ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 513 - (show annotations)
Mon May 3 11:13:37 2010 UTC (11 years, 2 months ago) by ph10
File size: 91010 byte(s)
Make \R and \X in a character class behave more like Perl
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 an error for this, by running it with the -w option.) There are at present
557 no other features controlled by this option. It can also be set by a (?X)
558 option setting within a pattern.
559 .sp
561 .sp
562 If this option is set, an unanchored pattern is required to match before or at
563 the first newline in the subject string, though the matched text may continue
564 over the newline.
565 .sp
567 .sp
568 If this option is set, PCRE's behaviour is changed in some ways so that it is
569 compatible with JavaScript rather than Perl. The changes are as follows:
570 .P
571 (1) A lone closing square bracket in a pattern causes a compile-time error,
572 because this is illegal in JavaScript (by default it is treated as a data
573 character). Thus, the pattern AB]CD becomes illegal when this option is set.
574 .P
575 (2) At run time, a back reference to an unset subpattern group matches an empty
576 string (by default this causes the current matching alternative to fail). A
577 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
578 an "a" in the subject), whereas it fails by default, for Perl compatibility.
579 .sp
581 .sp
582 By default, PCRE treats the subject string as consisting of a single line of
583 characters (even if it actually contains newlines). The "start of line"
584 metacharacter (^) matches only at the start of the string, while the "end of
585 line" metacharacter ($) matches only at the end of the string, or before a
586 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
587 Perl.
588 .P
589 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
590 match immediately following or immediately before internal newlines in the
591 subject string, respectively, as well as at the very start and end. This is
592 equivalent to Perl's /m option, and it can be changed within a pattern by a
593 (?m) option setting. If there are no newlines in a subject string, or no
594 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
595 .sp
601 .sp
602 These options override the default newline definition that was chosen when PCRE
603 was built. Setting the first or the second specifies that a newline is
604 indicated by a single character (CR or LF, respectively). Setting
605 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
606 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
607 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
608 that any Unicode newline sequence should be recognized. The Unicode newline
609 sequences are the three just mentioned, plus the single characters VT (vertical
610 tab, U+000B), FF (formfeed, U+000C), NEL (next line, U+0085), LS (line
611 separator, U+2028), and PS (paragraph separator, U+2029). The last two are
612 recognized only in UTF-8 mode.
613 .P
614 The newline setting in the options word uses three bits that are treated
615 as a number, giving eight possibilities. Currently only six are used (default
616 plus the five values above). This means that if you set more than one newline
617 option, the combination may or may not be sensible. For example,
619 other combinations may yield unused numbers and cause an error.
620 .P
621 The only time that a line break is specially recognized when compiling a
622 pattern is if PCRE_EXTENDED is set, and an unescaped # outside a character
623 class is encountered. This indicates a comment that lasts until after the next
624 line break sequence. In other circumstances, line break sequences are treated
625 as literal data, except that in PCRE_EXTENDED mode, both CR and LF are treated
626 as whitespace characters and are therefore ignored.
627 .P
628 The newline option that is set at compile time becomes the default that is used
629 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
630 .sp
632 .sp
633 If this option is set, it disables the use of numbered capturing parentheses in
634 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
635 were followed by ?: but named parentheses can still be used for capturing (and
636 they acquire numbers in the usual way). There is no equivalent of this option
637 in Perl.
638 .sp
640 .sp
641 This option inverts the "greediness" of the quantifiers so that they are not
642 greedy by default, but become greedy if followed by "?". It is not compatible
643 with Perl. It can also be set by a (?U) option setting within the pattern.
644 .sp
646 .sp
647 This option causes PCRE to regard both the pattern and the subject as strings
648 of UTF-8 characters instead of single-byte character strings. However, it is
649 available only when PCRE is built to include UTF-8 support. If not, the use
650 of this option provokes an error. Details of how this option changes the
651 behaviour of PCRE are given in the
652 .\" HTML <a href="pcre.html#utf8support">
653 .\" </a>
654 section on UTF-8 support
655 .\"
656 in the main
657 .\" HREF
658 \fBpcre\fP
659 .\"
660 page.
661 .sp
663 .sp
664 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
665 automatically checked. There is a discussion about the
666 .\" HTML <a href="pcre.html#utf8strings">
667 .\" </a>
668 validity of UTF-8 strings
669 .\"
670 in the main
671 .\" HREF
672 \fBpcre\fP
673 .\"
674 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_compile()\fP
675 returns an error. If you already know that your pattern is valid, and you want
676 to skip this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK
677 option. When it is set, the effect of passing an invalid UTF-8 string as a
678 pattern is undefined. It may cause your program to crash. Note that this option
679 can also be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress
680 the UTF-8 validity checking of subject strings.
681 .
682 .
684 .rs
685 .sp
686 The following table lists the error codes than may be returned by
687 \fBpcre_compile2()\fP, along with the error messages that may be returned by
688 both compiling functions. As PCRE has developed, some error codes have fallen
689 out of use. To avoid confusion, they have not been re-used.
690 .sp
691 0 no error
692 1 \e at end of pattern
693 2 \ec at end of pattern
694 3 unrecognized character follows \e
695 4 numbers out of order in {} quantifier
696 5 number too big in {} quantifier
697 6 missing terminating ] for character class
698 7 invalid escape sequence in character class
699 8 range out of order in character class
700 9 nothing to repeat
701 10 [this code is not in use]
702 11 internal error: unexpected repeat
703 12 unrecognized character after (? or (?-
704 13 POSIX named classes are supported only within a class
705 14 missing )
706 15 reference to non-existent subpattern
707 16 erroffset passed as NULL
708 17 unknown option bit(s) set
709 18 missing ) after comment
710 19 [this code is not in use]
711 20 regular expression is too large
712 21 failed to get memory
713 22 unmatched parentheses
714 23 internal error: code overflow
715 24 unrecognized character after (?<
716 25 lookbehind assertion is not fixed length
717 26 malformed number or name after (?(
718 27 conditional group contains more than two branches
719 28 assertion expected after (?(
720 29 (?R or (?[+-]digits must be followed by )
721 30 unknown POSIX class name
722 31 POSIX collating elements are not supported
723 32 this version of PCRE is not compiled with PCRE_UTF8 support
724 33 [this code is not in use]
725 34 character value in \ex{...} sequence is too large
726 35 invalid condition (?(0)
727 36 \eC not allowed in lookbehind assertion
728 37 PCRE does not support \eL, \el, \eN, \eU, or \eu
729 38 number after (?C is > 255
730 39 closing ) for (?C expected
731 40 recursive call could loop indefinitely
732 41 unrecognized character after (?P
733 42 syntax error in subpattern name (missing terminator)
734 43 two named subpatterns have the same name
735 44 invalid UTF-8 string
736 45 support for \eP, \ep, and \eX has not been compiled
737 46 malformed \eP or \ep sequence
738 47 unknown property name after \eP or \ep
739 48 subpattern name is too long (maximum 32 characters)
740 49 too many named subpatterns (maximum 10000)
741 50 [this code is not in use]
742 51 octal value is greater than \e377 (not in UTF-8 mode)
743 52 internal error: overran compiling workspace
744 53 internal error: previously-checked referenced subpattern not found
745 54 DEFINE group contains more than one branch
746 55 repeating a DEFINE group is not allowed
747 56 inconsistent NEWLINE options
748 57 \eg is not followed by a braced, angle-bracketed, or quoted
749 name/number or by a plain number
750 58 a numbered reference must not be zero
751 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
752 60 (*VERB) not recognized
753 61 number is too big
754 62 subpattern name expected
755 63 digit expected after (?+
756 64 ] is an invalid data character in JavaScript compatibility mode
757 65 different names for subpatterns of the same number are not allowed
758 66 (*MARK) must have an argument
759 .sp
760 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
761 be used if the limits were changed when PCRE was built.
762 .
763 .
765 .rs
766 .sp
767 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
768 .ti +5n
769 .B const char **\fIerrptr\fP);
770 .PP
771 If a compiled pattern is going to be used several times, it is worth spending
772 more time analyzing it in order to speed up the time taken for matching. The
773 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
774 argument. If studying the pattern produces additional information that will
775 help speed up matching, \fBpcre_study()\fP returns a pointer to a
776 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
777 results of the study.
778 .P
779 The returned value from \fBpcre_study()\fP can be passed directly to
780 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
781 also contains other fields that can be set by the caller before the block is
782 passed; these are described
783 .\" HTML <a href="#extradata">
784 .\" </a>
785 below
786 .\"
787 in the section on matching a pattern.
788 .P
789 If studying the pattern does not produce any useful information,
790 \fBpcre_study()\fP returns NULL. In that circumstance, if the calling program
791 wants to pass any of the other fields to \fBpcre_exec()\fP or
792 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block.
793 .P
794 The second argument of \fBpcre_study()\fP contains option bits. At present, no
795 options are defined, and this argument should always be zero.
796 .P
797 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
798 studying succeeds (even if no data is returned), the variable it points to is
799 set to NULL. Otherwise it is set to point to a textual error message. This is a
800 static string that is part of the library. You must not try to free it. You
801 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
802 sure that it has run successfully.
803 .P
804 This is a typical call to \fBpcre_study\fP():
805 .sp
806 pcre_extra *pe;
807 pe = pcre_study(
808 re, /* result of pcre_compile() */
809 0, /* no options exist */
810 &error); /* set to NULL or points to a message */
811 .sp
812 Studying a pattern does two things: first, a lower bound for the length of
813 subject string that is needed to match the pattern is computed. This does not
814 mean that there are any strings of that length that match, but it does
815 guarantee that no shorter strings match. The value is used by
816 \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP to avoid wasting time by trying to
817 match strings that are shorter than the lower bound. You can find out the value
818 in a calling program via the \fBpcre_fullinfo()\fP function.
819 .P
820 Studying a pattern is also useful for non-anchored patterns that do not have a
821 single fixed starting character. A bitmap of possible starting bytes is
822 created. This speeds up finding a position in the subject at which to start
823 matching.
824 .
825 .
826 .\" HTML <a name="localesupport"></a>
828 .rs
829 .sp
830 PCRE handles caseless matching, and determines whether characters are letters,
831 digits, or whatever, by reference to a set of tables, indexed by character
832 value. When running in UTF-8 mode, this applies only to characters with codes
833 less than 128. Higher-valued codes never match escapes such as \ew or \ed, but
834 can be tested with \ep if PCRE is built with Unicode character property
835 support. The use of locales with Unicode is discouraged. If you are handling
836 characters with codes greater than 128, you should either use UTF-8 and
837 Unicode, or use locales, but not try to mix the two.
838 .P
839 PCRE contains an internal set of tables that are used when the final argument
840 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
841 Normally, the internal tables recognize only ASCII characters. However, when
842 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
843 default "C" locale of the local system, which may cause them to be different.
844 .P
845 The internal tables can always be overridden by tables supplied by the
846 application that calls PCRE. These may be created in a different locale from
847 the default. As more and more applications change to using Unicode, the need
848 for this locale support is expected to die away.
849 .P
850 External tables are built by calling the \fBpcre_maketables()\fP function,
851 which has no arguments, in the relevant locale. The result can then be passed
852 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
853 example, to build and use tables that are appropriate for the French locale
854 (where accented characters with values greater than 128 are treated as letters),
855 the following code could be used:
856 .sp
857 setlocale(LC_CTYPE, "fr_FR");
858 tables = pcre_maketables();
859 re = pcre_compile(..., tables);
860 .sp
861 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
862 are using Windows, the name for the French locale is "french".
863 .P
864 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
865 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
866 that the memory containing the tables remains available for as long as it is
867 needed.
868 .P
869 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
870 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
871 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
872 pattern, compilation, studying and matching all happen in the same locale, but
873 different patterns can be compiled in different locales.
874 .P
875 It is possible to pass a table pointer or NULL (indicating the use of the
876 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
877 this facility could be used to match a pattern in a different locale from the
878 one in which it was compiled. Passing table pointers at run time is discussed
879 below in the section on matching a pattern.
880 .
881 .
883 .rs
884 .sp
885 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
886 .ti +5n
887 .B int \fIwhat\fP, void *\fIwhere\fP);
888 .PP
889 The \fBpcre_fullinfo()\fP function returns information about a compiled
890 pattern. It replaces the obsolete \fBpcre_info()\fP function, which is
891 nevertheless retained for backwards compability (and is documented below).
892 .P
893 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
894 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
895 the pattern was not studied. The third argument specifies which piece of
896 information is required, and the fourth argument is a pointer to a variable
897 to receive the data. The yield of the function is zero for success, or one of
898 the following negative numbers:
899 .sp
900 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
901 the argument \fIwhere\fP was NULL
902 PCRE_ERROR_BADMAGIC the "magic number" was not found
903 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
904 .sp
905 The "magic number" is placed at the start of each compiled pattern as an simple
906 check against passing an arbitrary memory pointer. Here is a typical call of
907 \fBpcre_fullinfo()\fP, to obtain the length of the compiled pattern:
908 .sp
909 int rc;
910 size_t length;
911 rc = pcre_fullinfo(
912 re, /* result of pcre_compile() */
913 pe, /* result of pcre_study(), or NULL */
914 PCRE_INFO_SIZE, /* what is required */
915 &length); /* where to put the data */
916 .sp
917 The possible values for the third argument are defined in \fBpcre.h\fP, and are
918 as follows:
919 .sp
921 .sp
922 Return the number of the highest back reference in the pattern. The fourth
923 argument should point to an \fBint\fP variable. Zero is returned if there are
924 no back references.
925 .sp
927 .sp
928 Return the number of capturing subpatterns in the pattern. The fourth argument
929 should point to an \fBint\fP variable.
930 .sp
932 .sp
933 Return a pointer to the internal default character tables within PCRE. The
934 fourth argument should point to an \fBunsigned char *\fP variable. This
935 information call is provided for internal use by the \fBpcre_study()\fP
936 function. External callers can cause PCRE to use its internal tables by passing
937 a NULL table pointer.
938 .sp
940 .sp
941 Return information about the first byte of any matched string, for a
942 non-anchored pattern. The fourth argument should point to an \fBint\fP
943 variable. (This option used to be called PCRE_INFO_FIRSTCHAR; the old name is
944 still recognized for backwards compatibility.)
945 .P
946 If there is a fixed first byte, for example, from a pattern such as
947 (cat|cow|coyote), its value is returned. Otherwise, if either
948 .sp
949 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
950 starts with "^", or
951 .sp
952 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
953 (if it were set, the pattern would be anchored),
954 .sp
955 -1 is returned, indicating that the pattern matches only at the start of a
956 subject string or after any newline within the string. Otherwise -2 is
957 returned. For anchored patterns, -2 is returned.
958 .sp
960 .sp
961 If the pattern was studied, and this resulted in the construction of a 256-bit
962 table indicating a fixed set of bytes for the first byte in any matching
963 string, a pointer to the table is returned. Otherwise NULL is returned. The
964 fourth argument should point to an \fBunsigned char *\fP variable.
965 .sp
967 .sp
968 Return 1 if the pattern contains any explicit matches for CR or LF characters,
969 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
970 explicit match is either a literal CR or LF character, or \er or \en.
971 .sp
973 .sp
974 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
975 0. The fourth argument should point to an \fBint\fP variable. (?J) and
976 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
977 .sp
979 .sp
980 Return the value of the rightmost literal byte that must exist in any matched
981 string, other than at its start, if such a byte has been recorded. The fourth
982 argument should point to an \fBint\fP variable. If there is no such byte, -1 is
983 returned. For anchored patterns, a last literal byte is recorded only if it
984 follows something of variable length. For example, for the pattern
985 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
986 is -1.
987 .sp
989 .sp
990 If the pattern was studied and a minimum length for matching subject strings
991 was computed, its value is returned. Otherwise the returned value is -1. The
992 value is a number of characters, not bytes (this may be relevant in UTF-8
993 mode). The fourth argument should point to an \fBint\fP variable. A
994 non-negative value is a lower bound to the length of any matching string. There
995 may not be any strings of that length that do actually match, but every string
996 that does match is at least that long.
997 .sp
1001 .sp
1002 PCRE supports the use of named as well as numbered capturing parentheses. The
1003 names are just an additional way of identifying the parentheses, which still
1004 acquire numbers. Several convenience functions such as
1005 \fBpcre_get_named_substring()\fP are provided for extracting captured
1006 substrings by name. It is also possible to extract the data directly, by first
1007 converting the name to a number in order to access the correct pointers in the
1008 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1009 you need to use the name-to-number map, which is described by these three
1010 values.
1011 .P
1012 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1013 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1014 entry; both of these return an \fBint\fP value. The entry size depends on the
1015 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1016 entry of the table (a pointer to \fBchar\fP). The first two bytes of each entry
1017 are the number of the capturing parenthesis, most significant byte first. The
1018 rest of the entry is the corresponding name, zero terminated.
1019 .P
1020 The names are in alphabetical order. Duplicate names may appear if (?| is used
1021 to create multiple groups with the same number, as described in the
1022 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1023 .\" </a>
1024 section on duplicate subpattern numbers
1025 .\"
1026 in the
1027 .\" HREF
1028 \fBpcrepattern\fP
1029 .\"
1030 page. Duplicate names for subpatterns with different numbers are permitted only
1031 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1032 table in the order in which they were found in the pattern. In the absence of
1033 (?| this is the order of increasing number; when (?| is used this is not
1034 necessarily the case because later subpatterns may have lower numbers.
1035 .P
1036 As a simple example of the name/number table, consider the following pattern
1037 (assume PCRE_EXTENDED is set, so white space - including newlines - is
1038 ignored):
1039 .sp
1040 .\" JOIN
1041 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1042 (?<month>\ed\ed) - (?<day>\ed\ed) )
1043 .sp
1044 There are four named subpatterns, so the table has four entries, and each entry
1045 in the table is eight bytes long. The table is as follows, with non-printing
1046 bytes shows in hexadecimal, and undefined bytes shown as ??:
1047 .sp
1048 00 01 d a t e 00 ??
1049 00 05 d a y 00 ?? ??
1050 00 04 m o n t h 00
1051 00 02 y e a r 00 ??
1052 .sp
1053 When writing code to extract data from named subpatterns using the
1054 name-to-number map, remember that the length of the entries is likely to be
1055 different for each compiled pattern.
1056 .sp
1058 .sp
1059 Return 1 if the pattern can be used for partial matching with
1060 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1061 \fBint\fP variable. From release 8.00, this always returns 1, because the
1062 restrictions that previously applied to partial matching have been lifted. The
1063 .\" HREF
1064 \fBpcrepartial\fP
1065 .\"
1066 documentation gives details of partial matching.
1067 .sp
1069 .sp
1070 Return a copy of the options with which the pattern was compiled. The fourth
1071 argument should point to an \fBunsigned long int\fP variable. These option bits
1072 are those specified in the call to \fBpcre_compile()\fP, modified by any
1073 top-level option settings at the start of the pattern itself. In other words,
1074 they are the options that will be in force when matching starts. For example,
1075 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1077 .P
1078 A pattern is automatically anchored by PCRE if all of its top-level
1079 alternatives begin with one of the following:
1080 .sp
1081 ^ unless PCRE_MULTILINE is set
1082 \eA always
1083 \eG always
1084 .\" JOIN
1085 .* if PCRE_DOTALL is set and there are no back
1086 references to the subpattern in which .* appears
1087 .sp
1088 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1089 \fBpcre_fullinfo()\fP.
1090 .sp
1092 .sp
1093 Return the size of the compiled pattern, that is, the value that was passed as
1094 the argument to \fBpcre_malloc()\fP when PCRE was getting memory in which to
1095 place the compiled data. The fourth argument should point to a \fBsize_t\fP
1096 variable.
1097 .sp
1099 .sp
1100 Return the size of the data block pointed to by the \fIstudy_data\fP field in
1101 a \fBpcre_extra\fP block. That is, it is the value that was passed to
1102 \fBpcre_malloc()\fP when PCRE was getting memory into which to place the data
1103 created by \fBpcre_study()\fP. If \fBpcre_extra\fP is NULL, or there is no
1104 study data, zero is returned. The fourth argument should point to a
1105 \fBsize_t\fP variable.
1106 .
1107 .
1109 .rs
1110 .sp
1111 .B int pcre_info(const pcre *\fIcode\fP, int *\fIoptptr\fP, int
1112 .B *\fIfirstcharptr\fP);
1113 .PP
1114 The \fBpcre_info()\fP function is now obsolete because its interface is too
1115 restrictive to return all the available data about a compiled pattern. New
1116 programs should use \fBpcre_fullinfo()\fP instead. The yield of
1117 \fBpcre_info()\fP is the number of capturing subpatterns, or one of the
1118 following negative numbers:
1119 .sp
1120 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1121 PCRE_ERROR_BADMAGIC the "magic number" was not found
1122 .sp
1123 If the \fIoptptr\fP argument is not NULL, a copy of the options with which the
1124 pattern was compiled is placed in the integer it points to (see
1125 PCRE_INFO_OPTIONS above).
1126 .P
1127 If the pattern is not anchored and the \fIfirstcharptr\fP argument is not NULL,
1128 it is used to pass back information about the first character of any matched
1129 string (see PCRE_INFO_FIRSTBYTE above).
1130 .
1131 .
1133 .rs
1134 .sp
1135 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1136 .PP
1137 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1138 data block that contains a compiled pattern. It is provided for the benefit of
1139 applications that operate in an object-oriented manner, where different parts
1140 of the application may be using the same compiled pattern, but you want to free
1141 the block when they are all done.
1142 .P
1143 When a pattern is compiled, the reference count field is initialized to zero.
1144 It is changed only by calling this function, whose action is to add the
1145 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1146 function is the new value. However, the value of the count is constrained to
1147 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1148 it is forced to the appropriate limit value.
1149 .P
1150 Except when it is zero, the reference count is not correctly preserved if a
1151 pattern is compiled on one host and then transferred to a host whose byte-order
1152 is different. (This seems a highly unlikely scenario.)
1153 .
1154 .
1156 .rs
1157 .sp
1158 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1159 .ti +5n
1160 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1161 .ti +5n
1162 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1163 .P
1164 The function \fBpcre_exec()\fP is called to match a subject string against a
1165 compiled pattern, which is passed in the \fIcode\fP argument. If the
1166 pattern was studied, the result of the study should be passed in the
1167 \fIextra\fP argument. This function is the main matching facility of the
1168 library, and it operates in a Perl-like manner. For specialist use there is
1169 also an alternative matching function, which is described
1170 .\" HTML <a href="#dfamatch">
1171 .\" </a>
1172 below
1173 .\"
1174 in the section about the \fBpcre_dfa_exec()\fP function.
1175 .P
1176 In most applications, the pattern will have been compiled (and optionally
1177 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1178 possible to save compiled patterns and study data, and then use them later
1179 in different processes, possibly even on different hosts. For a discussion
1180 about this, see the
1181 .\" HREF
1182 \fBpcreprecompile\fP
1183 .\"
1184 documentation.
1185 .P
1186 Here is an example of a simple call to \fBpcre_exec()\fP:
1187 .sp
1188 int rc;
1189 int ovector[30];
1190 rc = pcre_exec(
1191 re, /* result of pcre_compile() */
1192 NULL, /* we didn't study the pattern */
1193 "some string", /* the subject string */
1194 11, /* the length of the subject string */
1195 0, /* start at offset 0 in the subject */
1196 0, /* default options */
1197 ovector, /* vector of integers for substring information */
1198 30); /* number of elements (NOT size in bytes) */
1199 .
1200 .\" HTML <a name="extradata"></a>
1201 .SS "Extra data for \fBpcre_exec()\fR"
1202 .rs
1203 .sp
1204 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1205 data block. The \fBpcre_study()\fP function returns such a block (when it
1206 doesn't return NULL), but you can also create one for yourself, and pass
1207 additional information in it. The \fBpcre_extra\fP block contains the following
1208 fields (not necessarily in this order):
1209 .sp
1210 unsigned long int \fIflags\fP;
1211 void *\fIstudy_data\fP;
1212 unsigned long int \fImatch_limit\fP;
1213 unsigned long int \fImatch_limit_recursion\fP;
1214 void *\fIcallout_data\fP;
1215 const unsigned char *\fItables\fP;
1216 unsigned char **\fImark\fP;
1217 .sp
1218 The \fIflags\fP field is a bitmap that specifies which of the other fields
1219 are set. The flag bits are:
1220 .sp
1227 .sp
1228 Other flag bits should be set to zero. The \fIstudy_data\fP field is set in the
1229 \fBpcre_extra\fP block that is returned by \fBpcre_study()\fP, together with
1230 the appropriate flag bit. You should not set this yourself, but you may add to
1231 the block by setting the other fields and their corresponding flag bits.
1232 .P
1233 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1234 vast amount of resources when running patterns that are not going to match,
1235 but which have a very large number of possibilities in their search trees. The
1236 classic example is a pattern that uses nested unlimited repeats.
1237 .P
1238 Internally, PCRE uses a function called \fBmatch()\fP which it calls repeatedly
1239 (sometimes recursively). The limit set by \fImatch_limit\fP is imposed on the
1240 number of times this function is called during a match, which has the effect of
1241 limiting the amount of backtracking that can take place. For patterns that are
1242 not anchored, the count restarts from zero for each position in the subject
1243 string.
1244 .P
1245 The default value for the limit can be set when PCRE is built; the default
1246 default is 10 million, which handles all but the most extreme cases. You can
1247 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1248 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1249 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1251 .P
1252 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1253 instead of limiting the total number of times that \fBmatch()\fP is called, it
1254 limits the depth of recursion. The recursion depth is a smaller number than the
1255 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1256 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1257 .P
1258 Limiting the recursion depth limits the amount of stack that can be used, or,
1259 when PCRE has been compiled to use memory on the heap instead of the stack, the
1260 amount of heap memory that can be used.
1261 .P
1262 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1263 built; the default default is the same value as the default for
1264 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1265 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1266 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1267 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1268 .P
1269 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1270 and is described in the
1271 .\" HREF
1272 \fBpcrecallout\fP
1273 .\"
1274 documentation.
1275 .P
1276 The \fItables\fP field is used to pass a character tables pointer to
1277 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1278 pattern. A non-NULL value is stored with the compiled pattern only if custom
1279 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1280 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1281 internal tables to be used. This facility is helpful when re-using patterns
1282 that have been saved after compiling with an external set of tables, because
1283 the external tables might be at a different address when \fBpcre_exec()\fP is
1284 called. See the
1285 .\" HREF
1286 \fBpcreprecompile\fP
1287 .\"
1288 documentation for a discussion of saving compiled patterns for later use.
1289 .P
1290 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1291 be set to point to a \fBchar *\fP variable. If the pattern contains any
1292 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1293 a name to pass back, a pointer to the name string (zero terminated) is placed
1294 in the variable pointed to by the \fImark\fP field. The names are within the
1295 compiled pattern; if you wish to retain such a name you must copy it before
1296 freeing the memory of a compiled pattern. If there is no name to pass back, the
1297 variable pointed to by the \fImark\fP field set to NULL. For details of the
1298 backtracking control verbs, see the section entitled
1299 .\" HTML <a href="pcrepattern#backtrackcontrol">
1300 .\" </a>
1301 "Backtracking control"
1302 .\"
1303 in the
1304 .\" HREF
1305 \fBpcrepattern\fP
1306 .\"
1307 documentation.
1308 .
1309 .
1310 .\" HTML <a name="execoptions"></a>
1311 .SS "Option bits for \fBpcre_exec()\fP"
1312 .rs
1313 .sp
1314 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1315 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1319 .sp
1321 .sp
1322 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1323 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1324 to be anchored by virtue of its contents, it cannot be made unachored at
1325 matching time.
1326 .sp
1329 .sp
1330 These options (which are mutually exclusive) control what the \eR escape
1331 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1332 match any Unicode newline sequence. These options override the choice that was
1333 made or defaulted when the pattern was compiled.
1334 .sp
1340 .sp
1341 These options override the newline definition that was chosen or defaulted when
1342 the pattern was compiled. For details, see the description of
1343 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1344 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1345 the way the match position is advanced after a match failure for an unanchored
1346 pattern.
1347 .P
1349 match attempt for an unanchored pattern fails when the current position is at a
1350 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1351 characters, the match position is advanced by two characters instead of one, in
1352 other words, to after the CRLF.
1353 .P
1354 The above rule is a compromise that makes the most common cases work as
1355 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1356 set), it does not match the string "\er\enA" because, after failing at the
1357 start, it skips both the CR and the LF before retrying. However, the pattern
1358 [\er\en]A does match that string, because it contains an explicit CR or LF
1359 reference, and so advances only by one character after the first failure.
1360 .P
1361 An explicit match for CR of LF is either a literal appearance of one of those
1362 characters, or one of the \er or \en escape sequences. Implicit matches such as
1363 [^X] do not count, nor does \es (which includes CR and LF in the characters
1364 that it matches).
1365 .P
1366 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1367 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1368 .sp
1370 .sp
1371 This option specifies that first character of the subject string is not the
1372 beginning of a line, so the circumflex metacharacter should not match before
1373 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1374 never to match. This option affects only the behaviour of the circumflex
1375 metacharacter. It does not affect \eA.
1376 .sp
1378 .sp
1379 This option specifies that the end of the subject string is not the end of a
1380 line, so the dollar metacharacter should not match it nor (except in multiline
1381 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1382 compile time) causes dollar never to match. This option affects only the
1383 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1384 .sp
1386 .sp
1387 An empty string is not considered to be a valid match if this option is set. If
1388 there are alternatives in the pattern, they are tried. If all the alternatives
1389 match the empty string, the entire match fails. For example, if the pattern
1390 .sp
1391 a?b?
1392 .sp
1393 is applied to a string not beginning with "a" or "b", it matches an empty
1394 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1395 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1396 .sp
1398 .sp
1399 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1400 the start of the subject is permitted. If the pattern is anchored, such a match
1401 can occur only if the pattern contains \eK.
1402 .P
1403 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1404 does make a special case of a pattern match of the empty string within its
1405 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1406 emulate Perl's behaviour after matching a null string by first trying the match
1407 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1408 if that fails, by advancing the starting offset (see below) and trying an
1409 ordinary match again. There is some code that demonstrates how to do this in
1410 the
1411 .\" HREF
1412 \fBpcredemo\fP
1413 .\"
1414 sample program.
1415 .sp
1417 .sp
1418 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1419 a match, in order to speed up the process. For example, if it is known that a
1420 match must start with a specific character, it searches the subject for that
1421 character, and fails immediately if it cannot find it, without actually running
1422 the main matching function. When callouts are in use, these optimizations can
1423 cause them to be skipped. This option disables the "start-up" optimizations,
1424 causing performance to suffer, but ensuring that the callouts do occur.
1425 .sp
1427 .sp
1428 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1429 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1430 The value of \fIstartoffset\fP is also checked to ensure that it points to the
1431 start of a UTF-8 character. There is a discussion about the validity of UTF-8
1432 strings in the
1433 .\" HTML <a href="pcre.html#utf8strings">
1434 .\" </a>
1435 section on UTF-8 support
1436 .\"
1437 in the main
1438 .\" HREF
1439 \fBpcre\fP
1440 .\"
1441 page. If an invalid UTF-8 sequence of bytes is found, \fBpcre_exec()\fP returns
1442 the error PCRE_ERROR_BADUTF8. If \fIstartoffset\fP contains an invalid value,
1443 PCRE_ERROR_BADUTF8_OFFSET is returned.
1444 .P
1445 If you already know that your subject is valid, and you want to skip these
1446 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1447 calling \fBpcre_exec()\fP. You might want to do this for the second and
1448 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1449 all the matches in a single subject string. However, you should be sure that
1450 the value of \fIstartoffset\fP points to the start of a UTF-8 character. When
1451 PCRE_NO_UTF8_CHECK is set, the effect of passing an invalid UTF-8 string as a
1452 subject, or a value of \fIstartoffset\fP that does not point to the start of a
1453 UTF-8 character, is undefined. Your program may crash.
1454 .sp
1457 .sp
1458 These options turn on the partial matching feature. For backwards
1459 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1460 occurs if the end of the subject string is reached successfully, but there are
1461 not enough subject characters to complete the match. If this happens when
1462 PCRE_PARTIAL_HARD is set, \fBpcre_exec()\fP immediately returns
1463 PCRE_ERROR_PARTIAL. Otherwise, if PCRE_PARTIAL_SOFT is set, matching continues
1464 by testing any other alternatives. Only if they all fail is PCRE_ERROR_PARTIAL
1465 returned (instead of PCRE_ERROR_NOMATCH). The portion of the string that
1466 was inspected when the partial match was found is set as the first matching
1467 string. There is a more detailed discussion in the
1468 .\" HREF
1469 \fBpcrepartial\fP
1470 .\"
1471 documentation.
1472 .
1473 .SS "The string to be matched by \fBpcre_exec()\fP"
1474 .rs
1475 .sp
1476 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1477 \fIsubject\fP, a length (in bytes) in \fIlength\fP, and a starting byte offset
1478 in \fIstartoffset\fP. In UTF-8 mode, the byte offset must point to the start of
1479 a UTF-8 character. Unlike the pattern string, the subject may contain binary
1480 zero bytes. When the starting offset is zero, the search for a match starts at
1481 the beginning of the subject, and this is by far the most common case.
1482 .P
1483 A non-zero starting offset is useful when searching for another match in the
1484 same subject by calling \fBpcre_exec()\fP again after a previous success.
1485 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1486 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1487 lookbehind. For example, consider the pattern
1488 .sp
1489 \eBiss\eB
1490 .sp
1491 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1492 the current position in the subject is not a word boundary.) When applied to
1493 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1494 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1495 subject, namely "issipi", it does not match, because \eB is always false at the
1496 start of the subject, which is deemed to be a word boundary. However, if
1497 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1498 set to 4, it finds the second occurrence of "iss" because it is able to look
1499 behind the starting point to discover that it is preceded by a letter.
1500 .P
1501 If a non-zero starting offset is passed when the pattern is anchored, one
1502 attempt to match at the given offset is made. This can only succeed if the
1503 pattern does not require the match to be at the start of the subject.
1504 .
1505 .SS "How \fBpcre_exec()\fP returns captured substrings"
1506 .rs
1507 .sp
1508 In general, a pattern matches a certain portion of the subject, and in
1509 addition, further substrings from the subject may be picked out by parts of the
1510 pattern. Following the usage in Jeffrey Friedl's book, this is called
1511 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1512 a fragment of a pattern that picks out a substring. PCRE supports several other
1513 kinds of parenthesized subpattern that do not cause substrings to be captured.
1514 .P
1515 Captured substrings are returned to the caller via a vector of integers whose
1516 address is passed in \fIovector\fP. The number of elements in the vector is
1517 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1518 argument is NOT the size of \fIovector\fP in bytes.
1519 .P
1520 The first two-thirds of the vector is used to pass back captured substrings,
1521 each substring using a pair of integers. The remaining third of the vector is
1522 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1523 and is not available for passing back information. The number passed in
1524 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1525 rounded down.
1526 .P
1527 When a match is successful, information about captured substrings is returned
1528 in pairs of integers, starting at the beginning of \fIovector\fP, and
1529 continuing up to two-thirds of its length at the most. The first element of
1530 each pair is set to the byte offset of the first character in a substring, and
1531 the second is set to the byte offset of the first character after the end of a
1532 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1533 mode. They are not character counts.
1534 .P
1535 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1536 portion of the subject string matched by the entire pattern. The next pair is
1537 used for the first capturing subpattern, and so on. The value returned by
1538 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1539 For example, if two substrings have been captured, the returned value is 3. If
1540 there are no capturing subpatterns, the return value from a successful match is
1541 1, indicating that just the first pair of offsets has been set.
1542 .P
1543 If a capturing subpattern is matched repeatedly, it is the last portion of the
1544 string that it matched that is returned.
1545 .P
1546 If the vector is too small to hold all the captured substring offsets, it is
1547 used as far as possible (up to two-thirds of its length), and the function
1548 returns a value of zero. If the substring offsets are not of interest,
1549 \fBpcre_exec()\fP may be called with \fIovector\fP passed as NULL and
1550 \fIovecsize\fP as zero. However, if the pattern contains back references and
1551 the \fIovector\fP is not big enough to remember the related substrings, PCRE
1552 has to get additional memory for use during matching. Thus it is usually
1553 advisable to supply an \fIovector\fP.
1554 .P
1555 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1556 subpatterns there are in a compiled pattern. The smallest size for
1557 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1558 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1559 .P
1560 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1561 the subject when subpattern \fIn\fP has not been used at all. For example, if
1562 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1563 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1564 happens, both values in the offset pairs corresponding to unused subpatterns
1565 are set to -1.
1566 .P
1567 Offset values that correspond to unused subpatterns at the end of the
1568 expression are also set to -1. For example, if the string "abc" is matched
1569 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1570 return from the function is 2, because the highest used capturing subpattern
1571 number is 1. However, you can refer to the offsets for the second and third
1572 capturing subpatterns if you wish (assuming the vector is large enough, of
1573 course).
1574 .P
1575 Some convenience functions are provided for extracting the captured substrings
1576 as separate strings. These are described below.
1577 .
1578 .\" HTML <a name="errorlist"></a>
1579 .SS "Error return values from \fBpcre_exec()\fP"
1580 .rs
1581 .sp
1582 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1583 defined in the header file:
1584 .sp
1586 .sp
1587 The subject string did not match the pattern.
1588 .sp
1590 .sp
1591 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1592 NULL and \fIovecsize\fP was not zero.
1593 .sp
1595 .sp
1596 An unrecognized bit was set in the \fIoptions\fP argument.
1597 .sp
1599 .sp
1600 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1601 the case when it is passed a junk pointer and to detect when a pattern that was
1602 compiled in an environment of one endianness is run in an environment with the
1603 other endianness. This is the error that PCRE gives when the magic number is
1604 not present.
1605 .sp
1607 .sp
1608 While running the pattern match, an unknown item was encountered in the
1609 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
1610 of the compiled pattern.
1611 .sp
1613 .sp
1614 If a pattern contains back references, but the \fIovector\fP that is passed to
1615 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
1616 gets a block of memory at the start of matching to use for this purpose. If the
1617 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
1618 automatically freed at the end of matching.
1619 .sp
1621 .sp
1622 This error is used by the \fBpcre_copy_substring()\fP,
1623 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
1624 below). It is never returned by \fBpcre_exec()\fP.
1625 .sp
1627 .sp
1628 The backtracking limit, as specified by the \fImatch_limit\fP field in a
1629 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
1630 above.
1631 .sp
1633 .sp
1634 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
1635 use by callout functions that want to yield a distinctive error code. See the
1636 .\" HREF
1637 \fBpcrecallout\fP
1638 .\"
1639 documentation for details.
1640 .sp
1642 .sp
1643 A string that contains an invalid UTF-8 byte sequence was passed as a subject.
1644 .sp
1646 .sp
1647 The UTF-8 byte sequence that was passed as a subject was valid, but the value
1648 of \fIstartoffset\fP did not point to the beginning of a UTF-8 character.
1649 .sp
1651 .sp
1652 The subject string did not match, but it did match partially. See the
1653 .\" HREF
1654 \fBpcrepartial\fP
1655 .\"
1656 documentation for details of partial matching.
1657 .sp
1659 .sp
1660 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
1661 option was used with a compiled pattern containing items that were not
1662 supported for partial matching. From release 8.00 onwards, there are no
1663 restrictions on partial matching.
1664 .sp
1666 .sp
1667 An unexpected internal error has occurred. This error could be caused by a bug
1668 in PCRE or by overwriting of the compiled pattern.
1669 .sp
1671 .sp
1672 This error is given if the value of the \fIovecsize\fP argument is negative.
1673 .sp
1675 .sp
1676 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
1677 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
1678 description above.
1679 .sp
1681 .sp
1682 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
1683 .P
1684 Error numbers -16 to -20 and -22 are not used by \fBpcre_exec()\fP.
1685 .
1686 .
1688 .rs
1689 .sp
1690 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1691 .ti +5n
1692 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
1693 .ti +5n
1694 .B int \fIbuffersize\fP);
1695 .PP
1696 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
1697 .ti +5n
1698 .B int \fIstringcount\fP, int \fIstringnumber\fP,
1699 .ti +5n
1700 .B const char **\fIstringptr\fP);
1701 .PP
1702 .B int pcre_get_substring_list(const char *\fIsubject\fP,
1703 .ti +5n
1704 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
1705 .PP
1706 Captured substrings can be accessed directly by using the offsets returned by
1707 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
1708 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
1709 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
1710 as new, separate, zero-terminated strings. These functions identify substrings
1711 by number. The next section describes functions for extracting named
1712 substrings.
1713 .P
1714 A substring that contains a binary zero is correctly extracted and has a
1715 further zero added on the end, but the result is not, of course, a C string.
1716 However, you can process such a string by referring to the length that is
1717 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
1718 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
1719 for handling strings containing binary zeros, because the end of the final
1720 string is not independently indicated.
1721 .P
1722 The first three arguments are the same for all three of these functions:
1723 \fIsubject\fP is the subject string that has just been successfully matched,
1724 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
1725 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
1726 captured by the match, including the substring that matched the entire regular
1727 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
1728 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
1729 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
1730 number of elements in the vector divided by three.
1731 .P
1732 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
1733 extract a single substring, whose number is given as \fIstringnumber\fP. A
1734 value of zero extracts the substring that matched the entire pattern, whereas
1735 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
1736 the string is placed in \fIbuffer\fP, whose length is given by
1737 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
1738 obtained via \fBpcre_malloc\fP, and its address is returned via
1739 \fIstringptr\fP. The yield of the function is the length of the string, not
1740 including the terminating zero, or one of these error codes:
1741 .sp
1743 .sp
1744 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
1745 memory failed for \fBpcre_get_substring()\fP.
1746 .sp
1748 .sp
1749 There is no substring whose number is \fIstringnumber\fP.
1750 .P
1751 The \fBpcre_get_substring_list()\fP function extracts all available substrings
1752 and builds a list of pointers to them. All this is done in a single block of
1753 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
1754 is returned via \fIlistptr\fP, which is also the start of the list of string
1755 pointers. The end of the list is marked by a NULL pointer. The yield of the
1756 function is zero if all went well, or the error code
1757 .sp
1759 .sp
1760 if the attempt to get the memory block failed.
1761 .P
1762 When any of these functions encounter a substring that is unset, which can
1763 happen when capturing subpattern number \fIn+1\fP matches some part of the
1764 subject, but subpattern \fIn\fP has not been used at all, they return an empty
1765 string. This can be distinguished from a genuine zero-length substring by
1766 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
1767 substrings.
1768 .P
1769 The two convenience functions \fBpcre_free_substring()\fP and
1770 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
1771 a previous call of \fBpcre_get_substring()\fP or
1772 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
1773 the function pointed to by \fBpcre_free\fP, which of course could be called
1774 directly from a C program. However, PCRE is used in some situations where it is
1775 linked via a special interface to another programming language that cannot use
1776 \fBpcre_free\fP directly; it is for these cases that the functions are
1777 provided.
1778 .
1779 .
1781 .rs
1782 .sp
1783 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
1784 .ti +5n
1785 .B const char *\fIname\fP);
1786 .PP
1787 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
1788 .ti +5n
1789 .B const char *\fIsubject\fP, int *\fIovector\fP,
1790 .ti +5n
1791 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1792 .ti +5n
1793 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
1794 .PP
1795 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
1796 .ti +5n
1797 .B const char *\fIsubject\fP, int *\fIovector\fP,
1798 .ti +5n
1799 .B int \fIstringcount\fP, const char *\fIstringname\fP,
1800 .ti +5n
1801 .B const char **\fIstringptr\fP);
1802 .PP
1803 To extract a substring by name, you first have to find associated number.
1804 For example, for this pattern
1805 .sp
1806 (a+)b(?<xxx>\ed+)...
1807 .sp
1808 the number of the subpattern called "xxx" is 2. If the name is known to be
1809 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
1810 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
1811 pattern, and the second is the name. The yield of the function is the
1812 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
1813 that name.
1814 .P
1815 Given the number, you can extract the substring directly, or use one of the
1816 functions described in the previous section. For convenience, there are also
1817 two functions that do the whole job.
1818 .P
1819 Most of the arguments of \fBpcre_copy_named_substring()\fP and
1820 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
1821 functions that extract by number. As these are described in the previous
1822 section, they are not re-described here. There are just two differences:
1823 .P
1824 First, instead of a substring number, a substring name is given. Second, there
1825 is an extra argument, given at the start, which is a pointer to the compiled
1826 pattern. This is needed in order to gain access to the name-to-number
1827 translation table.
1828 .P
1829 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
1830 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
1831 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
1832 the behaviour may not be what you want (see the next section).
1833 .P
1834 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
1835 subpatterns with the same number, as described in the
1836 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1837 .\" </a>
1838 section on duplicate subpattern numbers
1839 .\"
1840 in the
1841 .\" HREF
1842 \fBpcrepattern\fP
1843 .\"
1844 page, you cannot use names to distinguish the different subpatterns, because
1845 names are not included in the compiled code. The matching process uses only
1846 numbers. For this reason, the use of different names for subpatterns of the
1847 same number causes an error at compile time.
1848 .
1850 .rs
1851 .sp
1852 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
1853 .ti +5n
1854 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
1855 .PP
1856 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
1857 are not required to be unique. (Duplicate names are always allowed for
1858 subpatterns with the same number, created by using the (?| feature. Indeed, if
1859 such subpatterns are named, they are required to use the same names.)
1860 .P
1861 Normally, patterns with duplicate names are such that in any one match, only
1862 one of the named subpatterns participates. An example is shown in the
1863 .\" HREF
1864 \fBpcrepattern\fP
1865 .\"
1866 documentation.
1867 .P
1868 When duplicates are present, \fBpcre_copy_named_substring()\fP and
1869 \fBpcre_get_named_substring()\fP return the first substring corresponding to
1870 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
1871 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
1872 returns one of the numbers that are associated with the name, but it is not
1873 defined which it is.
1874 .P
1875 If you want to get full details of all captured substrings for a given name,
1876 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
1877 argument is the compiled pattern, and the second is the name. The third and
1878 fourth are pointers to variables which are updated by the function. After it
1879 has run, they point to the first and last entries in the name-to-number table
1880 for the given name. The function itself returns the length of each entry, or
1881 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
1882 described above in the section entitled \fIInformation about a pattern\fP.
1883 Given all the relevant entries for the name, you can extract each of their
1884 numbers, and hence the captured data, if any.
1885 .
1886 .
1888 .rs
1889 .sp
1890 The traditional matching function uses a similar algorithm to Perl, which stops
1891 when it finds the first match, starting at a given point in the subject. If you
1892 want to find all possible matches, or the longest possible match, consider
1893 using the alternative matching function (see below) instead. If you cannot use
1894 the alternative function, but still need to find all possible matches, you
1895 can kludge it up by making use of the callout facility, which is described in
1896 the
1897 .\" HREF
1898 \fBpcrecallout\fP
1899 .\"
1900 documentation.
1901 .P
1902 What you have to do is to insert a callout right at the end of the pattern.
1903 When your callout function is called, extract and save the current matched
1904 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
1905 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
1906 will yield PCRE_ERROR_NOMATCH.
1907 .
1908 .
1909 .\" HTML <a name="dfamatch"></a>
1911 .rs
1912 .sp
1913 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1914 .ti +5n
1915 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1916 .ti +5n
1917 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
1918 .ti +5n
1919 .B int *\fIworkspace\fP, int \fIwscount\fP);
1920 .P
1921 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
1922 a compiled pattern, using a matching algorithm that scans the subject string
1923 just once, and does not backtrack. This has different characteristics to the
1924 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
1925 patterns are not supported. Nevertheless, there are times when this kind of
1926 matching can be useful. For a discussion of the two matching algorithms, and a
1927 list of features that \fBpcre_dfa_exec()\fP does not support, see the
1928 .\" HREF
1929 \fBpcrematching\fP
1930 .\"
1931 documentation.
1932 .P
1933 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
1934 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
1935 different way, and this is described below. The other common arguments are used
1936 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
1937 here.
1938 .P
1939 The two additional arguments provide workspace for the function. The workspace
1940 vector should contain at least 20 elements. It is used for keeping track of
1941 multiple paths through the pattern tree. More workspace will be needed for
1942 patterns and subjects where there are a lot of potential matches.
1943 .P
1944 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
1945 .sp
1946 int rc;
1947 int ovector[10];
1948 int wspace[20];
1949 rc = pcre_dfa_exec(
1950 re, /* result of pcre_compile() */
1951 NULL, /* we didn't study the pattern */
1952 "some string", /* the subject string */
1953 11, /* the length of the subject string */
1954 0, /* start at offset 0 in the subject */
1955 0, /* default options */
1956 ovector, /* vector of integers for substring information */
1957 10, /* number of elements (NOT size in bytes) */
1958 wspace, /* working space vector */
1959 20); /* number of elements (NOT size in bytes) */
1960 .
1961 .SS "Option bits for \fBpcre_dfa_exec()\fP"
1962 .rs
1963 .sp
1964 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
1965 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1968 and PCRE_DFA_RESTART. All but the last four of these are exactly the same as
1969 for \fBpcre_exec()\fP, so their description is not repeated here.
1970 .sp
1973 .sp
1974 These have the same general effect as they do for \fBpcre_exec()\fP, but the
1975 details are slightly different. When PCRE_PARTIAL_HARD is set for
1976 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
1977 is reached and there is still at least one matching possibility that requires
1978 additional characters. This happens even if some complete matches have also
1979 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
1980 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
1981 there have been no complete matches, but there is still at least one matching
1982 possibility. The portion of the string that was inspected when the longest
1983 partial match was found is set as the first matching string in both cases.
1984 .sp
1986 .sp
1987 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
1988 soon as it has found one match. Because of the way the alternative algorithm
1989 works, this is necessarily the shortest possible match at the first possible
1990 matching point in the subject string.
1991 .sp
1993 .sp
1994 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
1995 again, with additional subject characters, and have it continue with the same
1996 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
1997 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
1998 before because data about the match so far is left in them after a partial
1999 match. There is more discussion of this facility in the
2000 .\" HREF
2001 \fBpcrepartial\fP
2002 .\"
2003 documentation.
2004 .
2005 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2006 .rs
2007 .sp
2008 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2009 substring in the subject. Note, however, that all the matches from one run of
2010 the function start at the same point in the subject. The shorter matches are
2011 all initial substrings of the longer matches. For example, if the pattern
2012 .sp
2013 <.*>
2014 .sp
2015 is matched against the string
2016 .sp
2017 This is <something> <something else> <something further> no more
2018 .sp
2019 the three matched strings are
2020 .sp
2021 <something>
2022 <something> <something else>
2023 <something> <something else> <something further>
2024 .sp
2025 On success, the yield of the function is a number greater than zero, which is
2026 the number of matched substrings. The substrings themselves are returned in
2027 \fIovector\fP. Each string uses two elements; the first is the offset to the
2028 start, and the second is the offset to the end. In fact, all the strings have
2029 the same start offset. (Space could have been saved by giving this only once,
2030 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2031 returns data, even though the meaning of the strings is different.)
2032 .P
2033 The strings are returned in reverse order of length; that is, the longest
2034 matching string is given first. If there were too many matches to fit into
2035 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2036 the longest matches.
2037 .
2038 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2039 .rs
2040 .sp
2041 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2042 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2043 described
2044 .\" HTML <a href="#errorlist">
2045 .\" </a>
2046 above.
2047 .\"
2048 There are in addition the following errors that are specific to
2049 \fBpcre_dfa_exec()\fP:
2050 .sp
2052 .sp
2053 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2054 that it does not support, for instance, the use of \eC or a back reference.
2055 .sp
2057 .sp
2058 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2059 uses a back reference for the condition, or a test for recursion in a specific
2060 group. These are not supported.
2061 .sp
2063 .sp
2064 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2065 block that contains a setting of the \fImatch_limit\fP field. This is not
2066 supported (it is meaningless).
2067 .sp
2069 .sp
2070 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2071 \fIworkspace\fP vector.
2072 .sp
2074 .sp
2075 When a recursive subpattern is processed, the matching function calls itself
2076 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2077 error is given if the output vector is not large enough. This should be
2078 extremely rare, as a vector of size 1000 is used.
2079 .
2080 .
2081 .SH "SEE ALSO"
2082 .rs
2083 .sp
2084 \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2085 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2086 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2087 .
2088 .
2090 .rs
2091 .sp
2092 .nf
2093 Philip Hazel
2094 University Computing Service
2095 Cambridge CB2 3QH, England.
2096 .fi
2097 .
2098 .
2100 .rs
2101 .sp
2102 .nf
2103 Last updated: 03 May 2010
2104 Copyright (c) 1997-2010 University of Cambridge.
2105 .fi


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

  ViewVC Help
Powered by ViewVC 1.1.5