ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 553 - (show annotations)
Fri Oct 22 15:57:50 2010 UTC (10 years, 9 months ago) by ph10
File size: 95443 byte(s)
Change the way PCRE_PARTIAL_HARD handles \z, \Z, \b, \B, and $.
3 PCRE - Perl-compatible regular expressions
5 .rs
6 .sp
7 .B #include <pcre.h>
8 .PP
9 .SM
10 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
11 .ti +5n
12 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
13 .ti +5n
14 .B const unsigned char *\fItableptr\fP);
15 .PP
16 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
17 .ti +5n
18 .B int *\fIerrorcodeptr\fP,
19 .ti +5n
20 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
21 .ti +5n
22 .B const unsigned char *\fItableptr\fP);
23 .PP
24 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
25 .ti +5n
26 .B const char **\fIerrptr\fP);
27 .PP
28 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
29 .ti +5n
30 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
31 .ti +5n
32 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
33 .PP
34 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
35 .ti +5n
36 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
37 .ti +5n
38 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
39 .ti +5n
40 .B int *\fIworkspace\fP, int \fIwscount\fP);
41 .PP
42 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
43 .ti +5n
44 .B const char *\fIsubject\fP, int *\fIovector\fP,
45 .ti +5n
46 .B int \fIstringcount\fP, const char *\fIstringname\fP,
47 .ti +5n
48 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
49 .PP
50 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
51 .ti +5n
52 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
53 .ti +5n
54 .B int \fIbuffersize\fP);
55 .PP
56 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
57 .ti +5n
58 .B const char *\fIsubject\fP, int *\fIovector\fP,
59 .ti +5n
60 .B int \fIstringcount\fP, const char *\fIstringname\fP,
61 .ti +5n
62 .B const char **\fIstringptr\fP);
63 .PP
64 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
65 .ti +5n
66 .B const char *\fIname\fP);
67 .PP
68 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
69 .ti +5n
70 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
71 .PP
72 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
73 .ti +5n
74 .B int \fIstringcount\fP, int \fIstringnumber\fP,
75 .ti +5n
76 .B const char **\fIstringptr\fP);
77 .PP
78 .B int pcre_get_substring_list(const char *\fIsubject\fP,
79 .ti +5n
80 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
81 .PP
82 .B void pcre_free_substring(const char *\fIstringptr\fP);
83 .PP
84 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
85 .PP
86 .B const unsigned char *pcre_maketables(void);
87 .PP
88 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
89 .ti +5n
90 .B int \fIwhat\fP, void *\fIwhere\fP);
91 .PP
92 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
93 .B *\fIfirstcharptr\fP);
94 .PP
95 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
96 .PP
97 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
98 .PP
99 .B char *pcre_version(void);
100 .PP
101 .B void *(*pcre_malloc)(size_t);
102 .PP
103 .B void (*pcre_free)(void *);
104 .PP
105 .B void *(*pcre_stack_malloc)(size_t);
106 .PP
107 .B void (*pcre_stack_free)(void *);
108 .PP
109 .B int (*pcre_callout)(pcre_callout_block *);
110 .
111 .
113 .rs
114 .sp
115 PCRE has its own native API, which is described in this document. There are
116 also some wrapper functions that correspond to the POSIX regular expression
117 API. These are described in the
118 .\" HREF
119 \fBpcreposix\fP
120 .\"
121 documentation. Both of these APIs define a set of C function calls. A C++
122 wrapper is distributed with PCRE. It is documented in the
123 .\" HREF
124 \fBpcrecpp\fP
125 .\"
126 page.
127 .P
128 The native API C function prototypes are defined in the header file
129 \fBpcre.h\fP, and on Unix systems the library itself is called \fBlibpcre\fP.
130 It can normally be accessed by adding \fB-lpcre\fP to the command for linking
131 an application that uses PCRE. The header file defines the macros PCRE_MAJOR
132 and PCRE_MINOR to contain the major and minor release numbers for the library.
133 Applications can use these to include support for different releases of PCRE.
134 .P
135 In a Windows environment, if you want to statically link an application program
136 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
137 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
138 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
139 \fB__declspec(dllimport)\fP, with unwanted results.
140 .P
141 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
142 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
143 in a Perl-compatible manner. A sample program that demonstrates the simplest
144 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
145 source distribution. A listing of this program is given in the
146 .\" HREF
147 \fBpcredemo\fP
148 .\"
149 documentation, and the
150 .\" HREF
151 \fBpcresample\fP
152 .\"
153 documentation describes how to compile and run it.
154 .P
155 A second matching function, \fBpcre_dfa_exec()\fP, which is not
156 Perl-compatible, is also provided. This uses a different algorithm for the
157 matching. The alternative algorithm finds all possible matches (at a given
158 point in the subject), and scans the subject just once (unless there are
159 lookbehind assertions). However, this algorithm does not return captured
160 substrings. A description of the two matching algorithms and their advantages
161 and disadvantages is given in the
162 .\" HREF
163 \fBpcrematching\fP
164 .\"
165 documentation.
166 .P
167 In addition to the main compiling and matching functions, there are convenience
168 functions for extracting captured substrings from a subject string that is
169 matched by \fBpcre_exec()\fP. They are:
170 .sp
171 \fBpcre_copy_substring()\fP
172 \fBpcre_copy_named_substring()\fP
173 \fBpcre_get_substring()\fP
174 \fBpcre_get_named_substring()\fP
175 \fBpcre_get_substring_list()\fP
176 \fBpcre_get_stringnumber()\fP
177 \fBpcre_get_stringtable_entries()\fP
178 .sp
179 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
180 provided, to free the memory used for extracted strings.
181 .P
182 The function \fBpcre_maketables()\fP is used to build a set of character tables
183 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
184 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
185 specialist use. Most commonly, no special tables are passed, in which case
186 internal tables that are generated when PCRE is built are used.
187 .P
188 The function \fBpcre_fullinfo()\fP is used to find out information about a
189 compiled pattern; \fBpcre_info()\fP is an obsolete version that returns only
190 some of the available information, but is retained for backwards compatibility.
191 The function \fBpcre_version()\fP returns a pointer to a string containing the
192 version of PCRE and its date of release.
193 .P
194 The function \fBpcre_refcount()\fP maintains a reference count in a data block
195 containing a compiled pattern. This is provided for the benefit of
196 object-oriented applications.
197 .P
198 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
199 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
200 respectively. PCRE calls the memory management functions via these variables,
201 so a calling program can replace them if it wishes to intercept the calls. This
202 should be done before calling any PCRE functions.
203 .P
204 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
205 indirections to memory management functions. These special functions are used
206 only when PCRE is compiled to use the heap for remembering data, instead of
207 recursive function calls, when running the \fBpcre_exec()\fP function. See the
208 .\" HREF
209 \fBpcrebuild\fP
210 .\"
211 documentation for details of how to do this. It is a non-standard way of
212 building PCRE, for use in environments that have limited stacks. Because of the
213 greater use of memory management, it runs more slowly. Separate functions are
214 provided so that special-purpose external code can be used for this case. When
215 used, these functions are always called in a stack-like manner (last obtained,
216 first freed), and always for memory blocks of the same size. There is a
217 discussion about PCRE's stack usage in the
218 .\" HREF
219 \fBpcrestack\fP
220 .\"
221 documentation.
222 .P
223 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
224 by the caller to a "callout" function, which PCRE will then call at specified
225 points during a matching operation. Details are given in the
226 .\" HREF
227 \fBpcrecallout\fP
228 .\"
229 documentation.
230 .
231 .
232 .\" HTML <a name="newlines"></a>
234 .rs
235 .sp
236 PCRE supports five different conventions for indicating line breaks in
237 strings: a single CR (carriage return) character, a single LF (linefeed)
238 character, the two-character sequence CRLF, any of the three preceding, or any
239 Unicode newline sequence. The Unicode newline sequences are the three just
240 mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
241 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
242 (paragraph separator, U+2029).
243 .P
244 Each of the first three conventions is used by at least one operating system as
245 its standard newline sequence. When PCRE is built, a default can be specified.
246 The default default is LF, which is the Unix standard. When PCRE is run, the
247 default can be overridden, either when a pattern is compiled, or when it is
248 matched.
249 .P
250 At compile time, the newline convention can be specified by the \fIoptions\fP
251 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
252 start of the pattern itself; this overrides any other settings. See the
253 .\" HREF
254 \fBpcrepattern\fP
255 .\"
256 page for details of the special character sequences.
257 .P
258 In the PCRE documentation the word "newline" is used to mean "the character or
259 pair of characters that indicate a line break". The choice of newline
260 convention affects the handling of the dot, circumflex, and dollar
261 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
262 recognized line ending sequence, the match position advancement for a
263 non-anchored pattern. There is more detail about this in the
264 .\" HTML <a href="#execoptions">
265 .\" </a>
266 section on \fBpcre_exec()\fP options
267 .\"
268 below.
269 .P
270 The choice of newline convention does not affect the interpretation of
271 the \en or \er escape sequences, nor does it affect what \eR matches, which is
272 controlled in a similar way, but by separate options.
273 .
274 .
276 .rs
277 .sp
278 The PCRE functions can be used in multi-threading applications, with the
279 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
280 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
281 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
282 .P
283 The compiled form of a regular expression is not altered during matching, so
284 the same compiled pattern can safely be used by several threads at once.
285 .
286 .
288 .rs
289 .sp
290 The compiled form of a regular expression can be saved and re-used at a later
291 time, possibly by a different program, and even on a host other than the one on
292 which it was compiled. Details are given in the
293 .\" HREF
294 \fBpcreprecompile\fP
295 .\"
296 documentation. However, compiling a regular expression with one version of PCRE
297 for use with a different version is not guaranteed to work and may cause
298 crashes.
299 .
300 .
302 .rs
303 .sp
304 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
305 .PP
306 The function \fBpcre_config()\fP makes it possible for a PCRE client to
307 discover which optional features have been compiled into the PCRE library. The
308 .\" HREF
309 \fBpcrebuild\fP
310 .\"
311 documentation has more details about these optional features.
312 .P
313 The first argument for \fBpcre_config()\fP is an integer, specifying which
314 information is required; the second argument is a pointer to a variable into
315 which the information is placed. The following information is available:
316 .sp
318 .sp
319 The output is an integer that is set to one if UTF-8 support is available;
320 otherwise it is set to zero.
321 .sp
323 .sp
324 The output is an integer that is set to one if support for Unicode character
325 properties is available; otherwise it is set to zero.
326 .sp
328 .sp
329 The output is an integer whose value specifies the default character sequence
330 that is recognized as meaning "newline". The four values that are supported
331 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
332 Though they are derived from ASCII, the same values are returned in EBCDIC
333 environments. The default should normally correspond to the standard sequence
334 for your operating system.
335 .sp
337 .sp
338 The output is an integer whose value indicates what character sequences the \eR
339 escape sequence matches by default. A value of 0 means that \eR matches any
340 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
341 or CRLF. The default can be overridden when a pattern is compiled or matched.
342 .sp
344 .sp
345 The output is an integer that contains the number of bytes used for internal
346 linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
347 allow larger regular expressions to be compiled, at the expense of slower
348 matching. The default value of 2 is sufficient for all but the most massive
349 patterns, since it allows the compiled pattern to be up to 64K in size.
350 .sp
352 .sp
353 The output is an integer that contains the threshold above which the POSIX
354 interface uses \fBmalloc()\fP for output vectors. Further details are given in
355 the
356 .\" HREF
357 \fBpcreposix\fP
358 .\"
359 documentation.
360 .sp
362 .sp
363 The output is a long integer that gives the default limit for the number of
364 internal matching function calls in a \fBpcre_exec()\fP execution. Further
365 details are given with \fBpcre_exec()\fP below.
366 .sp
368 .sp
369 The output is a long integer that gives the default limit for the depth of
370 recursion when calling the internal matching function in a \fBpcre_exec()\fP
371 execution. Further details are given with \fBpcre_exec()\fP below.
372 .sp
374 .sp
375 The output is an integer that is set to one if internal recursion when running
376 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
377 to remember their state. This is the usual way that PCRE is compiled. The
378 output is zero if PCRE was compiled to use blocks of data on the heap instead
379 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
380 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
381 avoiding the use of the stack.
382 .
383 .
385 .rs
386 .sp
387 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
388 .ti +5n
389 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
390 .ti +5n
391 .B const unsigned char *\fItableptr\fP);
392 .sp
393 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
394 .ti +5n
395 .B int *\fIerrorcodeptr\fP,
396 .ti +5n
397 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
398 .ti +5n
399 .B const unsigned char *\fItableptr\fP);
400 .P
401 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
402 called to compile a pattern into an internal form. The only difference between
403 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
404 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
405 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
406 information applies equally to \fBpcre_compile2()\fP.
407 .P
408 The pattern is a C string terminated by a binary zero, and is passed in the
409 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
410 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
411 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
412 for a structure whose contents are not externally defined. It is up to the
413 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
414 .P
415 Although the compiled code of a PCRE regex is relocatable, that is, it does not
416 depend on memory location, the complete \fBpcre\fP data block is not
417 fully relocatable, because it may contain a copy of the \fItableptr\fP
418 argument, which is an address (see below).
419 .P
420 The \fIoptions\fP argument contains various bit settings that affect the
421 compilation. It should be zero if no options are required. The available
422 options are described below. Some of them (in particular, those that are
423 compatible with Perl, but some others as well) can also be set and unset from
424 within the pattern (see the detailed description in the
425 .\" HREF
426 \fBpcrepattern\fP
427 .\"
428 documentation). For those options that can be different in different parts of
429 the pattern, the contents of the \fIoptions\fP argument specifies their
430 settings at the start of compilation and execution. The PCRE_ANCHORED,
431 PCRE_BSR_\fIxxx\fP, and PCRE_NEWLINE_\fIxxx\fP options can be set at the time
432 of matching as well as at compile time.
433 .P
434 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
435 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
436 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
437 error message. This is a static string that is part of the library. You must
438 not try to free it. The byte offset from the start of the pattern to the
439 character that was being processed when the error was discovered is placed in
440 the variable pointed to by \fIerroffset\fP, which must not be NULL. If it is,
441 an immediate error is given. Some errors are not detected until checks are
442 carried out when the whole pattern has been scanned; in this case the offset is
443 set to the end of the pattern.
444 .P
445 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
446 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
447 returned via this argument in the event of an error. This is in addition to the
448 textual error message. Error codes and messages are listed below.
449 .P
450 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
451 character tables that are built when PCRE is compiled, using the default C
452 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
453 call to \fBpcre_maketables()\fP. This value is stored with the compiled
454 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
455 passed to it. For more discussion, see the section on locale support below.
456 .P
457 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
458 .sp
459 pcre *re;
460 const char *error;
461 int erroffset;
462 re = pcre_compile(
463 "^A.*Z", /* the pattern */
464 0, /* default options */
465 &error, /* for error message */
466 &erroffset, /* for error offset */
467 NULL); /* use default character tables */
468 .sp
469 The following names for option bits are defined in the \fBpcre.h\fP header
470 file:
471 .sp
473 .sp
474 If this bit is set, the pattern is forced to be "anchored", that is, it is
475 constrained to match only at the first matching point in the string that is
476 being searched (the "subject string"). This effect can also be achieved by
477 appropriate constructs in the pattern itself, which is the only way to do it in
478 Perl.
479 .sp
481 .sp
482 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
483 all with number 255, before each pattern item. For discussion of the callout
484 facility, see the
485 .\" HREF
486 \fBpcrecallout\fP
487 .\"
488 documentation.
489 .sp
492 .sp
493 These options (which are mutually exclusive) control what the \eR escape
494 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
495 match any Unicode newline sequence. The default is specified when PCRE is
496 built. It can be overridden from within the pattern, or by setting an option
497 when a compiled pattern is matched.
498 .sp
500 .sp
501 If this bit is set, letters in the pattern match both upper and lower case
502 letters. It is equivalent to Perl's /i option, and it can be changed within a
503 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
504 concept of case for characters whose values are less than 128, so caseless
505 matching is always possible. For characters with higher values, the concept of
506 case is supported if PCRE is compiled with Unicode property support, but not
507 otherwise. If you want to use caseless matching for characters 128 and above,
508 you must ensure that PCRE is compiled with Unicode property support as well as
509 with UTF-8 support.
510 .sp
512 .sp
513 If this bit is set, a dollar metacharacter in the pattern matches only at the
514 end of the subject string. Without this option, a dollar also matches
515 immediately before a newline at the end of the string (but not before any other
516 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
517 There is no equivalent to this option in Perl, and no way to set it within a
518 pattern.
519 .sp
521 .sp
522 If this bit is set, a dot metacharater in the pattern matches all characters,
523 including those that indicate newline. Without it, a dot does not match when
524 the current position is at a newline. This option is equivalent to Perl's /s
525 option, and it can be changed within a pattern by a (?s) option setting. A
526 negative class such as [^a] always matches newline characters, independent of
527 the setting of this option.
528 .sp
530 .sp
531 If this bit is set, names used to identify capturing subpatterns need not be
532 unique. This can be helpful for certain types of pattern when it is known that
533 only one instance of the named subpattern can ever be matched. There are more
534 details of named subpatterns below; see also the
535 .\" HREF
536 \fBpcrepattern\fP
537 .\"
538 documentation.
539 .sp
541 .sp
542 If this bit is set, whitespace data characters in the pattern are totally
543 ignored except when escaped or inside a character class. Whitespace does not
544 include the VT character (code 11). In addition, characters between an
545 unescaped # outside a character class and the next newline, inclusive, are also
546 ignored. This is equivalent to Perl's /x option, and it can be changed within a
547 pattern by a (?x) option setting.
548 .P
549 This option makes it possible to include comments inside complicated patterns.
550 Note, however, that this applies only to data characters. Whitespace characters
551 may never appear within special character sequences in a pattern, for example
552 within the sequence (?( which introduces a conditional subpattern.
553 .sp
555 .sp
556 This option was invented in order to turn on additional functionality of PCRE
557 that is incompatible with Perl, but it is currently of very little use. When
558 set, any backslash in a pattern that is followed by a letter that has no
559 special meaning causes an error, thus reserving these combinations for future
560 expansion. By default, as in Perl, a backslash followed by a letter with no
561 special meaning is treated as a literal. (Perl can, however, be persuaded to
562 give an error for this, by running it with the -w option.) There are at present
563 no other features controlled by this option. It can also be set by a (?X)
564 option setting within a pattern.
565 .sp
567 .sp
568 If this option is set, an unanchored pattern is required to match before or at
569 the first newline in the subject string, though the matched text may continue
570 over the newline.
571 .sp
573 .sp
574 If this option is set, PCRE's behaviour is changed in some ways so that it is
575 compatible with JavaScript rather than Perl. The changes are as follows:
576 .P
577 (1) A lone closing square bracket in a pattern causes a compile-time error,
578 because this is illegal in JavaScript (by default it is treated as a data
579 character). Thus, the pattern AB]CD becomes illegal when this option is set.
580 .P
581 (2) At run time, a back reference to an unset subpattern group matches an empty
582 string (by default this causes the current matching alternative to fail). A
583 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
584 an "a" in the subject), whereas it fails by default, for Perl compatibility.
585 .sp
587 .sp
588 By default, PCRE treats the subject string as consisting of a single line of
589 characters (even if it actually contains newlines). The "start of line"
590 metacharacter (^) matches only at the start of the string, while the "end of
591 line" metacharacter ($) matches only at the end of the string, or before a
592 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
593 Perl.
594 .P
595 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
596 match immediately following or immediately before internal newlines in the
597 subject string, respectively, as well as at the very start and end. This is
598 equivalent to Perl's /m option, and it can be changed within a pattern by a
599 (?m) option setting. If there are no newlines in a subject string, or no
600 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
601 .sp
607 .sp
608 These options override the default newline definition that was chosen when PCRE
609 was built. Setting the first or the second specifies that a newline is
610 indicated by a single character (CR or LF, respectively). Setting
611 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
612 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
613 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
614 that any Unicode newline sequence should be recognized. The Unicode newline
615 sequences are the three just mentioned, plus the single characters VT (vertical
616 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
617 separator, U+2028), and PS (paragraph separator, U+2029). The last two are
618 recognized only in UTF-8 mode.
619 .P
620 The newline setting in the options word uses three bits that are treated
621 as a number, giving eight possibilities. Currently only six are used (default
622 plus the five values above). This means that if you set more than one newline
623 option, the combination may or may not be sensible. For example,
625 other combinations may yield unused numbers and cause an error.
626 .P
627 The only time that a line break is specially recognized when compiling a
628 pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
629 class is encountered. This indicates a comment that lasts until after the next
630 line break sequence. In other circumstances, line break sequences are treated
631 as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
632 as whitespace characters and are therefore ignored.
633 .P
634 The newline option that is set at compile time becomes the default that is used
635 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
636 .sp
638 .sp
639 If this option is set, it disables the use of numbered capturing parentheses in
640 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
641 were followed by ?: but named parentheses can still be used for capturing (and
642 they acquire numbers in the usual way). There is no equivalent of this option
643 in Perl.
644 .sp
646 .sp
647 This option changes the way PCRE processes \eb, \ed, \es, \ew, and some of the
648 POSIX character classes. By default, only ASCII characters are recognized, but
649 if PCRE_UCP is set, Unicode properties are used instead to classify characters.
650 More details are given in the section on
651 .\" HTML <a href="pcre.html#genericchartypes">
652 .\" </a>
653 generic character types
654 .\"
655 in the
656 .\" HREF
657 \fBpcrepattern\fP
658 .\"
659 page. If you set PCRE_UCP, matching one of the items it affects takes much
660 longer. The option is available only if PCRE has been compiled with Unicode
661 property support.
662 .sp
664 .sp
665 This option inverts the "greediness" of the quantifiers so that they are not
666 greedy by default, but become greedy if followed by "?". It is not compatible
667 with Perl. It can also be set by a (?U) option setting within the pattern.
668 .sp
670 .sp
671 This option causes PCRE to regard both the pattern and the subject as strings
672 of UTF-8 characters instead of single-byte character strings. However, it is
673 available only when PCRE is built to include UTF-8 support. If not, the use
674 of this option provokes an error. Details of how this option changes the
675 behaviour of PCRE are given in the
676 .\" HTML <a href="pcre.html#utf8support">
677 .\" </a>
678 section on UTF-8 support
679 .\"
680 in the main
681 .\" HREF
682 \fBpcre\fP
683 .\"
684 page.
685 .sp
687 .sp
688 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
689 automatically checked. There is a discussion about the
690 .\" HTML <a href="pcre.html#utf8strings">
691 .\" </a>
692 validity of UTF-8 strings
693 .\"
694 in the main
695 .\" HREF
696 \fBpcre\fP
697 .\"
698 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
699 returns an error. If you already know that your pattern is valid, and you want
700 to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
701 option. When it is set, the effect of passing an invalid UTF-8 string as a
702 pattern is undefined. It may cause your program to crash. Note that this option
703 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
704 the UTF-8 validity checking of subject strings.
705 .
706 .
708 .rs
709 .sp
710 The following table lists the error codes than may be returned by
711 \fBpcre_compile2()\fP, along with the error messages that may be returned by
712 both compiling functions. As PCRE has developed, some error codes have fallen
713 out of use. To avoid confusion, they have not been re-used.
714 .sp
715 0 no error
716 1 \e at end of pattern
717 2 \ec at end of pattern
718 3 unrecognized character follows \e
719 4 numbers out of order in {} quantifier
720 5 number too big in {} quantifier
721 6 missing terminating ] for character class
722 7 invalid escape sequence in character class
723 8 range out of order in character class
724 9 nothing to repeat
725 10 [this code is not in use]
726 11 internal error: unexpected repeat
727 12 unrecognized character after (? or (?-
728 13 POSIX named classes are supported only within a class
729 14 missing )
730 15 reference to non-existent subpattern
731 16 erroffset passed as NULL
732 17 unknown option bit(s) set
733 18 missing ) after comment
734 19 [this code is not in use]
735 20 regular expression is too large
736 21 failed to get memory
737 22 unmatched parentheses
738 23 internal error: code overflow
739 24 unrecognized character after (?<
740 25 lookbehind assertion is not fixed length
741 26 malformed number or name after (?(
742 27 conditional group contains more than two branches
743 28 assertion expected after (?(
744 29 (?R or (?[+-]digits must be followed by )
745 30 unknown POSIX class name
746 31 POSIX collating elements are not supported
747 32 this version of PCRE is not compiled with PCRE_UTF8 support
748 33 [this code is not in use]
749 34 character value in \ex{...} sequence is too large
750 35 invalid condition (?(0)
751 36 \eC not allowed in lookbehind assertion
752 37 PCRE does not support \eL, \el, \eN, \eU, or \eu
753 38 number after (?C is > 255
754 39 closing ) for (?C expected
755 40 recursive call could loop indefinitely
756 41 unrecognized character after (?P
757 42 syntax error in subpattern name (missing terminator)
758 43 two named subpatterns have the same name
759 44 invalid UTF-8 string
760 45 support for \eP, \ep, and \eX has not been compiled
761 46 malformed \eP or \ep sequence
762 47 unknown property name after \eP or \ep
763 48 subpattern name is too long (maximum 32 characters)
764 49 too many named subpatterns (maximum 10000)
765 50 [this code is not in use]
766 51 octal value is greater than \e377 (not in UTF-8 mode)
767 52 internal error: overran compiling workspace
768 53 internal error: previously-checked referenced subpattern
769 not found
770 54 DEFINE group contains more than one branch
771 55 repeating a DEFINE group is not allowed
772 56 inconsistent NEWLINE options
773 57 \eg is not followed by a braced, angle-bracketed, or quoted
774 name/number or by a plain number
775 58 a numbered reference must not be zero
776 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
777 60 (*VERB) not recognized
778 61 number is too big
779 62 subpattern name expected
780 63 digit expected after (?+
781 64 ] is an invalid data character in JavaScript compatibility mode
782 65 different names for subpatterns of the same number are
783 not allowed
784 66 (*MARK) must have an argument
785 67 this version of PCRE is not compiled with PCRE_UCP support
786 .sp
787 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
788 be used if the limits were changed when PCRE was built.
789 .
790 .
792 .rs
793 .sp
794 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
795 .ti +5n
796 .B const char **\fIerrptr\fP);
797 .PP
798 If a compiled pattern is going to be used several times, it is worth spending
799 more time analyzing it in order to speed up the time taken for matching. The
800 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
801 argument. If studying the pattern produces additional information that will
802 help speed up matching, \fBpcre_study()\fP returns a pointer to a
803 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
804 results of the study.
805 .P
806 The returned value from \fBpcre_study()\fP can be passed directly to
807 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
808 also contains other fields that can be set by the caller before the block is
809 passed; these are described
810 .\" HTML <a href="#extradata">
811 .\" </a>
812 below
813 .\"
814 in the section on matching a pattern.
815 .P
816 If studying the pattern does not produce any useful information,
817 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
818 wants to pass any of the other fields to \fBpcre_exec()\fP or
819 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
820 .P
821 The second argument of \fBpcre_study()\fP contains option bits. At present, no
822 options are defined, and this argument should always be zero.
823 .P
824 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
825 studying succeeds (even if no data is returned), the variable it points to is
826 set to NULL. Otherwise it is set to point to a textual error message. This is a
827 static string that is part of the library. You must not try to free it. You
828 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
829 sure that it has run successfully.
830 .P
831 This is a typical call to \fBpcre_study\fP():
832 .sp
833 pcre_extra *pe;
834 pe = pcre_study(
835 re, /* result of pcre_compile() */
836 0, /* no options exist */
837 &error); /* set to NULL or points to a message */
838 .sp
839 Studying a pattern does two things: first, a lower bound for the length of
840 subject string that is needed to match the pattern is computed. This does not
841 mean that there are any strings of that length that match, but it does
842 guarantee that no shorter strings match. The value is used by
843 \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
844 match strings that are shorter than the lower bound. You can find out the value
845 in a calling program via the \fBpcre_fullinfo()\fP function.
846 .P
847 Studying a pattern is also useful for non-anchored patterns that do not have a
848 single fixed starting character. A bitmap of possible starting bytes is
849 created. This speeds up finding a position in the subject at which to start
850 matching.
851 .P
852 The two optimizations just described can be disabled by setting the
853 PCRE_NO_START_OPTIMIZE option when calling \fBpcre_exec()\fP or
854 \fBpcre_dfa_exec()\fP. You might want to do this if your pattern contains
855 callouts, or make use of (*MARK), and you make use of these in cases where
856 matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
857 .\" HTML <a href="#execoptions">
858 .\" </a>
859 below.
860 .\"
861 .
862 .
863 .\" HTML <a name="localesupport"></a>
865 .rs
866 .sp
867 PCRE handles caseless matching, and determines whether characters are letters,
868 digits, or whatever, by reference to a set of tables, indexed by character
869 value. When running in UTF-8 mode, this applies only to characters with codes
870 less than 128. By default, higher-valued codes never match escapes such as \ew
871 or \ed, but they can be tested with \ep if PCRE is built with Unicode character
872 property support. Alternatively, the PCRE_UCP option can be set at compile
873 time; this causes \ew and friends to use Unicode property support instead of
874 built-in tables. The use of locales with Unicode is discouraged. If you are
875 handling characters with codes greater than 128, you should either use UTF-8
876 and Unicode, or use locales, but not try to mix the two.
877 .P
878 PCRE contains an internal set of tables that are used when the final argument
879 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
880 Normally, the internal tables recognize only ASCII characters. However, when
881 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
882 default "C" locale of the local system, which may cause them to be different.
883 .P
884 The internal tables can always be overridden by tables supplied by the
885 application that calls PCRE. These may be created in a different locale from
886 the default. As more and more applications change to using Unicode, the need
887 for this locale support is expected to die away.
888 .P
889 External tables are built by calling the \fBpcre_maketables()\fP function,
890 which has no arguments, in the relevant locale. The result can then be passed
891 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
892 example, to build and use tables that are appropriate for the French locale
893 (where accented characters with values greater than 128 are treated as letters),
894 the following code could be used:
895 .sp
896 setlocale(LC_CTYPE, "fr_FR");
897 tables = pcre_maketables();
898 re = pcre_compile(..., tables);
899 .sp
900 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
901 are using Windows, the name for the French locale is "french".
902 .P
903 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
904 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
905 that the memory containing the tables remains available for as long as it is
906 needed.
907 .P
908 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
909 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
910 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
911 pattern, compilation, studying and matching all happen in the same locale, but
912 different patterns can be compiled in different locales.
913 .P
914 It is possible to pass a table pointer or NULL (indicating the use of the
915 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
916 this facility could be used to match a pattern in a different locale from the
917 one in which it was compiled. Passing table pointers at run time is discussed
918 below in the section on matching a pattern.
919 .
920 .
922 .rs
923 .sp
924 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
925 .ti +5n
926 .B int \fIwhat\fP, void *\fIwhere\fP);
927 .PP
928 The \fBpcre_fullinfo()\fP function returns information about a compiled
929 pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
930 nevertheless retained for backwards compability (and is documented below).
931 .P
932 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
933 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
934 the pattern was not studied. The third argument specifies which piece of
935 information is required, and the fourth argument is a pointer to a variable
936 to receive the data. The yield of the function is zero for success, or one of
937 the following negative numbers:
938 .sp
939 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
940 the argument \fIwhere\fP was NULL
941 PCRE_ERROR_BADMAGIC the "magic number" was not found
942 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
943 .sp
944 The "magic number" is placed at the start of each compiled pattern as an simple
945 check against passing an arbitrary memory pointer. Here is a typical call of
946 \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
947 .sp
948 int rc;
949 size_t length;
950 rc = pcre_fullinfo(
951 re, /* result of pcre_compile() */
952 pe, /* result of pcre_study(), or NULL */
953 PCRE_INFO_SIZE, /* what is required */
954 &length); /* where to put the data */
955 .sp
956 The possible values for the third argument are defined in \fBpcre.h\fP, and are
957 as follows:
958 .sp
960 .sp
961 Return the number of the highest back reference in the pattern. The fourth
962 argument should point to an \fBint\fP variable. Zero is returned if there are
963 no back references.
964 .sp
966 .sp
967 Return the number of capturing subpatterns in the pattern. The fourth argument
968 should point to an \fBint\fP variable.
969 .sp
971 .sp
972 Return a pointer to the internal default character tables within PCRE. The
973 fourth argument should point to an \fBunsigned char *\fP variable. This
974 information call is provided for internal use by the \fBpcre_study()\fP
975 function. External callers can cause PCRE to use its internal tables by passing
976 a NULL table pointer.
977 .sp
979 .sp
980 Return information about the first byte of any matched string, for a
981 non-anchored pattern. The fourth argument should point to an \fBint\fP
982 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
983 still recognized for backwards compatibility.)
984 .P
985 If there is a fixed first byte, for example, from a pattern such as
986 (cat|cow|coyote), its value is returned. Otherwise, if either
987 .sp
988 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
989 starts with "^", or
990 .sp
991 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
992 (if it were set, the pattern would be anchored),
993 .sp
994 -1 is returned, indicating that the pattern matches only at the start of a
995 subject string or after any newline within the string. Otherwise -2 is
996 returned. For anchored patterns, -2 is returned.
997 .sp
999 .sp
1000 If the pattern was studied, and this resulted in the construction of a 256-bit
1001 table indicating a fixed set of bytes for the first byte in any matching
1002 string, a pointer to the table is returned. Otherwise NULL is returned. The
1003 fourth argument should point to an \fBunsigned char *\fP variable.
1004 .sp
1006 .sp
1007 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1008 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1009 explicit match is either a literal CR or LF character, or \er or \en.
1010 .sp
1012 .sp
1013 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1014 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1015 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1016 .sp
1018 .sp
1019 Return the value of the rightmost literal byte that must exist in any matched
1020 string, other than at its start, if such a byte has been recorded. The fourth
1021 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
1022 returned. For anchored patterns, a last literal byte is recorded only if it
1023 follows something of variable length. For example, for the pattern
1024 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1025 is -1.
1026 .sp
1028 .sp
1029 If the pattern was studied and a minimum length for matching subject strings
1030 was computed, its value is returned. Otherwise the returned value is -1. The
1031 value is a number of characters, not bytes (this may be relevant in UTF-8
1032 mode). The fourth argument should point to an \fBint\fP variable. A
1033 non-negative value is a lower bound to the length of any matching string. There
1034 may not be any strings of that length that do actually match, but every string
1035 that does match is at least that long.
1036 .sp
1040 .sp
1041 PCRE supports the use of named as well as numbered capturing parentheses. The
1042 names are just an additional way of identifying the parentheses, which still
1043 acquire numbers. Several convenience functions such as
1044 \fBpcre_get_named_substring()\fP are provided for extracting captured
1045 substrings by name. It is also possible to extract the data directly, by first
1046 converting the name to a number in order to access the correct pointers in the
1047 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1048 you need to use the name-to-number map, which is described by these three
1049 values.
1050 .P
1051 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1052 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1053 entry; both of these return an \fBint\fP value. The entry size depends on the
1054 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1055 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1056 are the number of the capturing parenthesis, most significant byte first. The
1057 rest of the entry is the corresponding name, zero terminated.
1058 .P
1059 The names are in alphabetical order. Duplicate names may appear if (?| is used
1060 to create multiple groups with the same number, as described in the
1061 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1062 .\" </a>
1063 section on duplicate subpattern numbers
1064 .\"
1065 in the
1066 .\" HREF
1067 \fBpcrepattern\fP
1068 .\"
1069 page. Duplicate names for subpatterns with different numbers are permitted only
1070 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1071 table in the order in which they were found in the pattern. In the absence of
1072 (?| this is the order of increasing number; when (?| is used this is not
1073 necessarily the case because later subpatterns may have lower numbers.
1074 .P
1075 As a simple example of the name/number table, consider the following pattern
1076 (assume PCRE_EXTENDED is set, so white space - including newlines - is
1077 ignored):
1078 .sp
1079 .\" JOIN
1080 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1081 (?<month>\ed\ed) - (?<day>\ed\ed) )
1082 .sp
1083 There are four named subpatterns, so the table has four entries, and each entry
1084 in the table is eight bytes long. The table is as follows, with non-printing
1085 bytes shows in hexadecimal, and undefined bytes shown as ??:
1086 .sp
1087 00 01 d a t e 00 ??
1088 00 05 d a y 00 ?? ??
1089 00 04 m o n t h 00
1090 00 02 y e a r 00 ??
1091 .sp
1092 When writing code to extract data from named subpatterns using the
1093 name-to-number map, remember that the length of the entries is likely to be
1094 different for each compiled pattern.
1095 .sp
1097 .sp
1098 Return 1 if the pattern can be used for partial matching with
1099 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1100 \fBint\fP variable. From release 8.00, this always returns 1, because the
1101 restrictions that previously applied to partial matching have been lifted. The
1102 .\" HREF
1103 \fBpcrepartial\fP
1104 .\"
1105 documentation gives details of partial matching.
1106 .sp
1108 .sp
1109 Return a copy of the options with which the pattern was compiled. The fourth
1110 argument should point to an \fBunsigned long int\fP variable. These option bits
1111 are those specified in the call to \fBpcre_compile()\fP, modified by any
1112 top-level option settings at the start of the pattern itself. In other words,
1113 they are the options that will be in force when matching starts. For example,
1114 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1116 .P
1117 A pattern is automatically anchored by PCRE if all of its top-level
1118 alternatives begin with one of the following:
1119 .sp
1120 ^ unless PCRE_MULTILINE is set
1121 \eA always
1122 \eG always
1123 .\" JOIN
1124 .* if PCRE_DOTALL is set and there are no back
1125 references to the subpattern in which .* appears
1126 .sp
1127 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1128 \fBpcre_fullinfo()\fP.
1129 .sp
1131 .sp
1132 Return the size of the compiled pattern, that is, the value that was passed as
1133 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1134 place the compiled data. The fourth argument should point to a \fBsize_t\fP
1135 variable.
1136 .sp
1138 .sp
1139 Return the size of the data block pointed to by the \fIstudy_data\fP field in
1140 a \fBpcre_extra\fP block. That is, it is the value that was passed to
1141 \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
1142 created by \fBpcre_study()\fP. If \fBpcre_extra\fP is NULL, or there is no
1143 study data, zero is returned. The fourth argument should point to a
1144 \fBsize_t\fP variable.
1145 .
1146 .
1148 .rs
1149 .sp
1150 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1151 .B *\fIfirstcharptr\fP);
1152 .PP
1153 The \fBpcre_info()\fP function is now obsolete because its interface is too
1154 restrictive to return all the available data about a compiled pattern. New
1155 programs should use \fBpcre_fullinfo()\fP instead. The yield of
1156 \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1157 following negative numbers:
1158 .sp
1159 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1160 PCRE_ERROR_BADMAGIC the "magic number" was not found
1161 .sp
1162 If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1163 pattern was compiled is placed in the integer it points to (see
1164 PCRE_INFO_OPTIONS above).
1165 .P
1166 If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1167 it is used to pass back information about the first character of any matched
1168 string (see PCRE_INFO_FIRSTBYTE above).
1169 .
1170 .
1172 .rs
1173 .sp
1174 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1175 .PP
1176 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1177 data block that contains a compiled pattern. It is provided for the benefit of
1178 applications that operate in an object-oriented manner, where different parts
1179 of the application may be using the same compiled pattern, but you want to free
1180 the block when they are all done.
1181 .P
1182 When a pattern is compiled, the reference count field is initialized to zero.
1183 It is changed only by calling this function, whose action is to add the
1184 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1185 function is the new value. However, the value of the count is constrained to
1186 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1187 it is forced to the appropriate limit value.
1188 .P
1189 Except when it is zero, the reference count is not correctly preserved if a
1190 pattern is compiled on one host and then transferred to a host whose byte-order
1191 is different. (This seems a highly unlikely scenario.)
1192 .
1193 .
1195 .rs
1196 .sp
1197 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1198 .ti +5n
1199 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1200 .ti +5n
1201 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1202 .P
1203 The function \fBpcre_exec()\fP is called to match a subject string against a
1204 compiled pattern, which is passed in the \fIcode\fP argument. If the
1205 pattern was studied, the result of the study should be passed in the
1206 \fIextra\fP argument. This function is the main matching facility of the
1207 library, and it operates in a Perl-like manner. For specialist use there is
1208 also an alternative matching function, which is described
1209 .\" HTML <a href="#dfamatch">
1210 .\" </a>
1211 below
1212 .\"
1213 in the section about the \fBpcre_dfa_exec()\fP function.
1214 .P
1215 In most applications, the pattern will have been compiled (and optionally
1216 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1217 possible to save compiled patterns and study data, and then use them later
1218 in different processes, possibly even on different hosts. For a discussion
1219 about this, see the
1220 .\" HREF
1221 \fBpcreprecompile\fP
1222 .\"
1223 documentation.
1224 .P
1225 Here is an example of a simple call to \fBpcre_exec()\fP:
1226 .sp
1227 int rc;
1228 int ovector[30];
1229 rc = pcre_exec(
1230 re, /* result of pcre_compile() */
1231 NULL, /* we didn't study the pattern */
1232 "some string", /* the subject string */
1233 11, /* the length of the subject string */
1234 0, /* start at offset 0 in the subject */
1235 0, /* default options */
1236 ovector, /* vector of integers for substring information */
1237 30); /* number of elements (NOT size in bytes) */
1238 .
1239 .\" HTML <a name="extradata"></a>
1240 .SS "Extra data for \fBpcre_exec()\fR"
1241 .rs
1242 .sp
1243 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1244 data block. The \fBpcre_study()\fP function returns such a block (when it
1245 doesn't return NULL), but you can also create one for yourself, and pass
1246 additional information in it. The \fBpcre_extra\fP block contains the following
1247 fields (not necessarily in this order):
1248 .sp
1249 unsigned long int \fIflags\fP;
1250 void *\fIstudy_data\fP;
1251 unsigned long int \fImatch_limit\fP;
1252 unsigned long int \fImatch_limit_recursion\fP;
1253 void *\fIcallout_data\fP;
1254 const unsigned char *\fItables\fP;
1255 unsigned char **\fImark\fP;
1256 .sp
1257 The \fIflags\fP field is a bitmap that specifies which of the other fields
1258 are set. The flag bits are:
1259 .sp
1266 .sp
1267 Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1268 \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1269 the appropriate flag bit. You should not set this yourself, but you may add to
1270 the block by setting the other fields and their corresponding flag bits.
1271 .P
1272 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1273 vast amount of resources when running patterns that are not going to match,
1274 but which have a very large number of possibilities in their search trees. The
1275 classic example is a pattern that uses nested unlimited repeats.
1276 .P
1277 Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1278 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1279 number of times this function is called during a match, which has the effect of
1280 limiting the amount of backtracking that can take place. For patterns that are
1281 not anchored, the count restarts from zero for each position in the subject
1282 string.
1283 .P
1284 The default value for the limit can be set when PCRE is built; the default
1285 default is 10 million, which handles all but the most extreme cases. You can
1286 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1287 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1288 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1290 .P
1291 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1292 instead of limiting the total number of times that \fBmatch()\fP is called, it
1293 limits the depth of recursion. The recursion depth is a smaller number than the
1294 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1295 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1296 .P
1297 Limiting the recursion depth limits the amount of stack that can be used, or,
1298 when PCRE has been compiled to use memory on the heap instead of the stack, the
1299 amount of heap memory that can be used.
1300 .P
1301 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1302 built; the default default is the same value as the default for
1303 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1304 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1305 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1306 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1307 .P
1308 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1309 and is described in the
1310 .\" HREF
1311 \fBpcrecallout\fP
1312 .\"
1313 documentation.
1314 .P
1315 The \fItables\fP field is used to pass a character tables pointer to
1316 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1317 pattern. A non-NULL value is stored with the compiled pattern only if custom
1318 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1319 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1320 internal tables to be used. This facility is helpful when re-using patterns
1321 that have been saved after compiling with an external set of tables, because
1322 the external tables might be at a different address when \fBpcre_exec()\fP is
1323 called. See the
1324 .\" HREF
1325 \fBpcreprecompile\fP
1326 .\"
1327 documentation for a discussion of saving compiled patterns for later use.
1328 .P
1329 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1330 be set to point to a \fBchar *\fP variable. If the pattern contains any
1331 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1332 a name to pass back, a pointer to the name string (zero terminated) is placed
1333 in the variable pointed to by the \fImark\fP field. The names are within the
1334 compiled pattern; if you wish to retain such a name you must copy it before
1335 freeing the memory of a compiled pattern. If there is no name to pass back, the
1336 variable pointed to by the \fImark\fP field set to NULL. For details of the
1337 backtracking control verbs, see the section entitled
1338 .\" HTML <a href="pcrepattern#backtrackcontrol">
1339 .\" </a>
1340 "Backtracking control"
1341 .\"
1342 in the
1343 .\" HREF
1344 \fBpcrepattern\fP
1345 .\"
1346 documentation.
1347 .
1348 .
1349 .\" HTML <a name="execoptions"></a>
1350 .SS "Option bits for \fBpcre_exec()\fP"
1351 .rs
1352 .sp
1353 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1354 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1358 .sp
1360 .sp
1361 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1362 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1363 to be anchored by virtue of its contents, it cannot be made unachored at
1364 matching time.
1365 .sp
1368 .sp
1369 These options (which are mutually exclusive) control what the \eR escape
1370 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1371 match any Unicode newline sequence. These options override the choice that was
1372 made or defaulted when the pattern was compiled.
1373 .sp
1379 .sp
1380 These options override the newline definition that was chosen or defaulted when
1381 the pattern was compiled. For details, see the description of
1382 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1383 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1384 the way the match position is advanced after a match failure for an unanchored
1385 pattern.
1386 .P
1388 match attempt for an unanchored pattern fails when the current position is at a
1389 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1390 characters, the match position is advanced by two characters instead of one, in
1391 other words, to after the CRLF.
1392 .P
1393 The above rule is a compromise that makes the most common cases work as
1394 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1395 set), it does not match the string "\er\enA" because, after failing at the
1396 start, it skips both the CR and the LF before retrying. However, the pattern
1397 [\er\en]A does match that string, because it contains an explicit CR or LF
1398 reference, and so advances only by one character after the first failure.
1399 .P
1400 An explicit match for CR of LF is either a literal appearance of one of those
1401 characters, or one of the \er or \en escape sequences. Implicit matches such as
1402 [^X] do not count, nor does \es (which includes CR and LF in the characters
1403 that it matches).
1404 .P
1405 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1406 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1407 .sp
1409 .sp
1410 This option specifies that first character of the subject string is not the
1411 beginning of a line, so the circumflex metacharacter should not match before
1412 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1413 never to match. This option affects only the behaviour of the circumflex
1414 metacharacter. It does not affect \eA.
1415 .sp
1417 .sp
1418 This option specifies that the end of the subject string is not the end of a
1419 line, so the dollar metacharacter should not match it nor (except in multiline
1420 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1421 compile time) causes dollar never to match. This option affects only the
1422 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1423 .sp
1425 .sp
1426 An empty string is not considered to be a valid match if this option is set. If
1427 there are alternatives in the pattern, they are tried. If all the alternatives
1428 match the empty string, the entire match fails. For example, if the pattern
1429 .sp
1430 a?b?
1431 .sp
1432 is applied to a string not beginning with "a" or "b", it matches an empty
1433 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1434 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1435 .sp
1437 .sp
1438 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1439 the start of the subject is permitted. If the pattern is anchored, such a match
1440 can occur only if the pattern contains \eK.
1441 .P
1442 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1443 does make a special case of a pattern match of the empty string within its
1444 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1445 emulate Perl's behaviour after matching a null string by first trying the match
1446 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1447 if that fails, by advancing the starting offset (see below) and trying an
1448 ordinary match again. There is some code that demonstrates how to do this in
1449 the
1450 .\" HREF
1451 \fBpcredemo\fP
1452 .\"
1453 sample program.
1454 .sp
1456 .sp
1457 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1458 a match, in order to speed up the process. For example, if it is known that an
1459 unanchored match must start with a specific character, it searches the subject
1460 for that character, and fails immediately if it cannot find it, without
1461 actually running the main matching function. This means that a special item
1462 such as (*COMMIT) at the start of a pattern is not considered until after a
1463 suitable starting point for the match has been found. When callouts or (*MARK)
1464 items are in use, these "start-up" optimizations can cause them to be skipped
1465 if the pattern is never actually used. The start-up optimizations are in effect
1466 a pre-scan of the subject that takes place before the pattern is run.
1467 .P
1468 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1469 causing performance to suffer, but ensuring that in cases where the result is
1470 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1471 are considered at every possible starting position in the subject string.
1472 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1473 Consider the pattern
1474 .sp
1476 .sp
1477 When this is compiled, PCRE records the fact that a match must start with the
1478 character "A". Suppose the subject string is "DEFABC". The start-up
1479 optimization scans along the subject, finds "A" and runs the first match
1480 attempt from there. The (*COMMIT) item means that the pattern must match the
1481 current starting position, which in this case, it does. However, if the same
1482 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1483 subject string does not happen. The first match attempt is run starting from
1484 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1485 the overall result is "no match". If the pattern is studied, more start-up
1486 optimizations may be used. For example, a minimum length for the subject may be
1487 recorded. Consider the pattern
1488 .sp
1489 (*MARK:A)(X|Y)
1490 .sp
1491 The minimum length for a match is one character. If the subject is "ABC", there
1492 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1493 If the pattern is studied, the final attempt does not take place, because PCRE
1494 knows that the subject is too short, and so the (*MARK) is never encountered.
1495 In this case, studying the pattern does not affect the overall match result,
1496 which is still "no match", but it does affect the auxiliary information that is
1497 returned.
1498 .sp
1500 .sp
1501 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1502 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1503 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1504 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1505 strings in the
1506 .\" HTML <a href="pcre.html#utf8strings">
1507 .\" </a>
1508 section on UTF-8 support
1509 .\"
1510 in the main
1511 .\" HREF
1512 \fBpcre\fP
1513 .\"
1514 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1515 the error PCRE_ERROR_BADUTF8. If \fIstartoffset\fP contains an invalid value,
1516 PCRE_ERROR_BADUTF8_OFFSET is returned.
1517 .P
1518 If you already know that your subject is valid, and you want to skip these
1519 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1520 calling \fBpcre_exec()\fP. You might want to do this for the second and
1521 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1522 all the matches in a single subject string. However, you should be sure that
1523 the value of \fIstartoffset\fP points to the start of a UTF-8 character. When
1524 PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
1525 subject, or a value of \fIstartoffset\fP that does not point to the start of a
1526 UTF-8 character, is undefined. Your program may crash.
1527 .sp
1530 .sp
1531 These options turn on the partial matching feature. For backwards
1532 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1533 occurs if the end of the subject string is reached successfully, but there are
1534 not enough subject characters to complete the match. If this happens when
1535 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1536 testing any remaining alternatives. Only if no complete match can be found is
1537 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1538 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1539 but only if no complete match can be found.
1540 .P
1541 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1542 partial match is found, \fBpcre_exec()\fP immediately returns
1543 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1544 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1545 important that an alternative complete match.
1546 .P
1547 In both cases, the portion of the string that was inspected when the partial
1548 match was found is set as the first matching string. There is a more detailed
1549 discussion of partial and multi-segment matching, with examples, in the
1550 .\" HREF
1551 \fBpcrepartial\fP
1552 .\"
1553 documentation.
1554 .
1555 .SS "The string to be matched by \fBpcre_exec()\fP"
1556 .rs
1557 .sp
1558 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1559 \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1560 in \fIstartoffset\fP. In UTF-8 mode, the byte offset must point to the start of
1561 a UTF-8 character. Unlike the pattern string, the subject may contain binary
1562 zero bytes. When the starting offset is zero, the search for a match starts at
1563 the beginning of the subject, and this is by far the most common case.
1564 .P
1565 A non-zero starting offset is useful when searching for another match in the
1566 same subject by calling \fBpcre_exec()\fP again after a previous success.
1567 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1568 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1569 lookbehind. For example, consider the pattern
1570 .sp
1571 \eBiss\eB
1572 .sp
1573 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1574 the current position in the subject is not a word boundary.) When applied to
1575 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1576 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1577 subject, namely "issipi", it does not match, because \eB is always false at the
1578 start of the subject, which is deemed to be a word boundary. However, if
1579 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1580 set to 4, it finds the second occurrence of "iss" because it is able to look
1581 behind the starting point to discover that it is preceded by a letter.
1582 .P
1583 If a non-zero starting offset is passed when the pattern is anchored, one
1584 attempt to match at the given offset is made. This can only succeed if the
1585 pattern does not require the match to be at the start of the subject.
1586 .
1587 .SS "How \fBpcre_exec()\fP returns captured substrings"
1588 .rs
1589 .sp
1590 In general, a pattern matches a certain portion of the subject, and in
1591 addition, further substrings from the subject may be picked out by parts of the
1592 pattern. Following the usage in Jeffrey Friedl's book, this is called
1593 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1594 a fragment of a pattern that picks out a substring. PCRE supports several other
1595 kinds of parenthesized subpattern that do not cause substrings to be captured.
1596 .P
1597 Captured substrings are returned to the caller via a vector of integers whose
1598 address is passed in \fIovector\fP. The number of elements in the vector is
1599 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1600 argument is NOT the size of \fIovector\fP in bytes.
1601 .P
1602 The first two-thirds of the vector is used to pass back captured substrings,
1603 each substring using a pair of integers. The remaining third of the vector is
1604 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1605 and is not available for passing back information. The number passed in
1606 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1607 rounded down.
1608 .P
1609 When a match is successful, information about captured substrings is returned
1610 in pairs of integers, starting at the beginning of \fIovector\fP, and
1611 continuing up to two-thirds of its length at the most. The first element of
1612 each pair is set to the byte offset of the first character in a substring, and
1613 the second is set to the byte offset of the first character after the end of a
1614 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1615 mode. They are not character counts.
1616 .P
1617 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1618 portion of the subject string matched by the entire pattern. The next pair is
1619 used for the first capturing subpattern, and so on. The value returned by
1620 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1621 For example, if two substrings have been captured, the returned value is 3. If
1622 there are no capturing subpatterns, the return value from a successful match is
1623 1, indicating that just the first pair of offsets has been set.
1624 .P
1625 If a capturing subpattern is matched repeatedly, it is the last portion of the
1626 string that it matched that is returned.
1627 .P
1628 If the vector is too small to hold all the captured substring offsets, it is
1629 used as far as possible (up to two-thirds of its length), and the function
1630 returns a value of zero. If the substring offsets are not of interest,
1631 \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1632 \fIovecsize\fP as zero. However, if the pattern contains back references and
1633 the \fIovector\fP is not big enough to remember the related substrings, PCRE
1634 has to get additional memory for use during matching. Thus it is usually
1635 advisable to supply an \fIovector\fP.
1636 .P
1637 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1638 subpatterns there are in a compiled pattern. The smallest size for
1639 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1640 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1641 .P
1642 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1643 the subject when subpattern \fIn\fP has not been used at all. For example, if
1644 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1645 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1646 happens, both values in the offset pairs corresponding to unused subpatterns
1647 are set to -1.
1648 .P
1649 Offset values that correspond to unused subpatterns at the end of the
1650 expression are also set to -1. For example, if the string "abc" is matched
1651 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1652 return from the function is 2, because the highest used capturing subpattern
1653 number is 1. However, you can refer to the offsets for the second and third
1654 capturing subpatterns if you wish (assuming the vector is large enough, of
1655 course).
1656 .P
1657 Some convenience functions are provided for extracting the captured substrings
1658 as separate strings. These are described below.
1659 .
1660 .\" HTML <a name="errorlist"></a>
1661 .SS "Error return values from \fBpcre_exec()\fP"
1662 .rs
1663 .sp
1664 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1665 defined in the header file:
1666 .sp
1668 .sp
1669 The subject string did not match the pattern.
1670 .sp
1672 .sp
1673 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1674 NULL and \fIovecsize\fP was not zero.
1675 .sp
1677 .sp
1678 An unrecognized bit was set in the \fIoptions\fP argument.
1679 .sp
1681 .sp
1682 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1683 the case when it is passed a junk pointer and to detect when a pattern that was
1684 compiled in an environment of one endianness is run in an environment with the
1685 other endianness. This is the error that PCRE gives when the magic number is
1686 not present.
1687 .sp
1689 .sp
1690 While running the pattern match, an unknown item was encountered in the
1691 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1692 of the compiled pattern.
1693 .sp
1695 .sp
1696 If a pattern contains back references, but the \fIovector\fP that is passed to
1697 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1698 gets a block of memory at the start of matching to use for this purpose. If the
1699 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1700 automatically freed at the end of matching.
1701 .P
1702 This error is also given if \fBpcre_stack_malloc()\fP fails in
1703 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
1704 \fB--disable-stack-for-recursion\fP.
1705 .sp
1707 .sp
1708 This error is used by the \fBpcre_copy_substring()\fP,
1709 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1710 below). It is never returned by \fBpcre_exec()\fP.
1711 .sp
1713 .sp
1714 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1715 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1716 above.
1717 .sp
1719 .sp
1720 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1721 use by callout functions that want to yield a distinctive error code. See the
1722 .\" HREF
1723 \fBpcrecallout\fP
1724 .\"
1725 documentation for details.
1726 .sp
1728 .sp
1729 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1730 .sp
1732 .sp
1733 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1734 of \fIstartoffset\fP did not point to the beginning of a UTF-8 character.
1735 .sp
1737 .sp
1738 The subject string did not match, but it did match partially. See the
1739 .\" HREF
1740 \fBpcrepartial\fP
1741 .\"
1742 documentation for details of partial matching.
1743 .sp
1745 .sp
1746 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1747 option was used with a compiled pattern containing items that were not
1748 supported for partial matching. From release 8.00 onwards, there are no
1749 restrictions on partial matching.
1750 .sp
1752 .sp
1753 An unexpected internal error has occurred. This error could be caused by a bug
1754 in PCRE or by overwriting of the compiled pattern.
1755 .sp
1757 .sp
1758 This error is given if the value of the \fIovecsize\fP argument is negative.
1759 .sp
1761 .sp
1762 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1763 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1764 description above.
1765 .sp
1767 .sp
1768 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1769 .P
1770 Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1771 .
1772 .
1774 .rs
1775 .sp
1776 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1777 .ti +5n
1778 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1779 .ti +5n
1780 .B int \fIbuffersize\fP);
1781 .PP
1782 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1783 .ti +5n
1784 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1785 .ti +5n
1786 .B const char **\fIstringptr\fP);
1787 .PP
1788 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1789 .ti +5n
1790 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1791 .PP
1792 Captured substrings can be accessed directly by using the offsets returned by
1793 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1794 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1795 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1796 as new, separate, zero-terminated strings. These functions identify substrings
1797 by number. The next section describes functions for extracting named
1798 substrings.
1799 .P
1800 A substring that contains a binary zero is correctly extracted and has a
1801 further zero added on the end, but the result is not, of course, a C string.
1802 However, you can process such a string by referring to the length that is
1803 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1804 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1805 for handling strings containing binary zeros, because the end of the final
1806 string is not independently indicated.
1807 .P
1808 The first three arguments are the same for all three of these functions:
1809 \fIsubject\fP is the subject string that has just been successfully matched,
1810 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
1811 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
1812 captured by the match, including the substring that matched the entire regular
1813 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
1814 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
1815 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
1816 number of elements in the vector divided by three.
1817 .P
1818 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1819 extract a single substring, whose number is given as \fIstringnumber\fP. A
1820 value of zero extracts the substring that matched the entire pattern, whereas
1821 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1822 the string is placed in \fIbuffer\fP, whose length is given by
1823 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
1824 obtained via \fBpcre_malloc\fP, and its address is returned via
1825 \fIstringptr\fP. The yield of the function is the length of the string, not
1826 including the terminating zero, or one of these error codes:
1827 .sp
1829 .sp
1830 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
1831 memory failed for \fBpcre_get_substring()\fP.
1832 .sp
1834 .sp
1835 There is no substring whose number is \fIstringnumber\fP.
1836 .P
1837 The \fBpcre_get_substring_list()\fP function extracts all available substrings
1838 and builds a list of pointers to them. All this is done in a single block of
1839 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
1840 is returned via \fIlistptr\fP, which is also the start of the list of string
1841 pointers. The end of the list is marked by a NULL pointer. The yield of the
1842 function is zero if all went well, or the error code
1843 .sp
1845 .sp
1846 if the attempt to get the memory block failed.
1847 .P
1848 When any of these functions encounter a substring that is unset, which can
1849 happen when capturing subpattern number \fIn+1\fP matches some part of the
1850 subject, but subpattern \fIn\fP has not been used at all, they return an empty
1851 string. This can be distinguished from a genuine zero-length substring by
1852 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1853 substrings.
1854 .P
1855 The two convenience functions \fBpcre_free_substring()\fP and
1856 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
1857 a previous call of \fBpcre_get_substring()\fP or
1858 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1859 the function pointed to by \fBpcre_free\fP, which of course could be called
1860 directly from a C program. However, PCRE is used in some situations where it is
1861 linked via a special interface to another programming language that cannot use
1862 \fBpcre_free\fP directly; it is for these cases that the functions are
1863 provided.
1864 .
1865 .
1867 .rs
1868 .sp
1869 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1870 .ti +5n
1871 .B const char *\fIname\fP);
1872 .PP
1873 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1874 .ti +5n
1875 .B const char *\fIsubject\fP, int *\fIovector\fP,
1876 .ti +5n
1877 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1878 .ti +5n
1879 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1880 .PP
1881 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
1882 .ti +5n
1883 .B const char *\fIsubject\fP, int *\fIovector\fP,
1884 .ti +5n
1885 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1886 .ti +5n
1887 .B const char **\fIstringptr\fP);
1888 .PP
1889 To extract a substring by name, you first have to find associated number.
1890 For example, for this pattern
1891 .sp
1892 (a+)b(?<xxx>\ed+)...
1893 .sp
1894 the number of the subpattern called "xxx" is 2. If the name is known to be
1895 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1896 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1897 pattern, and the second is the name. The yield of the function is the
1898 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1899 that name.
1900 .P
1901 Given the number, you can extract the substring directly, or use one of the
1902 functions described in the previous section. For convenience, there are also
1903 two functions that do the whole job.
1904 .P
1905 Most of the arguments of \fBpcre_copy_named_substring()\fP and
1906 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1907 functions that extract by number. As these are described in the previous
1908 section, they are not re-described here. There are just two differences:
1909 .P
1910 First, instead of a substring number, a substring name is given. Second, there
1911 is an extra argument, given at the start, which is a pointer to the compiled
1912 pattern. This is needed in order to gain access to the name-to-number
1913 translation table.
1914 .P
1915 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1916 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1917 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1918 the behaviour may not be what you want (see the next section).
1919 .P
1920 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
1921 subpatterns with the same number, as described in the
1922 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1923 .\" </a>
1924 section on duplicate subpattern numbers
1925 .\"
1926 in the
1927 .\" HREF
1928 \fBpcrepattern\fP
1929 .\"
1930 page, you cannot use names to distinguish the different subpatterns, because
1931 names are not included in the compiled code. The matching process uses only
1932 numbers. For this reason, the use of different names for subpatterns of the
1933 same number causes an error at compile time.
1934 .
1936 .rs
1937 .sp
1938 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1939 .ti +5n
1940 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
1941 .PP
1942 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1943 are not required to be unique. (Duplicate names are always allowed for
1944 subpatterns with the same number, created by using the (?| feature. Indeed, if
1945 such subpatterns are named, they are required to use the same names.)
1946 .P
1947 Normally, patterns with duplicate names are such that in any one match, only
1948 one of the named subpatterns participates. An example is shown in the
1949 .\" HREF
1950 \fBpcrepattern\fP
1951 .\"
1952 documentation.
1953 .P
1954 When duplicates are present, \fBpcre_copy_named_substring()\fP and
1955 \fBpcre_get_named_substring()\fP return the first substring corresponding to
1956 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
1957 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
1958 returns one of the numbers that are associated with the name, but it is not
1959 defined which it is.
1960 .P
1961 If you want to get full details of all captured substrings for a given name,
1962 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
1963 argument is the compiled pattern, and the second is the name. The third and
1964 fourth are pointers to variables which are updated by the function. After it
1965 has run, they point to the first and last entries in the name-to-number table
1966 for the given name. The function itself returns the length of each entry, or
1967 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
1968 described above in the section entitled \fIInformation about a pattern\fP.
1969 Given all the relevant entries for the name, you can extract each of their
1970 numbers, and hence the captured data, if any.
1971 .
1972 .
1974 .rs
1975 .sp
1976 The traditional matching function uses a similar algorithm to Perl, which stops
1977 when it finds the first match, starting at a given point in the subject. If you
1978 want to find all possible matches, or the longest possible match, consider
1979 using the alternative matching function (see below) instead. If you cannot use
1980 the alternative function, but still need to find all possible matches, you
1981 can kludge it up by making use of the callout facility, which is described in
1982 the
1983 .\" HREF
1984 \fBpcrecallout\fP
1985 .\"
1986 documentation.
1987 .P
1988 What you have to do is to insert a callout right at the end of the pattern.
1989 When your callout function is called, extract and save the current matched
1990 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
1991 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
1992 will yield PCRE_ERROR_NOMATCH.
1993 .
1994 .
1995 .\" HTML <a name="dfamatch"></a>
1997 .rs
1998 .sp
1999 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2000 .ti +5n
2001 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2002 .ti +5n
2003 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2004 .ti +5n
2005 .B int *\fIworkspace\fP, int \fIwscount\fP);
2006 .P
2007 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2008 a compiled pattern, using a matching algorithm that scans the subject string
2009 just once, and does not backtrack. This has different characteristics to the
2010 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2011 patterns are not supported. Nevertheless, there are times when this kind of
2012 matching can be useful. For a discussion of the two matching algorithms, and a
2013 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2014 .\" HREF
2015 \fBpcrematching\fP
2016 .\"
2017 documentation.
2018 .P
2019 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2020 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2021 different way, and this is described below. The other common arguments are used
2022 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2023 here.
2024 .P
2025 The two additional arguments provide workspace for the function. The workspace
2026 vector should contain at least 20 elements. It is used for keeping track of
2027 multiple paths through the pattern tree. More workspace will be needed for
2028 patterns and subjects where there are a lot of potential matches.
2029 .P
2030 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2031 .sp
2032 int rc;
2033 int ovector[10];
2034 int wspace[20];
2035 rc = pcre_dfa_exec(
2036 re, /* result of pcre_compile() */
2037 NULL, /* we didn't study the pattern */
2038 "some string", /* the subject string */
2039 11, /* the length of the subject string */
2040 0, /* start at offset 0 in the subject */
2041 0, /* default options */
2042 ovector, /* vector of integers for substring information */
2043 10, /* number of elements (NOT size in bytes) */
2044 wspace, /* working space vector */
2045 20); /* number of elements (NOT size in bytes) */
2046 .
2047 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2048 .rs
2049 .sp
2050 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2051 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2055 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2056 so their description is not repeated here.
2057 .sp
2060 .sp
2061 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2062 details are slightly different. When PCRE_PARTIAL_HARD is set for
2063 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2064 is reached and there is still at least one matching possibility that requires
2065 additional characters. This happens even if some complete matches have also
2066 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2067 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2068 there have been no complete matches, but there is still at least one matching
2069 possibility. The portion of the string that was inspected when the longest
2070 partial match was found is set as the first matching string in both cases.
2071 There is a more detailed discussion of partial and multi-segment matching, with
2072 examples, in the
2073 .\" HREF
2074 \fBpcrepartial\fP
2075 .\"
2076 documentation.
2077 .sp
2079 .sp
2080 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2081 soon as it has found one match. Because of the way the alternative algorithm
2082 works, this is necessarily the shortest possible match at the first possible
2083 matching point in the subject string.
2084 .sp
2086 .sp
2087 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2088 again, with additional subject characters, and have it continue with the same
2089 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2090 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2091 before because data about the match so far is left in them after a partial
2092 match. There is more discussion of this facility in the
2093 .\" HREF
2094 \fBpcrepartial\fP
2095 .\"
2096 documentation.
2097 .
2098 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2099 .rs
2100 .sp
2101 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2102 substring in the subject. Note, however, that all the matches from one run of
2103 the function start at the same point in the subject. The shorter matches are
2104 all initial substrings of the longer matches. For example, if the pattern
2105 .sp
2106 <.*>
2107 .sp
2108 is matched against the string
2109 .sp
2110 This is <something> <something else> <something further> no more
2111 .sp
2112 the three matched strings are
2113 .sp
2114 <something>
2115 <something> <something else>
2116 <something> <something else> <something further>
2117 .sp
2118 On success, the yield of the function is a number greater than zero, which is
2119 the number of matched substrings. The substrings themselves are returned in
2120 \fIovector\fP. Each string uses two elements; the first is the offset to the
2121 start, and the second is the offset to the end. In fact, all the strings have
2122 the same start offset. (Space could have been saved by giving this only once,
2123 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2124 returns data, even though the meaning of the strings is different.)
2125 .P
2126 The strings are returned in reverse order of length; that is, the longest
2127 matching string is given first. If there were too many matches to fit into
2128 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2129 the longest matches.
2130 .
2131 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2132 .rs
2133 .sp
2134 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2135 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2136 described
2137 .\" HTML <a href="#errorlist">
2138 .\" </a>
2139 above.
2140 .\"
2141 There are in addition the following errors that are specific to
2142 \fBpcre_dfa_exec()\fP:
2143 .sp
2145 .sp
2146 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2147 that it does not support, for instance, the use of \eC or a back reference.
2148 .sp
2150 .sp
2151 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2152 uses a back reference for the condition, or a test for recursion in a specific
2153 group. These are not supported.
2154 .sp
2156 .sp
2157 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2158 block that contains a setting of the \fImatch_limit\fP field. This is not
2159 supported (it is meaningless).
2160 .sp
2162 .sp
2163 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2164 \fIworkspace\fP vector.
2165 .sp
2167 .sp
2168 When a recursive subpattern is processed, the matching function calls itself
2169 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2170 error is given if the output vector is not large enough. This should be
2171 extremely rare, as a vector of size 1000 is used.
2172 .
2173 .
2174 .SH "SEE ALSO"
2175 .rs
2176 .sp
2177 \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2178 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2179 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2180 .
2181 .
2183 .rs
2184 .sp
2185 .nf
2186 Philip Hazel
2187 University Computing Service
2188 Cambridge CB2 3QH, England.
2189 .fi
2190 .
2191 .
2193 .rs
2194 .sp
2195 .nf
2196 Last updated: 22 October 2010
2197 Copyright (c) 1997-2010 University of Cambridge.
2198 .fi


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

  ViewVC Help
Powered by ViewVC 1.1.5