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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 53 - (hide annotations) (download)
Sat Feb 24 21:39:42 2007 UTC (7 years, 1 month ago) by nigel
File size: 89461 byte(s)
Load pcre-3.5 into code/trunk.

1 nigel 41 .TH PCRE 3
2     .SH NAME
3     pcre - Perl-compatible regular expressions.
4     .SH SYNOPSIS
5     .B #include <pcre.h>
6     .PP
7     .SM
8     .br
9     .B pcre *pcre_compile(const char *\fIpattern\fR, int \fIoptions\fR,
10     .ti +5n
11     .B const char **\fIerrptr\fR, int *\fIerroffset\fR,
12     .ti +5n
13     .B const unsigned char *\fItableptr\fR);
14     .PP
15     .br
16     .B pcre_extra *pcre_study(const pcre *\fIcode\fR, int \fIoptions\fR,
17     .ti +5n
18     .B const char **\fIerrptr\fR);
19     .PP
20     .br
21     .B int pcre_exec(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"
22     .ti +5n
23     .B "const char *\fIsubject\fR," int \fIlength\fR, int \fIstartoffset\fR,
24     .ti +5n
25     .B int \fIoptions\fR, int *\fIovector\fR, int \fIovecsize\fR);
26     .PP
27     .br
28     .B int pcre_copy_substring(const char *\fIsubject\fR, int *\fIovector\fR,
29     .ti +5n
30     .B int \fIstringcount\fR, int \fIstringnumber\fR, char *\fIbuffer\fR,
31     .ti +5n
32     .B int \fIbuffersize\fR);
33     .PP
34     .br
35     .B int pcre_get_substring(const char *\fIsubject\fR, int *\fIovector\fR,
36     .ti +5n
37     .B int \fIstringcount\fR, int \fIstringnumber\fR,
38     .ti +5n
39     .B const char **\fIstringptr\fR);
40     .PP
41     .br
42     .B int pcre_get_substring_list(const char *\fIsubject\fR,
43     .ti +5n
44     .B int *\fIovector\fR, int \fIstringcount\fR, "const char ***\fIlistptr\fR);"
45     .PP
46     .br
47 nigel 49 .B void pcre_free_substring(const char *\fIstringptr\fR);
48     .PP
49     .br
50     .B void pcre_free_substring_list(const char **\fIstringptr\fR);
51     .PP
52     .br
53 nigel 41 .B const unsigned char *pcre_maketables(void);
54     .PP
55     .br
56 nigel 43 .B int pcre_fullinfo(const pcre *\fIcode\fR, "const pcre_extra *\fIextra\fR,"
57     .ti +5n
58     .B int \fIwhat\fR, void *\fIwhere\fR);
59     .PP
60     .br
61 nigel 41 .B int pcre_info(const pcre *\fIcode\fR, int *\fIoptptr\fR, int
62     .B *\fIfirstcharptr\fR);
63     .PP
64     .br
65     .B char *pcre_version(void);
66     .PP
67     .br
68     .B void *(*pcre_malloc)(size_t);
69     .PP
70     .br
71     .B void (*pcre_free)(void *);
72    
73    
74    
75     .SH DESCRIPTION
76     The PCRE library is a set of functions that implement regular expression
77     pattern matching using the same syntax and semantics as Perl 5, with just a few
78 nigel 43 differences (see below). The current implementation corresponds to Perl 5.005,
79 nigel 49 with some additional features from later versions. This includes some
80     experimental, incomplete support for UTF-8 encoded strings. Details of exactly
81     what is and what is not supported are given below.
82 nigel 41
83     PCRE has its own native API, which is described in this document. There is also
84 nigel 43 a set of wrapper functions that correspond to the POSIX regular expression API.
85     These are described in the \fBpcreposix\fR documentation.
86 nigel 41
87     The native API function prototypes are defined in the header file \fBpcre.h\fR,
88     and on Unix systems the library itself is called \fBlibpcre.a\fR, so can be
89     accessed by adding \fB-lpcre\fR to the command for linking an application which
90 nigel 43 calls it. The header file defines the macros PCRE_MAJOR and PCRE_MINOR to
91     contain the major and minor release numbers for the library. Applications can
92     use these to include support for different releases.
93 nigel 41
94     The functions \fBpcre_compile()\fR, \fBpcre_study()\fR, and \fBpcre_exec()\fR
95 nigel 53 are used for compiling and matching regular expressions. A sample program that
96     demonstrates the simplest way of using them is given in the file
97     \fIpcredemo.c\fR. The last section of this man page describes how to run it.
98 nigel 49
99     The functions \fBpcre_copy_substring()\fR, \fBpcre_get_substring()\fR, and
100 nigel 41 \fBpcre_get_substring_list()\fR are convenience functions for extracting
101 nigel 49 captured substrings from a matched subject string; \fBpcre_free_substring()\fR
102     and \fBpcre_free_substring_list()\fR are also provided, to free the memory used
103     for extracted strings.
104 nigel 41
105 nigel 49 The function \fBpcre_maketables()\fR is used (optionally) to build a set of
106     character tables in the current locale for passing to \fBpcre_compile()\fR.
107    
108 nigel 43 The function \fBpcre_fullinfo()\fR is used to find out information about a
109     compiled pattern; \fBpcre_info()\fR is an obsolete version which returns only
110     some of the available information, but is retained for backwards compatibility.
111     The function \fBpcre_version()\fR returns a pointer to a string containing the
112     version of PCRE and its date of release.
113 nigel 41
114     The global variables \fBpcre_malloc\fR and \fBpcre_free\fR initially contain
115     the entry points of the standard \fBmalloc()\fR and \fBfree()\fR functions
116     respectively. PCRE calls the memory management functions via these variables,
117     so a calling program can replace them if it wishes to intercept the calls. This
118     should be done before calling any PCRE functions.
119    
120    
121     .SH MULTI-THREADING
122     The PCRE functions can be used in multi-threading applications, with the
123     proviso that the memory management functions pointed to by \fBpcre_malloc\fR
124     and \fBpcre_free\fR are shared by all threads.
125    
126     The compiled form of a regular expression is not altered during matching, so
127     the same compiled pattern can safely be used by several threads at once.
128    
129    
130     .SH COMPILING A PATTERN
131     The function \fBpcre_compile()\fR is called to compile a pattern into an
132     internal form. The pattern is a C string terminated by a binary zero, and
133     is passed in the argument \fIpattern\fR. A pointer to a single block of memory
134 nigel 53 that is obtained via \fBpcre_malloc\fR is returned. This contains the compiled
135     code and related data. The \fBpcre\fR type is defined for the returned block;
136     this is a typedef for a structure whose contents are not externally defined. It
137     is up to the caller to free the memory when it is no longer required.
138    
139     Although the compiled code of a PCRE regex is relocatable, that is, it does not
140     depend on memory location, the complete \fBpcre\fR data block is not
141     fully relocatable, because it contains a copy of the \fItableptr\fR argument,
142     which is an address (see below).
143    
144 nigel 41 The size of a compiled pattern is roughly proportional to the length of the
145     pattern string, except that each character class (other than those containing
146     just a single character, negated or not) requires 33 bytes, and repeat
147     quantifiers with a minimum greater than one or a bounded maximum cause the
148     relevant portions of the compiled pattern to be replicated.
149 nigel 53
150 nigel 41 The \fIoptions\fR argument contains independent bits that affect the
151     compilation. It should be zero if no options are required. Some of the options,
152     in particular, those that are compatible with Perl, can also be set and unset
153     from within the pattern (see the detailed description of regular expressions
154     below). For these options, the contents of the \fIoptions\fR argument specifies
155     their initial settings at the start of compilation and execution. The
156     PCRE_ANCHORED option can be set at the time of matching as well as at compile
157     time.
158 nigel 53
159 nigel 41 If \fIerrptr\fR is NULL, \fBpcre_compile()\fR returns NULL immediately.
160     Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fR returns
161     NULL, and sets the variable pointed to by \fIerrptr\fR to point to a textual
162     error message. The offset from the start of the pattern to the character where
163     the error was discovered is placed in the variable pointed to by
164     \fIerroffset\fR, which must not be NULL. If it is, an immediate error is given.
165 nigel 53
166 nigel 41 If the final argument, \fItableptr\fR, is NULL, PCRE uses a default set of
167     character tables which are built when it is compiled, using the default C
168     locale. Otherwise, \fItableptr\fR must be the result of a call to
169     \fBpcre_maketables()\fR. See the section on locale support below.
170 nigel 53
171     This code fragment shows a typical straightforward call to \fBpcre_compile()\fR:
172    
173     pcre *re;
174     const char *error;
175     int erroffset;
176     re = pcre_compile(
177     "^A.*Z", /* the pattern */
178     0, /* default options */
179     &error, /* for error message */
180     &erroffset, /* for error offset */
181     NULL); /* use default character tables */
182    
183 nigel 41 The following option bits are defined in the header file:
184    
185     PCRE_ANCHORED
186    
187     If this bit is set, the pattern is forced to be "anchored", that is, it is
188     constrained to match only at the start of the string which is being searched
189     (the "subject string"). This effect can also be achieved by appropriate
190     constructs in the pattern itself, which is the only way to do it in Perl.
191    
192     PCRE_CASELESS
193    
194     If this bit is set, letters in the pattern match both upper and lower case
195     letters. It is equivalent to Perl's /i option.
196    
197     PCRE_DOLLAR_ENDONLY
198    
199     If this bit is set, a dollar metacharacter in the pattern matches only at the
200     end of the subject string. Without this option, a dollar also matches
201     immediately before the final character if it is a newline (but not before any
202     other newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is
203     set. There is no equivalent to this option in Perl.
204    
205     PCRE_DOTALL
206    
207     If this bit is set, a dot metacharater in the pattern matches all characters,
208     including newlines. Without it, newlines are excluded. This option is
209     equivalent to Perl's /s option. A negative class such as [^a] always matches a
210     newline character, independent of the setting of this option.
211    
212     PCRE_EXTENDED
213    
214     If this bit is set, whitespace data characters in the pattern are totally
215     ignored except when escaped or inside a character class, and characters between
216     an unescaped # outside a character class and the next newline character,
217     inclusive, are also ignored. This is equivalent to Perl's /x option, and makes
218     it possible to include comments inside complicated patterns. Note, however,
219     that this applies only to data characters. Whitespace characters may never
220     appear within special character sequences in a pattern, for example within the
221     sequence (?( which introduces a conditional subpattern.
222    
223     PCRE_EXTRA
224    
225 nigel 43 This option was invented in order to turn on additional functionality of PCRE
226     that is incompatible with Perl, but it is currently of very little use. When
227     set, any backslash in a pattern that is followed by a letter that has no
228 nigel 41 special meaning causes an error, thus reserving these combinations for future
229     expansion. By default, as in Perl, a backslash followed by a letter with no
230     special meaning is treated as a literal. There are at present no other features
231 nigel 43 controlled by this option. It can also be set by a (?X) option setting within a
232     pattern.
233 nigel 41
234     PCRE_MULTILINE
235    
236     By default, PCRE treats the subject string as consisting of a single "line" of
237     characters (even if it actually contains several newlines). The "start of line"
238     metacharacter (^) matches only at the start of the string, while the "end of
239     line" metacharacter ($) matches only at the end of the string, or before a
240     terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
241     Perl.
242    
243     When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
244     match immediately following or immediately before any newline in the subject
245     string, respectively, as well as at the very start and end. This is equivalent
246     to Perl's /m option. If there are no "\\n" characters in a subject string, or
247     no occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no
248     effect.
249    
250     PCRE_UNGREEDY
251    
252     This option inverts the "greediness" of the quantifiers so that they are not
253     greedy by default, but become greedy if followed by "?". It is not compatible
254     with Perl. It can also be set by a (?U) option setting within the pattern.
255    
256 nigel 49 PCRE_UTF8
257 nigel 41
258 nigel 49 This option causes PCRE to regard both the pattern and the subject as strings
259     of UTF-8 characters instead of just byte strings. However, it is available only
260     if PCRE has been built to include UTF-8 support. If not, the use of this option
261     provokes an error. Support for UTF-8 is new, experimental, and incomplete.
262     Details of exactly what it entails are given below.
263    
264    
265 nigel 41 .SH STUDYING A PATTERN
266     When a pattern is going to be used several times, it is worth spending more
267     time analyzing it in order to speed up the time taken for matching. The
268     function \fBpcre_study()\fR takes a pointer to a compiled pattern as its first
269 nigel 53 argument, and returns a pointer to a \fBpcre_extra\fR block (another typedef
270     for a structure with hidden contents) containing additional information about
271     the pattern; this can be passed to \fBpcre_exec()\fR. If no additional
272     information is available, NULL is returned.
273 nigel 41
274     The second argument contains option bits. At present, no options are defined
275     for \fBpcre_study()\fR, and this argument should always be zero.
276    
277     The third argument for \fBpcre_study()\fR is a pointer to an error message. If
278     studying succeeds (even if no data is returned), the variable it points to is
279     set to NULL. Otherwise it points to a textual error message.
280    
281 nigel 53 This is a typical call to \fBpcre_study\fR():
282    
283     pcre_extra *pe;
284     pe = pcre_study(
285     re, /* result of pcre_compile() */
286     0, /* no options exist */
287     &error); /* set to NULL or points to a message */
288    
289 nigel 41 At present, studying a pattern is useful only for non-anchored patterns that do
290     not have a single fixed starting character. A bitmap of possible starting
291     characters is created.
292    
293    
294     .SH LOCALE SUPPORT
295     PCRE handles caseless matching, and determines whether characters are letters,
296     digits, or whatever, by reference to a set of tables. The library contains a
297     default set of tables which is created in the default C locale when PCRE is
298     compiled. This is used when the final argument of \fBpcre_compile()\fR is NULL,
299     and is sufficient for many applications.
300    
301     An alternative set of tables can, however, be supplied. Such tables are built
302     by calling the \fBpcre_maketables()\fR function, which has no arguments, in the
303     relevant locale. The result can then be passed to \fBpcre_compile()\fR as often
304     as necessary. For example, to build and use tables that are appropriate for the
305     French locale (where accented characters with codes greater than 128 are
306     treated as letters), the following code could be used:
307    
308     setlocale(LC_CTYPE, "fr");
309     tables = pcre_maketables();
310     re = pcre_compile(..., tables);
311    
312     The tables are built in memory that is obtained via \fBpcre_malloc\fR. The
313     pointer that is passed to \fBpcre_compile\fR is saved with the compiled
314     pattern, and the same tables are used via this pointer by \fBpcre_study()\fR
315     and \fBpcre_exec()\fR. Thus for any single pattern, compilation, studying and
316     matching all happen in the same locale, but different patterns can be compiled
317     in different locales. It is the caller's responsibility to ensure that the
318     memory containing the tables remains available for as long as it is needed.
319    
320    
321     .SH INFORMATION ABOUT A PATTERN
322 nigel 43 The \fBpcre_fullinfo()\fR function returns information about a compiled
323     pattern. It replaces the obsolete \fBpcre_info()\fR function, which is
324     nevertheless retained for backwards compability (and is documented below).
325 nigel 41
326 nigel 43 The first argument for \fBpcre_fullinfo()\fR is a pointer to the compiled
327     pattern. The second argument is the result of \fBpcre_study()\fR, or NULL if
328     the pattern was not studied. The third argument specifies which piece of
329     information is required, while the fourth argument is a pointer to a variable
330     to receive the data. The yield of the function is zero for success, or one of
331     the following negative numbers:
332    
333 nigel 41 PCRE_ERROR_NULL the argument \fIcode\fR was NULL
334 nigel 43 the argument \fIwhere\fR was NULL
335 nigel 41 PCRE_ERROR_BADMAGIC the "magic number" was not found
336 nigel 43 PCRE_ERROR_BADOPTION the value of \fIwhat\fR was invalid
337 nigel 41
338 nigel 53 Here is a typical call of \fBpcre_fullinfo()\fR, to obtain the length of the
339     compiled pattern:
340    
341     int rc;
342     unsigned long int length;
343     rc = pcre_fullinfo(
344     re, /* result of pcre_compile() */
345     pe, /* result of pcre_study(), or NULL */
346     PCRE_INFO_SIZE, /* what is required */
347     &length); /* where to put the data */
348    
349 nigel 43 The possible values for the third argument are defined in \fBpcre.h\fR, and are
350     as follows:
351    
352     PCRE_INFO_OPTIONS
353    
354     Return a copy of the options with which the pattern was compiled. The fourth
355 nigel 53 argument should point to an \fBunsigned long int\fR variable. These option bits
356 nigel 41 are those specified in the call to \fBpcre_compile()\fR, modified by any
357     top-level option settings within the pattern itself, and with the PCRE_ANCHORED
358 nigel 43 bit forcibly set if the form of the pattern implies that it can match only at
359     the start of a subject string.
360 nigel 41
361 nigel 43 PCRE_INFO_SIZE
362 nigel 41
363 nigel 43 Return the size of the compiled pattern, that is, the value that was passed as
364     the argument to \fBpcre_malloc()\fR when PCRE was getting memory in which to
365     place the compiled data. The fourth argument should point to a \fBsize_t\fR
366     variable.
367    
368     PCRE_INFO_CAPTURECOUNT
369    
370     Return the number of capturing subpatterns in the pattern. The fourth argument
371     should point to an \fbint\fR variable.
372    
373     PCRE_INFO_BACKREFMAX
374    
375     Return the number of the highest back reference in the pattern. The fourth
376     argument should point to an \fBint\fR variable. Zero is returned if there are
377     no back references.
378    
379     PCRE_INFO_FIRSTCHAR
380    
381     Return information about the first character of any matched string, for a
382     non-anchored pattern. If there is a fixed first character, e.g. from a pattern
383 nigel 47 such as (cat|cow|coyote), it is returned in the integer pointed to by
384 nigel 43 \fIwhere\fR. Otherwise, if either
385    
386 nigel 41 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
387     starts with "^", or
388    
389     (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
390     (if it were set, the pattern would be anchored),
391    
392 nigel 47 -1 is returned, indicating that the pattern matches only at the start of a
393     subject string or after any "\\n" within the string. Otherwise -2 is returned.
394     For anchored patterns, -2 is returned.
395 nigel 41
396 nigel 43 PCRE_INFO_FIRSTTABLE
397 nigel 41
398 nigel 43 If the pattern was studied, and this resulted in the construction of a 256-bit
399     table indicating a fixed set of characters for the first character in any
400     matching string, a pointer to the table is returned. Otherwise NULL is
401     returned. The fourth argument should point to an \fBunsigned char *\fR
402     variable.
403    
404     PCRE_INFO_LASTLITERAL
405    
406     For a non-anchored pattern, return the value of the rightmost literal character
407     which must exist in any matched string, other than at its start. The fourth
408     argument should point to an \fBint\fR variable. If there is no such character,
409     or if the pattern is anchored, -1 is returned. For example, for the pattern
410     /a\\d+z\\d+/ the returned value is 'z'.
411    
412     The \fBpcre_info()\fR function is now obsolete because its interface is too
413     restrictive to return all the available data about a compiled pattern. New
414     programs should use \fBpcre_fullinfo()\fR instead. The yield of
415     \fBpcre_info()\fR is the number of capturing subpatterns, or one of the
416     following negative numbers:
417    
418     PCRE_ERROR_NULL the argument \fIcode\fR was NULL
419     PCRE_ERROR_BADMAGIC the "magic number" was not found
420    
421     If the \fIoptptr\fR argument is not NULL, a copy of the options with which the
422     pattern was compiled is placed in the integer it points to (see
423     PCRE_INFO_OPTIONS above).
424    
425     If the pattern is not anchored and the \fIfirstcharptr\fR argument is not NULL,
426     it is used to pass back information about the first character of any matched
427     string (see PCRE_INFO_FIRSTCHAR above).
428    
429    
430 nigel 41 .SH MATCHING A PATTERN
431     The function \fBpcre_exec()\fR is called to match a subject string against a
432     pre-compiled pattern, which is passed in the \fIcode\fR argument. If the
433     pattern has been studied, the result of the study should be passed in the
434     \fIextra\fR argument. Otherwise this must be NULL.
435    
436 nigel 53 Here is an example of a simple call to \fBpcre_exec()\fR:
437    
438     int rc;
439     int ovector[30];
440     rc = pcre_exec(
441     re, /* result of pcre_compile() */
442     NULL, /* we didn't study the pattern */
443     "some string", /* the subject string */
444     11, /* the length of the subject string */
445     0, /* start at offset 0 in the subject */
446     0, /* default options */
447     ovector, /* vector for substring information */
448     30); /* number of elements in the vector */
449    
450 nigel 41 The PCRE_ANCHORED option can be passed in the \fIoptions\fR argument, whose
451     unused bits must be zero. However, if a pattern was compiled with
452     PCRE_ANCHORED, or turned out to be anchored by virtue of its contents, it
453     cannot be made unachored at matching time.
454    
455     There are also three further options that can be set only at matching time:
456    
457     PCRE_NOTBOL
458    
459     The first character of the string is not the beginning of a line, so the
460     circumflex metacharacter should not match before it. Setting this without
461     PCRE_MULTILINE (at compile time) causes circumflex never to match.
462    
463     PCRE_NOTEOL
464    
465     The end of the string is not the end of a line, so the dollar metacharacter
466     should not match it nor (except in multiline mode) a newline immediately before
467     it. Setting this without PCRE_MULTILINE (at compile time) causes dollar never
468     to match.
469    
470     PCRE_NOTEMPTY
471    
472     An empty string is not considered to be a valid match if this option is set. If
473     there are alternatives in the pattern, they are tried. If all the alternatives
474     match the empty string, the entire match fails. For example, if the pattern
475    
476     a?b?
477    
478     is applied to a string not beginning with "a" or "b", it matches the empty
479     string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
480     valid, so PCRE searches further into the string for occurrences of "a" or "b".
481    
482     Perl has no direct equivalent of PCRE_NOTEMPTY, but it does make a special case
483     of a pattern match of the empty string within its \fBsplit()\fR function, and
484     when using the /g modifier. It is possible to emulate Perl's behaviour after
485     matching a null string by first trying the match again at the same offset with
486     PCRE_NOTEMPTY set, and then if that fails by advancing the starting offset (see
487     below) and trying an ordinary match again.
488    
489     The subject string is passed as a pointer in \fIsubject\fR, a length in
490     \fIlength\fR, and a starting offset in \fIstartoffset\fR. Unlike the pattern
491 nigel 53 string, the subject may contain binary zero characters. When the starting
492     offset is zero, the search for a match starts at the beginning of the subject,
493     and this is by far the most common case.
494 nigel 41
495     A non-zero starting offset is useful when searching for another match in the
496     same subject by calling \fBpcre_exec()\fR again after a previous success.
497     Setting \fIstartoffset\fR differs from just passing over a shortened string and
498     setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
499     lookbehind. For example, consider the pattern
500    
501     \\Biss\\B
502    
503     which finds occurrences of "iss" in the middle of words. (\\B matches only if
504     the current position in the subject is not a word boundary.) When applied to
505     the string "Mississipi" the first call to \fBpcre_exec()\fR finds the first
506     occurrence. If \fBpcre_exec()\fR is called again with just the remainder of the
507     subject, namely "issipi", it does not match, because \\B is always false at the
508     start of the subject, which is deemed to be a word boundary. However, if
509     \fBpcre_exec()\fR is passed the entire string again, but with \fIstartoffset\fR
510     set to 4, it finds the second occurrence of "iss" because it is able to look
511     behind the starting point to discover that it is preceded by a letter.
512    
513     If a non-zero starting offset is passed when the pattern is anchored, one
514     attempt to match at the given offset is tried. This can only succeed if the
515     pattern does not require the match to be at the start of the subject.
516    
517     In general, a pattern matches a certain portion of the subject, and in
518     addition, further substrings from the subject may be picked out by parts of the
519     pattern. Following the usage in Jeffrey Friedl's book, this is called
520     "capturing" in what follows, and the phrase "capturing subpattern" is used for
521     a fragment of a pattern that picks out a substring. PCRE supports several other
522     kinds of parenthesized subpattern that do not cause substrings to be captured.
523    
524     Captured substrings are returned to the caller via a vector of integer offsets
525     whose address is passed in \fIovector\fR. The number of elements in the vector
526     is passed in \fIovecsize\fR. The first two-thirds of the vector is used to pass
527     back captured substrings, each substring using a pair of integers. The
528     remaining third of the vector is used as workspace by \fBpcre_exec()\fR while
529     matching capturing subpatterns, and is not available for passing back
530     information. The length passed in \fIovecsize\fR should always be a multiple of
531     three. If it is not, it is rounded down.
532    
533     When a match has been successful, information about captured substrings is
534     returned in pairs of integers, starting at the beginning of \fIovector\fR, and
535     continuing up to two-thirds of its length at the most. The first element of a
536     pair is set to the offset of the first character in a substring, and the second
537     is set to the offset of the first character after the end of a substring. The
538     first pair, \fIovector[0]\fR and \fIovector[1]\fR, identify the portion of the
539     subject string matched by the entire pattern. The next pair is used for the
540     first capturing subpattern, and so on. The value returned by \fBpcre_exec()\fR
541     is the number of pairs that have been set. If there are no capturing
542     subpatterns, the return value from a successful match is 1, indicating that
543     just the first pair of offsets has been set.
544    
545     Some convenience functions are provided for extracting the captured substrings
546     as separate strings. These are described in the following section.
547    
548     It is possible for an capturing subpattern number \fIn+1\fR to match some
549     part of the subject when subpattern \fIn\fR has not been used at all. For
550     example, if the string "abc" is matched against the pattern (a|(z))(bc)
551     subpatterns 1 and 3 are matched, but 2 is not. When this happens, both offset
552     values corresponding to the unused subpattern are set to -1.
553    
554     If a capturing subpattern is matched repeatedly, it is the last portion of the
555     string that it matched that gets returned.
556    
557     If the vector is too small to hold all the captured substrings, it is used as
558     far as possible (up to two-thirds of its length), and the function returns a
559     value of zero. In particular, if the substring offsets are not of interest,
560     \fBpcre_exec()\fR may be called with \fIovector\fR passed as NULL and
561     \fIovecsize\fR as zero. However, if the pattern contains back references and
562     the \fIovector\fR isn't big enough to remember the related substrings, PCRE has
563     to get additional memory for use during matching. Thus it is usually advisable
564     to supply an \fIovector\fR.
565    
566     Note that \fBpcre_info()\fR can be used to find out how many capturing
567     subpatterns there are in a compiled pattern. The smallest size for
568     \fIovector\fR that will allow for \fIn\fR captured substrings in addition to
569     the offsets of the substring matched by the whole pattern is (\fIn\fR+1)*3.
570    
571     If \fBpcre_exec()\fR fails, it returns a negative number. The following are
572     defined in the header file:
573    
574     PCRE_ERROR_NOMATCH (-1)
575    
576     The subject string did not match the pattern.
577    
578     PCRE_ERROR_NULL (-2)
579    
580     Either \fIcode\fR or \fIsubject\fR was passed as NULL, or \fIovector\fR was
581     NULL and \fIovecsize\fR was not zero.
582    
583     PCRE_ERROR_BADOPTION (-3)
584    
585     An unrecognized bit was set in the \fIoptions\fR argument.
586    
587     PCRE_ERROR_BADMAGIC (-4)
588    
589     PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
590     the case when it is passed a junk pointer. This is the error it gives when the
591     magic number isn't present.
592    
593     PCRE_ERROR_UNKNOWN_NODE (-5)
594    
595     While running the pattern match, an unknown item was encountered in the
596     compiled pattern. This error could be caused by a bug in PCRE or by overwriting
597     of the compiled pattern.
598    
599     PCRE_ERROR_NOMEMORY (-6)
600    
601     If a pattern contains back references, but the \fIovector\fR that is passed to
602     \fBpcre_exec()\fR is not big enough to remember the referenced substrings, PCRE
603     gets a block of memory at the start of matching to use for this purpose. If the
604     call via \fBpcre_malloc()\fR fails, this error is given. The memory is freed at
605     the end of matching.
606    
607    
608     .SH EXTRACTING CAPTURED SUBSTRINGS
609     Captured substrings can be accessed directly by using the offsets returned by
610     \fBpcre_exec()\fR in \fIovector\fR. For convenience, the functions
611     \fBpcre_copy_substring()\fR, \fBpcre_get_substring()\fR, and
612     \fBpcre_get_substring_list()\fR are provided for extracting captured substrings
613     as new, separate, zero-terminated strings. A substring that contains a binary
614     zero is correctly extracted and has a further zero added on the end, but the
615     result does not, of course, function as a C string.
616    
617     The first three arguments are the same for all three functions: \fIsubject\fR
618     is the subject string which has just been successfully matched, \fIovector\fR
619     is a pointer to the vector of integer offsets that was passed to
620     \fBpcre_exec()\fR, and \fIstringcount\fR is the number of substrings that
621     were captured by the match, including the substring that matched the entire
622     regular expression. This is the value returned by \fBpcre_exec\fR if it
623     is greater than zero. If \fBpcre_exec()\fR returned zero, indicating that it
624 nigel 47 ran out of space in \fIovector\fR, the value passed as \fIstringcount\fR should
625     be the size of the vector divided by three.
626 nigel 41
627     The functions \fBpcre_copy_substring()\fR and \fBpcre_get_substring()\fR
628     extract a single substring, whose number is given as \fIstringnumber\fR. A
629     value of zero extracts the substring that matched the entire pattern, while
630     higher values extract the captured substrings. For \fBpcre_copy_substring()\fR,
631     the string is placed in \fIbuffer\fR, whose length is given by
632 nigel 49 \fIbuffersize\fR, while for \fBpcre_get_substring()\fR a new block of memory is
633 nigel 41 obtained via \fBpcre_malloc\fR, and its address is returned via
634     \fIstringptr\fR. The yield of the function is the length of the string, not
635     including the terminating zero, or one of
636    
637     PCRE_ERROR_NOMEMORY (-6)
638    
639     The buffer was too small for \fBpcre_copy_substring()\fR, or the attempt to get
640     memory failed for \fBpcre_get_substring()\fR.
641    
642     PCRE_ERROR_NOSUBSTRING (-7)
643    
644     There is no substring whose number is \fIstringnumber\fR.
645    
646     The \fBpcre_get_substring_list()\fR function extracts all available substrings
647     and builds a list of pointers to them. All this is done in a single block of
648     memory which is obtained via \fBpcre_malloc\fR. The address of the memory block
649     is returned via \fIlistptr\fR, which is also the start of the list of string
650     pointers. The end of the list is marked by a NULL pointer. The yield of the
651     function is zero if all went well, or
652    
653     PCRE_ERROR_NOMEMORY (-6)
654    
655     if the attempt to get the memory block failed.
656    
657     When any of these functions encounter a substring that is unset, which can
658     happen when capturing subpattern number \fIn+1\fR matches some part of the
659     subject, but subpattern \fIn\fR has not been used at all, they return an empty
660     string. This can be distinguished from a genuine zero-length substring by
661     inspecting the appropriate offset in \fIovector\fR, which is negative for unset
662     substrings.
663    
664 nigel 49 The two convenience functions \fBpcre_free_substring()\fR and
665     \fBpcre_free_substring_list()\fR can be used to free the memory returned by
666     a previous call of \fBpcre_get_substring()\fR or
667     \fBpcre_get_substring_list()\fR, respectively. They do nothing more than call
668     the function pointed to by \fBpcre_free\fR, which of course could be called
669     directly from a C program. However, PCRE is used in some situations where it is
670     linked via a special interface to another programming language which cannot use
671     \fBpcre_free\fR directly; it is for these cases that the functions are
672     provided.
673 nigel 41
674    
675     .SH LIMITATIONS
676     There are some size limitations in PCRE but it is hoped that they will never in
677     practice be relevant.
678     The maximum length of a compiled pattern is 65539 (sic) bytes.
679     All values in repeating quantifiers must be less than 65536.
680 nigel 53 There maximum number of capturing subpatterns is 65535.
681     There is no limit to the number of non-capturing subpatterns, but the maximum
682     depth of nesting of all kinds of parenthesized subpattern, including capturing
683 nigel 41 subpatterns, assertions, and other types of subpattern, is 200.
684    
685     The maximum length of a subject string is the largest positive number that an
686     integer variable can hold. However, PCRE uses recursion to handle subpatterns
687     and indefinite repetition. This means that the available stack space may limit
688     the size of a subject string that can be processed by certain patterns.
689    
690    
691     .SH DIFFERENCES FROM PERL
692     The differences described here are with respect to Perl 5.005.
693    
694     1. By default, a whitespace character is any character that the C library
695     function \fBisspace()\fR recognizes, though it is possible to compile PCRE with
696     alternative character type tables. Normally \fBisspace()\fR matches space,
697     formfeed, newline, carriage return, horizontal tab, and vertical tab. Perl 5
698     no longer includes vertical tab in its set of whitespace characters. The \\v
699     escape that was in the Perl documentation for a long time was never in fact
700     recognized. However, the character itself was treated as whitespace at least
701     up to 5.002. In 5.004 and 5.005 it does not match \\s.
702    
703     2. PCRE does not allow repeat quantifiers on lookahead assertions. Perl permits
704     them, but they do not mean what you might think. For example, (?!a){3} does
705     not assert that the next three characters are not "a". It just asserts that the
706     next character is not "a" three times.
707    
708     3. Capturing subpatterns that occur inside negative lookahead assertions are
709     counted, but their entries in the offsets vector are never set. Perl sets its
710     numerical variables from any such patterns that are matched before the
711     assertion fails to match something (thereby succeeding), but only if the
712     negative lookahead assertion contains just one branch.
713    
714     4. Though binary zero characters are supported in the subject string, they are
715     not allowed in a pattern string because it is passed as a normal C string,
716     terminated by zero. The escape sequence "\\0" can be used in the pattern to
717     represent a binary zero.
718    
719     5. The following Perl escape sequences are not supported: \\l, \\u, \\L, \\U,
720     \\E, \\Q. In fact these are implemented by Perl's general string-handling and
721     are not part of its pattern matching engine.
722    
723     6. The Perl \\G assertion is not supported as it is not relevant to single
724     pattern matches.
725    
726 nigel 43 7. Fairly obviously, PCRE does not support the (?{code}) and (?p{code})
727     constructions. However, there is some experimental support for recursive
728     patterns using the non-Perl item (?R).
729 nigel 41
730     8. There are at the time of writing some oddities in Perl 5.005_02 concerned
731     with the settings of captured strings when part of a pattern is repeated. For
732     example, matching "aba" against the pattern /^(a(b)?)+$/ sets $2 to the value
733     "b", but matching "aabbaa" against /^(aa(bb)?)+$/ leaves $2 unset. However, if
734 nigel 47 the pattern is changed to /^(aa(b(b))?)+$/ then $2 (and $3) are set.
735 nigel 41
736     In Perl 5.004 $2 is set in both cases, and that is also true of PCRE. If in the
737     future Perl changes to a consistent state that is different, PCRE may change to
738     follow.
739    
740     9. Another as yet unresolved discrepancy is that in Perl 5.005_02 the pattern
741     /^(a)?(?(1)a|b)+$/ matches the string "a", whereas in PCRE it does not.
742     However, in both Perl and PCRE /^(a)?a/ matched against "a" leaves $1 unset.
743    
744     10. PCRE provides some extensions to the Perl regular expression facilities:
745    
746     (a) Although lookbehind assertions must match fixed length strings, each
747     alternative branch of a lookbehind assertion can match a different length of
748     string. Perl 5.005 requires them all to have the same length.
749    
750     (b) If PCRE_DOLLAR_ENDONLY is set and PCRE_MULTILINE is not set, the $ meta-
751     character matches only at the very end of the string.
752    
753     (c) If PCRE_EXTRA is set, a backslash followed by a letter with no special
754     meaning is faulted.
755    
756     (d) If PCRE_UNGREEDY is set, the greediness of the repetition quantifiers is
757     inverted, that is, by default they are not greedy, but if followed by a
758     question mark they are.
759    
760     (e) PCRE_ANCHORED can be used to force a pattern to be tried only at the start
761     of the subject.
762    
763     (f) The PCRE_NOTBOL, PCRE_NOTEOL, and PCRE_NOTEMPTY options for
764     \fBpcre_exec()\fR have no Perl equivalents.
765    
766 nigel 43 (g) The (?R) construct allows for recursive pattern matching (Perl 5.6 can do
767     this using the (?p{code}) construct, which PCRE cannot of course support.)
768 nigel 41
769 nigel 43
770 nigel 41 .SH REGULAR EXPRESSION DETAILS
771     The syntax and semantics of the regular expressions supported by PCRE are
772     described below. Regular expressions are also described in the Perl
773     documentation and in a number of other books, some of which have copious
774     examples. Jeffrey Friedl's "Mastering Regular Expressions", published by
775 nigel 49 O'Reilly (ISBN 1-56592-257), covers them in great detail.
776 nigel 41
777 nigel 49 The description here is intended as reference documentation. The basic
778     operation of PCRE is on strings of bytes. However, there is the beginnings of
779     some support for UTF-8 character strings. To use this support you must
780     configure PCRE to include it, and then call \fBpcre_compile()\fR with the
781     PCRE_UTF8 option. How this affects the pattern matching is described in the
782     final section of this document.
783    
784 nigel 41 A regular expression is a pattern that is matched against a subject string from
785     left to right. Most characters stand for themselves in a pattern, and match the
786     corresponding characters in the subject. As a trivial example, the pattern
787    
788     The quick brown fox
789    
790     matches a portion of a subject string that is identical to itself. The power of
791     regular expressions comes from the ability to include alternatives and
792     repetitions in the pattern. These are encoded in the pattern by the use of
793     \fImeta-characters\fR, which do not stand for themselves but instead are
794     interpreted in some special way.
795    
796     There are two different sets of meta-characters: those that are recognized
797     anywhere in the pattern except within square brackets, and those that are
798     recognized in square brackets. Outside square brackets, the meta-characters are
799     as follows:
800    
801     \\ general escape character with several uses
802     ^ assert start of subject (or line, in multiline mode)
803     $ assert end of subject (or line, in multiline mode)
804     . match any character except newline (by default)
805     [ start character class definition
806     | start of alternative branch
807     ( start subpattern
808     ) end subpattern
809     ? extends the meaning of (
810     also 0 or 1 quantifier
811     also quantifier minimizer
812     * 0 or more quantifier
813     + 1 or more quantifier
814     { start min/max quantifier
815    
816     Part of a pattern that is in square brackets is called a "character class". In
817     a character class the only meta-characters are:
818    
819     \\ general escape character
820     ^ negate the class, but only if the first character
821     - indicates character range
822     ] terminates the character class
823    
824     The following sections describe the use of each of the meta-characters.
825    
826    
827     .SH BACKSLASH
828     The backslash character has several uses. Firstly, if it is followed by a
829     non-alphameric character, it takes away any special meaning that character may
830     have. This use of backslash as an escape character applies both inside and
831     outside character classes.
832    
833     For example, if you want to match a "*" character, you write "\\*" in the
834     pattern. This applies whether or not the following character would otherwise be
835     interpreted as a meta-character, so it is always safe to precede a
836     non-alphameric with "\\" to specify that it stands for itself. In particular,
837     if you want to match a backslash, you write "\\\\".
838    
839     If a pattern is compiled with the PCRE_EXTENDED option, whitespace in the
840     pattern (other than in a character class) and characters between a "#" outside
841     a character class and the next newline character are ignored. An escaping
842     backslash can be used to include a whitespace or "#" character as part of the
843     pattern.
844    
845     A second use of backslash provides a way of encoding non-printing characters
846     in patterns in a visible manner. There is no restriction on the appearance of
847     non-printing characters, apart from the binary zero that terminates a pattern,
848     but when a pattern is being prepared by text editing, it is usually easier to
849     use one of the following escape sequences than the binary character it
850     represents:
851    
852     \\a alarm, that is, the BEL character (hex 07)
853     \\cx "control-x", where x is any character
854     \\e escape (hex 1B)
855     \\f formfeed (hex 0C)
856     \\n newline (hex 0A)
857     \\r carriage return (hex 0D)
858     \\t tab (hex 09)
859     \\xhh character with hex code hh
860     \\ddd character with octal code ddd, or backreference
861    
862     The precise effect of "\\cx" is as follows: if "x" is a lower case letter, it
863     is converted to upper case. Then bit 6 of the character (hex 40) is inverted.
864     Thus "\\cz" becomes hex 1A, but "\\c{" becomes hex 3B, while "\\c;" becomes hex
865     7B.
866    
867     After "\\x", up to two hexadecimal digits are read (letters can be in upper or
868     lower case).
869    
870     After "\\0" up to two further octal digits are read. In both cases, if there
871     are fewer than two digits, just those that are present are used. Thus the
872     sequence "\\0\\x\\07" specifies two binary zeros followed by a BEL character.
873     Make sure you supply two digits after the initial zero if the character that
874     follows is itself an octal digit.
875    
876     The handling of a backslash followed by a digit other than 0 is complicated.
877     Outside a character class, PCRE reads it and any following digits as a decimal
878     number. If the number is less than 10, or if there have been at least that many
879     previous capturing left parentheses in the expression, the entire sequence is
880     taken as a \fIback reference\fR. A description of how this works is given
881     later, following the discussion of parenthesized subpatterns.
882    
883     Inside a character class, or if the decimal number is greater than 9 and there
884     have not been that many capturing subpatterns, PCRE re-reads up to three octal
885     digits following the backslash, and generates a single byte from the least
886     significant 8 bits of the value. Any subsequent digits stand for themselves.
887     For example:
888    
889     \\040 is another way of writing a space
890     \\40 is the same, provided there are fewer than 40
891     previous capturing subpatterns
892     \\7 is always a back reference
893     \\11 might be a back reference, or another way of
894     writing a tab
895     \\011 is always a tab
896     \\0113 is a tab followed by the character "3"
897     \\113 is the character with octal code 113 (since there
898     can be no more than 99 back references)
899     \\377 is a byte consisting entirely of 1 bits
900     \\81 is either a back reference, or a binary zero
901     followed by the two characters "8" and "1"
902    
903     Note that octal values of 100 or greater must not be introduced by a leading
904     zero, because no more than three octal digits are ever read.
905    
906     All the sequences that define a single byte value can be used both inside and
907     outside character classes. In addition, inside a character class, the sequence
908     "\\b" is interpreted as the backspace character (hex 08). Outside a character
909     class it has a different meaning (see below).
910    
911     The third use of backslash is for specifying generic character types:
912    
913     \\d any decimal digit
914     \\D any character that is not a decimal digit
915     \\s any whitespace character
916     \\S any character that is not a whitespace character
917     \\w any "word" character
918     \\W any "non-word" character
919    
920     Each pair of escape sequences partitions the complete set of characters into
921     two disjoint sets. Any given character matches one, and only one, of each pair.
922    
923     A "word" character is any letter or digit or the underscore character, that is,
924     any character which can be part of a Perl "word". The definition of letters and
925     digits is controlled by PCRE's character tables, and may vary if locale-
926     specific matching is taking place (see "Locale support" above). For example, in
927     the "fr" (French) locale, some character codes greater than 128 are used for
928     accented letters, and these are matched by \\w.
929    
930     These character type sequences can appear both inside and outside character
931     classes. They each match one character of the appropriate type. If the current
932     matching point is at the end of the subject string, all of them fail, since
933     there is no character to match.
934    
935     The fourth use of backslash is for certain simple assertions. An assertion
936     specifies a condition that has to be met at a particular point in a match,
937     without consuming any characters from the subject string. The use of
938     subpatterns for more complicated assertions is described below. The backslashed
939     assertions are
940    
941     \\b word boundary
942     \\B not a word boundary
943     \\A start of subject (independent of multiline mode)
944     \\Z end of subject or newline at end (independent of multiline mode)
945     \\z end of subject (independent of multiline mode)
946    
947     These assertions may not appear in character classes (but note that "\\b" has a
948     different meaning, namely the backspace character, inside a character class).
949    
950     A word boundary is a position in the subject string where the current character
951     and the previous character do not both match \\w or \\W (i.e. one matches
952     \\w and the other matches \\W), or the start or end of the string if the
953     first or last character matches \\w, respectively.
954    
955     The \\A, \\Z, and \\z assertions differ from the traditional circumflex and
956     dollar (described below) in that they only ever match at the very start and end
957     of the subject string, whatever options are set. They are not affected by the
958     PCRE_NOTBOL or PCRE_NOTEOL options. If the \fIstartoffset\fR argument of
959     \fBpcre_exec()\fR is non-zero, \\A can never match. The difference between \\Z
960     and \\z is that \\Z matches before a newline that is the last character of the
961     string as well as at the end of the string, whereas \\z matches only at the
962     end.
963    
964    
965     .SH CIRCUMFLEX AND DOLLAR
966     Outside a character class, in the default matching mode, the circumflex
967     character is an assertion which is true only if the current matching point is
968     at the start of the subject string. If the \fIstartoffset\fR argument of
969     \fBpcre_exec()\fR is non-zero, circumflex can never match. Inside a character
970     class, circumflex has an entirely different meaning (see below).
971    
972     Circumflex need not be the first character of the pattern if a number of
973     alternatives are involved, but it should be the first thing in each alternative
974     in which it appears if the pattern is ever to match that branch. If all
975     possible alternatives start with a circumflex, that is, if the pattern is
976     constrained to match only at the start of the subject, it is said to be an
977     "anchored" pattern. (There are also other constructs that can cause a pattern
978     to be anchored.)
979    
980     A dollar character is an assertion which is true only if the current matching
981     point is at the end of the subject string, or immediately before a newline
982     character that is the last character in the string (by default). Dollar need
983     not be the last character of the pattern if a number of alternatives are
984     involved, but it should be the last item in any branch in which it appears.
985     Dollar has no special meaning in a character class.
986    
987     The meaning of dollar can be changed so that it matches only at the very end of
988     the string, by setting the PCRE_DOLLAR_ENDONLY option at compile or matching
989     time. This does not affect the \\Z assertion.
990    
991     The meanings of the circumflex and dollar characters are changed if the
992     PCRE_MULTILINE option is set. When this is the case, they match immediately
993     after and immediately before an internal "\\n" character, respectively, in
994     addition to matching at the start and end of the subject string. For example,
995     the pattern /^abc$/ matches the subject string "def\\nabc" in multiline mode,
996     but not otherwise. Consequently, patterns that are anchored in single line mode
997     because all branches start with "^" are not anchored in multiline mode, and a
998     match for circumflex is possible when the \fIstartoffset\fR argument of
999     \fBpcre_exec()\fR is non-zero. The PCRE_DOLLAR_ENDONLY option is ignored if
1000     PCRE_MULTILINE is set.
1001    
1002     Note that the sequences \\A, \\Z, and \\z can be used to match the start and
1003     end of the subject in both modes, and if all branches of a pattern start with
1004 nigel 53 \\A it is always anchored, whether PCRE_MULTILINE is set or not.
1005 nigel 41
1006    
1007     .SH FULL STOP (PERIOD, DOT)
1008     Outside a character class, a dot in the pattern matches any one character in
1009     the subject, including a non-printing character, but not (by default) newline.
1010 nigel 47 If the PCRE_DOTALL option is set, dots match newlines as well. The handling of
1011     dot is entirely independent of the handling of circumflex and dollar, the only
1012     relationship being that they both involve newline characters. Dot has no
1013     special meaning in a character class.
1014 nigel 41
1015    
1016     .SH SQUARE BRACKETS
1017     An opening square bracket introduces a character class, terminated by a closing
1018     square bracket. A closing square bracket on its own is not special. If a
1019     closing square bracket is required as a member of the class, it should be the
1020     first data character in the class (after an initial circumflex, if present) or
1021     escaped with a backslash.
1022    
1023     A character class matches a single character in the subject; the character must
1024     be in the set of characters defined by the class, unless the first character in
1025     the class is a circumflex, in which case the subject character must not be in
1026     the set defined by the class. If a circumflex is actually required as a member
1027     of the class, ensure it is not the first character, or escape it with a
1028     backslash.
1029    
1030     For example, the character class [aeiou] matches any lower case vowel, while
1031     [^aeiou] matches any character that is not a lower case vowel. Note that a
1032     circumflex is just a convenient notation for specifying the characters which
1033     are in the class by enumerating those that are not. It is not an assertion: it
1034     still consumes a character from the subject string, and fails if the current
1035     pointer is at the end of the string.
1036    
1037     When caseless matching is set, any letters in a class represent both their
1038     upper case and lower case versions, so for example, a caseless [aeiou] matches
1039     "A" as well as "a", and a caseless [^aeiou] does not match "A", whereas a
1040     caseful version would.
1041    
1042     The newline character is never treated in any special way in character classes,
1043     whatever the setting of the PCRE_DOTALL or PCRE_MULTILINE options is. A class
1044     such as [^a] will always match a newline.
1045    
1046     The minus (hyphen) character can be used to specify a range of characters in a
1047     character class. For example, [d-m] matches any letter between d and m,
1048     inclusive. If a minus character is required in a class, it must be escaped with
1049     a backslash or appear in a position where it cannot be interpreted as
1050     indicating a range, typically as the first or last character in the class.
1051    
1052     It is not possible to have the literal character "]" as the end character of a
1053     range. A pattern such as [W-]46] is interpreted as a class of two characters
1054     ("W" and "-") followed by a literal string "46]", so it would match "W46]" or
1055     "-46]". However, if the "]" is escaped with a backslash it is interpreted as
1056     the end of range, so [W-\\]46] is interpreted as a single class containing a
1057     range followed by two separate characters. The octal or hexadecimal
1058     representation of "]" can also be used to end a range.
1059    
1060     Ranges operate in ASCII collating sequence. They can also be used for
1061     characters specified numerically, for example [\\000-\\037]. If a range that
1062     includes letters is used when caseless matching is set, it matches the letters
1063     in either case. For example, [W-c] is equivalent to [][\\^_`wxyzabc], matched
1064     caselessly, and if character tables for the "fr" locale are in use,
1065     [\\xc8-\\xcb] matches accented E characters in both cases.
1066    
1067     The character types \\d, \\D, \\s, \\S, \\w, and \\W may also appear in a
1068     character class, and add the characters that they match to the class. For
1069     example, [\\dABCDEF] matches any hexadecimal digit. A circumflex can
1070     conveniently be used with the upper case character types to specify a more
1071     restricted set of characters than the matching lower case type. For example,
1072     the class [^\\W_] matches any letter or digit, but not underscore.
1073    
1074     All non-alphameric characters other than \\, -, ^ (at the start) and the
1075     terminating ] are non-special in character classes, but it does no harm if they
1076     are escaped.
1077    
1078    
1079 nigel 43 .SH POSIX CHARACTER CLASSES
1080     Perl 5.6 (not yet released at the time of writing) is going to support the
1081     POSIX notation for character classes, which uses names enclosed by [: and :]
1082     within the enclosing square brackets. PCRE supports this notation. For example,
1083    
1084     [01[:alpha:]%]
1085    
1086     matches "0", "1", any alphabetic character, or "%". The supported class names
1087     are
1088    
1089     alnum letters and digits
1090     alpha letters
1091     ascii character codes 0 - 127
1092     cntrl control characters
1093     digit decimal digits (same as \\d)
1094     graph printing characters, excluding space
1095     lower lower case letters
1096     print printing characters, including space
1097     punct printing characters, excluding letters and digits
1098     space white space (same as \\s)
1099     upper upper case letters
1100     word "word" characters (same as \\w)
1101     xdigit hexadecimal digits
1102    
1103     The names "ascii" and "word" are Perl extensions. Another Perl extension is
1104     negation, which is indicated by a ^ character after the colon. For example,
1105    
1106     [12[:^digit:]]
1107    
1108 nigel 53 matches "1", "2", or any non-digit. PCRE (and Perl) also recognize the POSIX
1109 nigel 43 syntax [.ch.] and [=ch=] where "ch" is a "collating element", but these are not
1110     supported, and an error is given if they are encountered.
1111    
1112    
1113 nigel 41 .SH VERTICAL BAR
1114     Vertical bar characters are used to separate alternative patterns. For example,
1115     the pattern
1116    
1117     gilbert|sullivan
1118    
1119     matches either "gilbert" or "sullivan". Any number of alternatives may appear,
1120     and an empty alternative is permitted (matching the empty string).
1121     The matching process tries each alternative in turn, from left to right,
1122     and the first one that succeeds is used. If the alternatives are within a
1123     subpattern (defined below), "succeeds" means matching the rest of the main
1124     pattern as well as the alternative in the subpattern.
1125    
1126    
1127     .SH INTERNAL OPTION SETTING
1128     The settings of PCRE_CASELESS, PCRE_MULTILINE, PCRE_DOTALL, and PCRE_EXTENDED
1129     can be changed from within the pattern by a sequence of Perl option letters
1130     enclosed between "(?" and ")". The option letters are
1131    
1132     i for PCRE_CASELESS
1133     m for PCRE_MULTILINE
1134     s for PCRE_DOTALL
1135     x for PCRE_EXTENDED
1136    
1137     For example, (?im) sets caseless, multiline matching. It is also possible to
1138     unset these options by preceding the letter with a hyphen, and a combined
1139     setting and unsetting such as (?im-sx), which sets PCRE_CASELESS and
1140     PCRE_MULTILINE while unsetting PCRE_DOTALL and PCRE_EXTENDED, is also
1141     permitted. If a letter appears both before and after the hyphen, the option is
1142     unset.
1143    
1144     The scope of these option changes depends on where in the pattern the setting
1145     occurs. For settings that are outside any subpattern (defined below), the
1146     effect is the same as if the options were set or unset at the start of
1147     matching. The following patterns all behave in exactly the same way:
1148    
1149     (?i)abc
1150     a(?i)bc
1151     ab(?i)c
1152     abc(?i)
1153    
1154     which in turn is the same as compiling the pattern abc with PCRE_CASELESS set.
1155     In other words, such "top level" settings apply to the whole pattern (unless
1156     there are other changes inside subpatterns). If there is more than one setting
1157     of the same option at top level, the rightmost setting is used.
1158    
1159     If an option change occurs inside a subpattern, the effect is different. This
1160     is a change of behaviour in Perl 5.005. An option change inside a subpattern
1161     affects only that part of the subpattern that follows it, so
1162    
1163     (a(?i)b)c
1164    
1165     matches abc and aBc and no other strings (assuming PCRE_CASELESS is not used).
1166     By this means, options can be made to have different settings in different
1167     parts of the pattern. Any changes made in one alternative do carry on
1168     into subsequent branches within the same subpattern. For example,
1169    
1170     (a(?i)b|c)
1171    
1172     matches "ab", "aB", "c", and "C", even though when matching "C" the first
1173     branch is abandoned before the option setting. This is because the effects of
1174     option settings happen at compile time. There would be some very weird
1175     behaviour otherwise.
1176    
1177     The PCRE-specific options PCRE_UNGREEDY and PCRE_EXTRA can be changed in the
1178     same way as the Perl-compatible options by using the characters U and X
1179     respectively. The (?X) flag setting is special in that it must always occur
1180     earlier in the pattern than any of the additional features it turns on, even
1181     when it is at top level. It is best put at the start.
1182    
1183    
1184     .SH SUBPATTERNS
1185     Subpatterns are delimited by parentheses (round brackets), which can be nested.
1186     Marking part of a pattern as a subpattern does two things:
1187    
1188     1. It localizes a set of alternatives. For example, the pattern
1189    
1190     cat(aract|erpillar|)
1191    
1192     matches one of the words "cat", "cataract", or "caterpillar". Without the
1193     parentheses, it would match "cataract", "erpillar" or the empty string.
1194    
1195     2. It sets up the subpattern as a capturing subpattern (as defined above).
1196     When the whole pattern matches, that portion of the subject string that matched
1197     the subpattern is passed back to the caller via the \fIovector\fR argument of
1198     \fBpcre_exec()\fR. Opening parentheses are counted from left to right (starting
1199     from 1) to obtain the numbers of the capturing subpatterns.
1200    
1201     For example, if the string "the red king" is matched against the pattern
1202    
1203     the ((red|white) (king|queen))
1204    
1205     the captured substrings are "red king", "red", and "king", and are numbered 1,
1206 nigel 53 2, and 3, respectively.
1207 nigel 41
1208     The fact that plain parentheses fulfil two functions is not always helpful.
1209     There are often times when a grouping subpattern is required without a
1210     capturing requirement. If an opening parenthesis is followed by "?:", the
1211     subpattern does not do any capturing, and is not counted when computing the
1212     number of any subsequent capturing subpatterns. For example, if the string "the
1213     white queen" is matched against the pattern
1214    
1215     the ((?:red|white) (king|queen))
1216    
1217     the captured substrings are "white queen" and "queen", and are numbered 1 and
1218     2. The maximum number of captured substrings is 99, and the maximum number of
1219     all subpatterns, both capturing and non-capturing, is 200.
1220    
1221     As a convenient shorthand, if any option settings are required at the start of
1222     a non-capturing subpattern, the option letters may appear between the "?" and
1223     the ":". Thus the two patterns
1224    
1225     (?i:saturday|sunday)
1226     (?:(?i)saturday|sunday)
1227    
1228     match exactly the same set of strings. Because alternative branches are tried
1229     from left to right, and options are not reset until the end of the subpattern
1230     is reached, an option setting in one branch does affect subsequent branches, so
1231     the above patterns match "SUNDAY" as well as "Saturday".
1232    
1233    
1234     .SH REPETITION
1235     Repetition is specified by quantifiers, which can follow any of the following
1236     items:
1237    
1238     a single character, possibly escaped
1239     the . metacharacter
1240     a character class
1241     a back reference (see next section)
1242     a parenthesized subpattern (unless it is an assertion - see below)
1243    
1244     The general repetition quantifier specifies a minimum and maximum number of
1245     permitted matches, by giving the two numbers in curly brackets (braces),
1246     separated by a comma. The numbers must be less than 65536, and the first must
1247     be less than or equal to the second. For example:
1248    
1249     z{2,4}
1250    
1251     matches "zz", "zzz", or "zzzz". A closing brace on its own is not a special
1252     character. If the second number is omitted, but the comma is present, there is
1253     no upper limit; if the second number and the comma are both omitted, the
1254     quantifier specifies an exact number of required matches. Thus
1255    
1256     [aeiou]{3,}
1257    
1258     matches at least 3 successive vowels, but may match many more, while
1259    
1260     \\d{8}
1261    
1262     matches exactly 8 digits. An opening curly bracket that appears in a position
1263     where a quantifier is not allowed, or one that does not match the syntax of a
1264     quantifier, is taken as a literal character. For example, {,6} is not a
1265     quantifier, but a literal string of four characters.
1266    
1267     The quantifier {0} is permitted, causing the expression to behave as if the
1268     previous item and the quantifier were not present.
1269    
1270     For convenience (and historical compatibility) the three most common
1271     quantifiers have single-character abbreviations:
1272    
1273     * is equivalent to {0,}
1274     + is equivalent to {1,}
1275     ? is equivalent to {0,1}
1276    
1277     It is possible to construct infinite loops by following a subpattern that can
1278     match no characters with a quantifier that has no upper limit, for example:
1279    
1280     (a?)*
1281    
1282     Earlier versions of Perl and PCRE used to give an error at compile time for
1283     such patterns. However, because there are cases where this can be useful, such
1284     patterns are now accepted, but if any repetition of the subpattern does in fact
1285     match no characters, the loop is forcibly broken.
1286    
1287     By default, the quantifiers are "greedy", that is, they match as much as
1288     possible (up to the maximum number of permitted times), without causing the
1289     rest of the pattern to fail. The classic example of where this gives problems
1290     is in trying to match comments in C programs. These appear between the
1291     sequences /* and */ and within the sequence, individual * and / characters may
1292     appear. An attempt to match C comments by applying the pattern
1293    
1294     /\\*.*\\*/
1295    
1296     to the string
1297    
1298     /* first command */ not comment /* second comment */
1299    
1300 nigel 51 fails, because it matches the entire string owing to the greediness of the .*
1301 nigel 41 item.
1302    
1303 nigel 47 However, if a quantifier is followed by a question mark, it ceases to be
1304 nigel 41 greedy, and instead matches the minimum number of times possible, so the
1305     pattern
1306    
1307     /\\*.*?\\*/
1308    
1309     does the right thing with the C comments. The meaning of the various
1310     quantifiers is not otherwise changed, just the preferred number of matches.
1311     Do not confuse this use of question mark with its use as a quantifier in its
1312     own right. Because it has two uses, it can sometimes appear doubled, as in
1313    
1314     \\d??\\d
1315    
1316     which matches one digit by preference, but can match two if that is the only
1317     way the rest of the pattern matches.
1318    
1319 nigel 47 If the PCRE_UNGREEDY option is set (an option which is not available in Perl),
1320     the quantifiers are not greedy by default, but individual ones can be made
1321 nigel 41 greedy by following them with a question mark. In other words, it inverts the
1322     default behaviour.
1323    
1324     When a parenthesized subpattern is quantified with a minimum repeat count that
1325     is greater than 1 or with a limited maximum, more store is required for the
1326     compiled pattern, in proportion to the size of the minimum or maximum.
1327    
1328     If a pattern starts with .* or .{0,} and the PCRE_DOTALL option (equivalent
1329 nigel 47 to Perl's /s) is set, thus allowing the . to match newlines, the pattern is
1330     implicitly anchored, because whatever follows will be tried against every
1331 nigel 41 character position in the subject string, so there is no point in retrying the
1332     overall match at any position after the first. PCRE treats such a pattern as
1333     though it were preceded by \\A. In cases where it is known that the subject
1334     string contains no newlines, it is worth setting PCRE_DOTALL when the pattern
1335     begins with .* in order to obtain this optimization, or alternatively using ^
1336     to indicate anchoring explicitly.
1337    
1338     When a capturing subpattern is repeated, the value captured is the substring
1339     that matched the final iteration. For example, after
1340    
1341     (tweedle[dume]{3}\\s*)+
1342    
1343     has matched "tweedledum tweedledee" the value of the captured substring is
1344     "tweedledee". However, if there are nested capturing subpatterns, the
1345     corresponding captured values may have been set in previous iterations. For
1346     example, after
1347    
1348     /(a|(b))+/
1349    
1350     matches "aba" the value of the second captured substring is "b".
1351    
1352    
1353     .SH BACK REFERENCES
1354     Outside a character class, a backslash followed by a digit greater than 0 (and
1355     possibly further digits) is a back reference to a capturing subpattern earlier
1356     (i.e. to its left) in the pattern, provided there have been that many previous
1357     capturing left parentheses.
1358    
1359     However, if the decimal number following the backslash is less than 10, it is
1360     always taken as a back reference, and causes an error only if there are not
1361     that many capturing left parentheses in the entire pattern. In other words, the
1362     parentheses that are referenced need not be to the left of the reference for
1363     numbers less than 10. See the section entitled "Backslash" above for further
1364     details of the handling of digits following a backslash.
1365    
1366     A back reference matches whatever actually matched the capturing subpattern in
1367     the current subject string, rather than anything matching the subpattern
1368     itself. So the pattern
1369    
1370     (sens|respons)e and \\1ibility
1371    
1372     matches "sense and sensibility" and "response and responsibility", but not
1373     "sense and responsibility". If caseful matching is in force at the time of the
1374 nigel 47 back reference, the case of letters is relevant. For example,
1375 nigel 41
1376     ((?i)rah)\\s+\\1
1377    
1378     matches "rah rah" and "RAH RAH", but not "RAH rah", even though the original
1379     capturing subpattern is matched caselessly.
1380    
1381     There may be more than one back reference to the same subpattern. If a
1382 nigel 47 subpattern has not actually been used in a particular match, any back
1383 nigel 41 references to it always fail. For example, the pattern
1384    
1385     (a|(bc))\\2
1386    
1387     always fails if it starts to match "a" rather than "bc". Because there may be
1388     up to 99 back references, all digits following the backslash are taken
1389     as part of a potential back reference number. If the pattern continues with a
1390 nigel 47 digit character, some delimiter must be used to terminate the back reference.
1391     If the PCRE_EXTENDED option is set, this can be whitespace. Otherwise an empty
1392     comment can be used.
1393 nigel 41
1394     A back reference that occurs inside the parentheses to which it refers fails
1395     when the subpattern is first used, so, for example, (a\\1) never matches.
1396     However, such references can be useful inside repeated subpatterns. For
1397     example, the pattern
1398    
1399     (a|b\\1)+
1400    
1401 nigel 49 matches any number of "a"s and also "aba", "ababbaa" etc. At each iteration of
1402 nigel 41 the subpattern, the back reference matches the character string corresponding
1403     to the previous iteration. In order for this to work, the pattern must be such
1404     that the first iteration does not need to match the back reference. This can be
1405     done using alternation, as in the example above, or by a quantifier with a
1406     minimum of zero.
1407    
1408    
1409     .SH ASSERTIONS
1410     An assertion is a test on the characters following or preceding the current
1411     matching point that does not actually consume any characters. The simple
1412     assertions coded as \\b, \\B, \\A, \\Z, \\z, ^ and $ are described above. More
1413     complicated assertions are coded as subpatterns. There are two kinds: those
1414     that look ahead of the current position in the subject string, and those that
1415     look behind it.
1416    
1417     An assertion subpattern is matched in the normal way, except that it does not
1418     cause the current matching position to be changed. Lookahead assertions start
1419     with (?= for positive assertions and (?! for negative assertions. For example,
1420    
1421     \\w+(?=;)
1422    
1423     matches a word followed by a semicolon, but does not include the semicolon in
1424     the match, and
1425    
1426     foo(?!bar)
1427    
1428     matches any occurrence of "foo" that is not followed by "bar". Note that the
1429     apparently similar pattern
1430    
1431     (?!foo)bar
1432    
1433     does not find an occurrence of "bar" that is preceded by something other than
1434     "foo"; it finds any occurrence of "bar" whatsoever, because the assertion
1435     (?!foo) is always true when the next three characters are "bar". A
1436     lookbehind assertion is needed to achieve this effect.
1437    
1438     Lookbehind assertions start with (?<= for positive assertions and (?<! for
1439     negative assertions. For example,
1440    
1441     (?<!foo)bar
1442    
1443     does find an occurrence of "bar" that is not preceded by "foo". The contents of
1444     a lookbehind assertion are restricted such that all the strings it matches must
1445     have a fixed length. However, if there are several alternatives, they do not
1446     all have to have the same fixed length. Thus
1447    
1448     (?<=bullock|donkey)
1449    
1450     is permitted, but
1451    
1452     (?<!dogs?|cats?)
1453    
1454     causes an error at compile time. Branches that match different length strings
1455     are permitted only at the top level of a lookbehind assertion. This is an
1456     extension compared with Perl 5.005, which requires all branches to match the
1457     same length of string. An assertion such as
1458    
1459     (?<=ab(c|de))
1460    
1461     is not permitted, because its single top-level branch can match two different
1462     lengths, but it is acceptable if rewritten to use two top-level branches:
1463    
1464     (?<=abc|abde)
1465    
1466     The implementation of lookbehind assertions is, for each alternative, to
1467     temporarily move the current position back by the fixed width and then try to
1468     match. If there are insufficient characters before the current position, the
1469     match is deemed to fail. Lookbehinds in conjunction with once-only subpatterns
1470     can be particularly useful for matching at the ends of strings; an example is
1471     given at the end of the section on once-only subpatterns.
1472    
1473     Several assertions (of any sort) may occur in succession. For example,
1474    
1475     (?<=\\d{3})(?<!999)foo
1476    
1477     matches "foo" preceded by three digits that are not "999". Notice that each of
1478     the assertions is applied independently at the same point in the subject
1479     string. First there is a check that the previous three characters are all
1480 nigel 47 digits, and then there is a check that the same three characters are not "999".
1481 nigel 41 This pattern does \fInot\fR match "foo" preceded by six characters, the first
1482     of which are digits and the last three of which are not "999". For example, it
1483     doesn't match "123abcfoo". A pattern to do that is
1484    
1485     (?<=\\d{3}...)(?<!999)foo
1486    
1487     This time the first assertion looks at the preceding six characters, checking
1488     that the first three are digits, and then the second assertion checks that the
1489     preceding three characters are not "999".
1490    
1491     Assertions can be nested in any combination. For example,
1492    
1493     (?<=(?<!foo)bar)baz
1494    
1495     matches an occurrence of "baz" that is preceded by "bar" which in turn is not
1496     preceded by "foo", while
1497    
1498     (?<=\\d{3}(?!999)...)foo
1499    
1500     is another pattern which matches "foo" preceded by three digits and any three
1501     characters that are not "999".
1502    
1503     Assertion subpatterns are not capturing subpatterns, and may not be repeated,
1504     because it makes no sense to assert the same thing several times. If any kind
1505     of assertion contains capturing subpatterns within it, these are counted for
1506     the purposes of numbering the capturing subpatterns in the whole pattern.
1507     However, substring capturing is carried out only for positive assertions,
1508     because it does not make sense for negative assertions.
1509    
1510     Assertions count towards the maximum of 200 parenthesized subpatterns.
1511    
1512    
1513     .SH ONCE-ONLY SUBPATTERNS
1514     With both maximizing and minimizing repetition, failure of what follows
1515     normally causes the repeated item to be re-evaluated to see if a different
1516     number of repeats allows the rest of the pattern to match. Sometimes it is
1517     useful to prevent this, either to change the nature of the match, or to cause
1518     it fail earlier than it otherwise might, when the author of the pattern knows
1519     there is no point in carrying on.
1520    
1521     Consider, for example, the pattern \\d+foo when applied to the subject line
1522    
1523     123456bar
1524    
1525     After matching all 6 digits and then failing to match "foo", the normal
1526     action of the matcher is to try again with only 5 digits matching the \\d+
1527     item, and then with 4, and so on, before ultimately failing. Once-only
1528     subpatterns provide the means for specifying that once a portion of the pattern
1529     has matched, it is not to be re-evaluated in this way, so the matcher would
1530     give up immediately on failing to match "foo" the first time. The notation is
1531     another kind of special parenthesis, starting with (?> as in this example:
1532    
1533     (?>\\d+)bar
1534    
1535     This kind of parenthesis "locks up" the part of the pattern it contains once
1536     it has matched, and a failure further into the pattern is prevented from
1537     backtracking into it. Backtracking past it to previous items, however, works as
1538     normal.
1539    
1540     An alternative description is that a subpattern of this type matches the string
1541     of characters that an identical standalone pattern would match, if anchored at
1542     the current point in the subject string.
1543    
1544     Once-only subpatterns are not capturing subpatterns. Simple cases such as the
1545     above example can be thought of as a maximizing repeat that must swallow
1546     everything it can. So, while both \\d+ and \\d+? are prepared to adjust the
1547     number of digits they match in order to make the rest of the pattern match,
1548     (?>\\d+) can only match an entire sequence of digits.
1549    
1550     This construction can of course contain arbitrarily complicated subpatterns,
1551     and it can be nested.
1552    
1553     Once-only subpatterns can be used in conjunction with lookbehind assertions to
1554     specify efficient matching at the end of the subject string. Consider a simple
1555     pattern such as
1556    
1557     abcd$
1558    
1559 nigel 43 when applied to a long string which does not match. Because matching proceeds
1560     from left to right, PCRE will look for each "a" in the subject and then see if
1561     what follows matches the rest of the pattern. If the pattern is specified as
1562 nigel 41
1563     ^.*abcd$
1564    
1565 nigel 47 the initial .* matches the entire string at first, but when this fails (because
1566     there is no following "a"), it backtracks to match all but the last character,
1567     then all but the last two characters, and so on. Once again the search for "a"
1568     covers the entire string, from right to left, so we are no better off. However,
1569     if the pattern is written as
1570 nigel 41
1571     ^(?>.*)(?<=abcd)
1572    
1573 nigel 47 there can be no backtracking for the .* item; it can match only the entire
1574 nigel 41 string. The subsequent lookbehind assertion does a single test on the last four
1575     characters. If it fails, the match fails immediately. For long strings, this
1576     approach makes a significant difference to the processing time.
1577    
1578 nigel 43 When a pattern contains an unlimited repeat inside a subpattern that can itself
1579     be repeated an unlimited number of times, the use of a once-only subpattern is
1580     the only way to avoid some failing matches taking a very long time indeed.
1581     The pattern
1582 nigel 41
1583 nigel 43 (\\D+|<\\d+>)*[!?]
1584    
1585     matches an unlimited number of substrings that either consist of non-digits, or
1586     digits enclosed in <>, followed by either ! or ?. When it matches, it runs
1587     quickly. However, if it is applied to
1588    
1589     aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa
1590    
1591     it takes a long time before reporting failure. This is because the string can
1592     be divided between the two repeats in a large number of ways, and all have to
1593     be tried. (The example used [!?] rather than a single character at the end,
1594     because both PCRE and Perl have an optimization that allows for fast failure
1595     when a single character is used. They remember the last single character that
1596     is required for a match, and fail early if it is not present in the string.)
1597     If the pattern is changed to
1598    
1599     ((?>\\D+)|<\\d+>)*[!?]
1600    
1601     sequences of non-digits cannot be broken, and failure happens quickly.
1602    
1603    
1604 nigel 41 .SH CONDITIONAL SUBPATTERNS
1605     It is possible to cause the matching process to obey a subpattern
1606     conditionally or to choose between two alternative subpatterns, depending on
1607     the result of an assertion, or whether a previous capturing subpattern matched
1608     or not. The two possible forms of conditional subpattern are
1609    
1610     (?(condition)yes-pattern)
1611     (?(condition)yes-pattern|no-pattern)
1612    
1613     If the condition is satisfied, the yes-pattern is used; otherwise the
1614     no-pattern (if present) is used. If there are more than two alternatives in the
1615     subpattern, a compile-time error occurs.
1616    
1617     There are two kinds of condition. If the text between the parentheses consists
1618 nigel 47 of a sequence of digits, the condition is satisfied if the capturing subpattern
1619 nigel 51 of that number has previously matched. The number must be greater than zero.
1620     Consider the following pattern, which contains non-significant white space to
1621     make it more readable (assume the PCRE_EXTENDED option) and to divide it into
1622     three parts for ease of discussion:
1623 nigel 41
1624     ( \\( )? [^()]+ (?(1) \\) )
1625    
1626     The first part matches an optional opening parenthesis, and if that
1627     character is present, sets it as the first captured substring. The second part
1628     matches one or more characters that are not parentheses. The third part is a
1629     conditional subpattern that tests whether the first set of parentheses matched
1630     or not. If they did, that is, if subject started with an opening parenthesis,
1631     the condition is true, and so the yes-pattern is executed and a closing
1632     parenthesis is required. Otherwise, since no-pattern is not present, the
1633     subpattern matches nothing. In other words, this pattern matches a sequence of
1634     non-parentheses, optionally enclosed in parentheses.
1635    
1636     If the condition is not a sequence of digits, it must be an assertion. This may
1637     be a positive or negative lookahead or lookbehind assertion. Consider this
1638     pattern, again containing non-significant white space, and with the two
1639     alternatives on the second line:
1640    
1641     (?(?=[^a-z]*[a-z])
1642     \\d{2}-[a-z]{3}-\\d{2} | \\d{2}-\\d{2}-\\d{2} )
1643    
1644     The condition is a positive lookahead assertion that matches an optional
1645     sequence of non-letters followed by a letter. In other words, it tests for the
1646     presence of at least one letter in the subject. If a letter is found, the
1647     subject is matched against the first alternative; otherwise it is matched
1648     against the second. This pattern matches strings in one of the two forms
1649     dd-aaa-dd or dd-dd-dd, where aaa are letters and dd are digits.
1650    
1651    
1652     .SH COMMENTS
1653     The sequence (?# marks the start of a comment which continues up to the next
1654     closing parenthesis. Nested parentheses are not permitted. The characters
1655     that make up a comment play no part in the pattern matching at all.
1656    
1657     If the PCRE_EXTENDED option is set, an unescaped # character outside a
1658     character class introduces a comment that continues up to the next newline
1659     character in the pattern.
1660    
1661    
1662 nigel 43 .SH RECURSIVE PATTERNS
1663     Consider the problem of matching a string in parentheses, allowing for
1664     unlimited nested parentheses. Without the use of recursion, the best that can
1665     be done is to use a pattern that matches up to some fixed depth of nesting. It
1666     is not possible to handle an arbitrary nesting depth. Perl 5.6 has provided an
1667     experimental facility that allows regular expressions to recurse (amongst other
1668     things). It does this by interpolating Perl code in the expression at run time,
1669     and the code can refer to the expression itself. A Perl pattern to solve the
1670     parentheses problem can be created like this:
1671    
1672     $re = qr{\\( (?: (?>[^()]+) | (?p{$re}) )* \\)}x;
1673    
1674     The (?p{...}) item interpolates Perl code at run time, and in this case refers
1675     recursively to the pattern in which it appears. Obviously, PCRE cannot support
1676     the interpolation of Perl code. Instead, the special item (?R) is provided for
1677     the specific case of recursion. This PCRE pattern solves the parentheses
1678     problem (assume the PCRE_EXTENDED option is set so that white space is
1679     ignored):
1680    
1681     \\( ( (?>[^()]+) | (?R) )* \\)
1682    
1683     First it matches an opening parenthesis. Then it matches any number of
1684     substrings which can either be a sequence of non-parentheses, or a recursive
1685     match of the pattern itself (i.e. a correctly parenthesized substring). Finally
1686     there is a closing parenthesis.
1687    
1688     This particular example pattern contains nested unlimited repeats, and so the
1689     use of a once-only subpattern for matching strings of non-parentheses is
1690     important when applying the pattern to strings that do not match. For example,
1691     when it is applied to
1692    
1693     (aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa()
1694    
1695     it yields "no match" quickly. However, if a once-only subpattern is not used,
1696     the match runs for a very long time indeed because there are so many different
1697     ways the + and * repeats can carve up the subject, and all have to be tested
1698     before failure can be reported.
1699    
1700     The values set for any capturing subpatterns are those from the outermost level
1701     of the recursion at which the subpattern value is set. If the pattern above is
1702     matched against
1703    
1704     (ab(cd)ef)
1705    
1706     the value for the capturing parentheses is "ef", which is the last value taken
1707     on at the top level. If additional parentheses are added, giving
1708    
1709     \\( ( ( (?>[^()]+) | (?R) )* ) \\)
1710     ^ ^
1711     ^ ^
1712 nigel 47 the string they capture is "ab(cd)ef", the contents of the top level
1713 nigel 43 parentheses. If there are more than 15 capturing parentheses in a pattern, PCRE
1714     has to obtain extra memory to store data during a recursion, which it does by
1715     using \fBpcre_malloc\fR, freeing it via \fBpcre_free\fR afterwards. If no
1716     memory can be obtained, it saves data for the first 15 capturing parentheses
1717     only, as there is no way to give an out-of-memory error from within a
1718     recursion.
1719    
1720    
1721 nigel 41 .SH PERFORMANCE
1722     Certain items that may appear in patterns are more efficient than others. It is
1723     more efficient to use a character class like [aeiou] than a set of alternatives
1724     such as (a|e|i|o|u). In general, the simplest construction that provides the
1725     required behaviour is usually the most efficient. Jeffrey Friedl's book
1726     contains a lot of discussion about optimizing regular expressions for efficient
1727     performance.
1728    
1729     When a pattern begins with .* and the PCRE_DOTALL option is set, the pattern is
1730     implicitly anchored by PCRE, since it can match only at the start of a subject
1731     string. However, if PCRE_DOTALL is not set, PCRE cannot make this optimization,
1732     because the . metacharacter does not then match a newline, and if the subject
1733     string contains newlines, the pattern may match from the character immediately
1734     following one of them instead of from the very start. For example, the pattern
1735    
1736     (.*) second
1737    
1738     matches the subject "first\\nand second" (where \\n stands for a newline
1739     character) with the first captured substring being "and". In order to do this,
1740     PCRE has to retry the match starting after every newline in the subject.
1741    
1742     If you are using such a pattern with subject strings that do not contain
1743     newlines, the best performance is obtained by setting PCRE_DOTALL, or starting
1744     the pattern with ^.* to indicate explicit anchoring. That saves PCRE from
1745     having to scan along the subject looking for a newline to restart at.
1746    
1747     Beware of patterns that contain nested indefinite repeats. These can take a
1748     long time to run when applied to a string that does not match. Consider the
1749     pattern fragment
1750    
1751     (a+)*
1752    
1753     This can match "aaaa" in 33 different ways, and this number increases very
1754     rapidly as the string gets longer. (The * repeat can match 0, 1, 2, 3, or 4
1755     times, and for each of those cases other than 0, the + repeats can match
1756     different numbers of times.) When the remainder of the pattern is such that the
1757     entire match is going to fail, PCRE has in principle to try every possible
1758     variation, and this can take an extremely long time.
1759    
1760     An optimization catches some of the more simple cases such as
1761    
1762     (a+)*b
1763    
1764     where a literal character follows. Before embarking on the standard matching
1765     procedure, PCRE checks that there is a "b" later in the subject string, and if
1766     there is not, it fails the match immediately. However, when there is no
1767     following literal this optimization cannot be used. You can see the difference
1768     by comparing the behaviour of
1769    
1770     (a+)*\\d
1771    
1772     with the pattern above. The former gives a failure almost instantly when
1773     applied to a whole line of "a" characters, whereas the latter takes an
1774     appreciable time with strings longer than about 20 characters.
1775    
1776 nigel 49
1777     .SH UTF-8 SUPPORT
1778     Starting at release 3.3, PCRE has some support for character strings encoded
1779     in the UTF-8 format. This is incomplete, and is regarded as experimental. In
1780     order to use it, you must configure PCRE to include UTF-8 support in the code,
1781     and, in addition, you must call \fBpcre_compile()\fR with the PCRE_UTF8 option
1782     flag. When you do this, both the pattern and any subject strings that are
1783     matched against it are treated as UTF-8 strings instead of just strings of
1784     bytes, but only in the cases that are mentioned below.
1785    
1786     If you compile PCRE with UTF-8 support, but do not use it at run time, the
1787     library will be a bit bigger, but the additional run time overhead is limited
1788     to testing the PCRE_UTF8 flag in several places, so should not be very large.
1789    
1790     PCRE assumes that the strings it is given contain valid UTF-8 codes. It does
1791     not diagnose invalid UTF-8 strings. If you pass invalid UTF-8 strings to PCRE,
1792     the results are undefined.
1793    
1794     Running with PCRE_UTF8 set causes these changes in the way PCRE works:
1795    
1796     1. In a pattern, the escape sequence \\x{...}, where the contents of the braces
1797     is a string of hexadecimal digits, is interpreted as a UTF-8 character whose
1798     code number is the given hexadecimal number, for example: \\x{1234}. This
1799     inserts from one to six literal bytes into the pattern, using the UTF-8
1800     encoding. If a non-hexadecimal digit appears between the braces, the item is
1801     not recognized.
1802    
1803     2. The original hexadecimal escape sequence, \\xhh, generates a two-byte UTF-8
1804     character if its value is greater than 127.
1805    
1806     3. Repeat quantifiers are NOT correctly handled if they follow a multibyte
1807     character. For example, \\x{100}* and \\xc3+ do not work. If you want to
1808     repeat such characters, you must enclose them in non-capturing parentheses,
1809     for example (?:\\x{100}), at present.
1810    
1811     4. The dot metacharacter matches one UTF-8 character instead of a single byte.
1812    
1813     5. Unlike literal UTF-8 characters, the dot metacharacter followed by a
1814     repeat quantifier does operate correctly on UTF-8 characters instead of
1815     single bytes.
1816    
1817     4. Although the \\x{...} escape is permitted in a character class, characters
1818     whose values are greater than 255 cannot be included in a class.
1819    
1820     5. A class is matched against a UTF-8 character instead of just a single byte,
1821     but it can match only characters whose values are less than 256. Characters
1822     with greater values always fail to match a class.
1823    
1824     6. Repeated classes work correctly on multiple characters.
1825    
1826     7. Classes containing just a single character whose value is greater than 127
1827     (but less than 256), for example, [\\x80] or [^\\x{93}], do not work because
1828     these are optimized into single byte matches. In the first case, of course,
1829     the class brackets are just redundant.
1830    
1831     8. Lookbehind assertions move backwards in the subject by a fixed number of
1832     characters instead of a fixed number of bytes. Simple cases have been tested
1833     to work correctly, but there may be hidden gotchas herein.
1834    
1835     9. The character types such as \\d and \\w do not work correctly with UTF-8
1836     characters. They continue to test a single byte.
1837    
1838     10. Anything not explicitly mentioned here continues to work in bytes rather
1839     than in characters.
1840    
1841     The following UTF-8 features of Perl 5.6 are not implemented:
1842    
1843     1. The escape sequence \\C to match a single byte.
1844    
1845     2. The use of Unicode tables and properties and escapes \\p, \\P, and \\X.
1846    
1847 nigel 53
1848     .SH SAMPLE PROGRAM
1849     The code below is a simple, complete demonstration program, to get you started
1850     with using PCRE. This code is also supplied in the file \fIpcredemo.c\fR in the
1851     PCRE distribution.
1852    
1853     The program compiles the regular expression that is its first argument, and
1854     matches it against the subject string in its second argument. No options are
1855     set, and default character tables are used. If matching succeeds, the program
1856     outputs the portion of the subject that matched, together with the contents of
1857     any captured substrings.
1858    
1859     On a Unix system that has PCRE installed in \fI/usr/local\fR, you can compile
1860     the demonstration program using a command like this:
1861    
1862     gcc -o pcredemo pcredemo.c -I/usr/local/include -L/usr/local/lib -lpcre
1863    
1864     Then you can run simple tests like this:
1865    
1866     ./pcredemo 'cat|dog' 'the cat sat on the mat'
1867    
1868     Note that there is a much more comprehensive test program, called
1869     \fBpcretest\fR, which supports many more facilities for testing regular
1870     expressions. The \fBpcredemo\fR program is provided as a simple coding example.
1871    
1872     On some operating systems (e.g. Solaris) you may get an error like this when
1873     you try to run \fBpcredemo\fR:
1874    
1875     ld.so.1: a.out: fatal: libpcre.so.0: open failed: No such file or directory
1876    
1877     This is caused by the way shared library support works on those systems. You
1878     need to add
1879    
1880     -R/usr/local/lib
1881    
1882     to the compile command to get round this problem. Here's the code:
1883    
1884     #include <stdio.h>
1885     #include <string.h>
1886     #include <pcre.h>
1887    
1888     #define OVECCOUNT 30 /* should be a multiple of 3 */
1889    
1890     int main(int argc, char **argv)
1891     {
1892     pcre *re;
1893     const char *error;
1894     int erroffset;
1895     int ovector[OVECCOUNT];
1896     int rc, i;
1897    
1898     if (argc != 3)
1899     {
1900     printf("Two arguments required: a regex and a "
1901     "subject string\\n");
1902     return 1;
1903     }
1904    
1905     /* Compile the regular expression in the first argument */
1906    
1907     re = pcre_compile(
1908     argv[1], /* the pattern */
1909     0, /* default options */
1910     &error, /* for error message */
1911     &erroffset, /* for error offset */
1912     NULL); /* use default character tables */
1913    
1914     /* Compilation failed: print the error message and exit */
1915    
1916     if (re == NULL)
1917     {
1918     printf("PCRE compilation failed at offset %d: %s\\n",
1919     erroffset, error);
1920     return 1;
1921     }
1922    
1923     /* Compilation succeeded: match the subject in the second
1924     argument */
1925    
1926     rc = pcre_exec(
1927     re, /* the compiled pattern */
1928     NULL, /* we didn't study the pattern */
1929     argv[2], /* the subject string */
1930     (int)strlen(argv[2]), /* the length of the subject */
1931     0, /* start at offset 0 in the subject */
1932     0, /* default options */
1933     ovector, /* vector for substring information */
1934     OVECCOUNT); /* number of elements in the vector */
1935    
1936     /* Matching failed: handle error cases */
1937    
1938     if (rc < 0)
1939     {
1940     switch(rc)
1941     {
1942     case PCRE_ERROR_NOMATCH: printf("No match\\n"); break;
1943     /*
1944     Handle other special cases if you like
1945     */
1946     default: printf("Matching error %d\\n", rc); break;
1947     }
1948     return 1;
1949     }
1950    
1951     /* Match succeded */
1952    
1953     printf("Match succeeded\\n");
1954    
1955     /* The output vector wasn't big enough */
1956    
1957     if (rc == 0)
1958     {
1959     rc = OVECCOUNT/3;
1960     printf("ovector only has room for %d captured "
1961     substrings\\n", rc - 1);
1962     }
1963    
1964     /* Show substrings stored in the output vector */
1965    
1966     for (i = 0; i < rc; i++)
1967     {
1968     char *substring_start = argv[2] + ovector[2*i];
1969     int substring_length = ovector[2*i+1] - ovector[2*i];
1970     printf("%2d: %.*s\\n", i, substring_length,
1971     substring_start);
1972     }
1973    
1974     return 0;
1975     }
1976    
1977    
1978 nigel 41 .SH AUTHOR
1979     Philip Hazel <ph10@cam.ac.uk>
1980     .br
1981     University Computing Service,
1982     .br
1983     New Museums Site,
1984     .br
1985     Cambridge CB2 3QG, England.
1986     .br
1987     Phone: +44 1223 334714
1988    
1989 nigel 53 Last updated: 15 August 2001
1990 nigel 41 .br
1991 nigel 53 Copyright (c) 1997-2001 University of Cambridge.

webmaster@exim.org
ViewVC Help
Powered by ViewVC 1.1.12