in the pattern with the given number. The following pattern matches
1840"abcabc" or "defabc":
1841<pre>
1842  /(?|(abc)|(def))(?1)/
1843</pre>
1844A relative reference such as (?-1) is no different: it is just a convenient way
1845of computing an absolute group number.
1846</P>
1847<P>
1848If a
1849<a href="#conditions">condition test</a>
1850for a group's having matched refers to a non-unique number, the test is
1851true if any group with that number has matched.
1852</P>
1853<P>
1854An alternative approach to using this "branch reset" feature is to use
1855duplicate named groups, as described in the next section.
1856</P>
1857<br><a name="SEC16" href="#TOC1">NAMED CAPTURE GROUPS</a><br>
1858<P>
1859Identifying capture groups by number is simple, but it can be very hard to keep
1860track of the numbers in complicated patterns. Furthermore, if an expression is
1861modified, the numbers may change. To help with this difficulty, PCRE2 supports
1862the naming of capture groups. This feature was not added to Perl until release
18635.10. Python had the feature earlier, and PCRE1 introduced it at release 4.0,
1864using the Python syntax. PCRE2 supports both the Perl and the Python syntax.
1865</P>
1866<P>
1867In PCRE2, a capture group can be named in one of three ways: (?&#60;name&#62;...) or
1868(?'name'...) as in Perl, or (?P&#60;name&#62;...) as in Python. Names may be up to 32
1869code units long. When PCRE2_UTF is not set, they may contain only ASCII
1870alphanumeric characters and underscores, but must start with a non-digit. When
1871PCRE2_UTF is set, the syntax of group names is extended to allow any Unicode
1872letter or Unicode decimal digit. In other words, group names must match one of
1873these patterns:
1874<pre>
1875  ^[_A-Za-z][_A-Za-z0-9]*\z   when PCRE2_UTF is not set
1876  ^[_\p{L}][_\p{L}\p{Nd}]*\z  when PCRE2_UTF is set
1877</pre>
1878References to capture groups from other parts of the pattern, such as
1879<a href="#backreferences">backreferences,</a>
1880<a href="#recursion">recursion,</a>
1881and
1882<a href="#conditions">conditions,</a>
1883can all be made by name as well as by number.
1884</P>
1885<P>
1886Named capture groups are allocated numbers as well as names, exactly as
1887if the names were not present. In both PCRE2 and Perl, capture groups
1888are primarily identified by numbers; any names are just aliases for these
1889numbers. The PCRE2 API provides function calls for extracting the complete
1890name-to-number translation table from a compiled pattern, as well as
1891convenience functions for extracting captured substrings by name.
1892</P>
1893<P>
1894<b>Warning:</b> When more than one capture group has the same number, as
1895described in the previous section, a name given to one of them applies to all
1896of them. Perl allows identically numbered groups to have different names.
1897Consider this pattern, where there are two capture groups, both numbered 1:
1898<pre>
1899  (?|(?&#60;AA&#62;aa)|(?&#60;BB&#62;bb))
1900</pre>
1901Perl allows this, with both names AA and BB as aliases of group 1. Thus, after
1902a successful match, both names yield the same value (either "aa" or "bb").
1903</P>
1904<P>
1905In an attempt to reduce confusion, PCRE2 does not allow the same group number
1906to be associated with more than one name. The example above provokes a
1907compile-time error. However, there is still scope for confusion. Consider this
1908pattern:
1909<pre>
1910  (?|(?&#60;AA&#62;aa)|(bb))
1911</pre>
1912Although the second group number 1 is not explicitly named, the name AA is
1913still an alias for any group 1. Whether the pattern matches "aa" or "bb", a
1914reference by name to group AA yields the matched string.
1915</P>
1916<P>
1917By default, a name must be unique within a pattern, except that duplicate names
1918are permitted for groups with the same number, for example:
1919<pre>
1920  (?|(?&#60;AA&#62;aa)|(?&#60;AA&#62;bb))
1921</pre>
1922The duplicate name constraint can be disabled by setting the PCRE2_DUPNAMES
1923option at compile time, or by the use of (?J) within the pattern, as described
1924in the section entitled
1925<a href="#internaloptions">"Internal Option Setting"</a>
1926above.
1927</P>
1928<P>
1929Duplicate names can be useful for patterns where only one instance of the named
1930capture group can match. Suppose you want to match the name of a weekday,
1931either as a 3-letter abbreviation or as the full name, and in both cases you
1932want to extract the abbreviation. This pattern (ignoring the line breaks) does
1933the job:
1934<pre>
1935  (?J)
1936  (?&#60;DN&#62;Mon|Fri|Sun)(?:day)?|
1937  (?&#60;DN&#62;Tue)(?:sday)?|
1938  (?&#60;DN&#62;Wed)(?:nesday)?|
1939  (?&#60;DN&#62;Thu)(?:rsday)?|
1940  (?&#60;DN&#62;Sat)(?:urday)?
1941</pre>
1942There are five capture groups, but only one is ever set after a match. The
1943convenience functions for extracting the data by name returns the substring for
1944the first (and in this example, the only) group of that name that matched. This
1945saves searching to find which numbered group it was. (An alternative way of
1946solving this problem is to use a "branch reset" group, as described in the
1947previous section.)
1948</P>
1949<P>
1950If you make a backreference to a non-unique named group from elsewhere in the
1951pattern, the groups to which the name refers are checked in the order in which
1952they appear in the overall pattern. The first one that is set is used for the
1953reference. For example, this pattern matches both "foofoo" and "barbar" but not
1954"foobar" or "barfoo":
1955<pre>
1956  (?J)(?:(?&#60;n&#62;foo)|(?&#60;n&#62;bar))\k&#60;n&#62;
1957
1958</PRE>
1959</P>
1960<P>
1961If you make a subroutine call to a non-unique named group, the one that
1962corresponds to the first occurrence of the name is used. In the absence of
1963duplicate numbers this is the one with the lowest number.
1964</P>
1965<P>
1966If you use a named reference in a condition
1967test (see the
1968<a href="#conditions">section about conditions</a>
1969below), either to check whether a capture group has matched, or to check for
1970recursion, all groups with the same name are tested. If the condition is true
1971for any one of them, the overall condition is true. This is the same behaviour
1972as testing by number. For further details of the interfaces for handling named
1973capture groups, see the
1974<a href="pcre2api.html"><b>pcre2api</b></a>
1975documentation.
1976</P>
1977<br><a name="SEC17" href="#TOC1">REPETITION</a><br>
1978<P>
1979Repetition is specified by quantifiers, which can follow any of the following
1980items:
1981<pre>
1982  a literal data character
1983  the dot metacharacter
1984  the \C escape sequence
1985  the \R escape sequence
1986  the \X escape sequence
1987  an escape such as \d or \pL that matches a single character
1988  a character class
1989  a backreference
1990  a parenthesized group (including lookaround assertions)
1991  a subroutine call (recursive or otherwise)
1992</pre>
1993The general repetition quantifier specifies a minimum and maximum number of
1994permitted matches, by giving the two numbers in curly brackets (braces),
1995separated by a comma. The numbers must be less than 65536, and the first must
1996be less than or equal to the second. For example,
1997<pre>
1998  z{2,4}
1999</pre>
2000matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
2001character. If the second number is omitted, but the comma is present, there is
2002no upper limit; if the second number and the comma are both omitted, the
2003quantifier specifies an exact number of required matches. Thus
2004<pre>
2005  [aeiou]{3,}
2006</pre>
2007matches at least 3 successive vowels, but may match many more, whereas
2008<pre>
2009  \d{8}
2010</pre>
2011matches exactly 8 digits. An opening curly bracket that appears in a position
2012where a quantifier is not allowed, or one that does not match the syntax of a
2013quantifier, is taken as a literal character. For example, {,6} is not a
2014quantifier, but a literal string of four characters.
2015</P>
2016<P>
2017In UTF modes, quantifiers apply to characters rather than to individual code
2018units. Thus, for example, \x{100}{2} matches two characters, each of
2019which is represented by a two-byte sequence in a UTF-8 string. Similarly,
2020\X{3} matches three Unicode extended grapheme clusters, each of which may be
2021several code units long (and they may be of different lengths).
2022</P>
2023<P>
2024The quantifier {0} is permitted, causing the expression to behave as if the
2025previous item and the quantifier were not present. This may be useful for
2026capture groups that are referenced as
2027<a href="#groupsassubroutines">subroutines</a>
2028from elsewhere in the pattern (but see also the section entitled
2029<a href="#subdefine">"Defining capture groups for use by reference only"</a>
2030below). Except for parenthesized groups, items that have a {0} quantifier are
2031omitted from the compiled pattern.
2032</P>
2033<P>
2034For convenience, the three most common quantifiers have single-character
2035abbreviations:
2036<pre>
2037  *    is equivalent to {0,}
2038  +    is equivalent to {1,}
2039  ?    is equivalent to {0,1}
2040</pre>
2041It is possible to construct infinite loops by following a group that can match
2042no characters with a quantifier that has no upper limit, for example:
2043<pre>
2044  (a?)*
2045</pre>
2046Earlier versions of Perl and PCRE1 used to give an error at compile time for
2047such patterns. However, because there are cases where this can be useful, such
2048patterns are now accepted, but whenever an iteration of such a group matches no
2049characters, matching moves on to the next item in the pattern instead of
2050repeatedly matching an empty string. This does not prevent backtracking into
2051any of the iterations if a subsequent item fails to match.
2052</P>
2053<P>
2054By default, quantifiers are "greedy", that is, they match as much as possible
2055(up to the maximum number of permitted times), without causing the rest of the
2056pattern to fail. The classic example of where this gives problems is in trying
2057to match comments in C programs. These appear between /* and */ and within the
2058comment, individual * and / characters may appear. An attempt to match C
2059comments by applying the pattern
2060<pre>
2061  /\*.*\*/
2062</pre>
2063to the string
2064<pre>
2065  /* first comment */  not comment  /* second comment */
2066</pre>
2067fails, because it matches the entire string owing to the greediness of the .*
2068item. However, if a quantifier is followed by a question mark, it ceases to be
2069greedy, and instead matches the minimum number of times possible, so the
2070pattern
2071<pre>
2072  /\*.*?\*/
2073</pre>
2074does the right thing with the C comments. The meaning of the various
2075quantifiers is not otherwise changed, just the preferred number of matches.
2076Do not confuse this use of question mark with its use as a quantifier in its
2077own right. Because it has two uses, it can sometimes appear doubled, as in
2078<pre>
2079  \d??\d
2080</pre>
2081which matches one digit by preference, but can match two if that is the only
2082way the rest of the pattern matches.
2083</P>
2084<P>
2085If the PCRE2_UNGREEDY option is set (an option that is not available in Perl),
2086the quantifiers are not greedy by default, but individual ones can be made
2087greedy by following them with a question mark. In other words, it inverts the
2088default behaviour.
2089</P>
2090<P>
2091When a parenthesized group is quantified with a minimum repeat count that
2092is greater than 1 or with a limited maximum, more memory is required for the
2093compiled pattern, in proportion to the size of the minimum or maximum.
2094</P>
2095<P>
2096If a pattern starts with .* or .{0,} and the PCRE2_DOTALL option (equivalent
2097to Perl's /s) is set, thus allowing the dot to match newlines, the pattern is
2098implicitly anchored, because whatever follows will be tried against every
2099character position in the subject string, so there is no point in retrying the
2100overall match at any position after the first. PCRE2 normally treats such a
2101pattern as though it were preceded by \A.
2102</P>
2103<P>
2104In cases where it is known that the subject string contains no newlines, it is
2105worth setting PCRE2_DOTALL in order to obtain this optimization, or
2106alternatively, using ^ to indicate anchoring explicitly.
2107</P>
2108<P>
2109However, there are some cases where the optimization cannot be used. When .*
2110is inside capturing parentheses that are the subject of a backreference
2111elsewhere in the pattern, a match at the start may fail where a later one
2112succeeds. Consider, for example:
2113<pre>
2114  (.*)abc\1
2115</pre>
2116If the subject is "xyz123abc123" the match point is the fourth character. For
2117this reason, such a pattern is not implicitly anchored.
2118</P>
2119<P>
2120Another case where implicit anchoring is not applied is when the leading .* is
2121inside an atomic group. Once again, a match at the start may fail where a later
2122one succeeds. Consider this pattern:
2123<pre>
2124  (?&#62;.*?a)b
2125</pre>
2126It matches "ab" in the subject "aab". The use of the backtracking control verbs
2127(*PRUNE) and (*SKIP) also disable this optimization, and there is an option,
2128PCRE2_NO_DOTSTAR_ANCHOR, to do so explicitly.
2129</P>
2130<P>
2131When a capture group is repeated, the value captured is the substring that
2132matched the final iteration. For example, after
2133<pre>
2134  (tweedle[dume]{3}\s*)+
2135</pre>
2136has matched "tweedledum tweedledee" the value of the captured substring is
2137"tweedledee". However, if there are nested capture groups, the corresponding
2138captured values may have been set in previous iterations. For example, after
2139<pre>
2140  (a|(b))+
2141</pre>
2142matches "aba" the value of the second captured substring is "b".
2143<a name="atomicgroup"></a></P>
2144<br><a name="SEC18" href="#TOC1">ATOMIC GROUPING AND POSSESSIVE QUANTIFIERS</a><br>
2145<P>
2146With both maximizing ("greedy") and minimizing ("ungreedy" or "lazy")
2147repetition, failure of what follows normally causes the repeated item to be
2148re-evaluated to see if a different number of repeats allows the rest of the
2149pattern to match. Sometimes it is useful to prevent this, either to change the
2150nature of the match, or to cause it fail earlier than it otherwise might, when
2151the author of the pattern knows there is no point in carrying on.
2152</P>
2153<P>
2154Consider, for example, the pattern \d+foo when applied to the subject line
2155<pre>
2156  123456bar
2157</pre>
2158After matching all 6 digits and then failing to match "foo", the normal
2159action of the matcher is to try again with only 5 digits matching the \d+
2160item, and then with 4, and so on, before ultimately failing. "Atomic grouping"
2161(a term taken from Jeffrey Friedl's book) provides the means for specifying
2162that once a group has matched, it is not to be re-evaluated in this way.
2163</P>
2164<P>
2165If we use atomic grouping for the previous example, the matcher gives up
2166immediately on failing to match "foo" the first time. The notation is a kind of
2167special parenthesis, starting with (?&#62; as in this example:
2168<pre>
2169  (?&#62;\d+)foo
2170</pre>
2171Perl 5.28 introduced an experimental alphabetic form starting with (* which may
2172be easier to remember:
2173<pre>
2174  (*atomic:\d+)foo
2175</pre>
2176This kind of parenthesized group "locks up" the  part of the pattern it
2177contains once it has matched, and a failure further into the pattern is
2178prevented from backtracking into it. Backtracking past it to previous items,
2179however, works as normal.
2180</P>
2181<P>
2182An alternative description is that a group of this type matches exactly the
2183string of characters that an identical standalone pattern would match, if
2184anchored at the current point in the subject string.
2185</P>
2186<P>
2187Atomic groups are not capture groups. Simple cases such as the above example
2188can be thought of as a maximizing repeat that must swallow everything it can.
2189So, while both \d+ and \d+? are prepared to adjust the number of digits they
2190match in order to make the rest of the pattern match, (?&#62;\d+) can only match
2191an entire sequence of digits.
2192</P>
2193<P>
2194Atomic groups in general can of course contain arbitrarily complicated
2195expressions, and can be nested. However, when the contents of an atomic
2196group is just a single repeated item, as in the example above, a simpler
2197notation, called a "possessive quantifier" can be used. This consists of an
2198additional + character following a quantifier. Using this notation, the
2199previous example can be rewritten as
2200<pre>
2201  \d++foo
2202</pre>
2203Note that a possessive quantifier can be used with an entire group, for
2204example:
2205<pre>
2206  (abc|xyz){2,3}+
2207</pre>
2208Possessive quantifiers are always greedy; the setting of the PCRE2_UNGREEDY
2209option is ignored. They are a convenient notation for the simpler forms of
2210atomic group. However, there is no difference in the meaning of a possessive
2211quantifier and the equivalent atomic group, though there may be a performance
2212difference; possessive quantifiers should be slightly faster.
2213</P>
2214<P>
2215The possessive quantifier syntax is an extension to the Perl 5.8 syntax.
2216Jeffrey Friedl originated the idea (and the name) in the first edition of his
2217book. Mike McCloskey liked it, so implemented it when he built Sun's Java
2218package, and PCRE1 copied it from there. It found its way into Perl at release
22195.10.
2220</P>
2221<P>
2222PCRE2 has an optimization that automatically "possessifies" certain simple
2223pattern constructs. For example, the sequence A+B is treated as A++B because
2224there is no point in backtracking into a sequence of A's when B must follow.
2225This feature can be disabled by the PCRE2_NO_AUTOPOSSESS option, or starting
2226the pattern with (*NO_AUTO_POSSESS).
2227</P>
2228<P>
2229When a pattern contains an unlimited repeat inside a group that can itself be
2230repeated an unlimited number of times, the use of an atomic group is the only
2231way to avoid some failing matches taking a very long time indeed. The pattern
2232<pre>
2233  (\D+|&#60;\d+&#62;)*[!?]
2234</pre>
2235matches an unlimited number of substrings that either consist of non-digits, or
2236digits enclosed in &#60;&#62;, followed by either ! or ?. When it matches, it runs
2237quickly. However, if it is applied to
2238<pre>
2239  aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
2240</pre>
2241it takes a long time before reporting failure. This is because the string can
2242be divided between the internal \D+ repeat and the external * repeat in a
2243large number of ways, and all have to be tried. (The example uses [!?] rather
2244than a single character at the end, because both PCRE2 and Perl have an
2245optimization that allows for fast failure when a single character is used. They
2246remember the last single character that is required for a match, and fail early
2247if it is not present in the string.) If the pattern is changed so that it uses
2248an atomic group, like this:
2249<pre>
2250  ((?&#62;\D+)|&#60;\d+&#62;)*[!?]
2251</pre>
2252sequences of non-digits cannot be broken, and failure happens quickly.
2253<a name="backreferences"></a></P>
2254<br><a name="SEC19" href="#TOC1">BACKREFERENCES</a><br>
2255<P>
2256Outside a character class, a backslash followed by a digit greater than 0 (and
2257possibly further digits) is a backreference to a capture group earlier (that
2258is, to its left) in the pattern, provided there have been that many previous
2259capture groups.
2260</P>
2261<P>
2262However, if the decimal number following the backslash is less than 8, it is
2263always taken as a backreference, and causes an error only if there are not that
2264many capture groups in the entire pattern. In other words, the group that is
2265referenced need not be to the left of the reference for numbers less than 8. A
2266"forward backreference" of this type can make sense when a repetition is
2267involved and the group to the right has participated in an earlier iteration.
2268</P>
2269<P>
2270It is not possible to have a numerical "forward backreference" to a group whose
2271number is 8 or more using this syntax because a sequence such as \50 is
2272interpreted as a character defined in octal. See the subsection entitled
2273"Non-printing characters"
2274<a href="#digitsafterbackslash">above</a>
2275for further details of the handling of digits following a backslash. Other
2276forms of backreferencing do not suffer from this restriction. In particular,
2277there is no problem when named capture groups are used (see below).
2278</P>
2279<P>
2280Another way of avoiding the ambiguity inherent in the use of digits following a
2281backslash is to use the \g escape sequence. This escape must be followed by a
2282signed or unsigned number, optionally enclosed in braces. These examples are
2283all identical:
2284<pre>
2285  (ring), \1
2286  (ring), \g1
2287  (ring), \g{1}
2288</pre>
2289An unsigned number specifies an absolute reference without the ambiguity that
2290is present in the older syntax. It is also useful when literal digits follow
2291the reference. A signed number is a relative reference. Consider this example:
2292<pre>
2293  (abc(def)ghi)\g{-1}
2294</pre>
2295The sequence \g{-1} is a reference to the most recently started capture group
2296before \g, that is, is it equivalent to \2 in this example. Similarly,
2297\g{-2} would be equivalent to \1. The use of relative references can be
2298helpful in long patterns, and also in patterns that are created by joining
2299together fragments that contain references within themselves.
2300</P>
2301<P>
2302The sequence \g{+1} is a reference to the next capture group. This kind of
2303forward reference can be useful in patterns that repeat. Perl does not support
2304the use of + in this way.
2305</P>
2306<P>
2307A backreference matches whatever actually most recently matched the capture
2308group in the current subject string, rather than anything at all that matches
2309the group (see
2310<a href="#groupsassubroutines">"Groups as subroutines"</a>
2311below for a way of doing that). So the pattern
2312<pre>
2313  (sens|respons)e and \1ibility
2314</pre>
2315matches "sense and sensibility" and "response and responsibility", but not
2316"sense and responsibility". If caseful matching is in force at the time of the
2317backreference, the case of letters is relevant. For example,
2318<pre>
2319  ((?i)rah)\s+\1
2320</pre>
2321matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
2322capture group is matched caselessly.
2323</P>
2324<P>
2325There are several different ways of writing backreferences to named capture
2326groups. The .NET syntax \k{name} and the Perl syntax \k&#60;name&#62; or \k'name'
2327are supported, as is the Python syntax (?P=name). Perl 5.10's unified
2328backreference syntax, in which \g can be used for both numeric and named
2329references, is also supported. We could rewrite the above example in any of the
2330following ways:
2331<pre>
2332  (?&#60;p1&#62;(?i)rah)\s+\k&#60;p1&#62;
2333  (?'p1'(?i)rah)\s+\k{p1}
2334  (?P&#60;p1&#62;(?i)rah)\s+(?P=p1)
2335  (?&#60;p1&#62;(?i)rah)\s+\g{p1}
2336</pre>
2337A capture group that is referenced by name may appear in the pattern before or
2338after the reference.
2339</P>
2340<P>
2341There may be more than one backreference to the same group. If a group has not
2342actually been used in a particular match, backreferences to it always fail by
2343default. For example, the pattern
2344<pre>
2345  (a|(bc))\2
2346</pre>
2347always fails if it starts to match "a" rather than "bc". However, if the
2348PCRE2_MATCH_UNSET_BACKREF option is set at compile time, a backreference to an
2349unset value matches an empty string.
2350</P>
2351<P>
2352Because there may be many capture groups in a pattern, all digits following a
2353backslash are taken as part of a potential backreference number. If the pattern
2354continues with a digit character, some delimiter must be used to terminate the
2355backreference. If the PCRE2_EXTENDED or PCRE2_EXTENDED_MORE option is set, this
2356can be white space. Otherwise, the \g{} syntax or an empty comment (see
2357<a href="#comments">"Comments"</a>
2358below) can be used.
2359</P>
2360<br><b>
2361Recursive backreferences
2362</b><br>
2363<P>
2364A backreference that occurs inside the group to which it refers fails when the
2365group is first used, so, for example, (a\1) never matches. However, such
2366references can be useful inside repeated groups. For example, the pattern
2367<pre>
2368  (a|b\1)+
2369</pre>
2370matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
2371the group, the backreference matches the character string corresponding to the
2372previous iteration. In order for this to work, the pattern must be such that
2373the first iteration does not need to match the backreference. This can be done
2374using alternation, as in the example above, or by a quantifier with a minimum
2375of zero.
2376</P>
2377<P>
2378For versions of PCRE2 less than 10.25, backreferences of this type used to
2379cause the group that they reference to be treated as an
2380<a href="#atomicgroup">atomic group.</a>
2381This restriction no longer applies, and backtracking into such groups can occur
2382as normal.
2383<a name="bigassertions"></a></P>
2384<br><a name="SEC20" href="#TOC1">ASSERTIONS</a><br>
2385<P>
2386An assertion is a test on the characters following or preceding the current
2387matching point that does not consume any characters. The simple assertions
2388coded as \b, \B, \A, \G, \Z, \z, ^ and $ are described
2389<a href="#smallassertions">above.</a>
2390</P>
2391<P>
2392More complicated assertions are coded as parenthesized groups. There are two
2393kinds: those that look ahead of the current position in the subject string, and
2394those that look behind it, and in each case an assertion may be positive (must
2395match for the assertion to be true) or negative (must not match for the
2396assertion to be true). An assertion group is matched in the normal way,
2397and if it is true, matching continues after it, but with the matching position
2398in the subject string reset to what it was before the assertion was processed.
2399</P>
2400<P>
2401The Perl-compatible lookaround assertions are atomic. If an assertion is true,
2402but there is a subsequent matching failure, there is no backtracking into the
2403assertion. However, there are some cases where non-atomic assertions can be
2404useful. PCRE2 has some support for these, described in the section entitled
2405<a href="#nonatomicassertions">"Non-atomic assertions"</a>
2406below, but they are not Perl-compatible.
2407</P>
2408<P>
2409A lookaround assertion may appear as the condition in a
2410<a href="#conditions">conditional group</a>
2411(see below). In this case, the result of matching the assertion determines
2412which branch of the condition is followed.
2413</P>
2414<P>
2415Assertion groups are not capture groups. If an assertion contains capture
2416groups within it, these are counted for the purposes of numbering the capture
2417groups in the whole pattern. Within each branch of an assertion, locally
2418captured substrings may be referenced in the usual way. For example, a sequence
2419such as (.)\g{-1} can be used to check that two adjacent characters are the
2420same.
2421</P>
2422<P>
2423When a branch within an assertion fails to match, any substrings that were
2424captured are discarded (as happens with any pattern branch that fails to
2425match). A negative assertion is true only when all its branches fail to match;
2426this means that no captured substrings are ever retained after a successful
2427negative assertion. When an assertion contains a matching branch, what happens
2428depends on the type of assertion.
2429</P>
2430<P>
2431For a positive assertion, internally captured substrings in the successful
2432branch are retained, and matching continues with the next pattern item after
2433the assertion. For a negative assertion, a matching branch means that the
2434assertion is not true. If such an assertion is being used as a condition in a
2435<a href="#conditions">conditional group</a>
2436(see below), captured substrings are retained, because matching continues with
2437the "no" branch of the condition. For other failing negative assertions,
2438control passes to the previous backtracking point, thus discarding any captured
2439strings within the assertion.
2440</P>
2441<P>
2442Most assertion groups may be repeated; though it makes no sense to assert the
2443same thing several times, the side effect of capturing in positive assertions
2444may occasionally be useful. However, an assertion that forms the condition for
2445a conditional group may not be quantified. PCRE2 used to restrict the
2446repetition of assertions, but from release 10.35 the only restriction is that
2447an unlimited maximum repetition is changed to be one more than the minimum. For
2448example, {3,} is treated as {3,4}.
2449</P>
2450<br><b>
2451Alphabetic assertion names
2452</b><br>
2453<P>
2454Traditionally, symbolic sequences such as (?= and (?&#60;= have been used to
2455specify lookaround assertions. Perl 5.28 introduced some experimental
2456alphabetic alternatives which might be easier to remember. They all start with
2457(* instead of (? and must be written using lower case letters. PCRE2 supports
2458the following synonyms:
2459<pre>
2460  (*positive_lookahead:  or (*pla: is the same as (?=
2461  (*negative_lookahead:  or (*nla: is the same as (?!
2462  (*positive_lookbehind: or (*plb: is the same as (?&#60;=
2463  (*negative_lookbehind: or (*nlb: is the same as (?&#60;!
2464</pre>
2465For example, (*pla:foo) is the same assertion as (?=foo). In the following
2466sections, the various assertions are described using the original symbolic
2467forms.
2468</P>
2469<br><b>
2470Lookahead assertions
2471</b><br>
2472<P>
2473Lookahead assertions start with (?= for positive assertions and (?! for
2474negative assertions. For example,
2475<pre>
2476  \w+(?=;)
2477</pre>
2478matches a word followed by a semicolon, but does not include the semicolon in
2479the match, and
2480<pre>
2481  foo(?!bar)
2482</pre>
2483matches any occurrence of "foo" that is not followed by "bar". Note that the
2484apparently similar pattern
2485<pre>
2486  (?!foo)bar
2487</pre>
2488does not find an occurrence of "bar" that is preceded by something other than
2489"foo"; it finds any occurrence of "bar" whatsoever, because the assertion
2490(?!foo) is always true when the next three characters are "bar". A
2491lookbehind assertion is needed to achieve the other effect.
2492</P>
2493<P>
2494If you want to force a matching failure at some point in a pattern, the most
2495convenient way to do it is with (?!) because an empty string always matches, so
2496an assertion that requires there not to be an empty string must always fail.
2497The backtracking control verb (*FAIL) or (*F) is a synonym for (?!).
2498<a name="lookbehind"></a></P>
2499<br><b>
2500Lookbehind assertions
2501</b><br>
2502<P>
2503Lookbehind assertions start with (?&#60;= for positive assertions and (?&#60;! for
2504negative assertions. For example,
2505<pre>
2506  (?&#60;!foo)bar
2507</pre>
2508does find an occurrence of "bar" that is not preceded by "foo". The contents of
2509a lookbehind assertion are restricted such that all the strings it matches must
2510have a fixed length. However, if there are several top-level alternatives, they
2511do not all have to have the same fixed length. Thus
2512<pre>
2513  (?&#60;=bullock|donkey)
2514</pre>
2515is permitted, but
2516<pre>
2517  (?&#60;!dogs?|cats?)
2518</pre>
2519causes an error at compile time. Branches that match different length strings
2520are permitted only at the top level of a lookbehind assertion. This is an
2521extension compared with Perl, which requires all branches to match the same
2522length of string. An assertion such as
2523<pre>
2524  (?&#60;=ab(c|de))
2525</pre>
2526is not permitted, because its single top-level branch can match two different
2527lengths, but it is acceptable to PCRE2 if rewritten to use two top-level
2528branches:
2529<pre>
2530  (?&#60;=abc|abde)
2531</pre>
2532In some cases, the escape sequence \K
2533<a href="#resetmatchstart">(see above)</a>
2534can be used instead of a lookbehind assertion to get round the fixed-length
2535restriction.
2536</P>
2537<P>
2538The implementation of lookbehind assertions is, for each alternative, to
2539temporarily move the current position back by the fixed length and then try to
2540match. If there are insufficient characters before the current position, the
2541assertion fails.
2542</P>
2543<P>
2544In UTF-8 and UTF-16 modes, PCRE2 does not allow the \C escape (which matches a
2545single code unit even in a UTF mode) to appear in lookbehind assertions,
2546because it makes it impossible to calculate the length of the lookbehind. The
2547\X and \R escapes, which can match different numbers of code units, are never
2548permitted in lookbehinds.
2549</P>
2550<P>
2551<a href="#groupsassubroutines">"Subroutine"</a>
2552calls (see below) such as (?2) or (?&X) are permitted in lookbehinds, as long
2553as the called capture group matches a fixed-length string. However,
2554<a href="#recursion">recursion,</a>
2555that is, a "subroutine" call into a group that is already active,
2556is not supported.
2557</P>
2558<P>
2559Perl does not support backreferences in lookbehinds. PCRE2 does support them,
2560but only if certain conditions are met. The PCRE2_MATCH_UNSET_BACKREF option
2561must not be set, there must be no use of (?| in the pattern (it creates
2562duplicate group numbers), and if the backreference is by name, the name
2563must be unique. Of course, the referenced group must itself match a fixed
2564length substring. The following pattern matches words containing at least two
2565characters that begin and end with the same character:
2566<pre>
2567   \b(\w)\w++(?&#60;=\1)
2568</PRE>
2569</P>
2570<P>
2571Possessive quantifiers can be used in conjunction with lookbehind assertions to
2572specify efficient matching of fixed-length strings at the end of subject
2573strings. Consider a simple pattern such as
2574<pre>
2575  abcd$
2576</pre>
2577when applied to a long string that does not match. Because matching proceeds
2578from left to right, PCRE2 will look for each "a" in the subject and then see if
2579what follows matches the rest of the pattern. If the pattern is specified as
2580<pre>
2581  ^.*abcd$
2582</pre>
2583the initial .* matches the entire string at first, but when this fails (because
2584there is no following "a"), it backtracks to match all but the last character,
2585then all but the last two characters, and so on. Once again the search for "a"
2586covers the entire string, from right to left, so we are no better off. However,
2587if the pattern is written as
2588<pre>
2589  ^.*+(?&#60;=abcd)
2590</pre>
2591there can be no backtracking for the .*+ item because of the possessive
2592quantifier; it can match only the entire string. The subsequent lookbehind
2593assertion does a single test on the last four characters. If it fails, the
2594match fails immediately. For long strings, this approach makes a significant
2595difference to the processing time.
2596</P>
2597<br><b>
2598Using multiple assertions
2599</b><br>
2600<P>
2601Several assertions (of any sort) may occur in succession. For example,
2602<pre>
2603  (?&#60;=\d{3})(?&#60;!999)foo
2604</pre>
2605matches "foo" preceded by three digits that are not "999". Notice that each of
2606the assertions is applied independently at the same point in the subject
2607string. First there is a check that the previous three characters are all
2608digits, and then there is a check that the same three characters are not "999".
2609This pattern does <i>not</i> match "foo" preceded by six characters, the first
2610of which are digits and the last three of which are not "999". For example, it
2611doesn't match "123abcfoo". A pattern to do that is
2612<pre>
2613  (?&#60;=\d{3}...)(?&#60;!999)foo
2614</pre>
2615This time the first assertion looks at the preceding six characters, checking
2616that the first three are digits, and then the second assertion checks that the
2617preceding three characters are not "999".
2618</P>
2619<P>
2620Assertions can be nested in any combination. For example,
2621<pre>
2622  (?&#60;=(?&#60;!foo)bar)baz
2623</pre>
2624matches an occurrence of "baz" that is preceded by "bar" which in turn is not
2625preceded by "foo", while
2626<pre>
2627  (?&#60;=\d{3}(?!999)...)foo
2628</pre>
2629is another pattern that matches "foo" preceded by three digits and any three
2630characters that are not "999".
2631<a name="nonatomicassertions"></a></P>
2632<br><a name="SEC21" href="#TOC1">NON-ATOMIC ASSERTIONS</a><br>
2633<P>
2634The traditional Perl-compatible lookaround assertions are atomic. That is, if
2635an assertion is true, but there is a subsequent matching failure, there is no
2636backtracking into the assertion. However, there are some cases where non-atomic
2637positive assertions can be useful. PCRE2 provides these using the following
2638syntax:
2639<pre>
2640  (*non_atomic_positive_lookahead:  or (*napla: or (?*
2641  (*non_atomic_positive_lookbehind: or (*naplb: or (?&#60;*
2642</pre>
2643Consider the problem of finding the right-most word in a string that also
2644appears earlier in the string, that is, it must appear at least twice in total.
2645This pattern returns the required result as captured substring 1:
2646<pre>
2647  ^(?x)(*napla: .* \b(\w++)) (?&#62; .*? \b\1\b ){2}
2648</pre>
2649For a subject such as "word1 word2 word3 word2 word3 word4" the result is
2650"word3". How does it work? At the start, ^(?x) anchors the pattern and sets the
2651"x" option, which causes white space (introduced for readability) to be
2652ignored. Inside the assertion, the greedy .* at first consumes the entire
2653string, but then has to backtrack until the rest of the assertion can match a
2654word, which is captured by group 1. In other words, when the assertion first
2655succeeds, it captures the right-most word in the string.
2656</P>
2657<P>
2658The current matching point is then reset to the start of the subject, and the
2659rest of the pattern match checks for two occurrences of the captured word,
2660using an ungreedy .*? to scan from the left. If this succeeds, we are done, but
2661if the last word in the string does not occur twice, this part of the pattern
2662fails. If a traditional atomic lookhead (?= or (*pla: had been used, the
2663assertion could not be re-entered, and the whole match would fail. The pattern
2664would succeed only if the very last word in the subject was found twice.
2665</P>
2666<P>
2667Using a non-atomic lookahead, however, means that when the last word does not
2668occur twice in the string, the lookahead can backtrack and find the second-last
2669word, and so on, until either the match succeeds, or all words have been
2670tested.
2671</P>
2672<P>
2673Two conditions must be met for a non-atomic assertion to be useful: the
2674contents of one or more capturing groups must change after a backtrack into the
2675assertion, and there must be a backreference to a changed group later in the
2676pattern. If this is not the case, the rest of the pattern match fails exactly
2677as before because nothing has changed, so using a non-atomic assertion just
2678wastes resources.
2679</P>
2680<P>
2681There is one exception to backtracking into a non-atomic assertion. If an
2682(*ACCEPT) control verb is triggered, the assertion succeeds atomically. That
2683is, a subsequent match failure cannot backtrack into the assertion.
2684</P>
2685<P>
2686Non-atomic assertions are not supported by the alternative matching function
2687<b>pcre2_dfa_match()</b>. They are supported by JIT, but only if they do not
2688contain any control verbs such as (*ACCEPT). (This may change in future). Note
2689that assertions that appear as conditions for
2690<a href="#conditions">conditional groups</a>
2691(see below) must be atomic.
2692</P>
2693<br><a name="SEC22" href="#TOC1">SCRIPT RUNS</a><br>
2694<P>
2695In concept, a script run is a sequence of characters that are all from the same
2696Unicode script such as Latin or Greek. However, because some scripts are
2697commonly used together, and because some diacritical and other marks are used
2698with multiple scripts, it is not that simple. There is a full description of
2699the rules that PCRE2 uses in the section entitled
2700<a href="pcre2unicode.html#scriptruns">"Script Runs"</a>
2701in the
2702<a href="pcre2unicode.html"><b>pcre2unicode</b></a>
2703documentation.
2704</P>
2705<P>
2706If part of a pattern is enclosed between (*script_run: or (*sr: and a closing
2707parenthesis, it fails if the sequence of characters that it matches are not a
2708script run. After a failure, normal backtracking occurs. Script runs can be
2709used to detect spoofing attacks using characters that look the same, but are
2710from different scripts. The string "paypal.com" is an infamous example, where
2711the letters could be a mixture of Latin and Cyrillic. This pattern ensures that
2712the matched characters in a sequence of non-spaces that follow white space are
2713a script run:
2714<pre>
2715  \s+(*sr:\S+)
2716</pre>
2717To be sure that they are all from the Latin script (for example), a lookahead
2718can be used:
2719<pre>
2720  \s+(?=\p{Latin})(*sr:\S+)
2721</pre>
2722This works as long as the first character is expected to be a character in that
2723script, and not (for example) punctuation, which is allowed with any script. If
2724this is not the case, a more creative lookahead is needed. For example, if
2725digits, underscore, and dots are permitted at the start:
2726<pre>
2727  \s+(?=[0-9_.]*\p{Latin})(*sr:\S+)
2728
2729</PRE>
2730</P>
2731<P>
2732In many cases, backtracking into a script run pattern fragment is not
2733desirable. The script run can employ an atomic group to prevent this. Because
2734this is a common requirement, a shorthand notation is provided by
2735(*atomic_script_run: or (*asr:
2736<pre>
2737  (*asr:...) is the same as (*sr:(?&#62;...))
2738</pre>
2739Note that the atomic group is inside the script run. Putting it outside would
2740not prevent backtracking into the script run pattern.
2741</P>
2742<P>
2743Support for script runs is not available if PCRE2 is compiled without Unicode
2744support. A compile-time error is given if any of the above constructs is
2745encountered. Script runs are not supported by the alternate matching function,
2746<b>pcre2_dfa_match()</b> because they use the same mechanism as capturing
2747parentheses.
2748</P>
2749<P>
2750<b>Warning:</b> The (*ACCEPT) control verb
2751<a href="#acceptverb">(see below)</a>
2752should not be used within a script run group, because it causes an immediate
2753exit from the group, bypassing the script run checking.
2754<a name="conditions"></a></P>
2755<br><a name="SEC23" href="#TOC1">CONDITIONAL GROUPS</a><br>
2756<P>
2757It is possible to cause the matching process to obey a pattern fragment
2758conditionally or to choose between two alternative fragments, depending on
2759the result of an assertion, or whether a specific capture group has
2760already been matched. The two possible forms of conditional group are:
2761<pre>
2762  (?(condition)yes-pattern)
2763  (?(condition)yes-pattern|no-pattern)
2764</pre>
2765If the condition is satisfied, the yes-pattern is used; otherwise the
2766no-pattern (if present) is used. An absent no-pattern is equivalent to an empty
2767string (it always matches). If there are more than two alternatives in the
2768group, a compile-time error occurs. Each of the two alternatives may itself
2769contain nested groups of any form, including conditional groups; the
2770restriction to two alternatives applies only at the level of the condition
2771itself. This pattern fragment is an example where the alternatives are complex:
2772<pre>
2773  (?(1) (A|B|C) | (D | (?(2)E|F) | E) )
2774
2775</PRE>
2776</P>
2777<P>
2778There are five kinds of condition: references to capture groups, references to
2779recursion, two pseudo-conditions called DEFINE and VERSION, and assertions.
2780</P>
2781<br><b>
2782Checking for a used capture group by number
2783</b><br>
2784<P>
2785If the text between the parentheses consists of a sequence of digits, the
2786condition is true if a capture group of that number has previously matched. If
2787there is more than one capture group with the same number (see the earlier
2788<a href="#recursion">section about duplicate group numbers),</a>
2789the condition is true if any of them have matched. An alternative notation is
2790to precede the digits with a plus or minus sign. In this case, the group number
2791is relative rather than absolute. The most recently opened capture group can be
2792referenced by (?(-1), the next most recent by (?(-2), and so on. Inside loops
2793it can also make sense to refer to subsequent groups. The next capture group
2794can be referenced as (?(+1), and so on. (The value zero in any of these forms
2795is not used; it provokes a compile-time error.)
2796</P>
2797<P>
2798Consider the following pattern, which contains non-significant white space to
2799make it more readable (assume the PCRE2_EXTENDED option) and to divide it into
2800three parts for ease of discussion:
2801<pre>
2802  ( \( )?    [^()]+    (?(1) \) )
2803</pre>
2804The first part matches an optional opening parenthesis, and if that
2805character is present, sets it as the first captured substring. The second part
2806matches one or more characters that are not parentheses. The third part is a
2807conditional group that tests whether or not the first capture group
2808matched. If it did, that is, if subject started with an opening parenthesis,
2809the condition is true, and so the yes-pattern is executed and a closing
2810parenthesis is required. Otherwise, since no-pattern is not present, the
2811conditional group matches nothing. In other words, this pattern matches a
2812sequence of non-parentheses, optionally enclosed in parentheses.
2813</P>
2814<P>
2815If you were embedding this pattern in a larger one, you could use a relative
2816reference:
2817<pre>
2818  ...other stuff... ( \( )?    [^()]+    (?(-1) \) ) ...
2819</pre>
2820This makes the fragment independent of the parentheses in the larger pattern.
2821</P>
2822<br><b>
2823Checking for a used capture group by name
2824</b><br>
2825<P>
2826Perl uses the syntax (?(&#60;name&#62;)...) or (?('name')...) to test for a used
2827capture group by name. For compatibility with earlier versions of PCRE1, which
2828had this facility before Perl, the syntax (?(name)...) is also recognized.
2829Note, however, that undelimited names consisting of the letter R followed by
2830digits are ambiguous (see the following section). Rewriting the above example
2831to use a named group gives this:
2832<pre>
2833  (?&#60;OPEN&#62; \( )?    [^()]+    (?(&#60;OPEN&#62;) \) )
2834</pre>
2835If the name used in a condition of this kind is a duplicate, the test is
2836applied to all groups of the same name, and is true if any one of them has
2837matched.
2838</P>
2839<br><b>
2840Checking for pattern recursion
2841</b><br>
2842<P>
2843"Recursion" in this sense refers to any subroutine-like call from one part of
2844the pattern to another, whether or not it is actually recursive. See the
2845sections entitled
2846<a href="#recursion">"Recursive patterns"</a>
2847and
2848<a href="#groupsassubroutines">"Groups as subroutines"</a>
2849below for details of recursion and subroutine calls.
2850</P>
2851<P>
2852If a condition is the string (R), and there is no capture group with the name
2853R, the condition is true if matching is currently in a recursion or subroutine
2854call to the whole pattern or any capture group. If digits follow the letter R,
2855and there is no group with that name, the condition is true if the most recent
2856call is into a group with the given number, which must exist somewhere in the
2857overall pattern. This is a contrived example that is equivalent to a+b:
2858<pre>
2859  ((?(R1)a+|(?1)b))
2860</pre>
2861However, in both cases, if there is a capture group with a matching name, the
2862condition tests for its being set, as described in the section above, instead
2863of testing for recursion. For example, creating a group with the name R1 by
2864adding (?&#60;R1&#62;) to the above pattern completely changes its meaning.
2865</P>
2866<P>
2867If a name preceded by ampersand follows the letter R, for example:
2868<pre>
2869  (?(R&name)...)
2870</pre>
2871the condition is true if the most recent recursion is into a group of that name
2872(which must exist within the pattern).
2873</P>
2874<P>
2875This condition does not check the entire recursion stack. It tests only the
2876current level. If the name used in a condition of this kind is a duplicate, the
2877test is applied to all groups of the same name, and is true if any one of
2878them is the most recent recursion.
2879</P>
2880<P>
2881At "top level", all these recursion test conditions are false.
2882<a name="subdefine"></a></P>
2883<br><b>
2884Defining capture groups for use by reference only
2885</b><br>
2886<P>
2887If the condition is the string (DEFINE), the condition is always false, even if
2888there is a group with the name DEFINE. In this case, there may be only one
2889alternative in the rest of the conditional group. It is always skipped if
2890control reaches this point in the pattern; the idea of DEFINE is that it can be
2891used to define subroutines that can be referenced from elsewhere. (The use of
2892<a href="#groupsassubroutines">subroutines</a>
2893is described below.) For example, a pattern to match an IPv4 address such as
2894"192.168.23.245" could be written like this (ignore white space and line
2895breaks):
2896<pre>
2897  (?(DEFINE) (?&#60;byte&#62; 2[0-4]\d | 25[0-5] | 1\d\d | [1-9]?\d) )
2898  \b (?&byte) (\.(?&byte)){3} \b
2899</pre>
2900The first part of the pattern is a DEFINE group inside which a another group
2901named "byte" is defined. This matches an individual component of an IPv4
2902address (a number less than 256). When matching takes place, this part of the
2903pattern is skipped because DEFINE acts like a false condition. The rest of the
2904pattern uses references to the named group to match the four dot-separated
2905components of an IPv4 address, insisting on a word boundary at each end.
2906</P>
2907<br><b>
2908Checking the PCRE2 version
2909</b><br>
2910<P>
2911Programs that link with a PCRE2 library can check the version by calling
2912<b>pcre2_config()</b> with appropriate arguments. Users of applications that do
2913not have access to the underlying code cannot do this. A special "condition"
2914called VERSION exists to allow such users to discover which version of PCRE2
2915they are dealing with by using this condition to match a string such as
2916"yesno". VERSION must be followed either by "=" or "&#62;=" and a version number.
2917For example:
2918<pre>
2919  (?(VERSION&#62;=10.4)yes|no)
2920</pre>
2921This pattern matches "yes" if the PCRE2 version is greater or equal to 10.4, or
2922"no" otherwise. The fractional part of the version number may not contain more
2923than two digits.
2924</P>
2925<br><b>
2926Assertion conditions
2927</b><br>
2928<P>
2929If the condition is not in any of the above formats, it must be a parenthesized
2930assertion. This may be a positive or negative lookahead or lookbehind
2931assertion. However, it must be a traditional atomic assertion, not one of the
2932PCRE2-specific
2933<a href="#nonatomicassertions">non-atomic assertions.</a>
2934</P>
2935<P>
2936Consider this pattern, again containing non-significant white space, and with
2937the two alternatives on the second line:
2938<pre>
2939  (?(?=[^a-z]*[a-z])
2940  \d{2}-[a-z]{3}-\d{2}  |  \d{2}-\d{2}-\d{2} )
2941</pre>
2942The condition is a positive lookahead assertion that matches an optional
2943sequence of non-letters followed by a letter. In other words, it tests for the
2944presence of at least one letter in the subject. If a letter is found, the
2945subject is matched against the first alternative; otherwise it is matched
2946against the second. This pattern matches strings in one of the two forms
2947dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
2948</P>
2949<P>
2950When an assertion that is a condition contains capture groups, any
2951capturing that occurs in a matching branch is retained afterwards, for both
2952positive and negative assertions, because matching always continues after the
2953assertion, whether it succeeds or fails. (Compare non-conditional assertions,
2954for which captures are retained only for positive assertions that succeed.)
2955<a name="comments"></a></P>
2956<br><a name="SEC24" href="#TOC1">COMMENTS</a><br>
2957<P>
2958There are two ways of including comments in patterns that are processed by
2959PCRE2. In both cases, the start of the comment must not be in a character
2960class, nor in the middle of any other sequence of related characters such as
2961(?: or a group name or number. The characters that make up a comment play
2962no part in the pattern matching.
2963</P>
2964<P>
2965The sequence (?# marks the start of a comment that continues up to the next
2966closing parenthesis. Nested parentheses are not permitted. If the
2967PCRE2_EXTENDED or PCRE2_EXTENDED_MORE option is set, an unescaped # character
2968also introduces a comment, which in this case continues to immediately after
2969the next newline character or character sequence in the pattern. Which
2970characters are interpreted as newlines is controlled by an option passed to the
2971compiling function or by a special sequence at the start of the pattern, as
2972described in the section entitled
2973<a href="#newlines">"Newline conventions"</a>
2974above. Note that the end of this type of comment is a literal newline sequence
2975in the pattern; escape sequences that happen to represent a newline do not
2976count. For example, consider this pattern when PCRE2_EXTENDED is set, and the
2977default newline convention (a single linefeed character) is in force:
2978<pre>
2979  abc #comment \n still comment
2980</pre>
2981On encountering the # character, <b>pcre2_compile()</b> skips along, looking for
2982a newline in the pattern. The sequence \n is still literal at this stage, so
2983it does not terminate the comment. Only an actual character with the code value
29840x0a (the default newline) does so.
2985<a name="recursion"></a></P>
2986<br><a name="SEC25" href="#TOC1">RECURSIVE PATTERNS</a><br>
2987<P>
2988Consider the problem of matching a string in parentheses, allowing for
2989unlimited nested parentheses. Without the use of recursion, the best that can
2990be done is to use a pattern that matches up to some fixed depth of nesting. It
2991is not possible to handle an arbitrary nesting depth.
2992</P>
2993<P>
2994For some time, Perl has provided a facility that allows regular expressions to
2995recurse (amongst other things). It does this by interpolating Perl code in the
2996expression at run time, and the code can refer to the expression itself. A Perl
2997pattern using code interpolation to solve the parentheses problem can be
2998created like this:
2999<pre>
3000  $re = qr{\( (?: (?&#62;[^()]+) | (?p{$re}) )* \)}x;
3001</pre>
3002The (?p{...}) item interpolates Perl code at run time, and in this case refers
3003recursively to the pattern in which it appears.
3004</P>
3005<P>
3006Obviously, PCRE2 cannot support the interpolation of Perl code. Instead, it
3007supports special syntax for recursion of the entire pattern, and also for
3008individual capture group recursion. After its introduction in PCRE1 and Python,
3009this kind of recursion was subsequently introduced into Perl at release 5.10.
3010</P>
3011<P>
3012A special item that consists of (? followed by a number greater than zero and a
3013closing parenthesis is a recursive subroutine call of the capture group of the
3014given number, provided that it occurs inside that group. (If not, it is a
3015<a href="#groupsassubroutines">non-recursive subroutine</a>
3016call, which is described in the next section.) The special item (?R) or (?0) is
3017a recursive call of the entire regular expression.
3018</P>
3019<P>
3020This PCRE2 pattern solves the nested parentheses problem (assume the
3021PCRE2_EXTENDED option is set so that white space is ignored):
3022<pre>
3023  \( ( [^()]++ | (?R) )* \)
3024</pre>
3025First it matches an opening parenthesis. Then it matches any number of
3026substrings which can either be a sequence of non-parentheses, or a recursive
3027match of the pattern itself (that is, a correctly parenthesized substring).
3028Finally there is a closing parenthesis. Note the use of a possessive quantifier
3029to avoid backtracking into sequences of non-parentheses.
3030</P>
3031<P>
3032If this were part of a larger pattern, you would not want to recurse the entire
3033pattern, so instead you could use this:
3034<pre>
3035  ( \( ( [^()]++ | (?1) )* \) )
3036</pre>
3037We have put the pattern into parentheses, and caused the recursion to refer to
3038them instead of the whole pattern.
3039</P>
3040<P>
3041In a larger pattern, keeping track of parenthesis numbers can be tricky. This
3042is made easier by the use of relative references. Instead of (?1) in the
3043pattern above you can write (?-2) to refer to the second most recently opened
3044parentheses preceding the recursion. In other words, a negative number counts
3045capturing parentheses leftwards from the point at which it is encountered.
3046</P>
3047<P>
3048Be aware however, that if
3049<a href="#dupgroupnumber">duplicate capture group numbers</a>
3050are in use, relative references refer to the earliest group with the
3051appropriate number. Consider, for example:
3052<pre>
3053  (?|(a)|(b)) (c) (?-2)
3054</pre>
3055The first two capture groups (a) and (b) are both numbered 1, and group (c)
3056is number 2. When the reference (?-2) is encountered, the second most recently
3057opened parentheses has the number 1, but it is the first such group (the (a)
3058group) to which the recursion refers. This would be the same if an absolute
3059reference (?1) was used. In other words, relative references are just a
3060shorthand for computing a group number.
3061</P>
3062<P>
3063It is also possible to refer to subsequent capture groups, by writing
3064references such as (?+2). However, these cannot be recursive because the
3065reference is not inside the parentheses that are referenced. They are always
3066<a href="#groupsassubroutines">non-recursive subroutine</a>
3067calls, as described in the next section.
3068</P>
3069<P>
3070An alternative approach is to use named parentheses. The Perl syntax for this
3071is (?&name); PCRE1's earlier syntax (?P&#62;name) is also supported. We could
3072rewrite the above example as follows:
3073<pre>
3074  (?&#60;pn&#62; \( ( [^()]++ | (?&pn) )* \) )
3075</pre>
3076If there is more than one group with the same name, the earliest one is
3077used.
3078</P>
3079<P>
3080The example pattern that we have been looking at contains nested unlimited
3081repeats, and so the use of a possessive quantifier for matching strings of
3082non-parentheses is important when applying the pattern to strings that do not
3083match. For example, when this pattern is applied to
3084<pre>
3085  (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
3086</pre>
3087it yields "no match" quickly. However, if a possessive quantifier is not used,
3088the match runs for a very long time indeed because there are so many different
3089ways the + and * repeats can carve up the subject, and all have to be tested
3090before failure can be reported.
3091</P>
3092<P>
3093At the end of a match, the values of capturing parentheses are those from
3094the outermost level. If you want to obtain intermediate values, a callout
3095function can be used (see below and the
3096<a href="pcre2callout.html"><b>pcre2callout</b></a>
3097documentation). If the pattern above is matched against
3098<pre>
3099  (ab(cd)ef)
3100</pre>
3101the value for the inner capturing parentheses (numbered 2) is "ef", which is
3102the last value taken on at the top level. If a capture group is not matched at
3103the top level, its final captured value is unset, even if it was (temporarily)
3104set at a deeper level during the matching process.
3105</P>
3106<P>
3107Do not confuse the (?R) item with the condition (R), which tests for recursion.
3108Consider this pattern, which matches text in angle brackets, allowing for
3109arbitrary nesting. Only digits are allowed in nested brackets (that is, when
3110recursing), whereas any characters are permitted at the outer level.
3111<pre>
3112  &#60; (?: (?(R) \d++  | [^&#60;&#62;]*+) | (?R)) * &#62;
3113</pre>
3114In this pattern, (?(R) is the start of a conditional group, with two different
3115alternatives for the recursive and non-recursive cases. The (?R) item is the
3116actual recursive call.
3117<a name="recursiondifference"></a></P>
3118<br><b>
3119Differences in recursion processing between PCRE2 and Perl
3120</b><br>
3121<P>
3122Some former differences between PCRE2 and Perl no longer exist.
3123</P>
3124<P>
3125Before release 10.30, recursion processing in PCRE2 differed from Perl in that
3126a recursive subroutine call was always treated as an atomic group. That is,
3127once it had matched some of the subject string, it was never re-entered, even
3128if it contained untried alternatives and there was a subsequent matching
3129failure. (Historical note: PCRE implemented recursion before Perl did.)
3130</P>
3131<P>
3132Starting with release 10.30, recursive subroutine calls are no longer treated
3133as atomic. That is, they can be re-entered to try unused alternatives if there
3134is a matching failure later in the pattern. This is now compatible with the way
3135Perl works. If you want a subroutine call to be atomic, you must explicitly
3136enclose it in an atomic group.
3137</P>
3138<P>
3139Supporting backtracking into recursions simplifies certain types of recursive
3140pattern. For example, this pattern matches palindromic strings:
3141<pre>
3142  ^((.)(?1)\2|.?)$
3143</pre>
3144The second branch in the group matches a single central character in the
3145palindrome when there are an odd number of characters, or nothing when there
3146are an even number of characters, but in order to work it has to be able to try
3147the second case when the rest of the pattern match fails. If you want to match
3148typical palindromic phrases, the pattern has to ignore all non-word characters,
3149which can be done like this:
3150<pre>
3151  ^\W*+((.)\W*+(?1)\W*+\2|\W*+.?)\W*+$
3152</pre>
3153If run with the PCRE2_CASELESS option, this pattern matches phrases such as "A
3154man, a plan, a canal: Panama!". Note the use of the possessive quantifier *+ to
3155avoid backtracking into sequences of non-word characters. Without this, PCRE2
3156takes a great deal longer (ten times or more) to match typical phrases, and
3157Perl takes so long that you think it has gone into a loop.
3158</P>
3159<P>
3160Another way in which PCRE2 and Perl used to differ in their recursion
3161processing is in the handling of captured values. Formerly in Perl, when a
3162group was called recursively or as a subroutine (see the next section), it
3163had no access to any values that were captured outside the recursion, whereas
3164in PCRE2 these values can be referenced. Consider this pattern:
3165<pre>
3166  ^(.)(\1|a(?2))
3167</pre>
3168This pattern matches "bab". The first capturing parentheses match "b", then in
3169the second group, when the backreference \1 fails to match "b", the second
3170alternative matches "a" and then recurses. In the recursion, \1 does now match
3171"b" and so the whole match succeeds. This match used to fail in Perl, but in
3172later versions (I tried 5.024) it now works.
3173<a name="groupsassubroutines"></a></P>
3174<br><a name="SEC26" href="#TOC1">GROUPS AS SUBROUTINES</a><br>
3175<P>
3176If the syntax for a recursive group call (either by number or by name) is used
3177outside the parentheses to which it refers, it operates a bit like a subroutine
3178in a programming language. More accurately, PCRE2 treats the referenced group
3179as an independent subpattern which it tries to match at the current matching
3180position. The called group may be defined before or after the reference. A
3181numbered reference can be absolute or relative, as in these examples:
3182<pre>
3183  (...(absolute)...)...(?2)...
3184  (...(relative)...)...(?-1)...
3185  (...(?+1)...(relative)...
3186</pre>
3187An earlier example pointed out that the pattern
3188<pre>
3189  (sens|respons)e and \1ibility
3190</pre>
3191matches "sense and sensibility" and "response and responsibility", but not
3192"sense and responsibility". If instead the pattern
3193<pre>
3194  (sens|respons)e and (?1)ibility
3195</pre>
3196is used, it does match "sense and responsibility" as well as the other two
3197strings. Another example is given in the discussion of DEFINE above.
3198</P>
3199<P>
3200Like recursions, subroutine calls used to be treated as atomic, but this
3201changed at PCRE2 release 10.30, so backtracking into subroutine calls can now
3202occur. However, any capturing parentheses that are set during the subroutine
3203call revert to their previous values afterwards.
3204</P>
3205<P>
3206Processing options such as case-independence are fixed when a group is
3207defined, so if it is used as a subroutine, such options cannot be changed for
3208different calls. For example, consider this pattern:
3209<pre>
3210  (abc)(?i:(?-1))
3211</pre>
3212It matches "abcabc". It does not match "abcABC" because the change of
3213processing option does not affect the called group.
3214</P>
3215<P>
3216The behaviour of
3217<a href="#backtrackcontrol">backtracking control verbs</a>
3218in groups when called as subroutines is described in the section entitled
3219<a href="#btsub">"Backtracking verbs in subroutines"</a>
3220below.
3221<a name="onigurumasubroutines"></a></P>
3222<br><a name="SEC27" href="#TOC1">ONIGURUMA SUBROUTINE SYNTAX</a><br>
3223<P>
3224For compatibility with Oniguruma, the non-Perl syntax \g followed by a name or
3225a number enclosed either in angle brackets or single quotes, is an alternative
3226syntax for calling a group as a subroutine, possibly recursively. Here are two
3227of the examples used above, rewritten using this syntax:
3228<pre>
3229  (?&#60;pn&#62; \( ( (?&#62;[^()]+) | \g&#60;pn&#62; )* \) )
3230  (sens|respons)e and \g'1'ibility
3231</pre>
3232PCRE2 supports an extension to Oniguruma: if a number is preceded by a
3233plus or a minus sign it is taken as a relative reference. For example:
3234<pre>
3235  (abc)(?i:\g&#60;-1&#62;)
3236</pre>
3237Note that \g{...} (Perl syntax) and \g&#60;...&#62; (Oniguruma syntax) are <i>not</i>
3238synonymous. The former is a backreference; the latter is a subroutine call.
3239</P>
3240<br><a name="SEC28" href="#TOC1">CALLOUTS</a><br>
3241<P>
3242Perl has a feature whereby using the sequence (?{...}) causes arbitrary Perl
3243code to be obeyed in the middle of matching a regular expression. This makes it
3244possible, amongst other things, to extract different substrings that match the
3245same pair of parentheses when there is a repetition.
3246</P>
3247<P>
3248PCRE2 provides a similar feature, but of course it cannot obey arbitrary Perl
3249code. The feature is called "callout". The caller of PCRE2 provides an external
3250function by putting its entry point in a match context using the function
3251<b>pcre2_set_callout()</b>, and then passing that context to <b>pcre2_match()</b>
3252or <b>pcre2_dfa_match()</b>. If no match context is passed, or if the callout
3253entry point is set to NULL, callouts are disabled.
3254</P>
3255<P>
3256Within a regular expression, (?C&#60;arg&#62;) indicates a point at which the external
3257function is to be called. There are two kinds of callout: those with a
3258numerical argument and those with a string argument. (?C) on its own with no
3259argument is treated as (?C0). A numerical argument allows the application to
3260distinguish between different callouts. String arguments were added for release
326110.20 to make it possible for script languages that use PCRE2 to embed short
3262scripts within patterns in a similar way to Perl.
3263</P>
3264<P>
3265During matching, when PCRE2 reaches a callout point, the external function is
3266called. It is provided with the number or string argument of the callout, the
3267position in the pattern, and one item of data that is also set in the match
3268block. The callout function may cause matching to proceed, to backtrack, or to
3269fail.
3270</P>
3271<P>
3272By default, PCRE2 implements a number of optimizations at matching time, and
3273one side-effect is that sometimes callouts are skipped. If you need all
3274possible callouts to happen, you need to set options that disable the relevant
3275optimizations. More details, including a complete description of the
3276programming interface to the callout function, are given in the
3277<a href="pcre2callout.html"><b>pcre2callout</b></a>
3278documentation.
3279</P>
3280<br><b>
3281Callouts with numerical arguments
3282</b><br>
3283<P>
3284If you just want to have a means of identifying different callout points, put a
3285number less than 256 after the letter C. For example, this pattern has two
3286callout points:
3287<pre>
3288  (?C1)abc(?C2)def
3289</pre>
3290If the PCRE2_AUTO_CALLOUT flag is passed to <b>pcre2_compile()</b>, numerical
3291callouts are automatically installed before each item in the pattern. They are
3292all numbered 255. If there is a conditional group in the pattern whose
3293condition is an assertion, an additional callout is inserted just before the
3294condition. An explicit callout may also be set at this position, as in this
3295example:
3296<pre>
3297  (?(?C9)(?=a)abc|def)
3298</pre>
3299Note that this applies only to assertion conditions, not to other types of
3300condition.
3301</P>
3302<br><b>
3303Callouts with string arguments
3304</b><br>
3305<P>
3306A delimited string may be used instead of a number as a callout argument. The
3307starting delimiter must be one of ` ' " ^ % # $ { and the ending delimiter is
3308the same as the start, except for {, where the ending delimiter is }. If the
3309ending delimiter is needed within the string, it must be doubled. For
3310example:
3311<pre>
3312  (?C'ab ''c'' d')xyz(?C{any text})pqr
3313</pre>
3314The doubling is removed before the string is passed to the callout function.
3315<a name="backtrackcontrol"></a></P>
3316<br><a name="SEC29" href="#TOC1">BACKTRACKING CONTROL</a><br>
3317<P>
3318There are a number of special "Backtracking Control Verbs" (to use Perl's
3319terminology) that modify the behaviour of backtracking during matching. They
3320are generally of the form (*VERB) or (*VERB:NAME). Some verbs take either form,
3321and may behave differently depending on whether or not a name argument is
3322present. The names are not required to be unique within the pattern.
3323</P>
3324<P>
3325By default, for compatibility with Perl, a name is any sequence of characters
3326that does not include a closing parenthesis. The name is not processed in
3327any way, and it is not possible to include a closing parenthesis in the name.
3328This can be changed by setting the PCRE2_ALT_VERBNAMES option, but the result
3329is no longer Perl-compatible.
3330</P>
3331<P>
3332When PCRE2_ALT_VERBNAMES is set, backslash processing is applied to verb names
3333and only an unescaped closing parenthesis terminates the name. However, the
3334only backslash items that are permitted are \Q, \E, and sequences such as
3335\x{100} that define character code points. Character type escapes such as \d
3336are faulted.
3337</P>
3338<P>
3339A closing parenthesis can be included in a name either as \) or between \Q
3340and \E. In addition to backslash processing, if the PCRE2_EXTENDED or
3341PCRE2_EXTENDED_MORE option is also set, unescaped whitespace in verb names is
3342skipped, and #-comments are recognized, exactly as in the rest of the pattern.
3343PCRE2_EXTENDED and PCRE2_EXTENDED_MORE do not affect verb names unless
3344PCRE2_ALT_VERBNAMES is also set.
3345</P>
3346<P>
3347The maximum length of a name is 255 in the 8-bit library and 65535 in the
334816-bit and 32-bit libraries. If the name is empty, that is, if the closing
3349parenthesis immediately follows the colon, the effect is as if the colon were
3350not there. Any number of these verbs may occur in a pattern. Except for
3351(*ACCEPT), they may not be quantified.
3352</P>
3353<P>
3354Since these verbs are specifically related to backtracking, most of them can be
3355used only when the pattern is to be matched using the traditional matching
3356function, because that uses a backtracking algorithm. With the exception of
3357(*FAIL), which behaves like a failing negative assertion, the backtracking
3358control verbs cause an error if encountered by the DFA matching function.
3359</P>
3360<P>
3361The behaviour of these verbs in
3362<a href="#btrepeat">repeated groups,</a>
3363<a href="#btassert">assertions,</a>
3364and in
3365<a href="#btsub">capture groups called as subroutines</a>
3366(whether or not recursively) is documented below.
3367<a name="nooptimize"></a></P>
3368<br><b>
3369Optimizations that affect backtracking verbs
3370</b><br>
3371<P>
3372PCRE2 contains some optimizations that are used to speed up matching by running
3373some checks at the start of each match attempt. For example, it may know the
3374minimum length of matching subject, or that a particular character must be
3375present. When one of these optimizations bypasses the running of a match, any
3376included backtracking verbs will not, of course, be processed. You can suppress
3377the start-of-match optimizations by setting the PCRE2_NO_START_OPTIMIZE option
3378when calling <b>pcre2_compile()</b>, or by starting the pattern with
3379(*NO_START_OPT). There is more discussion of this option in the section
3380entitled
3381<a href="pcre2api.html#compiling">"Compiling a pattern"</a>
3382in the
3383<a href="pcre2api.html"><b>pcre2api</b></a>
3384documentation.
3385</P>
3386<P>
3387Experiments with Perl suggest that it too has similar optimizations, and like
3388PCRE2, turning them off can change the result of a match.
3389<a name="acceptverb"></a></P>
3390<br><b>
3391Verbs that act immediately
3392</b><br>
3393<P>
3394The following verbs act as soon as they are encountered.
3395<pre>
3396   (*ACCEPT) or (*ACCEPT:NAME)
3397</pre>
3398This verb causes the match to end successfully, skipping the remainder of the
3399pattern. However, when it is inside a capture group that is called as a
3400subroutine, only that group is ended successfully. Matching then continues
3401at the outer level. If (*ACCEPT) in triggered in a positive assertion, the
3402assertion succeeds; in a negative assertion, the assertion fails.
3403</P>
3404<P>
3405If (*ACCEPT) is inside capturing parentheses, the data so far is captured. For
3406example:
3407<pre>
3408  A((?:A|B(*ACCEPT)|C)D)
3409</pre>
3410This matches "AB", "AAD", or "ACD"; when it matches "AB", "B" is captured by
3411the outer parentheses.
3412</P>
3413<P>
3414(*ACCEPT) is the only backtracking verb that is allowed to be quantified
3415because an ungreedy quantification with a minimum of zero acts only when a
3416backtrack happens. Consider, for example,
3417<pre>
3418  (A(*ACCEPT)??B)C
3419</pre>
3420where A, B, and C may be complex expressions. After matching "A", the matcher
3421processes "BC"; if that fails, causing a backtrack, (*ACCEPT) is triggered and
3422the match succeeds. In both cases, all but C is captured. Whereas (*COMMIT)
3423(see below) means "fail on backtrack", a repeated (*ACCEPT) of this type means
3424"succeed on backtrack".
3425</P>
3426<P>
3427<b>Warning:</b> (*ACCEPT) should not be used within a script run group, because
3428it causes an immediate exit from the group, bypassing the script run checking.
3429<pre>
3430  (*FAIL) or (*FAIL:NAME)
3431</pre>
3432This verb causes a matching failure, forcing backtracking to occur. It may be
3433abbreviated to (*F). It is equivalent to (?!) but easier to read. The Perl
3434documentation notes that it is probably useful only when combined with (?{}) or
3435(??{}). Those are, of course, Perl features that are not present in PCRE2. The
3436nearest equivalent is the callout feature, as for example in this pattern:
3437<pre>
3438  a+(?C)(*FAIL)
3439</pre>
3440A match with the string "aaaa" always fails, but the callout is taken before
3441each backtrack happens (in this example, 10 times).
3442</P>
3443<P>
3444(*ACCEPT:NAME) and (*FAIL:NAME) behave the same as (*MARK:NAME)(*ACCEPT) and
3445(*MARK:NAME)(*FAIL), respectively, that is, a (*MARK) is recorded just before
3446the verb acts.
3447</P>
3448<br><b>
3449Recording which path was taken
3450</b><br>
3451<P>
3452There is one verb whose main purpose is to track how a match was arrived at,
3453though it also has a secondary use in conjunction with advancing the match
3454starting point (see (*SKIP) below).
3455<pre>
3456  (*MARK:NAME) or (*:NAME)
3457</pre>
3458A name is always required with this verb. For all the other backtracking
3459control verbs, a NAME argument is optional.
3460</P>
3461<P>
3462When a match succeeds, the name of the last-encountered mark name on the
3463matching path is passed back to the caller as described in the section entitled
3464<a href="pcre2api.html#matchotherdata">"Other information about the match"</a>
3465in the
3466<a href="pcre2api.html"><b>pcre2api</b></a>
3467documentation. This applies to all instances of (*MARK) and other verbs,
3468including those inside assertions and atomic groups. However, there are
3469differences in those cases when (*MARK) is used in conjunction with (*SKIP) as
3470described below.
3471</P>
3472<P>
3473The mark name that was last encountered on the matching path is passed back. A
3474verb without a NAME argument is ignored for this purpose. Here is an example of
3475<b>pcre2test</b> output, where the "mark" modifier requests the retrieval and
3476outputting of (*MARK) data:
3477<pre>
3478    re&#62; /X(*MARK:A)Y|X(*MARK:B)Z/mark
3479  data&#62; XY
3480   0: XY
3481  MK: A
3482  XZ
3483   0: XZ
3484  MK: B
3485</pre>
3486The (*MARK) name is tagged with "MK:" in this output, and in this example it
3487indicates which of the two alternatives matched. This is a more efficient way
3488of obtaining this information than putting each alternative in its own
3489capturing parentheses.
3490</P>
3491<P>
3492If a verb with a name is encountered in a positive assertion that is true, the
3493name is recorded and passed back if it is the last-encountered. This does not
3494happen for negative assertions or failing positive assertions.
3495</P>
3496<P>
3497After a partial match or a failed match, the last encountered name in the
3498entire match process is returned. For example:
3499<pre>
3500    re&#62; /X(*MARK:A)Y|X(*MARK:B)Z/mark
3501  data&#62; XP
3502  No match, mark = B
3503</pre>
3504Note that in this unanchored example the mark is retained from the match
3505attempt that started at the letter "X" in the subject. Subsequent match
3506attempts starting at "P" and then with an empty string do not get as far as the
3507(*MARK) item, but nevertheless do not reset it.
3508</P>
3509<P>
3510If you are interested in (*MARK) values after failed matches, you should
3511probably set the PCRE2_NO_START_OPTIMIZE option
3512<a href="#nooptimize">(see above)</a>
3513to ensure that the match is always attempted.
3514</P>
3515<br><b>
3516Verbs that act after backtracking
3517</b><br>
3518<P>
3519The following verbs do nothing when they are encountered. Matching continues
3520with what follows, but if there is a subsequent match failure, causing a
3521backtrack to the verb, a failure is forced. That is, backtracking cannot pass
3522to the left of the verb. However, when one of these verbs appears inside an
3523atomic group or in a lookaround assertion that is true, its effect is confined
3524to that group, because once the group has been matched, there is never any
3525backtracking into it. Backtracking from beyond an assertion or an atomic group
3526ignores the entire group, and seeks a preceding backtracking point.
3527</P>
3528<P>
3529These verbs differ in exactly what kind of failure occurs when backtracking
3530reaches them. The behaviour described below is what happens when the verb is
3531not in a subroutine or an assertion. Subsequent sections cover these special
3532cases.
3533<pre>
3534  (*COMMIT) or (*COMMIT:NAME)
3535</pre>
3536This verb causes the whole match to fail outright if there is a later matching
3537failure that causes backtracking to reach it. Even if the pattern is
3538unanchored, no further attempts to find a match by advancing the starting point
3539take place. If (*COMMIT) is the only backtracking verb that is encountered,
3540once it has been passed <b>pcre2_match()</b> is committed to finding a match at
3541the current starting point, or not at all. For example:
3542<pre>
3543  a+(*COMMIT)b
3544</pre>
3545This matches "xxaab" but not "aacaab". It can be thought of as a kind of
3546dynamic anchor, or "I've started, so I must finish."
3547</P>
3548<P>
3549The behaviour of (*COMMIT:NAME) is not the same as (*MARK:NAME)(*COMMIT). It is
3550like (*MARK:NAME) in that the name is remembered for passing back to the
3551caller. However, (*SKIP:NAME) searches only for names that are set with
3552(*MARK), ignoring those set by any of the other backtracking verbs.
3553</P>
3554<P>
3555If there is more than one backtracking verb in a pattern, a different one that
3556follows (*COMMIT) may be triggered first, so merely passing (*COMMIT) during a
3557match does not always guarantee that a match must be at this starting point.
3558</P>
3559<P>
3560Note that (*COMMIT) at the start of a pattern is not the same as an anchor,
3561unless PCRE2's start-of-match optimizations are turned off, as shown in this
3562output from <b>pcre2test</b>:
3563<pre>
3564    re&#62; /(*COMMIT)abc/
3565  data&#62; xyzabc
3566   0: abc
3567  data&#62;
3568  re&#62; /(*COMMIT)abc/no_start_optimize
3569  data&#62; xyzabc
3570  No match
3571</pre>
3572For the first pattern, PCRE2 knows that any match must start with "a", so the
3573optimization skips along the subject to "a" before applying the pattern to the
3574first set of data. The match attempt then succeeds. The second pattern disables
3575the optimization that skips along to the first character. The pattern is now
3576applied starting at "x", and so the (*COMMIT) causes the match to fail without
3577trying any other starting points.
3578<pre>
3579  (*PRUNE) or (*PRUNE:NAME)
3580</pre>
3581This verb causes the match to fail at the current starting position in the
3582subject if there is a later matching failure that causes backtracking to reach
3583it. If the pattern is unanchored, the normal "bumpalong" advance to the next
3584starting character then happens. Backtracking can occur as usual to the left of
3585(*PRUNE), before it is reached, or when matching to the right of (*PRUNE), but
3586if there is no match to the right, backtracking cannot cross (*PRUNE). In
3587simple cases, the use of (*PRUNE) is just an alternative to an atomic group or
3588possessive quantifier, but there are some uses of (*PRUNE) that cannot be
3589expressed in any other way. In an anchored pattern (*PRUNE) has the same effect
3590as (*COMMIT).
3591</P>
3592<P>
3593The behaviour of (*PRUNE:NAME) is not the same as (*MARK:NAME)(*PRUNE). It is
3594like (*MARK:NAME) in that the name is remembered for passing back to the
3595caller. However, (*SKIP:NAME) searches only for names set with (*MARK),
3596ignoring those set by other backtracking verbs.
3597<pre>
3598  (*SKIP)
3599</pre>
3600This verb, when given without a name, is like (*PRUNE), except that if the
3601pattern is unanchored, the "bumpalong" advance is not to the next character,
3602but to the position in the subject where (*SKIP) was encountered. (*SKIP)
3603signifies that whatever text was matched leading up to it cannot be part of a
3604successful match if there is a later mismatch. Consider:
3605<pre>
3606  a+(*SKIP)b
3607</pre>
3608If the subject is "aaaac...", after the first match attempt fails (starting at
3609the first character in the string), the starting point skips on to start the
3610next attempt at "c". Note that a possessive quantifer does not have the same
3611effect as this example; although it would suppress backtracking during the
3612first match attempt, the second attempt would start at the second character
3613instead of skipping on to "c".
3614</P>
3615<P>
3616If (*SKIP) is used to specify a new starting position that is the same as the
3617starting position of the current match, or (by being inside a lookbehind)
3618earlier, the position specified by (*SKIP) is ignored, and instead the normal
3619"bumpalong" occurs.
3620<pre>
3621  (*SKIP:NAME)
3622</pre>
3623When (*SKIP) has an associated name, its behaviour is modified. When such a
3624(*SKIP) is triggered, the previous path through the pattern is searched for the
3625most recent (*MARK) that has the same name. If one is found, the "bumpalong"
3626advance is to the subject position that corresponds to that (*MARK) instead of
3627to where (*SKIP) was encountered. If no (*MARK) with a matching name is found,
3628the (*SKIP) is ignored.
3629</P>
3630<P>
3631The search for a (*MARK) name uses the normal backtracking mechanism, which
3632means that it does not see (*MARK) settings that are inside atomic groups or
3633assertions, because they are never re-entered by backtracking. Compare the
3634following <b>pcre2test</b> examples:
3635<pre>
3636    re&#62; /a(?&#62;(*MARK:X))(*SKIP:X)(*F)|(.)/
3637  data: abc
3638   0: a
3639   1: a
3640  data:
3641    re&#62; /a(?:(*MARK:X))(*SKIP:X)(*F)|(.)/
3642  data: abc
3643   0: b
3644   1: b
3645</pre>
3646In the first example, the (*MARK) setting is in an atomic group, so it is not
3647seen when (*SKIP:X) triggers, causing the (*SKIP) to be ignored. This allows
3648the second branch of the pattern to be tried at the first character position.
3649In the second example, the (*MARK) setting is not in an atomic group. This
3650allows (*SKIP:X) to find the (*MARK) when it backtracks, and this causes a new
3651matching attempt to start at the second character. This time, the (*MARK) is
3652never seen because "a" does not match "b", so the matcher immediately jumps to
3653the second branch of the pattern.
3654</P>
3655<P>
3656Note that (*SKIP:NAME) searches only for names set by (*MARK:NAME). It ignores
3657names that are set by other backtracking verbs.
3658<pre>
3659  (*THEN) or (*THEN:NAME)
3660</pre>
3661This verb causes a skip to the next innermost alternative when backtracking
3662reaches it. That is, it cancels any further backtracking within the current
3663alternative. Its name comes from the observation that it can be used for a
3664pattern-based if-then-else block:
3665<pre>
3666  ( COND1 (*THEN) FOO | COND2 (*THEN) BAR | COND3 (*THEN) BAZ ) ...
3667</pre>
3668If the COND1 pattern matches, FOO is tried (and possibly further items after
3669the end of the group if FOO succeeds); on failure, the matcher skips to the
3670second alternative and tries COND2, without backtracking into COND1. If that
3671succeeds and BAR fails, COND3 is tried. If subsequently BAZ fails, there are no
3672more alternatives, so there is a backtrack to whatever came before the entire
3673group. If (*THEN) is not inside an alternation, it acts like (*PRUNE).
3674</P>
3675<P>
3676The behaviour of (*THEN:NAME) is not the same as (*MARK:NAME)(*THEN). It is
3677like (*MARK:NAME) in that the name is remembered for passing back to the
3678caller. However, (*SKIP:NAME) searches only for names set with (*MARK),
3679ignoring those set by other backtracking verbs.
3680</P>
3681<P>
3682A group that does not contain a | character is just a part of the enclosing
3683alternative; it is not a nested alternation with only one alternative. The
3684effect of (*THEN) extends beyond such a group to the enclosing alternative.
3685Consider this pattern, where A, B, etc. are complex pattern fragments that do
3686not contain any | characters at this level:
3687<pre>
3688  A (B(*THEN)C) | D
3689</pre>
3690If A and B are matched, but there is a failure in C, matching does not
3691backtrack into A; instead it moves to the next alternative, that is, D.
3692However, if the group containing (*THEN) is given an alternative, it
3693behaves differently:
3694<pre>
3695  A (B(*THEN)C | (*FAIL)) | D
3696</pre>
3697The effect of (*THEN) is now confined to the inner group. After a failure in C,
3698matching moves to (*FAIL), which causes the whole group to fail because there
3699are no more alternatives to try. In this case, matching does backtrack into A.
3700</P>
3701<P>
3702Note that a conditional group is not considered as having two alternatives,
3703because only one is ever used. In other words, the | character in a conditional
3704group has a different meaning. Ignoring white space, consider:
3705<pre>
3706  ^.*? (?(?=a) a | b(*THEN)c )
3707</pre>
3708If the subject is "ba", this pattern does not match. Because .*? is ungreedy,
3709it initially matches zero characters. The condition (?=a) then fails, the
3710character "b" is matched, but "c" is not. At this point, matching does not
3711backtrack to .*? as might perhaps be expected from the presence of the |
3712character. The conditional group is part of the single alternative that
3713comprises the whole pattern, and so the match fails. (If there was a backtrack
3714into .*?, allowing it to match "b", the match would succeed.)
3715</P>
3716<P>
3717The verbs just described provide four different "strengths" of control when
3718subsequent matching fails. (*THEN) is the weakest, carrying on the match at the
3719next alternative. (*PRUNE) comes next, failing the match at the current
3720starting position, but allowing an advance to the next character (for an
3721unanchored pattern). (*SKIP) is similar, except that the advance may be more
3722than one character. (*COMMIT) is the strongest, causing the entire match to
3723fail.
3724</P>
3725<br><b>
3726More than one backtracking verb
3727</b><br>
3728<P>
3729If more than one backtracking verb is present in a pattern, the one that is
3730backtracked onto first acts. For example, consider this pattern, where A, B,
3731etc. are complex pattern fragments:
3732<pre>
3733  (A(*COMMIT)B(*THEN)C|ABD)
3734</pre>
3735If A matches but B fails, the backtrack to (*COMMIT) causes the entire match to
3736fail. However, if A and B match, but C fails, the backtrack to (*THEN) causes
3737the next alternative (ABD) to be tried. This behaviour is consistent, but is
3738not always the same as Perl's. It means that if two or more backtracking verbs
3739appear in succession, all the the last of them has no effect. Consider this
3740example:
3741<pre>
3742  ...(*COMMIT)(*PRUNE)...
3743</pre>
3744If there is a matching failure to the right, backtracking onto (*PRUNE) causes
3745it to be triggered, and its action is taken. There can never be a backtrack
3746onto (*COMMIT).
3747<a name="btrepeat"></a></P>
3748<br><b>
3749Backtracking verbs in repeated groups
3750</b><br>
3751<P>
3752PCRE2 sometimes differs from Perl in its handling of backtracking verbs in
3753repeated groups. For example, consider:
3754<pre>
3755  /(a(*COMMIT)b)+ac/
3756</pre>
3757If the subject is "abac", Perl matches unless its optimizations are disabled,
3758but PCRE2 always fails because the (*COMMIT) in the second repeat of the group
3759acts.
3760<a name="btassert"></a></P>
3761<br><b>
3762Backtracking verbs in assertions
3763</b><br>
3764<P>
3765(*FAIL) in any assertion has its normal effect: it forces an immediate
3766backtrack. The behaviour of the other backtracking verbs depends on whether or
3767not the assertion is standalone or acting as the condition in a conditional
3768group.
3769</P>
3770<P>
3771(*ACCEPT) in a standalone positive assertion causes the assertion to succeed
3772without any further processing; captured strings and a mark name (if set) are
3773retained. In a standalone negative assertion, (*ACCEPT) causes the assertion to
3774fail without any further processing; captured substrings and any mark name are
3775discarded.
3776</P>
3777<P>
3778If the assertion is a condition, (*ACCEPT) causes the condition to be true for
3779a positive assertion and false for a negative one; captured substrings are
3780retained in both cases.
3781</P>
3782<P>
3783The remaining verbs act only when a later failure causes a backtrack to
3784reach them. This means that, for the Perl-compatible assertions, their effect
3785is confined to the assertion, because Perl lookaround assertions are atomic. A
3786backtrack that occurs after such an assertion is complete does not jump back
3787into the assertion. Note in particular that a (*MARK) name that is set in an
3788assertion is not "seen" by an instance of (*SKIP:NAME) later in the pattern.
3789</P>
3790<P>
3791PCRE2 now supports non-atomic positive assertions, as described in the section
3792entitled
3793<a href="#nonatomicassertions">"Non-atomic assertions"</a>
3794above. These assertions must be standalone (not used as conditions). They are
3795not Perl-compatible. For these assertions, a later backtrack does jump back
3796into the assertion, and therefore verbs such as (*COMMIT) can be triggered by
3797backtracks from later in the pattern.
3798</P>
3799<P>
3800The effect of (*THEN) is not allowed to escape beyond an assertion. If there
3801are no more branches to try, (*THEN) causes a positive assertion to be false,
3802and a negative assertion to be true.
3803</P>
3804<P>
3805The other backtracking verbs are not treated specially if they appear in a
3806standalone positive assertion. In a conditional positive assertion,
3807backtracking (from within the assertion) into (*COMMIT), (*SKIP), or (*PRUNE)
3808causes the condition to be false. However, for both standalone and conditional
3809negative assertions, backtracking into (*COMMIT), (*SKIP), or (*PRUNE) causes
3810the assertion to be true, without considering any further alternative branches.
3811<a name="btsub"></a></P>
3812<br><b>
3813Backtracking verbs in subroutines
3814</b><br>
3815<P>
3816These behaviours occur whether or not the group is called recursively.
3817</P>
3818<P>
3819(*ACCEPT) in a group called as a subroutine causes the subroutine match to
3820succeed without any further processing. Matching then continues after the
3821subroutine call. Perl documents this behaviour. Perl's treatment of the other
3822verbs in subroutines is different in some cases.
3823</P>
3824<P>
3825(*FAIL) in a group called as a subroutine has its normal effect: it forces
3826an immediate backtrack.
3827</P>
3828<P>
3829(*COMMIT), (*SKIP), and (*PRUNE) cause the subroutine match to fail when
3830triggered by being backtracked to in a group called as a subroutine. There is
3831then a backtrack at the outer level.
3832</P>
3833<P>
3834(*THEN), when triggered, skips to the next alternative in the innermost
3835enclosing group that has alternatives (its normal behaviour). However, if there
3836is no such group within the subroutine's group, the subroutine match fails and
3837there is a backtrack at the outer level.
3838</P>
3839<br><a name="SEC30" href="#TOC1">SEE ALSO</a><br>
3840<P>
3841<b>pcre2api</b>(3), <b>pcre2callout</b>(3), <b>pcre2matching</b>(3),
3842<b>pcre2syntax</b>(3), <b>pcre2</b>(3).
3843</P>
3844<br><a name="SEC31" href="#TOC1">AUTHOR</a><br>
3845<P>
3846Philip Hazel
3847<br>
3848University Computing Service
3849<br>
3850Cambridge, England.
3851<br>
3852</P>
3853<br><a name="SEC32" href="#TOC1">REVISION</a><br>
3854<P>
3855Last updated: 06 October 2020
3856<br>
3857Copyright &copy; 1997-2020 University of Cambridge.
3858<br>
3859<p>
3860Return to the <a href="index.html">PCRE2 index page</a>.
3861</p>
3862