ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 510 - (show annotations)
Sat Mar 27 17:45:29 2010 UTC (10 years, 8 months ago) by ph10
File size: 90982 byte(s)
Error occurred while calculating annotation data.
Add support for *MARK and names for *PRUNE, *SKIP, *THEN.
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 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
136 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
137 in a Perl-compatible manner. A sample program that demonstrates the simplest
138 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
139 source distribution. A listing of this program is given in the
140 .\" HREF
141 \fBpcredemo\fP
142 .\"
143 documentation, and the
144 .\" HREF
145 \fBpcresample\fP
146 .\"
147 documentation describes how to compile and run it.
148 .P
149 A second matching function, \fBpcre_dfa_exec()\fP, which is not
150 Perl-compatible, is also provided. This uses a different algorithm for the
151 matching. The alternative algorithm finds all possible matches (at a given
152 point in the subject), and scans the subject just once (unless there are
153 lookbehind assertions). However, this algorithm does not return captured
154 substrings. A description of the two matching algorithms and their advantages
155 and disadvantages is given in the
156 .\" HREF
157 \fBpcrematching\fP
158 .\"
159 documentation.
160 .P
161 In addition to the main compiling and matching functions, there are convenience
162 functions for extracting captured substrings from a subject string that is
163 matched by \fBpcre_exec()\fP. They are:
164 .sp
165 \fBpcre_copy_substring()\fP
166 \fBpcre_copy_named_substring()\fP
167 \fBpcre_get_substring()\fP
168 \fBpcre_get_named_substring()\fP
169 \fBpcre_get_substring_list()\fP
170 \fBpcre_get_stringnumber()\fP
171 \fBpcre_get_stringtable_entries()\fP
172 .sp
173 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
174 provided, to free the memory used for extracted strings.
175 .P
176 The function \fBpcre_maketables()\fP is used to build a set of character tables
177 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
178 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
179 specialist use. Most commonly, no special tables are passed, in which case
180 internal tables that are generated when PCRE is built are used.
181 .P
182 The function \fBpcre_fullinfo()\fP is used to find out information about a
183 compiled pattern; \fBpcre_info()\fP is an obsolete version that returns only
184 some of the available information, but is retained for backwards compatibility.
185 The function \fBpcre_version()\fP returns a pointer to a string containing the
186 version of PCRE and its date of release.
187 .P
188 The function \fBpcre_refcount()\fP maintains a reference count in a data block
189 containing a compiled pattern. This is provided for the benefit of
190 object-oriented applications.
191 .P
192 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
193 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
194 respectively. PCRE calls the memory management functions via these variables,
195 so a calling program can replace them if it wishes to intercept the calls. This
196 should be done before calling any PCRE functions.
197 .P
198 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
199 indirections to memory management functions. These special functions are used
200 only when PCRE is compiled to use the heap for remembering data, instead of
201 recursive function calls, when running the \fBpcre_exec()\fP function. See the
202 .\" HREF
203 \fBpcrebuild\fP
204 .\"
205 documentation for details of how to do this. It is a non-standard way of
206 building PCRE, for use in environments that have limited stacks. Because of the
207 greater use of memory management, it runs more slowly. Separate functions are
208 provided so that special-purpose external code can be used for this case. When
209 used, these functions are always called in a stack-like manner (last obtained,
210 first freed), and always for memory blocks of the same size. There is a
211 discussion about PCRE's stack usage in the
212 .\" HREF
213 \fBpcrestack\fP
214 .\"
215 documentation.
216 .P
217 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
218 by the caller to a "callout" function, which PCRE will then call at specified
219 points during a matching operation. Details are given in the
220 .\" HREF
221 \fBpcrecallout\fP
222 .\"
223 documentation.
224 .
225 .
226 .\" HTML <a name="newlines"></a>
228 .rs
229 .sp
230 PCRE supports five different conventions for indicating line breaks in
231 strings: a single CR (carriage return) character, a single LF (linefeed)
232 character, the two-character sequence CRLF, any of the three preceding, or any
233 Unicode newline sequence. The Unicode newline sequences are the three just
234 mentioned, plus the single characters VT (vertical tab, U+000B), FF (formfeed,
235 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
236 (paragraph separator, U+2029).
237 .P
238 Each of the first three conventions is used by at least one operating system as
239 its standard newline sequence. When PCRE is built, a default can be specified.
240 The default default is LF, which is the Unix standard. When PCRE is run, the
241 default can be overridden, either when a pattern is compiled, or when it is
242 matched.
243 .P
244 At compile time, the newline convention can be specified by the \fIoptions\fP
245 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
246 start of the pattern itself; this overrides any other settings. See the
247 .\" HREF
248 \fBpcrepattern\fP
249 .\"
250 page for details of the special character sequences.
251 .P
252 In the PCRE documentation the word "newline" is used to mean "the character or
253 pair of characters that indicate a line break". The choice of newline
254 convention affects the handling of the dot, circumflex, and dollar
255 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
256 recognized line ending sequence, the match position advancement for a
257 non-anchored pattern. There is more detail about this in the
258 .\" HTML <a href="#execoptions">
259 .\" </a>
260 section on \fBpcre_exec()\fP options
261 .\"
262 below.
263 .P
264 The choice of newline convention does not affect the interpretation of
265 the \en or \er escape sequences, nor does it affect what \eR matches, which is
266 controlled in a similar way, but by separate options.
267 .
268 .
270 .rs
271 .sp
272 The PCRE functions can be used in multi-threading applications, with the
273 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
274 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
275 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
276 .P
277 The compiled form of a regular expression is not altered during matching, so
278 the same compiled pattern can safely be used by several threads at once.
279 .
280 .
282 .rs
283 .sp
284 The compiled form of a regular expression can be saved and re-used at a later
285 time, possibly by a different program, and even on a host other than the one on
286 which it was compiled. Details are given in the
287 .\" HREF
288 \fBpcreprecompile\fP
289 .\"
290 documentation. However, compiling a regular expression with one version of PCRE
291 for use with a different version is not guaranteed to work and may cause
292 crashes.
293 .
294 .
296 .rs
297 .sp
298 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
299 .PP
300 The function \fBpcre_config()\fP makes it possible for a PCRE client to
301 discover which optional features have been compiled into the PCRE library. The
302 .\" HREF
303 \fBpcrebuild\fP
304 .\"
305 documentation has more details about these optional features.
306 .P
307 The first argument for \fBpcre_config()\fP is an integer, specifying which
308 information is required; the second argument is a pointer to a variable into
309 which the information is placed. The following information is available:
310 .sp
312 .sp
313 The output is an integer that is set to one if UTF-8 support is available;
314 otherwise it is set to zero.
315 .sp
317 .sp
318 The output is an integer that is set to one if support for Unicode character
319 properties is available; otherwise it is set to zero.
320 .sp
322 .sp
323 The output is an integer whose value specifies the default character sequence
324 that is recognized as meaning "newline". The four values that are supported
325 are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for ANYCRLF, and -1 for ANY.
326 Though they are derived from ASCII, the same values are returned in EBCDIC
327 environments. The default should normally correspond to the standard sequence
328 for your operating system.
329 .sp
331 .sp
332 The output is an integer whose value indicates what character sequences the \eR
333 escape sequence matches by default. A value of 0 means that \eR matches any
334 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
335 or CRLF. The default can be overridden when a pattern is compiled or matched.
336 .sp
338 .sp
339 The output is an integer that contains the number of bytes used for internal
340 linkage in compiled regular expressions. The value is 2, 3, or 4. Larger values
341 allow larger regular expressions to be compiled, at the expense of slower
342 matching. The default value of 2 is sufficient for all but the most massive
343 patterns, since it allows the compiled pattern to be up to 64K in size.
344 .sp
346 .sp
347 The output is an integer that contains the threshold above which the POSIX
348 interface uses \fBmalloc()\fP for output vectors. Further details are given in
349 the
350 .\" HREF
351 \fBpcreposix\fP
352 .\"
353 documentation.
354 .sp
356 .sp
357 The output is a long integer that gives the default limit for the number of
358 internal matching function calls in a \fBpcre_exec()\fP execution. Further
359 details are given with \fBpcre_exec()\fP below.
360 .sp
362 .sp
363 The output is a long integer that gives the default limit for the depth of
364 recursion when calling the internal matching function in a \fBpcre_exec()\fP
365 execution. Further details are given with \fBpcre_exec()\fP below.
366 .sp
368 .sp
369 The output is an integer that is set to one if internal recursion when running
370 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
371 to remember their state. This is the usual way that PCRE is compiled. The
372 output is zero if PCRE was compiled to use blocks of data on the heap instead
373 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
374 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
375 avoiding the use of the stack.
376 .
377 .
379 .rs
380 .sp
381 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
382 .ti +5n
383 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
384 .ti +5n
385 .B const unsigned char *\fItableptr\fP);
386 .sp
387 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
388 .ti +5n
389 .B int *\fIerrorcodeptr\fP,
390 .ti +5n
391 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
392 .ti +5n
393 .B const unsigned char *\fItableptr\fP);
394 .P
395 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
396 called to compile a pattern into an internal form. The only difference between
397 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
398 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
399 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
400 information applies equally to \fBpcre_compile2()\fP.
401 .P
402 The pattern is a C string terminated by a binary zero, and is passed in the
403 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
404 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
405 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
406 for a structure whose contents are not externally defined. It is up to the
407 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
408 .P
409 Although the compiled code of a PCRE regex is relocatable, that is, it does not
410 depend on memory location, the complete \fBpcre\fP data block is not
411 fully relocatable, because it may contain a copy of the \fItableptr\fP
412 argument, which is an address (see below).
413 .P
414 The \fIoptions\fP argument contains various bit settings that affect the
415 compilation. It should be zero if no options are required. The available
416 options are described below. Some of them (in particular, those that are
417 compatible with Perl, but some others as well) can also be set and unset from
418 within the pattern (see the detailed description in the
419 .\" HREF
420 \fBpcrepattern\fP
421 .\"
422 documentation). For those options that can be different in different parts of
423 the pattern, the contents of the \fIoptions\fP argument specifies their
424 settings at the start of compilation and execution. The PCRE_ANCHORED,
425 PCRE_BSR_\fIxxx\fP, and PCRE_NEWLINE_\fIxxx\fP options can be set at the time
426 of matching as well as at compile time.
427 .P
428 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
429 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
430 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
431 error message. This is a static string that is part of the library. You must
432 not try to free it. The byte offset from the start of the pattern to the
433 character that was being processed when the error was discovered is placed in
434 the variable pointed to by \fIerroffset\fP, which must not be NULL. If it is,
435 an immediate error is given. Some errors are not detected until checks are
436 carried out when the whole pattern has been scanned; in this case the offset is
437 set to the end of the pattern.
438 .P
439 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
440 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
441 returned via this argument in the event of an error. This is in addition to the
442 textual error message. Error codes and messages are listed below.
443 .P
444 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
445 character tables that are built when PCRE is compiled, using the default C
446 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
447 call to \fBpcre_maketables()\fP. This value is stored with the compiled
448 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
449 passed to it. For more discussion, see the section on locale support below.
450 .P
451 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
452 .sp
453 pcre *re;
454 const char *error;
455 int erroffset;
456 re = pcre_compile(
457 "^A.*Z", /* the pattern */
458 0, /* default options */
459 &error, /* for error message */
460 &erroffset, /* for error offset */
461 NULL); /* use default character tables */
462 .sp
463 The following names for option bits are defined in the \fBpcre.h\fP header
464 file:
465 .sp
467 .sp
468 If this bit is set, the pattern is forced to be "anchored", that is, it is
469 constrained to match only at the first matching point in the string that is
470 being searched (the "subject string"). This effect can also be achieved by
471 appropriate constructs in the pattern itself, which is the only way to do it in
472 Perl.
473 .sp
475 .sp
476 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
477 all with number 255, before each pattern item. For discussion of the callout
478 facility, see the
479 .\" HREF
480 \fBpcrecallout\fP
481 .\"
482 documentation.
483 .sp
486 .sp
487 These options (which are mutually exclusive) control what the \eR escape
488 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
489 match any Unicode newline sequence. The default is specified when PCRE is
490 built. It can be overridden from within the pattern, or by setting an option
491 when a compiled pattern is matched.
492 .sp
494 .sp
495 If this bit is set, letters in the pattern match both upper and lower case
496 letters. It is equivalent to Perl's /i option, and it can be changed within a
497 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
498 concept of case for characters whose values are less than 128, so caseless
499 matching is always possible. For characters with higher values, the concept of
500 case is supported if PCRE is compiled with Unicode property support, but not
501 otherwise. If you want to use caseless matching for characters 128 and above,
502 you must ensure that PCRE is compiled with Unicode property support as well as
503 with UTF-8 support.
504 .sp
506 .sp
507 If this bit is set, a dollar metacharacter in the pattern matches only at the
508 end of the subject string. Without this option, a dollar also matches
509 immediately before a newline at the end of the string (but not before any other
510 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
511 There is no equivalent to this option in Perl, and no way to set it within a
512 pattern.
513 .sp
515 .sp
516 If this bit is set, a dot metacharater in the pattern matches all characters,
517 including those that indicate newline. Without it, a dot does not match when
518 the current position is at a newline. This option is equivalent to Perl's /s
519 option, and it can be changed within a pattern by a (?s) option setting. A
520 negative class such as [^a] always matches newline characters, independent of
521 the setting of this option.
522 .sp
524 .sp
525 If this bit is set, names used to identify capturing subpatterns need not be
526 unique. This can be helpful for certain types of pattern when it is known that
527 only one instance of the named subpattern can ever be matched. There are more
528 details of named subpatterns below; see also the
529 .\" HREF
530 \fBpcrepattern\fP
531 .\"
532 documentation.
533 .sp
535 .sp
536 If this bit is set, whitespace data characters in the pattern are totally
537 ignored except when escaped or inside a character class. Whitespace does not
538 include the VT character (code 11). In addition, characters between an
539 unescaped # outside a character class and the next newline, inclusive, are also
540 ignored. This is equivalent to Perl's /x option, and it can be changed within a
541 pattern by a (?x) option setting.
542 .P
543 This option makes it possible to include comments inside complicated patterns.
544 Note, however, that this applies only to data characters. Whitespace characters
545 may never appear within special character sequences in a pattern, for example
546 within the sequence (?( which introduces a conditional subpattern.
547 .sp
549 .sp
550 This option was invented in order to turn on additional functionality of PCRE
551 that is incompatible with Perl, but it is currently of very little use. When
552 set, any backslash in a pattern that is followed by a letter that has no
553 special meaning causes an error, thus reserving these combinations for future
554 expansion. By default, as in Perl, a backslash followed by a letter with no
555 special meaning is treated as a literal. (Perl can, however, be persuaded to
556 give a warning for this.) There are at present no other features controlled by
557 this option. It can also be set by a (?X) option setting within a pattern.
558 .sp
560 .sp
561 If this option is set, an unanchored pattern is required to match before or at
562 the first newline in the subject string, though the matched text may continue
563 over the newline.
564 .sp
566 .sp
567 If this option is set, PCRE's behaviour is changed in some ways so that it is
568 compatible with JavaScript rather than Perl. The changes are as follows:
569 .P
570 (1) A lone closing square bracket in a pattern causes a compile-time error,
571 because this is illegal in JavaScript (by default it is treated as a data
572 character). Thus, the pattern AB]CD becomes illegal when this option is set.
573 .P
574 (2) At run time, a back reference to an unset subpattern group matches an empty
575 string (by default this causes the current matching alternative to fail). A
576 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
577 an "a" in the subject), whereas it fails by default, for Perl compatibility.
578 .sp
580 .sp
581 By default, PCRE treats the subject string as consisting of a single line of
582 characters (even if it actually contains newlines). The "start of line"
583 metacharacter (^) matches only at the start of the string, while the "end of
584 line" metacharacter ($) matches only at the end of the string, or before a
585 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
586 Perl.
587 .P
588 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
589 match immediately following or immediately before internal newlines in the
590 subject string, respectively, as well as at the very start and end. This is
591 equivalent to Perl's /m option, and it can be changed within a pattern by a
592 (?m) option setting. If there are no newlines in a subject string, or no
593 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
594 .sp
600 .sp
601 These options override the default newline definition that was chosen when PCRE
602 was built. Setting the first or the second specifies that a newline is
603 indicated by a single character (CR or LF, respectively). Setting
604 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
605 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
606 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
607 that any Unicode newline sequence should be recognized. The Unicode newline
608 sequences are the three just mentioned, plus the single characters VT (vertical
609 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
610 separator, U+2028), and PS (paragraph separator, U+2029). The last two are
611 recognized only in UTF-8 mode.
612 .P
613 The newline setting in the options word uses three bits that are treated
614 as a number, giving eight possibilities. Currently only six are used (default
615 plus the five values above). This means that if you set more than one newline
616 option, the combination may or may not be sensible. For example,
618 other combinations may yield unused numbers and cause an error.
619 .P
620 The only time that a line break is specially recognized when compiling a
621 pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
622 class is encountered. This indicates a comment that lasts until after the next
623 line break sequence. In other circumstances, line break sequences are treated
624 as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
625 as whitespace characters and are therefore ignored.
626 .P
627 The newline option that is set at compile time becomes the default that is used
628 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
629 .sp
631 .sp
632 If this option is set, it disables the use of numbered capturing parentheses in
633 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
634 were followed by ?: but named parentheses can still be used for capturing (and
635 they acquire numbers in the usual way). There is no equivalent of this option
636 in Perl.
637 .sp
639 .sp
640 This option inverts the "greediness" of the quantifiers so that they are not
641 greedy by default, but become greedy if followed by "?". It is not compatible
642 with Perl. It can also be set by a (?U) option setting within the pattern.
643 .sp
645 .sp
646 This option causes PCRE to regard both the pattern and the subject as strings
647 of UTF-8 characters instead of single-byte character strings. However, it is
648 available only when PCRE is built to include UTF-8 support. If not, the use
649 of this option provokes an error. Details of how this option changes the
650 behaviour of PCRE are given in the
651 .\" HTML <a href="pcre.html#utf8support">
652 .\" </a>
653 section on UTF-8 support
654 .\"
655 in the main
656 .\" HREF
657 \fBpcre\fP
658 .\"
659 page.
660 .sp
662 .sp
663 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
664 automatically checked. There is a discussion about the
665 .\" HTML <a href="pcre.html#utf8strings">
666 .\" </a>
667 validity of UTF-8 strings
668 .\"
669 in the main
670 .\" HREF
671 \fBpcre\fP
672 .\"
673 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
674 returns an error. If you already know that your pattern is valid, and you want
675 to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
676 option. When it is set, the effect of passing an invalid UTF-8 string as a
677 pattern is undefined. It may cause your program to crash. Note that this option
678 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
679 the UTF-8 validity checking of subject strings.
680 .
681 .
683 .rs
684 .sp
685 The following table lists the error codes than may be returned by
686 \fBpcre_compile2()\fP, along with the error messages that may be returned by
687 both compiling functions. As PCRE has developed, some error codes have fallen
688 out of use. To avoid confusion, they have not been re-used.
689 .sp
690 0 no error
691 1 \e at end of pattern
692 2 \ec at end of pattern
693 3 unrecognized character follows \e
694 4 numbers out of order in {} quantifier
695 5 number too big in {} quantifier
696 6 missing terminating ] for character class
697 7 invalid escape sequence in character class
698 8 range out of order in character class
699 9 nothing to repeat
700 10 [this code is not in use]
701 11 internal error: unexpected repeat
702 12 unrecognized character after (? or (?-
703 13 POSIX named classes are supported only within a class
704 14 missing )
705 15 reference to non-existent subpattern
706 16 erroffset passed as NULL
707 17 unknown option bit(s) set
708 18 missing ) after comment
709 19 [this code is not in use]
710 20 regular expression is too large
711 21 failed to get memory
712 22 unmatched parentheses
713 23 internal error: code overflow
714 24 unrecognized character after (?<
715 25 lookbehind assertion is not fixed length
716 26 malformed number or name after (?(
717 27 conditional group contains more than two branches
718 28 assertion expected after (?(
719 29 (?R or (?[+-]digits must be followed by )
720 30 unknown POSIX class name
721 31 POSIX collating elements are not supported
722 32 this version of PCRE is not compiled with PCRE_UTF8 support
723 33 [this code is not in use]
724 34 character value in \ex{...} sequence is too large
725 35 invalid condition (?(0)
726 36 \eC not allowed in lookbehind assertion
727 37 PCRE does not support \eL, \el, \eN, \eU, or \eu
728 38 number after (?C is > 255
729 39 closing ) for (?C expected
730 40 recursive call could loop indefinitely
731 41 unrecognized character after (?P
732 42 syntax error in subpattern name (missing terminator)
733 43 two named subpatterns have the same name
734 44 invalid UTF-8 string
735 45 support for \eP, \ep, and \eX has not been compiled
736 46 malformed \eP or \ep sequence
737 47 unknown property name after \eP or \ep
738 48 subpattern name is too long (maximum 32 characters)
739 49 too many named subpatterns (maximum 10000)
740 50 [this code is not in use]
741 51 octal value is greater than \e377 (not in UTF-8 mode)
742 52 internal error: overran compiling workspace
743 53 internal error: previously-checked referenced subpattern not found
744 54 DEFINE group contains more than one branch
745 55 repeating a DEFINE group is not allowed
746 56 inconsistent NEWLINE options
747 57 \eg is not followed by a braced, angle-bracketed, or quoted
748 name/number or by a plain number
749 58 a numbered reference must not be zero
750 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
751 60 (*VERB) not recognized
752 61 number is too big
753 62 subpattern name expected
754 63 digit expected after (?+
755 64 ] is an invalid data character in JavaScript compatibility mode
756 65 different names for subpatterns of the same number are not allowed
757 66 (*MARK) must have an argument
758 .sp
759 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
760 be used if the limits were changed when PCRE was built.
761 .
762 .
764 .rs
765 .sp
766 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
767 .ti +5n
768 .B const char **\fIerrptr\fP);
769 .PP
770 If a compiled pattern is going to be used several times, it is worth spending
771 more time analyzing it in order to speed up the time taken for matching. The
772 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
773 argument. If studying the pattern produces additional information that will
774 help speed up matching, \fBpcre_study()\fP returns a pointer to a
775 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
776 results of the study.
777 .P
778 The returned value from \fBpcre_study()\fP can be passed directly to
779 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
780 also contains other fields that can be set by the caller before the block is
781 passed; these are described
782 .\" HTML <a href="#extradata">
783 .\" </a>
784 below
785 .\"
786 in the section on matching a pattern.
787 .P
788 If studying the pattern does not produce any useful information,
789 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
790 wants to pass any of the other fields to \fBpcre_exec()\fP or
791 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
792 .P
793 The second argument of \fBpcre_study()\fP contains option bits. At present, no
794 options are defined, and this argument should always be zero.
795 .P
796 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
797 studying succeeds (even if no data is returned), the variable it points to is
798 set to NULL. Otherwise it is set to point to a textual error message. This is a
799 static string that is part of the library. You must not try to free it. You
800 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
801 sure that it has run successfully.
802 .P
803 This is a typical call to \fBpcre_study\fP():
804 .sp
805 pcre_extra *pe;
806 pe = pcre_study(
807 re, /* result of pcre_compile() */
808 0, /* no options exist */
809 &error); /* set to NULL or points to a message */
810 .sp
811 Studying a pattern does two things: first, a lower bound for the length of
812 subject string that is needed to match the pattern is computed. This does not
813 mean that there are any strings of that length that match, but it does
814 guarantee that no shorter strings match. The value is used by
815 \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
816 match strings that are shorter than the lower bound. You can find out the value
817 in a calling program via the \fBpcre_fullinfo()\fP function.
818 .P
819 Studying a pattern is also useful for non-anchored patterns that do not have a
820 single fixed starting character. A bitmap of possible starting bytes is
821 created. This speeds up finding a position in the subject at which to start
822 matching.
823 .
824 .
825 .\" HTML <a name="localesupport"></a>
827 .rs
828 .sp
829 PCRE handles caseless matching, and determines whether characters are letters,
830 digits, or whatever, by reference to a set of tables, indexed by character
831 value. When running in UTF-8 mode, this applies only to characters with codes
832 less than 128. Higher-valued codes never match escapes such as \ew or \ed, but
833 can be tested with \ep if PCRE is built with Unicode character property
834 support. The use of locales with Unicode is discouraged. If you are handling
835 characters with codes greater than 128, you should either use UTF-8 and
836 Unicode, or use locales, but not try to mix the two.
837 .P
838 PCRE contains an internal set of tables that are used when the final argument
839 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
840 Normally, the internal tables recognize only ASCII characters. However, when
841 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
842 default "C" locale of the local system, which may cause them to be different.
843 .P
844 The internal tables can always be overridden by tables supplied by the
845 application that calls PCRE. These may be created in a different locale from
846 the default. As more and more applications change to using Unicode, the need
847 for this locale support is expected to die away.
848 .P
849 External tables are built by calling the \fBpcre_maketables()\fP function,
850 which has no arguments, in the relevant locale. The result can then be passed
851 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
852 example, to build and use tables that are appropriate for the French locale
853 (where accented characters with values greater than 128 are treated as letters),
854 the following code could be used:
855 .sp
856 setlocale(LC_CTYPE, "fr_FR");
857 tables = pcre_maketables();
858 re = pcre_compile(..., tables);
859 .sp
860 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
861 are using Windows, the name for the French locale is "french".
862 .P
863 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
864 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
865 that the memory containing the tables remains available for as long as it is
866 needed.
867 .P
868 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
869 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
870 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
871 pattern, compilation, studying and matching all happen in the same locale, but
872 different patterns can be compiled in different locales.
873 .P
874 It is possible to pass a table pointer or NULL (indicating the use of the
875 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
876 this facility could be used to match a pattern in a different locale from the
877 one in which it was compiled. Passing table pointers at run time is discussed
878 below in the section on matching a pattern.
879 .
880 .
882 .rs
883 .sp
884 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
885 .ti +5n
886 .B int \fIwhat\fP, void *\fIwhere\fP);
887 .PP
888 The \fBpcre_fullinfo()\fP function returns information about a compiled
889 pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
890 nevertheless retained for backwards compability (and is documented below).
891 .P
892 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
893 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
894 the pattern was not studied. The third argument specifies which piece of
895 information is required, and the fourth argument is a pointer to a variable
896 to receive the data. The yield of the function is zero for success, or one of
897 the following negative numbers:
898 .sp
899 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
900 the argument \fIwhere\fP was NULL
901 PCRE_ERROR_BADMAGIC the "magic number" was not found
902 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
903 .sp
904 The "magic number" is placed at the start of each compiled pattern as an simple
905 check against passing an arbitrary memory pointer. Here is a typical call of
906 \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
907 .sp
908 int rc;
909 size_t length;
910 rc = pcre_fullinfo(
911 re, /* result of pcre_compile() */
912 pe, /* result of pcre_study(), or NULL */
913 PCRE_INFO_SIZE, /* what is required */
914 &length); /* where to put the data */
915 .sp
916 The possible values for the third argument are defined in \fBpcre.h\fP, and are
917 as follows:
918 .sp
920 .sp
921 Return the number of the highest back reference in the pattern. The fourth
922 argument should point to an \fBint\fP variable. Zero is returned if there are
923 no back references.
924 .sp
926 .sp
927 Return the number of capturing subpatterns in the pattern. The fourth argument
928 should point to an \fBint\fP variable.
929 .sp
931 .sp
932 Return a pointer to the internal default character tables within PCRE. The
933 fourth argument should point to an \fBunsigned char *\fP variable. This
934 information call is provided for internal use by the \fBpcre_study()\fP
935 function. External callers can cause PCRE to use its internal tables by passing
936 a NULL table pointer.
937 .sp
939 .sp
940 Return information about the first byte of any matched string, for a
941 non-anchored pattern. The fourth argument should point to an \fBint\fP
942 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
943 still recognized for backwards compatibility.)
944 .P
945 If there is a fixed first byte, for example, from a pattern such as
946 (cat|cow|coyote), its value is returned. Otherwise, if either
947 .sp
948 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
949 starts with "^", or
950 .sp
951 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
952 (if it were set, the pattern would be anchored),
953 .sp
954 -1 is returned, indicating that the pattern matches only at the start of a
955 subject string or after any newline within the string. Otherwise -2 is
956 returned. For anchored patterns, -2 is returned.
957 .sp
959 .sp
960 If the pattern was studied, and this resulted in the construction of a 256-bit
961 table indicating a fixed set of bytes for the first byte in any matching
962 string, a pointer to the table is returned. Otherwise NULL is returned. The
963 fourth argument should point to an \fBunsigned char *\fP variable.
964 .sp
966 .sp
967 Return 1 if the pattern contains any explicit matches for CR or LF characters,
968 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
969 explicit match is either a literal CR or LF character, or \er or \en.
970 .sp
972 .sp
973 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
974 0. The fourth argument should point to an \fBint\fP variable. (?J) and
975 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
976 .sp
978 .sp
979 Return the value of the rightmost literal byte that must exist in any matched
980 string, other than at its start, if such a byte has been recorded. The fourth
981 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
982 returned. For anchored patterns, a last literal byte is recorded only if it
983 follows something of variable length. For example, for the pattern
984 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
985 is -1.
986 .sp
988 .sp
989 If the pattern was studied and a minimum length for matching subject strings
990 was computed, its value is returned. Otherwise the returned value is -1. The
991 value is a number of characters, not bytes (this may be relevant in UTF-8
992 mode). The fourth argument should point to an \fBint\fP variable. A
993 non-negative value is a lower bound to the length of any matching string. There
994 may not be any strings of that length that do actually match, but every string
995 that does match is at least that long.
996 .sp
1000 .sp
1001 PCRE supports the use of named as well as numbered capturing parentheses. The
1002 names are just an additional way of identifying the parentheses, which still
1003 acquire numbers. Several convenience functions such as
1004 \fBpcre_get_named_substring()\fP are provided for extracting captured
1005 substrings by name. It is also possible to extract the data directly, by first
1006 converting the name to a number in order to access the correct pointers in the
1007 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1008 you need to use the name-to-number map, which is described by these three
1009 values.
1010 .P
1011 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1012 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1013 entry; both of these return an \fBint\fP value. The entry size depends on the
1014 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1015 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1016 are the number of the capturing parenthesis, most significant byte first. The
1017 rest of the entry is the corresponding name, zero terminated.
1018 .P
1019 The names are in alphabetical order. Duplicate names may appear if (?| is used
1020 to create multiple groups with the same number, as described in the
1021 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1022 .\" </a>
1023 section on duplicate subpattern numbers
1024 .\"
1025 in the
1026 .\" HREF
1027 \fBpcrepattern\fP
1028 .\"
1029 page. Duplicate names for subpatterns with different numbers are permitted only
1030 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1031 table in the order in which they were found in the pattern. In the absence of
1032 (?| this is the order of increasing number; when (?| is used this is not
1033 necessarily the case because later subpatterns may have lower numbers.
1034 .P
1035 As a simple example of the name/number table, consider the following pattern
1036 (assume PCRE_EXTENDED is set, so white space - including newlines - is
1037 ignored):
1038 .sp
1039 .\" JOIN
1040 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1041 (?<month>\ed\ed) - (?<day>\ed\ed) )
1042 .sp
1043 There are four named subpatterns, so the table has four entries, and each entry
1044 in the table is eight bytes long. The table is as follows, with non-printing
1045 bytes shows in hexadecimal, and undefined bytes shown as ??:
1046 .sp
1047 00 01 d a t e 00 ??
1048 00 05 d a y 00 ?? ??
1049 00 04 m o n t h 00
1050 00 02 y e a r 00 ??
1051 .sp
1052 When writing code to extract data from named subpatterns using the
1053 name-to-number map, remember that the length of the entries is likely to be
1054 different for each compiled pattern.
1055 .sp
1057 .sp
1058 Return 1 if the pattern can be used for partial matching with
1059 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1060 \fBint\fP variable. From release 8.00, this always returns 1, because the
1061 restrictions that previously applied to partial matching have been lifted. The
1062 .\" HREF
1063 \fBpcrepartial\fP
1064 .\"
1065 documentation gives details of partial matching.
1066 .sp
1068 .sp
1069 Return a copy of the options with which the pattern was compiled. The fourth
1070 argument should point to an \fBunsigned long int\fP variable. These option bits
1071 are those specified in the call to \fBpcre_compile()\fP, modified by any
1072 top-level option settings at the start of the pattern itself. In other words,
1073 they are the options that will be in force when matching starts. For example,
1074 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1076 .P
1077 A pattern is automatically anchored by PCRE if all of its top-level
1078 alternatives begin with one of the following:
1079 .sp
1080 ^ unless PCRE_MULTILINE is set
1081 \eA always
1082 \eG always
1083 .\" JOIN
1084 .* if PCRE_DOTALL is set and there are no back
1085 references to the subpattern in which .* appears
1086 .sp
1087 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1088 \fBpcre_fullinfo()\fP.
1089 .sp
1091 .sp
1092 Return the size of the compiled pattern, that is, the value that was passed as
1093 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1094 place the compiled data. The fourth argument should point to a \fBsize_t\fP
1095 variable.
1096 .sp
1098 .sp
1099 Return the size of the data block pointed to by the \fIstudy_data\fP field in
1100 a \fBpcre_extra\fP block. That is, it is the value that was passed to
1101 \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
1102 created by \fBpcre_study()\fP. If \fBpcre_extra\fP is NULL, or there is no
1103 study data, zero is returned. The fourth argument should point to a
1104 \fBsize_t\fP variable.
1105 .
1106 .
1108 .rs
1109 .sp
1110 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1111 .B *\fIfirstcharptr\fP);
1112 .PP
1113 The \fBpcre_info()\fP function is now obsolete because its interface is too
1114 restrictive to return all the available data about a compiled pattern. New
1115 programs should use \fBpcre_fullinfo()\fP instead. The yield of
1116 \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1117 following negative numbers:
1118 .sp
1119 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1120 PCRE_ERROR_BADMAGIC the "magic number" was not found
1121 .sp
1122 If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1123 pattern was compiled is placed in the integer it points to (see
1124 PCRE_INFO_OPTIONS above).
1125 .P
1126 If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1127 it is used to pass back information about the first character of any matched
1128 string (see PCRE_INFO_FIRSTBYTE above).
1129 .
1130 .
1132 .rs
1133 .sp
1134 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1135 .PP
1136 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1137 data block that contains a compiled pattern. It is provided for the benefit of
1138 applications that operate in an object-oriented manner, where different parts
1139 of the application may be using the same compiled pattern, but you want to free
1140 the block when they are all done.
1141 .P
1142 When a pattern is compiled, the reference count field is initialized to zero.
1143 It is changed only by calling this function, whose action is to add the
1144 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1145 function is the new value. However, the value of the count is constrained to
1146 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1147 it is forced to the appropriate limit value.
1148 .P
1149 Except when it is zero, the reference count is not correctly preserved if a
1150 pattern is compiled on one host and then transferred to a host whose byte-order
1151 is different. (This seems a highly unlikely scenario.)
1152 .
1153 .
1155 .rs
1156 .sp
1157 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1158 .ti +5n
1159 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1160 .ti +5n
1161 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1162 .P
1163 The function \fBpcre_exec()\fP is called to match a subject string against a
1164 compiled pattern, which is passed in the \fIcode\fP argument. If the
1165 pattern was studied, the result of the study should be passed in the
1166 \fIextra\fP argument. This function is the main matching facility of the
1167 library, and it operates in a Perl-like manner. For specialist use there is
1168 also an alternative matching function, which is described
1169 .\" HTML <a href="#dfamatch">
1170 .\" </a>
1171 below
1172 .\"
1173 in the section about the \fBpcre_dfa_exec()\fP function.
1174 .P
1175 In most applications, the pattern will have been compiled (and optionally
1176 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1177 possible to save compiled patterns and study data, and then use them later
1178 in different processes, possibly even on different hosts. For a discussion
1179 about this, see the
1180 .\" HREF
1181 \fBpcreprecompile\fP
1182 .\"
1183 documentation.
1184 .P
1185 Here is an example of a simple call to \fBpcre_exec()\fP:
1186 .sp
1187 int rc;
1188 int ovector[30];
1189 rc = pcre_exec(
1190 re, /* result of pcre_compile() */
1191 NULL, /* we didn't study the pattern */
1192 "some string", /* the subject string */
1193 11, /* the length of the subject string */
1194 0, /* start at offset 0 in the subject */
1195 0, /* default options */
1196 ovector, /* vector of integers for substring information */
1197 30); /* number of elements (NOT size in bytes) */
1198 .
1199 .\" HTML <a name="extradata"></a>
1200 .SS "Extra data for \fBpcre_exec()\fR"
1201 .rs
1202 .sp
1203 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1204 data block. The \fBpcre_study()\fP function returns such a block (when it
1205 doesn't return NULL), but you can also create one for yourself, and pass
1206 additional information in it. The \fBpcre_extra\fP block contains the following
1207 fields (not necessarily in this order):
1208 .sp
1209 unsigned long int \fIflags\fP;
1210 void *\fIstudy_data\fP;
1211 unsigned long int \fImatch_limit\fP;
1212 unsigned long int \fImatch_limit_recursion\fP;
1213 void *\fIcallout_data\fP;
1214 const unsigned char *\fItables\fP;
1215 unsigned char **\fImark\fP;
1216 .sp
1217 The \fIflags\fP field is a bitmap that specifies which of the other fields
1218 are set. The flag bits are:
1219 .sp
1226 .sp
1227 Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1228 \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1229 the appropriate flag bit. You should not set this yourself, but you may add to
1230 the block by setting the other fields and their corresponding flag bits.
1231 .P
1232 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1233 vast amount of resources when running patterns that are not going to match,
1234 but which have a very large number of possibilities in their search trees. The
1235 classic example is a pattern that uses nested unlimited repeats.
1236 .P
1237 Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1238 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1239 number of times this function is called during a match, which has the effect of
1240 limiting the amount of backtracking that can take place. For patterns that are
1241 not anchored, the count restarts from zero for each position in the subject
1242 string.
1243 .P
1244 The default value for the limit can be set when PCRE is built; the default
1245 default is 10 million, which handles all but the most extreme cases. You can
1246 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1247 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1248 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1250 .P
1251 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1252 instead of limiting the total number of times that \fBmatch()\fP is called, it
1253 limits the depth of recursion. The recursion depth is a smaller number than the
1254 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1255 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1256 .P
1257 Limiting the recursion depth limits the amount of stack that can be used, or,
1258 when PCRE has been compiled to use memory on the heap instead of the stack, the
1259 amount of heap memory that can be used.
1260 .P
1261 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1262 built; the default default is the same value as the default for
1263 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1264 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1265 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1266 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1267 .P
1268 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1269 and is described in the
1270 .\" HREF
1271 \fBpcrecallout\fP
1272 .\"
1273 documentation.
1274 .P
1275 The \fItables\fP field is used to pass a character tables pointer to
1276 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1277 pattern. A non-NULL value is stored with the compiled pattern only if custom
1278 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1279 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1280 internal tables to be used. This facility is helpful when re-using patterns
1281 that have been saved after compiling with an external set of tables, because
1282 the external tables might be at a different address when \fBpcre_exec()\fP is
1283 called. See the
1284 .\" HREF
1285 \fBpcreprecompile\fP
1286 .\"
1287 documentation for a discussion of saving compiled patterns for later use.
1288 .P
1289 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1290 be set to point to a \fBchar *\fP variable. If the pattern contains any
1291 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1292 a name to pass back, a pointer to the name string (zero terminated) is placed
1293 in the variable pointed to by the \fImark\fP field. The names are within the
1294 compiled pattern; if you wish to retain such a name you must copy it before
1295 freeing the memory of a compiled pattern. If there is no name to pass back, the
1296 variable pointed to by the \fImark\fP field set to NULL. For details of the
1297 backtracking control verbs, see the section entitled
1298 .\" HTML <a href="pcrepattern#backtrackcontrol">
1299 .\" </a>
1300 "Backtracking control"
1301 .\"
1302 in the
1303 .\" HREF
1304 \fBpcrepattern\fP
1305 .\"
1306 documentation.
1307 .
1308 .
1309 .\" HTML <a name="execoptions"></a>
1310 .SS "Option bits for \fBpcre_exec()\fP"
1311 .rs
1312 .sp
1313 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1314 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1318 .sp
1320 .sp
1321 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1322 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1323 to be anchored by virtue of its contents, it cannot be made unachored at
1324 matching time.
1325 .sp
1328 .sp
1329 These options (which are mutually exclusive) control what the \eR escape
1330 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1331 match any Unicode newline sequence. These options override the choice that was
1332 made or defaulted when the pattern was compiled.
1333 .sp
1339 .sp
1340 These options override the newline definition that was chosen or defaulted when
1341 the pattern was compiled. For details, see the description of
1342 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1343 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1344 the way the match position is advanced after a match failure for an unanchored
1345 pattern.
1346 .P
1348 match attempt for an unanchored pattern fails when the current position is at a
1349 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1350 characters, the match position is advanced by two characters instead of one, in
1351 other words, to after the CRLF.
1352 .P
1353 The above rule is a compromise that makes the most common cases work as
1354 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1355 set), it does not match the string "\er\enA" because, after failing at the
1356 start, it skips both the CR and the LF before retrying. However, the pattern
1357 [\er\en]A does match that string, because it contains an explicit CR or LF
1358 reference, and so advances only by one character after the first failure.
1359 .P
1360 An explicit match for CR of LF is either a literal appearance of one of those
1361 characters, or one of the \er or \en escape sequences. Implicit matches such as
1362 [^X] do not count, nor does \es (which includes CR and LF in the characters
1363 that it matches).
1364 .P
1365 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1366 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1367 .sp
1369 .sp
1370 This option specifies that first character of the subject string is not the
1371 beginning of a line, so the circumflex metacharacter should not match before
1372 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1373 never to match. This option affects only the behaviour of the circumflex
1374 metacharacter. It does not affect \eA.
1375 .sp
1377 .sp
1378 This option specifies that the end of the subject string is not the end of a
1379 line, so the dollar metacharacter should not match it nor (except in multiline
1380 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1381 compile time) causes dollar never to match. This option affects only the
1382 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1383 .sp
1385 .sp
1386 An empty string is not considered to be a valid match if this option is set. If
1387 there are alternatives in the pattern, they are tried. If all the alternatives
1388 match the empty string, the entire match fails. For example, if the pattern
1389 .sp
1390 a?b?
1391 .sp
1392 is applied to a string not beginning with "a" or "b", it matches an empty
1393 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1394 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1395 .sp
1397 .sp
1398 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1399 the start of the subject is permitted. If the pattern is anchored, such a match
1400 can occur only if the pattern contains \eK.
1401 .P
1402 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1403 does make a special case of a pattern match of the empty string within its
1404 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1405 emulate Perl's behaviour after matching a null string by first trying the match
1406 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1407 if that fails, by advancing the starting offset (see below) and trying an
1408 ordinary match again. There is some code that demonstrates how to do this in
1409 the
1410 .\" HREF
1411 \fBpcredemo\fP
1412 .\"
1413 sample program.
1414 .sp
1416 .sp
1417 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1418 a match, in order to speed up the process. For example, if it is known that a
1419 match must start with a specific character, it searches the subject for that
1420 character, and fails immediately if it cannot find it, without actually running
1421 the main matching function. When callouts are in use, these optimizations can
1422 cause them to be skipped. This option disables the "start-up" optimizations,
1423 causing performance to suffer, but ensuring that the callouts do occur.
1424 .sp
1426 .sp
1427 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1428 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1429 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1430 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1431 strings in the
1432 .\" HTML <a href="pcre.html#utf8strings">
1433 .\" </a>
1434 section on UTF-8 support
1435 .\"
1436 in the main
1437 .\" HREF
1438 \fBpcre\fP
1439 .\"
1440 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1441 the error PCRE_ERROR_BADUTF8. If \fIstartoffset\fP contains an invalid value,
1442 PCRE_ERROR_BADUTF8_OFFSET is returned.
1443 .P
1444 If you already know that your subject is valid, and you want to skip these
1445 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1446 calling \fBpcre_exec()\fP. You might want to do this for the second and
1447 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1448 all the matches in a single subject string. However, you should be sure that
1449 the value of \fIstartoffset\fP points to the start of a UTF-8 character. When
1450 PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
1451 subject, or a value of \fIstartoffset\fP that does not point to the start of a
1452 UTF-8 character, is undefined. Your program may crash.
1453 .sp
1456 .sp
1457 These options turn on the partial matching feature. For backwards
1458 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1459 occurs if the end of the subject string is reached successfully, but there are
1460 not enough subject characters to complete the match. If this happens when
1461 PCRE_PARTIAL_HARD is set, \fBpcre_exec()\fP immediately returns
1462 PCRE_ERROR_PARTIAL. Otherwise, if PCRE_PARTIAL_SOFT is set, matching continues
1463 by testing any other alternatives. Only if they all fail is PCRE_ERROR_PARTIAL
1464 returned (instead of PCRE_ERROR_NOMATCH). The portion of the string that
1465 was inspected when the partial match was found is set as the first matching
1466 string. There is a more detailed discussion in the
1467 .\" HREF
1468 \fBpcrepartial\fP
1469 .\"
1470 documentation.
1471 .
1472 .SS "The string to be matched by \fBpcre_exec()\fP"
1473 .rs
1474 .sp
1475 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1476 \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1477 in \fIstartoffset\fP. In UTF-8 mode, the byte offset must point to the start of
1478 a UTF-8 character. Unlike the pattern string, the subject may contain binary
1479 zero bytes. When the starting offset is zero, the search for a match starts at
1480 the beginning of the subject, and this is by far the most common case.
1481 .P
1482 A non-zero starting offset is useful when searching for another match in the
1483 same subject by calling \fBpcre_exec()\fP again after a previous success.
1484 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1485 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1486 lookbehind. For example, consider the pattern
1487 .sp
1488 \eBiss\eB
1489 .sp
1490 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1491 the current position in the subject is not a word boundary.) When applied to
1492 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1493 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1494 subject, namely "issipi", it does not match, because \eB is always false at the
1495 start of the subject, which is deemed to be a word boundary. However, if
1496 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1497 set to 4, it finds the second occurrence of "iss" because it is able to look
1498 behind the starting point to discover that it is preceded by a letter.
1499 .P
1500 If a non-zero starting offset is passed when the pattern is anchored, one
1501 attempt to match at the given offset is made. This can only succeed if the
1502 pattern does not require the match to be at the start of the subject.
1503 .
1504 .SS "How \fBpcre_exec()\fP returns captured substrings"
1505 .rs
1506 .sp
1507 In general, a pattern matches a certain portion of the subject, and in
1508 addition, further substrings from the subject may be picked out by parts of the
1509 pattern. Following the usage in Jeffrey Friedl's book, this is called
1510 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1511 a fragment of a pattern that picks out a substring. PCRE supports several other
1512 kinds of parenthesized subpattern that do not cause substrings to be captured.
1513 .P
1514 Captured substrings are returned to the caller via a vector of integers whose
1515 address is passed in \fIovector\fP. The number of elements in the vector is
1516 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1517 argument is NOT the size of \fIovector\fP in bytes.
1518 .P
1519 The first two-thirds of the vector is used to pass back captured substrings,
1520 each substring using a pair of integers. The remaining third of the vector is
1521 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1522 and is not available for passing back information. The number passed in
1523 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1524 rounded down.
1525 .P
1526 When a match is successful, information about captured substrings is returned
1527 in pairs of integers, starting at the beginning of \fIovector\fP, and
1528 continuing up to two-thirds of its length at the most. The first element of
1529 each pair is set to the byte offset of the first character in a substring, and
1530 the second is set to the byte offset of the first character after the end of a
1531 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1532 mode. They are not character counts.
1533 .P
1534 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1535 portion of the subject string matched by the entire pattern. The next pair is
1536 used for the first capturing subpattern, and so on. The value returned by
1537 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1538 For example, if two substrings have been captured, the returned value is 3. If
1539 there are no capturing subpatterns, the return value from a successful match is
1540 1, indicating that just the first pair of offsets has been set.
1541 .P
1542 If a capturing subpattern is matched repeatedly, it is the last portion of the
1543 string that it matched that is returned.
1544 .P
1545 If the vector is too small to hold all the captured substring offsets, it is
1546 used as far as possible (up to two-thirds of its length), and the function
1547 returns a value of zero. If the substring offsets are not of interest,
1548 \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1549 \fIovecsize\fP as zero. However, if the pattern contains back references and
1550 the \fIovector\fP is not big enough to remember the related substrings, PCRE
1551 has to get additional memory for use during matching. Thus it is usually
1552 advisable to supply an \fIovector\fP.
1553 .P
1554 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1555 subpatterns there are in a compiled pattern. The smallest size for
1556 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1557 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1558 .P
1559 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1560 the subject when subpattern \fIn\fP has not been used at all. For example, if
1561 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1562 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1563 happens, both values in the offset pairs corresponding to unused subpatterns
1564 are set to -1.
1565 .P
1566 Offset values that correspond to unused subpatterns at the end of the
1567 expression are also set to -1. For example, if the string "abc" is matched
1568 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1569 return from the function is 2, because the highest used capturing subpattern
1570 number is 1. However, you can refer to the offsets for the second and third
1571 capturing subpatterns if you wish (assuming the vector is large enough, of
1572 course).
1573 .P
1574 Some convenience functions are provided for extracting the captured substrings
1575 as separate strings. These are described below.
1576 .
1577 .\" HTML <a name="errorlist"></a>
1578 .SS "Error return values from \fBpcre_exec()\fP"
1579 .rs
1580 .sp
1581 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1582 defined in the header file:
1583 .sp
1585 .sp
1586 The subject string did not match the pattern.
1587 .sp
1589 .sp
1590 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1591 NULL and \fIovecsize\fP was not zero.
1592 .sp
1594 .sp
1595 An unrecognized bit was set in the \fIoptions\fP argument.
1596 .sp
1598 .sp
1599 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1600 the case when it is passed a junk pointer and to detect when a pattern that was
1601 compiled in an environment of one endianness is run in an environment with the
1602 other endianness. This is the error that PCRE gives when the magic number is
1603 not present.
1604 .sp
1606 .sp
1607 While running the pattern match, an unknown item was encountered in the
1608 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1609 of the compiled pattern.
1610 .sp
1612 .sp
1613 If a pattern contains back references, but the \fIovector\fP that is passed to
1614 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1615 gets a block of memory at the start of matching to use for this purpose. If the
1616 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1617 automatically freed at the end of matching.
1618 .sp
1620 .sp
1621 This error is used by the \fBpcre_copy_substring()\fP,
1622 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1623 below). It is never returned by \fBpcre_exec()\fP.
1624 .sp
1626 .sp
1627 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1628 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1629 above.
1630 .sp
1632 .sp
1633 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1634 use by callout functions that want to yield a distinctive error code. See the
1635 .\" HREF
1636 \fBpcrecallout\fP
1637 .\"
1638 documentation for details.
1639 .sp
1641 .sp
1642 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1643 .sp
1645 .sp
1646 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1647 of \fIstartoffset\fP did not point to the beginning of a UTF-8 character.
1648 .sp
1650 .sp
1651 The subject string did not match, but it did match partially. See the
1652 .\" HREF
1653 \fBpcrepartial\fP
1654 .\"
1655 documentation for details of partial matching.
1656 .sp
1658 .sp
1659 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1660 option was used with a compiled pattern containing items that were not
1661 supported for partial matching. From release 8.00 onwards, there are no
1662 restrictions on partial matching.
1663 .sp
1665 .sp
1666 An unexpected internal error has occurred. This error could be caused by a bug
1667 in PCRE or by overwriting of the compiled pattern.
1668 .sp
1670 .sp
1671 This error is given if the value of the \fIovecsize\fP argument is negative.
1672 .sp
1674 .sp
1675 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1676 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1677 description above.
1678 .sp
1680 .sp
1681 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1682 .P
1683 Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1684 .
1685 .
1687 .rs
1688 .sp
1689 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1690 .ti +5n
1691 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1692 .ti +5n
1693 .B int \fIbuffersize\fP);
1694 .PP
1695 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1696 .ti +5n
1697 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1698 .ti +5n
1699 .B const char **\fIstringptr\fP);
1700 .PP
1701 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1702 .ti +5n
1703 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1704 .PP
1705 Captured substrings can be accessed directly by using the offsets returned by
1706 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1707 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1708 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1709 as new, separate, zero-terminated strings. These functions identify substrings
1710 by number. The next section describes functions for extracting named
1711 substrings.
1712 .P
1713 A substring that contains a binary zero is correctly extracted and has a
1714 further zero added on the end, but the result is not, of course, a C string.
1715 However, you can process such a string by referring to the length that is
1716 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1717 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1718 for handling strings containing binary zeros, because the end of the final
1719 string is not independently indicated.
1720 .P
1721 The first three arguments are the same for all three of these functions:
1722 \fIsubject\fP is the subject string that has just been successfully matched,
1723 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
1724 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
1725 captured by the match, including the substring that matched the entire regular
1726 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
1727 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
1728 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
1729 number of elements in the vector divided by three.
1730 .P
1731 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1732 extract a single substring, whose number is given as \fIstringnumber\fP. A
1733 value of zero extracts the substring that matched the entire pattern, whereas
1734 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1735 the string is placed in \fIbuffer\fP, whose length is given by
1736 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
1737 obtained via \fBpcre_malloc\fP, and its address is returned via
1738 \fIstringptr\fP. The yield of the function is the length of the string, not
1739 including the terminating zero, or one of these error codes:
1740 .sp
1742 .sp
1743 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
1744 memory failed for \fBpcre_get_substring()\fP.
1745 .sp
1747 .sp
1748 There is no substring whose number is \fIstringnumber\fP.
1749 .P
1750 The \fBpcre_get_substring_list()\fP function extracts all available substrings
1751 and builds a list of pointers to them. All this is done in a single block of
1752 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
1753 is returned via \fIlistptr\fP, which is also the start of the list of string
1754 pointers. The end of the list is marked by a NULL pointer. The yield of the
1755 function is zero if all went well, or the error code
1756 .sp
1758 .sp
1759 if the attempt to get the memory block failed.
1760 .P
1761 When any of these functions encounter a substring that is unset, which can
1762 happen when capturing subpattern number \fIn+1\fP matches some part of the
1763 subject, but subpattern \fIn\fP has not been used at all, they return an empty
1764 string. This can be distinguished from a genuine zero-length substring by
1765 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1766 substrings.
1767 .P
1768 The two convenience functions \fBpcre_free_substring()\fP and
1769 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
1770 a previous call of \fBpcre_get_substring()\fP or
1771 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1772 the function pointed to by \fBpcre_free\fP, which of course could be called
1773 directly from a C program. However, PCRE is used in some situations where it is
1774 linked via a special interface to another programming language that cannot use
1775 \fBpcre_free\fP directly; it is for these cases that the functions are
1776 provided.
1777 .
1778 .
1780 .rs
1781 .sp
1782 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1783 .ti +5n
1784 .B const char *\fIname\fP);
1785 .PP
1786 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1787 .ti +5n
1788 .B const char *\fIsubject\fP, int *\fIovector\fP,
1789 .ti +5n
1790 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1791 .ti +5n
1792 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1793 .PP
1794 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
1795 .ti +5n
1796 .B const char *\fIsubject\fP, int *\fIovector\fP,
1797 .ti +5n
1798 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1799 .ti +5n
1800 .B const char **\fIstringptr\fP);
1801 .PP
1802 To extract a substring by name, you first have to find associated number.
1803 For example, for this pattern
1804 .sp
1805 (a+)b(?<xxx>\ed+)...
1806 .sp
1807 the number of the subpattern called "xxx" is 2. If the name is known to be
1808 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1809 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1810 pattern, and the second is the name. The yield of the function is the
1811 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1812 that name.
1813 .P
1814 Given the number, you can extract the substring directly, or use one of the
1815 functions described in the previous section. For convenience, there are also
1816 two functions that do the whole job.
1817 .P
1818 Most of the arguments of \fBpcre_copy_named_substring()\fP and
1819 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1820 functions that extract by number. As these are described in the previous
1821 section, they are not re-described here. There are just two differences:
1822 .P
1823 First, instead of a substring number, a substring name is given. Second, there
1824 is an extra argument, given at the start, which is a pointer to the compiled
1825 pattern. This is needed in order to gain access to the name-to-number
1826 translation table.
1827 .P
1828 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1829 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1830 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1831 the behaviour may not be what you want (see the next section).
1832 .P
1833 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
1834 subpatterns with the same number, as described in the
1835 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1836 .\" </a>
1837 section on duplicate subpattern numbers
1838 .\"
1839 in the
1840 .\" HREF
1841 \fBpcrepattern\fP
1842 .\"
1843 page, you cannot use names to distinguish the different subpatterns, because
1844 names are not included in the compiled code. The matching process uses only
1845 numbers. For this reason, the use of different names for subpatterns of the
1846 same number causes an error at compile time.
1847 .
1849 .rs
1850 .sp
1851 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1852 .ti +5n
1853 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
1854 .PP
1855 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1856 are not required to be unique. (Duplicate names are always allowed for
1857 subpatterns with the same number, created by using the (?| feature. Indeed, if
1858 such subpatterns are named, they are required to use the same names.)
1859 .P
1860 Normally, patterns with duplicate names are such that in any one match, only
1861 one of the named subpatterns participates. An example is shown in the
1862 .\" HREF
1863 \fBpcrepattern\fP
1864 .\"
1865 documentation.
1866 .P
1867 When duplicates are present, \fBpcre_copy_named_substring()\fP and
1868 \fBpcre_get_named_substring()\fP return the first substring corresponding to
1869 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
1870 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
1871 returns one of the numbers that are associated with the name, but it is not
1872 defined which it is.
1873 .P
1874 If you want to get full details of all captured substrings for a given name,
1875 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
1876 argument is the compiled pattern, and the second is the name. The third and
1877 fourth are pointers to variables which are updated by the function. After it
1878 has run, they point to the first and last entries in the name-to-number table
1879 for the given name. The function itself returns the length of each entry, or
1880 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
1881 described above in the section entitled \fIInformation about a pattern\fP.
1882 Given all the relevant entries for the name, you can extract each of their
1883 numbers, and hence the captured data, if any.
1884 .
1885 .
1887 .rs
1888 .sp
1889 The traditional matching function uses a similar algorithm to Perl, which stops
1890 when it finds the first match, starting at a given point in the subject. If you
1891 want to find all possible matches, or the longest possible match, consider
1892 using the alternative matching function (see below) instead. If you cannot use
1893 the alternative function, but still need to find all possible matches, you
1894 can kludge it up by making use of the callout facility, which is described in
1895 the
1896 .\" HREF
1897 \fBpcrecallout\fP
1898 .\"
1899 documentation.
1900 .P
1901 What you have to do is to insert a callout right at the end of the pattern.
1902 When your callout function is called, extract and save the current matched
1903 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
1904 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
1905 will yield PCRE_ERROR_NOMATCH.
1906 .
1907 .
1908 .\" HTML <a name="dfamatch"></a>
1910 .rs
1911 .sp
1912 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1913 .ti +5n
1914 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1915 .ti +5n
1916 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
1917 .ti +5n
1918 .B int *\fIworkspace\fP, int \fIwscount\fP);
1919 .P
1920 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
1921 a compiled pattern, using a matching algorithm that scans the subject string
1922 just once, and does not backtrack. This has different characteristics to the
1923 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
1924 patterns are not supported. Nevertheless, there are times when this kind of
1925 matching can be useful. For a discussion of the two matching algorithms, and a
1926 list of features that \fBpcre_dfa_exec()\fP does not support, see the
1927 .\" HREF
1928 \fBpcrematching\fP
1929 .\"
1930 documentation.
1931 .P
1932 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
1933 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
1934 different way, and this is described below. The other common arguments are used
1935 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
1936 here.
1937 .P
1938 The two additional arguments provide workspace for the function. The workspace
1939 vector should contain at least 20 elements. It is used for keeping track of
1940 multiple paths through the pattern tree. More workspace will be needed for
1941 patterns and subjects where there are a lot of potential matches.
1942 .P
1943 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
1944 .sp
1945 int rc;
1946 int ovector[10];
1947 int wspace[20];
1948 rc = pcre_dfa_exec(
1949 re, /* result of pcre_compile() */
1950 NULL, /* we didn't study the pattern */
1951 "some string", /* the subject string */
1952 11, /* the length of the subject string */
1953 0, /* start at offset 0 in the subject */
1954 0, /* default options */
1955 ovector, /* vector of integers for substring information */
1956 10, /* number of elements (NOT size in bytes) */
1957 wspace, /* working space vector */
1958 20); /* number of elements (NOT size in bytes) */
1959 .
1960 .SS "Option bits for \fBpcre_dfa_exec()\fP"
1961 .rs
1962 .sp
1963 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
1964 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1967 and PCRE_DFA_RESTART. All but the last four of these are exactly the same as
1968 for \fBpcre_exec()\fP, so their description is not repeated here.
1969 .sp
1972 .sp
1973 These have the same general effect as they do for \fBpcre_exec()\fP, but the
1974 details are slightly different. When PCRE_PARTIAL_HARD is set for
1975 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
1976 is reached and there is still at least one matching possibility that requires
1977 additional characters. This happens even if some complete matches have also
1978 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
1979 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
1980 there have been no complete matches, but there is still at least one matching
1981 possibility. The portion of the string that was inspected when the longest
1982 partial match was found is set as the first matching string in both cases.
1983 .sp
1985 .sp
1986 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
1987 soon as it has found one match. Because of the way the alternative algorithm
1988 works, this is necessarily the shortest possible match at the first possible
1989 matching point in the subject string.
1990 .sp
1992 .sp
1993 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
1994 again, with additional subject characters, and have it continue with the same
1995 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
1996 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
1997 before because data about the match so far is left in them after a partial
1998 match. There is more discussion of this facility in the
1999 .\" HREF
2000 \fBpcrepartial\fP
2001 .\"
2002 documentation.
2003 .
2004 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2005 .rs
2006 .sp
2007 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2008 substring in the subject. Note, however, that all the matches from one run of
2009 the function start at the same point in the subject. The shorter matches are
2010 all initial substrings of the longer matches. For example, if the pattern
2011 .sp
2012 <.*>
2013 .sp
2014 is matched against the string
2015 .sp
2016 This is <something> <something else> <something further> no more
2017 .sp
2018 the three matched strings are
2019 .sp
2020 <something>
2021 <something> <something else>
2022 <something> <something else> <something further>
2023 .sp
2024 On success, the yield of the function is a number greater than zero, which is
2025 the number of matched substrings. The substrings themselves are returned in
2026 \fIovector\fP. Each string uses two elements; the first is the offset to the
2027 start, and the second is the offset to the end. In fact, all the strings have
2028 the same start offset. (Space could have been saved by giving this only once,
2029 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2030 returns data, even though the meaning of the strings is different.)
2031 .P
2032 The strings are returned in reverse order of length; that is, the longest
2033 matching string is given first. If there were too many matches to fit into
2034 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2035 the longest matches.
2036 .
2037 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2038 .rs
2039 .sp
2040 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2041 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2042 described
2043 .\" HTML <a href="#errorlist">
2044 .\" </a>
2045 above.
2046 .\"
2047 There are in addition the following errors that are specific to
2048 \fBpcre_dfa_exec()\fP:
2049 .sp
2051 .sp
2052 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2053 that it does not support, for instance, the use of \eC or a back reference.
2054 .sp
2056 .sp
2057 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2058 uses a back reference for the condition, or a test for recursion in a specific
2059 group. These are not supported.
2060 .sp
2062 .sp
2063 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2064 block that contains a setting of the \fImatch_limit\fP field. This is not
2065 supported (it is meaningless).
2066 .sp
2068 .sp
2069 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2070 \fIworkspace\fP vector.
2071 .sp
2073 .sp
2074 When a recursive subpattern is processed, the matching function calls itself
2075 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2076 error is given if the output vector is not large enough. This should be
2077 extremely rare, as a vector of size 1000 is used.
2078 .
2079 .
2080 .SH "SEE ALSO"
2081 .rs
2082 .sp
2083 \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2084 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2085 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2086 .
2087 .
2089 .rs
2090 .sp
2091 .nf
2092 Philip Hazel
2093 University Computing Service
2094 Cambridge CB2 3QH, England.
2095 .fi
2096 .
2097 .
2099 .rs
2100 .sp
2101 .nf
2102 Last updated: 26 March 2010
2103 Copyright (c) 1997-2010 University of Cambridge.
2104 .fi


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

  ViewVC Help
Powered by ViewVC 1.1.5