ViewVC logotype

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

Parent Directory Parent Directory | Revision Log Revision Log

Revision 1031 - (show annotations)
Sat Sep 8 15:59:01 2012 UTC (6 years, 10 months ago) by ph10
File size: 116749 byte(s)
Error occurred while calculating annotation data.
More documentation updates/tidies for EBCDIC environments.
1 .TH PCREAPI 3 "07 September 2012" "PCRE 8.32"
3 PCRE - Perl-compatible regular expressions
4 .sp
5 .B #include <pcre.h>
6 .
7 .
9 .rs
10 .sp
11 .SM
12 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
13 .ti +5n
14 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
15 .ti +5n
16 .B const unsigned char *\fItableptr\fP);
17 .PP
18 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
19 .ti +5n
20 .B int *\fIerrorcodeptr\fP,
21 .ti +5n
22 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
23 .ti +5n
24 .B const unsigned char *\fItableptr\fP);
25 .PP
26 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP,
27 .ti +5n
28 .B const char **\fIerrptr\fP);
29 .PP
30 .B void pcre_free_study(pcre_extra *\fIextra\fP);
31 .PP
32 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
33 .ti +5n
34 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
35 .ti +5n
36 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
37 .PP
38 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
39 .ti +5n
40 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
41 .ti +5n
42 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
43 .ti +5n
44 .B int *\fIworkspace\fP, int \fIwscount\fP);
45 .
46 .
48 .rs
49 .sp
50 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
51 .ti +5n
52 .B const char *\fIsubject\fP, int *\fIovector\fP,
53 .ti +5n
54 .B int \fIstringcount\fP, const char *\fIstringname\fP,
55 .ti +5n
56 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
57 .PP
58 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
59 .ti +5n
60 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
61 .ti +5n
62 .B int \fIbuffersize\fP);
63 .PP
64 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
65 .ti +5n
66 .B const char *\fIsubject\fP, int *\fIovector\fP,
67 .ti +5n
68 .B int \fIstringcount\fP, const char *\fIstringname\fP,
69 .ti +5n
70 .B const char **\fIstringptr\fP);
71 .PP
72 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
73 .ti +5n
74 .B const char *\fIname\fP);
75 .PP
76 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
77 .ti +5n
78 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
79 .PP
80 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
81 .ti +5n
82 .B int \fIstringcount\fP, int \fIstringnumber\fP,
83 .ti +5n
84 .B const char **\fIstringptr\fP);
85 .PP
86 .B int pcre_get_substring_list(const char *\fIsubject\fP,
87 .ti +5n
88 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
89 .PP
90 .B void pcre_free_substring(const char *\fIstringptr\fP);
91 .PP
92 .B void pcre_free_substring_list(const char **\fIstringptr\fP);
93 .
94 .
96 .rs
97 .sp
98 .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
99 .PP
100 .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
101 .PP
102 .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
103 .ti +5n
104 .B pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);
105 .PP
106 .B const unsigned char *pcre_maketables(void);
107 .PP
108 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
109 .ti +5n
110 .B int \fIwhat\fP, void *\fIwhere\fP);
111 .PP
112 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
113 .PP
114 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
115 .PP
116 .B const char *pcre_version(void);
117 .PP
118 .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
119 .ti +5n
120 .B pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);
121 .
122 .
124 .rs
125 .sp
126 .B void *(*pcre_malloc)(size_t);
127 .PP
128 .B void (*pcre_free)(void *);
129 .PP
130 .B void *(*pcre_stack_malloc)(size_t);
131 .PP
132 .B void (*pcre_stack_free)(void *);
133 .PP
134 .B int (*pcre_callout)(pcre_callout_block *);
135 .
136 .
138 .rs
139 .sp
140 From release 8.30, PCRE can be compiled as a library for handling 16-bit
141 character strings as well as, or instead of, the original library that handles
142 8-bit character strings. To avoid too much complication, this document
143 describes the 8-bit versions of the functions, with only occasional references
144 to the 16-bit library.
145 .P
146 The 16-bit functions operate in the same way as their 8-bit counterparts; they
147 just use different data types for their arguments and results, and their names
148 start with \fBpcre16_\fP instead of \fBpcre_\fP. For every option that has UTF8
149 in its name (for example, PCRE_UTF8), there is a corresponding 16-bit name with
150 UTF8 replaced by UTF16. This facility is in fact just cosmetic; the 16-bit
151 option names define the same bit values.
152 .P
153 References to bytes and UTF-8 in this document should be read as references to
154 16-bit data quantities and UTF-16 when using the 16-bit library, unless
155 specified otherwise. More details of the specific differences for the 16-bit
156 library are given in the
157 .\" HREF
158 \fBpcre16\fP
159 .\"
160 page.
161 .
162 .
164 .rs
165 .sp
166 PCRE has its own native API, which is described in this document. There are
167 also some wrapper functions (for the 8-bit library only) that correspond to the
168 POSIX regular expression API, but they do not give access to all the
169 functionality. They are described in the
170 .\" HREF
171 \fBpcreposix\fP
172 .\"
173 documentation. Both of these APIs define a set of C function calls. A C++
174 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
175 documented in the
176 .\" HREF
177 \fBpcrecpp\fP
178 .\"
179 page.
180 .P
181 The native API C function prototypes are defined in the header file
182 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
183 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
184 command for linking an application that uses PCRE. The header file defines the
185 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
186 for the library. Applications can use these to include support for different
187 releases of PCRE.
188 .P
189 In a Windows environment, if you want to statically link an application program
190 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
191 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
192 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
193 \fB__declspec(dllimport)\fP, with unwanted results.
194 .P
195 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
196 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
197 in a Perl-compatible manner. A sample program that demonstrates the simplest
198 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
199 source distribution. A listing of this program is given in the
200 .\" HREF
201 \fBpcredemo\fP
202 .\"
203 documentation, and the
204 .\" HREF
205 \fBpcresample\fP
206 .\"
207 documentation describes how to compile and run it.
208 .P
209 Just-in-time compiler support is an optional feature of PCRE that can be built
210 in appropriate hardware environments. It greatly speeds up the matching
211 performance of many patterns. Simple programs can easily request that it be
212 used if available, by setting an option that is ignored when it is not
213 relevant. More complicated programs might need to make use of the functions
214 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
215 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
216 These functions are discussed in the
217 .\" HREF
218 \fBpcrejit\fP
219 .\"
220 documentation.
221 .P
222 A second matching function, \fBpcre_dfa_exec()\fP, which is not
223 Perl-compatible, is also provided. This uses a different algorithm for the
224 matching. The alternative algorithm finds all possible matches (at a given
225 point in the subject), and scans the subject just once (unless there are
226 lookbehind assertions). However, this algorithm does not return captured
227 substrings. A description of the two matching algorithms and their advantages
228 and disadvantages is given in the
229 .\" HREF
230 \fBpcrematching\fP
231 .\"
232 documentation.
233 .P
234 In addition to the main compiling and matching functions, there are convenience
235 functions for extracting captured substrings from a subject string that is
236 matched by \fBpcre_exec()\fP. They are:
237 .sp
238 \fBpcre_copy_substring()\fP
239 \fBpcre_copy_named_substring()\fP
240 \fBpcre_get_substring()\fP
241 \fBpcre_get_named_substring()\fP
242 \fBpcre_get_substring_list()\fP
243 \fBpcre_get_stringnumber()\fP
244 \fBpcre_get_stringtable_entries()\fP
245 .sp
246 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
247 provided, to free the memory used for extracted strings.
248 .P
249 The function \fBpcre_maketables()\fP is used to build a set of character tables
250 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
251 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
252 specialist use. Most commonly, no special tables are passed, in which case
253 internal tables that are generated when PCRE is built are used.
254 .P
255 The function \fBpcre_fullinfo()\fP is used to find out information about a
256 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
257 string containing the version of PCRE and its date of release.
258 .P
259 The function \fBpcre_refcount()\fP maintains a reference count in a data block
260 containing a compiled pattern. This is provided for the benefit of
261 object-oriented applications.
262 .P
263 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
264 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
265 respectively. PCRE calls the memory management functions via these variables,
266 so a calling program can replace them if it wishes to intercept the calls. This
267 should be done before calling any PCRE functions.
268 .P
269 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
270 indirections to memory management functions. These special functions are used
271 only when PCRE is compiled to use the heap for remembering data, instead of
272 recursive function calls, when running the \fBpcre_exec()\fP function. See the
273 .\" HREF
274 \fBpcrebuild\fP
275 .\"
276 documentation for details of how to do this. It is a non-standard way of
277 building PCRE, for use in environments that have limited stacks. Because of the
278 greater use of memory management, it runs more slowly. Separate functions are
279 provided so that special-purpose external code can be used for this case. When
280 used, these functions are always called in a stack-like manner (last obtained,
281 first freed), and always for memory blocks of the same size. There is a
282 discussion about PCRE's stack usage in the
283 .\" HREF
284 \fBpcrestack\fP
285 .\"
286 documentation.
287 .P
288 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
289 by the caller to a "callout" function, which PCRE will then call at specified
290 points during a matching operation. Details are given in the
291 .\" HREF
292 \fBpcrecallout\fP
293 .\"
294 documentation.
295 .
296 .
297 .\" HTML <a name="newlines"></a>
299 .rs
300 .sp
301 PCRE supports five different conventions for indicating line breaks in
302 strings: a single CR (carriage return) character, a single LF (linefeed)
303 character, the two-character sequence CRLF, any of the three preceding, or any
304 Unicode newline sequence. The Unicode newline sequences are the three just
305 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
306 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
307 (paragraph separator, U+2029).
308 .P
309 Each of the first three conventions is used by at least one operating system as
310 its standard newline sequence. When PCRE is built, a default can be specified.
311 The default default is LF, which is the Unix standard. When PCRE is run, the
312 default can be overridden, either when a pattern is compiled, or when it is
313 matched.
314 .P
315 At compile time, the newline convention can be specified by the \fIoptions\fP
316 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
317 start of the pattern itself; this overrides any other settings. See the
318 .\" HREF
319 \fBpcrepattern\fP
320 .\"
321 page for details of the special character sequences.
322 .P
323 In the PCRE documentation the word "newline" is used to mean "the character or
324 pair of characters that indicate a line break". The choice of newline
325 convention affects the handling of the dot, circumflex, and dollar
326 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
327 recognized line ending sequence, the match position advancement for a
328 non-anchored pattern. There is more detail about this in the
329 .\" HTML <a href="#execoptions">
330 .\" </a>
331 section on \fBpcre_exec()\fP options
332 .\"
333 below.
334 .P
335 The choice of newline convention does not affect the interpretation of
336 the \en or \er escape sequences, nor does it affect what \eR matches, which is
337 controlled in a similar way, but by separate options.
338 .
339 .
341 .rs
342 .sp
343 The PCRE functions can be used in multi-threading applications, with the
344 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
345 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
346 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
347 .P
348 The compiled form of a regular expression is not altered during matching, so
349 the same compiled pattern can safely be used by several threads at once.
350 .P
351 If the just-in-time optimization feature is being used, it needs separate
352 memory stack areas for each thread. See the
353 .\" HREF
354 \fBpcrejit\fP
355 .\"
356 documentation for more details.
357 .
358 .
360 .rs
361 .sp
362 The compiled form of a regular expression can be saved and re-used at a later
363 time, possibly by a different program, and even on a host other than the one on
364 which it was compiled. Details are given in the
365 .\" HREF
366 \fBpcreprecompile\fP
367 .\"
368 documentation, which includes a description of the
369 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
370 expression with one version of PCRE for use with a different version is not
371 guaranteed to work and may cause crashes.
372 .
373 .
375 .rs
376 .sp
377 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
378 .PP
379 The function \fBpcre_config()\fP makes it possible for a PCRE client to
380 discover which optional features have been compiled into the PCRE library. The
381 .\" HREF
382 \fBpcrebuild\fP
383 .\"
384 documentation has more details about these optional features.
385 .P
386 The first argument for \fBpcre_config()\fP is an integer, specifying which
387 information is required; the second argument is a pointer to a variable into
388 which the information is placed. The returned value is zero on success, or the
389 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
390 not recognized. The following information is available:
391 .sp
393 .sp
394 The output is an integer that is set to one if UTF-8 support is available;
395 otherwise it is set to zero. If this option is given to the 16-bit version of
396 this function, \fBpcre16_config()\fP, the result is PCRE_ERROR_BADOPTION.
397 .sp
399 .sp
400 The output is an integer that is set to one if UTF-16 support is available;
401 otherwise it is set to zero. This value should normally be given to the 16-bit
402 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
403 version of this function, the result is PCRE_ERROR_BADOPTION.
404 .sp
406 .sp
407 The output is an integer that is set to one if support for Unicode character
408 properties is available; otherwise it is set to zero.
409 .sp
411 .sp
412 The output is an integer that is set to one if support for just-in-time
413 compiling is available; otherwise it is set to zero.
414 .sp
416 .sp
417 The output is a pointer to a zero-terminated "const char *" string. If JIT
418 support is available, the string contains the name of the architecture for
419 which the JIT compiler is configured, for example "x86 32bit (little endian +
420 unaligned)". If JIT support is not available, the result is NULL.
421 .sp
423 .sp
424 The output is an integer whose value specifies the default character sequence
425 that is recognized as meaning "newline". The values that are supported in
426 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
427 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
428 same values. However, the value for LF is normally 21, though some EBCDIC
429 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
430 default should normally correspond to the standard sequence for your operating
431 system.
432 .sp
434 .sp
435 The output is an integer whose value indicates what character sequences the \eR
436 escape sequence matches by default. A value of 0 means that \eR matches any
437 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
438 or CRLF. The default can be overridden when a pattern is compiled or matched.
439 .sp
441 .sp
442 The output is an integer that contains the number of bytes used for internal
443 linkage in compiled regular expressions. For the 8-bit library, the value can
444 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
445 a number of bytes. The default value of 2 is sufficient for all but the most
446 massive patterns, since it allows the compiled pattern to be up to 64K in size.
447 Larger values allow larger regular expressions to be compiled, at the expense
448 of slower matching.
449 .sp
451 .sp
452 The output is an integer that contains the threshold above which the POSIX
453 interface uses \fBmalloc()\fP for output vectors. Further details are given in
454 the
455 .\" HREF
456 \fBpcreposix\fP
457 .\"
458 documentation.
459 .sp
461 .sp
462 The output is a long integer that gives the default limit for the number of
463 internal matching function calls in a \fBpcre_exec()\fP execution. Further
464 details are given with \fBpcre_exec()\fP below.
465 .sp
467 .sp
468 The output is a long integer that gives the default limit for the depth of
469 recursion when calling the internal matching function in a \fBpcre_exec()\fP
470 execution. Further details are given with \fBpcre_exec()\fP below.
471 .sp
473 .sp
474 The output is an integer that is set to one if internal recursion when running
475 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
476 to remember their state. This is the usual way that PCRE is compiled. The
477 output is zero if PCRE was compiled to use blocks of data on the heap instead
478 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
479 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
480 avoiding the use of the stack.
481 .
482 .
484 .rs
485 .sp
486 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
487 .ti +5n
488 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
489 .ti +5n
490 .B const unsigned char *\fItableptr\fP);
491 .sp
492 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
493 .ti +5n
494 .B int *\fIerrorcodeptr\fP,
495 .ti +5n
496 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
497 .ti +5n
498 .B const unsigned char *\fItableptr\fP);
499 .P
500 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
501 called to compile a pattern into an internal form. The only difference between
502 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
503 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
504 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
505 information applies equally to \fBpcre_compile2()\fP.
506 .P
507 The pattern is a C string terminated by a binary zero, and is passed in the
508 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
509 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
510 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
511 for a structure whose contents are not externally defined. It is up to the
512 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
513 .P
514 Although the compiled code of a PCRE regex is relocatable, that is, it does not
515 depend on memory location, the complete \fBpcre\fP data block is not
516 fully relocatable, because it may contain a copy of the \fItableptr\fP
517 argument, which is an address (see below).
518 .P
519 The \fIoptions\fP argument contains various bit settings that affect the
520 compilation. It should be zero if no options are required. The available
521 options are described below. Some of them (in particular, those that are
522 compatible with Perl, but some others as well) can also be set and unset from
523 within the pattern (see the detailed description in the
524 .\" HREF
525 \fBpcrepattern\fP
526 .\"
527 documentation). For those options that can be different in different parts of
528 the pattern, the contents of the \fIoptions\fP argument specifies their
529 settings at the start of compilation and execution. The PCRE_ANCHORED,
531 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
532 compile time.
533 .P
534 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
535 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
536 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
537 error message. This is a static string that is part of the library. You must
538 not try to free it. Normally, the offset from the start of the pattern to the
539 byte that was being processed when the error was discovered is placed in the
540 variable pointed to by \fIerroffset\fP, which must not be NULL (if it is, an
541 immediate error is given). However, for an invalid UTF-8 string, the offset is
542 that of the first byte of the failing character.
543 .P
544 Some errors are not detected until the whole pattern has been scanned; in these
545 cases, the offset passed back is the length of the pattern. Note that the
546 offset is in bytes, not characters, even in UTF-8 mode. It may sometimes point
547 into the middle of a UTF-8 character.
548 .P
549 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
550 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
551 returned via this argument in the event of an error. This is in addition to the
552 textual error message. Error codes and messages are listed below.
553 .P
554 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
555 character tables that are built when PCRE is compiled, using the default C
556 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
557 call to \fBpcre_maketables()\fP. This value is stored with the compiled
558 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
559 passed to it. For more discussion, see the section on locale support below.
560 .P
561 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
562 .sp
563 pcre *re;
564 const char *error;
565 int erroffset;
566 re = pcre_compile(
567 "^A.*Z", /* the pattern */
568 0, /* default options */
569 &error, /* for error message */
570 &erroffset, /* for error offset */
571 NULL); /* use default character tables */
572 .sp
573 The following names for option bits are defined in the \fBpcre.h\fP header
574 file:
575 .sp
577 .sp
578 If this bit is set, the pattern is forced to be "anchored", that is, it is
579 constrained to match only at the first matching point in the string that is
580 being searched (the "subject string"). This effect can also be achieved by
581 appropriate constructs in the pattern itself, which is the only way to do it in
582 Perl.
583 .sp
585 .sp
586 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
587 all with number 255, before each pattern item. For discussion of the callout
588 facility, see the
589 .\" HREF
590 \fBpcrecallout\fP
591 .\"
592 documentation.
593 .sp
596 .sp
597 These options (which are mutually exclusive) control what the \eR escape
598 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
599 match any Unicode newline sequence. The default is specified when PCRE is
600 built. It can be overridden from within the pattern, or by setting an option
601 when a compiled pattern is matched.
602 .sp
604 .sp
605 If this bit is set, letters in the pattern match both upper and lower case
606 letters. It is equivalent to Perl's /i option, and it can be changed within a
607 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
608 concept of case for characters whose values are less than 128, so caseless
609 matching is always possible. For characters with higher values, the concept of
610 case is supported if PCRE is compiled with Unicode property support, but not
611 otherwise. If you want to use caseless matching for characters 128 and above,
612 you must ensure that PCRE is compiled with Unicode property support as well as
613 with UTF-8 support.
614 .sp
616 .sp
617 If this bit is set, a dollar metacharacter in the pattern matches only at the
618 end of the subject string. Without this option, a dollar also matches
619 immediately before a newline at the end of the string (but not before any other
620 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
621 There is no equivalent to this option in Perl, and no way to set it within a
622 pattern.
623 .sp
625 .sp
626 If this bit is set, a dot metacharacter in the pattern matches a character of
627 any value, including one that indicates a newline. However, it only ever
628 matches one character, even if newlines are coded as CRLF. Without this option,
629 a dot does not match when the current position is at a newline. This option is
630 equivalent to Perl's /s option, and it can be changed within a pattern by a
631 (?s) option setting. A negative class such as [^a] always matches newline
632 characters, independent of the setting of this option.
633 .sp
635 .sp
636 If this bit is set, names used to identify capturing subpatterns need not be
637 unique. This can be helpful for certain types of pattern when it is known that
638 only one instance of the named subpattern can ever be matched. There are more
639 details of named subpatterns below; see also the
640 .\" HREF
641 \fBpcrepattern\fP
642 .\"
643 documentation.
644 .sp
646 .sp
647 If this bit is set, white space data characters in the pattern are totally
648 ignored except when escaped or inside a character class. White space does not
649 include the VT character (code 11). In addition, characters between an
650 unescaped # outside a character class and the next newline, inclusive, are also
651 ignored. This is equivalent to Perl's /x option, and it can be changed within a
652 pattern by a (?x) option setting.
653 .P
654 Which characters are interpreted as newlines is controlled by the options
655 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
656 pattern, as described in the section entitled
657 .\" HTML <a href="pcrepattern.html#newlines">
658 .\" </a>
659 "Newline conventions"
660 .\"
661 in the \fBpcrepattern\fP documentation. Note that the end of this type of
662 comment is a literal newline sequence in the pattern; escape sequences that
663 happen to represent a newline do not count.
664 .P
665 This option makes it possible to include comments inside complicated patterns.
666 Note, however, that this applies only to data characters. White space characters
667 may never appear within special character sequences in a pattern, for example
668 within the sequence (?( that introduces a conditional subpattern.
669 .sp
671 .sp
672 This option was invented in order to turn on additional functionality of PCRE
673 that is incompatible with Perl, but it is currently of very little use. When
674 set, any backslash in a pattern that is followed by a letter that has no
675 special meaning causes an error, thus reserving these combinations for future
676 expansion. By default, as in Perl, a backslash followed by a letter with no
677 special meaning is treated as a literal. (Perl can, however, be persuaded to
678 give an error for this, by running it with the -w option.) There are at present
679 no other features controlled by this option. It can also be set by a (?X)
680 option setting within a pattern.
681 .sp
683 .sp
684 If this option is set, an unanchored pattern is required to match before or at
685 the first newline in the subject string, though the matched text may continue
686 over the newline.
687 .sp
689 .sp
690 If this option is set, PCRE's behaviour is changed in some ways so that it is
691 compatible with JavaScript rather than Perl. The changes are as follows:
692 .P
693 (1) A lone closing square bracket in a pattern causes a compile-time error,
694 because this is illegal in JavaScript (by default it is treated as a data
695 character). Thus, the pattern AB]CD becomes illegal when this option is set.
696 .P
697 (2) At run time, a back reference to an unset subpattern group matches an empty
698 string (by default this causes the current matching alternative to fail). A
699 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
700 an "a" in the subject), whereas it fails by default, for Perl compatibility.
701 .P
702 (3) \eU matches an upper case "U" character; by default \eU causes a compile
703 time error (Perl uses \eU to upper case subsequent characters).
704 .P
705 (4) \eu matches a lower case "u" character unless it is followed by four
706 hexadecimal digits, in which case the hexadecimal number defines the code point
707 to match. By default, \eu causes a compile time error (Perl uses it to upper
708 case the following character).
709 .P
710 (5) \ex matches a lower case "x" character unless it is followed by two
711 hexadecimal digits, in which case the hexadecimal number defines the code point
712 to match. By default, as in Perl, a hexadecimal number is always expected after
713 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
714 binary zero character followed by z).
715 .sp
717 .sp
718 By default, PCRE treats the subject string as consisting of a single line of
719 characters (even if it actually contains newlines). The "start of line"
720 metacharacter (^) matches only at the start of the string, while the "end of
721 line" metacharacter ($) matches only at the end of the string, or before a
722 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
723 Perl.
724 .P
725 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
726 match immediately following or immediately before internal newlines in the
727 subject string, respectively, as well as at the very start and end. This is
728 equivalent to Perl's /m option, and it can be changed within a pattern by a
729 (?m) option setting. If there are no newlines in a subject string, or no
730 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
731 .sp
737 .sp
738 These options override the default newline definition that was chosen when PCRE
739 was built. Setting the first or the second specifies that a newline is
740 indicated by a single character (CR or LF, respectively). Setting
741 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
742 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
743 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
744 that any Unicode newline sequence should be recognized.
745 .P
746 In an ASCII/Unicode environment, the Unicode newline sequences are the three
747 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
748 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
749 (paragraph separator, U+2029). For the 8-bit library, the last two are
750 recognized only in UTF-8 mode.
751 .P
752 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
753 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
754 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
755 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
756 less than 256. For more details, see the
757 .\" HREF
758 \fBpcrebuild\fP
759 .\"
760 documentation.
761 .P
762 The newline setting in the options word uses three bits that are treated
763 as a number, giving eight possibilities. Currently only six are used (default
764 plus the five values above). This means that if you set more than one newline
765 option, the combination may or may not be sensible. For example,
767 other combinations may yield unused numbers and cause an error.
768 .P
769 The only time that a line break in a pattern is specially recognized when
770 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
771 and so are ignored in this mode. Also, an unescaped # outside a character class
772 indicates a comment that lasts until after the next line break sequence. In
773 other circumstances, line break sequences in patterns are treated as literal
774 data.
775 .P
776 The newline option that is set at compile time becomes the default that is used
777 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
778 .sp
780 .sp
781 If this option is set, it disables the use of numbered capturing parentheses in
782 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
783 were followed by ?: but named parentheses can still be used for capturing (and
784 they acquire numbers in the usual way). There is no equivalent of this option
785 in Perl.
786 .sp
788 .sp
789 This is an option that acts at matching time; that is, it is really an option
790 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
791 it is remembered with the compiled pattern and assumed at matching time. For
792 details see the discussion of PCRE_NO_START_OPTIMIZE
793 .\" HTML <a href="#execoptions">
794 .\" </a>
795 below.
796 .\"
797 .sp
799 .sp
800 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
801 \ew, and some of the POSIX character classes. By default, only ASCII characters
802 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
803 classify characters. More details are given in the section on
804 .\" HTML <a href="pcre.html#genericchartypes">
805 .\" </a>
806 generic character types
807 .\"
808 in the
809 .\" HREF
810 \fBpcrepattern\fP
811 .\"
812 page. If you set PCRE_UCP, matching one of the items it affects takes much
813 longer. The option is available only if PCRE has been compiled with Unicode
814 property support.
815 .sp
817 .sp
818 This option inverts the "greediness" of the quantifiers so that they are not
819 greedy by default, but become greedy if followed by "?". It is not compatible
820 with Perl. It can also be set by a (?U) option setting within the pattern.
821 .sp
823 .sp
824 This option causes PCRE to regard both the pattern and the subject as strings
825 of UTF-8 characters instead of single-byte strings. However, it is available
826 only when PCRE is built to include UTF support. If not, the use of this option
827 provokes an error. Details of how this option changes the behaviour of PCRE are
828 given in the
829 .\" HREF
830 \fBpcreunicode\fP
831 .\"
832 page.
833 .sp
835 .sp
836 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8
837 string is automatically checked. There is a discussion about the
838 .\" HTML <a href="pcreunicode.html#utf8strings">
839 .\" </a>
840 validity of UTF-8 strings
841 .\"
842 in the
843 .\" HREF
844 \fBpcreunicode\fP
845 .\"
846 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
847 error. If you already know that your pattern is valid, and you want to skip
848 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
849 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
850 undefined. It may cause your program to crash. Note that this option can also
851 be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
852 validity checking of subject strings.
853 .
854 .
856 .rs
857 .sp
858 The following table lists the error codes than may be returned by
859 \fBpcre_compile2()\fP, along with the error messages that may be returned by
860 both compiling functions. Note that error messages are always 8-bit ASCII
861 strings, even in 16-bit mode. As PCRE has developed, some error codes have
862 fallen out of use. To avoid confusion, they have not been re-used.
863 .sp
864 0 no error
865 1 \e at end of pattern
866 2 \ec at end of pattern
867 3 unrecognized character follows \e
868 4 numbers out of order in {} quantifier
869 5 number too big in {} quantifier
870 6 missing terminating ] for character class
871 7 invalid escape sequence in character class
872 8 range out of order in character class
873 9 nothing to repeat
874 10 [this code is not in use]
875 11 internal error: unexpected repeat
876 12 unrecognized character after (? or (?-
877 13 POSIX named classes are supported only within a class
878 14 missing )
879 15 reference to non-existent subpattern
880 16 erroffset passed as NULL
881 17 unknown option bit(s) set
882 18 missing ) after comment
883 19 [this code is not in use]
884 20 regular expression is too large
885 21 failed to get memory
886 22 unmatched parentheses
887 23 internal error: code overflow
888 24 unrecognized character after (?<
889 25 lookbehind assertion is not fixed length
890 26 malformed number or name after (?(
891 27 conditional group contains more than two branches
892 28 assertion expected after (?(
893 29 (?R or (?[+-]digits must be followed by )
894 30 unknown POSIX class name
895 31 POSIX collating elements are not supported
896 32 this version of PCRE is compiled without UTF support
897 33 [this code is not in use]
898 34 character value in \ex{...} sequence is too large
899 35 invalid condition (?(0)
900 36 \eC not allowed in lookbehind assertion
901 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
902 38 number after (?C is > 255
903 39 closing ) for (?C expected
904 40 recursive call could loop indefinitely
905 41 unrecognized character after (?P
906 42 syntax error in subpattern name (missing terminator)
907 43 two named subpatterns have the same name
908 44 invalid UTF-8 string (specifically UTF-8)
909 45 support for \eP, \ep, and \eX has not been compiled
910 46 malformed \eP or \ep sequence
911 47 unknown property name after \eP or \ep
912 48 subpattern name is too long (maximum 32 characters)
913 49 too many named subpatterns (maximum 10000)
914 50 [this code is not in use]
915 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
916 52 internal error: overran compiling workspace
917 53 internal error: previously-checked referenced subpattern
918 not found
919 54 DEFINE group contains more than one branch
920 55 repeating a DEFINE group is not allowed
921 56 inconsistent NEWLINE options
922 57 \eg is not followed by a braced, angle-bracketed, or quoted
923 name/number or by a plain number
924 58 a numbered reference must not be zero
925 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
926 60 (*VERB) not recognized
927 61 number is too big
928 62 subpattern name expected
929 63 digit expected after (?+
930 64 ] is an invalid data character in JavaScript compatibility mode
931 65 different names for subpatterns of the same number are
932 not allowed
933 66 (*MARK) must have an argument
934 67 this version of PCRE is not compiled with Unicode property
935 support
936 68 \ec must be followed by an ASCII character
937 69 \ek is not followed by a braced, angle-bracketed, or quoted name
938 70 internal error: unknown opcode in find_fixedlength()
939 71 \eN is not supported in a class
940 72 too many forward references
941 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
942 74 invalid UTF-16 string (specifically UTF-16)
943 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
944 76 character value in \eu.... sequence is too large
945 .sp
946 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
947 be used if the limits were changed when PCRE was built.
948 .
949 .
950 .\" HTML <a name="studyingapattern"></a>
952 .rs
953 .sp
954 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
955 .ti +5n
956 .B const char **\fIerrptr\fP);
957 .PP
958 If a compiled pattern is going to be used several times, it is worth spending
959 more time analyzing it in order to speed up the time taken for matching. The
960 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
961 argument. If studying the pattern produces additional information that will
962 help speed up matching, \fBpcre_study()\fP returns a pointer to a
963 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
964 results of the study.
965 .P
966 The returned value from \fBpcre_study()\fP can be passed directly to
967 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
968 also contains other fields that can be set by the caller before the block is
969 passed; these are described
970 .\" HTML <a href="#extradata">
971 .\" </a>
972 below
973 .\"
974 in the section on matching a pattern.
975 .P
976 If studying the pattern does not produce any useful information,
977 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
978 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
979 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
980 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
981 returns a \fBpcre_extra\fP block even if studying did not find any additional
982 information. It may still return NULL, however, if an error occurs in
983 \fBpcre_study()\fP.
984 .P
985 The second argument of \fBpcre_study()\fP contains option bits. There are three
986 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
987 .sp
991 .sp
992 If any of these are set, and the just-in-time compiler is available, the
993 pattern is further compiled into machine code that executes much faster than
994 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
995 compiler is not available, these options are ignored. All undefined bits in the
996 \fIoptions\fP argument must be zero.
997 .P
998 JIT compilation is a heavyweight optimization. It can take some time for
999 patterns to be analyzed, and for one-off matches and simple patterns the
1000 benefit of faster execution might be offset by a much slower study time.
1001 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1002 handled, matching automatically falls back to the \fBpcre_exec()\fP
1003 interpreter. For more details, see the
1004 .\" HREF
1005 \fBpcrejit\fP
1006 .\"
1007 documentation.
1008 .P
1009 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1010 studying succeeds (even if no data is returned), the variable it points to is
1011 set to NULL. Otherwise it is set to point to a textual error message. This is a
1012 static string that is part of the library. You must not try to free it. You
1013 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1014 sure that it has run successfully.
1015 .P
1016 When you are finished with a pattern, you can free the memory used for the
1017 study data by calling \fBpcre_free_study()\fP. This function was added to the
1018 API for release 8.20. For earlier versions, the memory could be freed with
1019 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1020 where JIT optimization is not used, but it is advisable to change to the new
1021 function when convenient.
1022 .P
1023 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1024 real application there should be tests for errors):
1025 .sp
1026 int rc;
1027 pcre *re;
1028 pcre_extra *sd;
1029 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1030 sd = pcre_study(
1031 re, /* result of pcre_compile() */
1032 0, /* no options */
1033 &error); /* set to NULL or points to a message */
1034 rc = pcre_exec( /* see below for details of pcre_exec() options */
1035 re, sd, "subject", 7, 0, 0, ovector, 30);
1036 ...
1037 pcre_free_study(sd);
1038 pcre_free(re);
1039 .sp
1040 Studying a pattern does two things: first, a lower bound for the length of
1041 subject string that is needed to match the pattern is computed. This does not
1042 mean that there are any strings of that length that match, but it does
1043 guarantee that no shorter strings match. The value is used to avoid wasting
1044 time by trying to match strings that are shorter than the lower bound. You can
1045 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1046 .P
1047 Studying a pattern is also useful for non-anchored patterns that do not have a
1048 single fixed starting character. A bitmap of possible starting bytes is
1049 created. This speeds up finding a position in the subject at which to start
1050 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.)
1051 .P
1052 These two optimizations apply to both \fBpcre_exec()\fP and
1053 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1054 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1055 when calling \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP, but if this is done,
1056 JIT execution is also disabled. You might want to do this if your pattern
1057 contains callouts or (*MARK) and you want to make use of these facilities in
1058 cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
1059 .\" HTML <a href="#execoptions">
1060 .\" </a>
1061 below.
1062 .\"
1063 .
1064 .
1065 .\" HTML <a name="localesupport"></a>
1067 .rs
1068 .sp
1069 PCRE handles caseless matching, and determines whether characters are letters,
1070 digits, or whatever, by reference to a set of tables, indexed by character
1071 value. When running in UTF-8 mode, this applies only to characters
1072 with codes less than 128. By default, higher-valued codes never match escapes
1073 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1074 Unicode character property support. Alternatively, the PCRE_UCP option can be
1075 set at compile time; this causes \ew and friends to use Unicode property
1076 support instead of built-in tables. The use of locales with Unicode is
1077 discouraged. If you are handling characters with codes greater than 128, you
1078 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1079 two.
1080 .P
1081 PCRE contains an internal set of tables that are used when the final argument
1082 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1083 Normally, the internal tables recognize only ASCII characters. However, when
1084 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1085 default "C" locale of the local system, which may cause them to be different.
1086 .P
1087 The internal tables can always be overridden by tables supplied by the
1088 application that calls PCRE. These may be created in a different locale from
1089 the default. As more and more applications change to using Unicode, the need
1090 for this locale support is expected to die away.
1091 .P
1092 External tables are built by calling the \fBpcre_maketables()\fP function,
1093 which has no arguments, in the relevant locale. The result can then be passed
1094 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1095 example, to build and use tables that are appropriate for the French locale
1096 (where accented characters with values greater than 128 are treated as letters),
1097 the following code could be used:
1098 .sp
1099 setlocale(LC_CTYPE, "fr_FR");
1100 tables = pcre_maketables();
1101 re = pcre_compile(..., tables);
1102 .sp
1103 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1104 are using Windows, the name for the French locale is "french".
1105 .P
1106 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1107 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1108 that the memory containing the tables remains available for as long as it is
1109 needed.
1110 .P
1111 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1112 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1113 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1114 pattern, compilation, studying and matching all happen in the same locale, but
1115 different patterns can be compiled in different locales.
1116 .P
1117 It is possible to pass a table pointer or NULL (indicating the use of the
1118 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1119 this facility could be used to match a pattern in a different locale from the
1120 one in which it was compiled. Passing table pointers at run time is discussed
1121 below in the section on matching a pattern.
1122 .
1123 .
1124 .\" HTML <a name="infoaboutpattern"></a>
1126 .rs
1127 .sp
1128 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1129 .ti +5n
1130 .B int \fIwhat\fP, void *\fIwhere\fP);
1131 .PP
1132 The \fBpcre_fullinfo()\fP function returns information about a compiled
1133 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1134 library at version 8.30, after more than 10 years of obsolescence.
1135 .P
1136 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1137 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1138 the pattern was not studied. The third argument specifies which piece of
1139 information is required, and the fourth argument is a pointer to a variable
1140 to receive the data. The yield of the function is zero for success, or one of
1141 the following negative numbers:
1142 .sp
1143 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1144 the argument \fIwhere\fP was NULL
1145 PCRE_ERROR_BADMAGIC the "magic number" was not found
1146 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1147 endianness
1148 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1149 .sp
1150 The "magic number" is placed at the start of each compiled pattern as an simple
1151 check against passing an arbitrary memory pointer. The endianness error can
1152 occur if a compiled pattern is saved and reloaded on a different host. Here is
1153 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1154 pattern:
1155 .sp
1156 int rc;
1157 size_t length;
1158 rc = pcre_fullinfo(
1159 re, /* result of pcre_compile() */
1160 sd, /* result of pcre_study(), or NULL */
1161 PCRE_INFO_SIZE, /* what is required */
1162 &length); /* where to put the data */
1163 .sp
1164 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1165 as follows:
1166 .sp
1168 .sp
1169 Return the number of the highest back reference in the pattern. The fourth
1170 argument should point to an \fBint\fP variable. Zero is returned if there are
1171 no back references.
1172 .sp
1174 .sp
1175 Return the number of capturing subpatterns in the pattern. The fourth argument
1176 should point to an \fBint\fP variable.
1177 .sp
1179 .sp
1180 Return a pointer to the internal default character tables within PCRE. The
1181 fourth argument should point to an \fBunsigned char *\fP variable. This
1182 information call is provided for internal use by the \fBpcre_study()\fP
1183 function. External callers can cause PCRE to use its internal tables by passing
1184 a NULL table pointer.
1185 .sp
1187 .sp
1188 Return information about the first data unit of any matched string, for a
1189 non-anchored pattern. (The name of this option refers to the 8-bit library,
1190 where data units are bytes.) The fourth argument should point to an \fBint\fP
1191 variable.
1192 .P
1193 If there is a fixed first value, for example, the letter "c" from a pattern
1194 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1195 value is always less than 256; in the 16-bit library the value can be up to
1196 0xffff.
1197 .P
1198 If there is no fixed first value, and if either
1199 .sp
1200 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1201 starts with "^", or
1202 .sp
1203 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1204 (if it were set, the pattern would be anchored),
1205 .sp
1206 -1 is returned, indicating that the pattern matches only at the start of a
1207 subject string or after any newline within the string. Otherwise -2 is
1208 returned. For anchored patterns, -2 is returned.
1209 .sp
1211 .sp
1212 If the pattern was studied, and this resulted in the construction of a 256-bit
1213 table indicating a fixed set of values for the first data unit in any matching
1214 string, a pointer to the table is returned. Otherwise NULL is returned. The
1215 fourth argument should point to an \fBunsigned char *\fP variable.
1216 .sp
1218 .sp
1219 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1220 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1221 explicit match is either a literal CR or LF character, or \er or \en.
1222 .sp
1224 .sp
1225 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1226 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1227 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1228 .sp
1230 .sp
1231 Return 1 if the pattern was studied with one of the JIT options, and
1232 just-in-time compiling was successful. The fourth argument should point to an
1233 \fBint\fP variable. A return value of 0 means that JIT support is not available
1234 in this version of PCRE, or that the pattern was not studied with a JIT option,
1235 or that the JIT compiler could not handle this particular pattern. See the
1236 .\" HREF
1237 \fBpcrejit\fP
1238 .\"
1239 documentation for details of what can and cannot be handled.
1240 .sp
1242 .sp
1243 If the pattern was successfully studied with a JIT option, return the size of
1244 the JIT compiled code, otherwise return zero. The fourth argument should point
1245 to a \fBsize_t\fP variable.
1246 .sp
1248 .sp
1249 Return the value of the rightmost literal data unit that must exist in any
1250 matched string, other than at its start, if such a value has been recorded. The
1251 fourth argument should point to an \fBint\fP variable. If there is no such
1252 value, -1 is returned. For anchored patterns, a last literal value is recorded
1253 only if it follows something of variable length. For example, for the pattern
1254 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1255 is -1.
1256 .sp
1258 .sp
1259 Return the number of characters (NB not bytes) in the longest lookbehind
1260 assertion in the pattern. Note that the simple assertions \eb and \eB require a
1261 one-character lookbehind. This information is useful when doing multi-segment
1262 matching using the partial matching facilities.
1263 .sp
1265 .sp
1266 If the pattern was studied and a minimum length for matching subject strings
1267 was computed, its value is returned. Otherwise the returned value is -1. The
1268 value is a number of characters, which in UTF-8 mode may be different from the
1269 number of bytes. The fourth argument should point to an \fBint\fP variable. A
1270 non-negative value is a lower bound to the length of any matching string. There
1271 may not be any strings of that length that do actually match, but every string
1272 that does match is at least that long.
1273 .sp
1277 .sp
1278 PCRE supports the use of named as well as numbered capturing parentheses. The
1279 names are just an additional way of identifying the parentheses, which still
1280 acquire numbers. Several convenience functions such as
1281 \fBpcre_get_named_substring()\fP are provided for extracting captured
1282 substrings by name. It is also possible to extract the data directly, by first
1283 converting the name to a number in order to access the correct pointers in the
1284 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1285 you need to use the name-to-number map, which is described by these three
1286 values.
1287 .P
1288 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1289 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1290 entry; both of these return an \fBint\fP value. The entry size depends on the
1291 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1292 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1293 the first two bytes of each entry are the number of the capturing parenthesis,
1294 most significant byte first. In the 16-bit library, the pointer points to
1295 16-bit data units, the first of which contains the parenthesis number. The rest
1296 of the entry is the corresponding name, zero terminated.
1297 .P
1298 The names are in alphabetical order. Duplicate names may appear if (?| is used
1299 to create multiple groups with the same number, as described in the
1300 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1301 .\" </a>
1302 section on duplicate subpattern numbers
1303 .\"
1304 in the
1305 .\" HREF
1306 \fBpcrepattern\fP
1307 .\"
1308 page. Duplicate names for subpatterns with different numbers are permitted only
1309 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1310 table in the order in which they were found in the pattern. In the absence of
1311 (?| this is the order of increasing number; when (?| is used this is not
1312 necessarily the case because later subpatterns may have lower numbers.
1313 .P
1314 As a simple example of the name/number table, consider the following pattern
1315 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1316 space - including newlines - is ignored):
1317 .sp
1318 .\" JOIN
1319 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1320 (?<month>\ed\ed) - (?<day>\ed\ed) )
1321 .sp
1322 There are four named subpatterns, so the table has four entries, and each entry
1323 in the table is eight bytes long. The table is as follows, with non-printing
1324 bytes shows in hexadecimal, and undefined bytes shown as ??:
1325 .sp
1326 00 01 d a t e 00 ??
1327 00 05 d a y 00 ?? ??
1328 00 04 m o n t h 00
1329 00 02 y e a r 00 ??
1330 .sp
1331 When writing code to extract data from named subpatterns using the
1332 name-to-number map, remember that the length of the entries is likely to be
1333 different for each compiled pattern.
1334 .sp
1336 .sp
1337 Return 1 if the pattern can be used for partial matching with
1338 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1339 \fBint\fP variable. From release 8.00, this always returns 1, because the
1340 restrictions that previously applied to partial matching have been lifted. The
1341 .\" HREF
1342 \fBpcrepartial\fP
1343 .\"
1344 documentation gives details of partial matching.
1345 .sp
1347 .sp
1348 Return a copy of the options with which the pattern was compiled. The fourth
1349 argument should point to an \fBunsigned long int\fP variable. These option bits
1350 are those specified in the call to \fBpcre_compile()\fP, modified by any
1351 top-level option settings at the start of the pattern itself. In other words,
1352 they are the options that will be in force when matching starts. For example,
1353 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1355 .P
1356 A pattern is automatically anchored by PCRE if all of its top-level
1357 alternatives begin with one of the following:
1358 .sp
1359 ^ unless PCRE_MULTILINE is set
1360 \eA always
1361 \eG always
1362 .\" JOIN
1363 .* if PCRE_DOTALL is set and there are no back
1364 references to the subpattern in which .* appears
1365 .sp
1366 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1367 \fBpcre_fullinfo()\fP.
1368 .sp
1370 .sp
1371 Return the size of the compiled pattern in bytes (for both libraries). The
1372 fourth argument should point to a \fBsize_t\fP variable. This value does not
1373 include the size of the \fBpcre\fP structure that is returned by
1374 \fBpcre_compile()\fP. The value that is passed as the argument to
1375 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1376 place the compiled data is the value returned by this option plus the size of
1377 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1378 does not alter the value returned by this option.
1379 .sp
1381 .sp
1382 Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1383 field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1384 study data, zero is returned. The fourth argument should point to a
1385 \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1386 to record information that will speed up matching (see the section entitled
1387 .\" HTML <a href="#studyingapattern">
1388 .\" </a>
1389 "Studying a pattern"
1390 .\"
1391 above). The format of the \fIstudy_data\fP block is private, but its length
1392 is made available via this option so that it can be saved and restored (see the
1393 .\" HREF
1394 \fBpcreprecompile\fP
1395 .\"
1396 documentation for details).
1397 .
1398 .
1400 .rs
1401 .sp
1402 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1403 .PP
1404 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1405 data block that contains a compiled pattern. It is provided for the benefit of
1406 applications that operate in an object-oriented manner, where different parts
1407 of the application may be using the same compiled pattern, but you want to free
1408 the block when they are all done.
1409 .P
1410 When a pattern is compiled, the reference count field is initialized to zero.
1411 It is changed only by calling this function, whose action is to add the
1412 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1413 function is the new value. However, the value of the count is constrained to
1414 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1415 it is forced to the appropriate limit value.
1416 .P
1417 Except when it is zero, the reference count is not correctly preserved if a
1418 pattern is compiled on one host and then transferred to a host whose byte-order
1419 is different. (This seems a highly unlikely scenario.)
1420 .
1421 .
1423 .rs
1424 .sp
1425 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1426 .ti +5n
1427 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1428 .ti +5n
1429 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1430 .P
1431 The function \fBpcre_exec()\fP is called to match a subject string against a
1432 compiled pattern, which is passed in the \fIcode\fP argument. If the
1433 pattern was studied, the result of the study should be passed in the
1434 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1435 and \fIextra\fP arguments as many times as you like, in order to match
1436 different subject strings with the same pattern.
1437 .P
1438 This function is the main matching facility of the library, and it operates in
1439 a Perl-like manner. For specialist use there is also an alternative matching
1440 function, which is described
1441 .\" HTML <a href="#dfamatch">
1442 .\" </a>
1443 below
1444 .\"
1445 in the section about the \fBpcre_dfa_exec()\fP function.
1446 .P
1447 In most applications, the pattern will have been compiled (and optionally
1448 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1449 possible to save compiled patterns and study data, and then use them later
1450 in different processes, possibly even on different hosts. For a discussion
1451 about this, see the
1452 .\" HREF
1453 \fBpcreprecompile\fP
1454 .\"
1455 documentation.
1456 .P
1457 Here is an example of a simple call to \fBpcre_exec()\fP:
1458 .sp
1459 int rc;
1460 int ovector[30];
1461 rc = pcre_exec(
1462 re, /* result of pcre_compile() */
1463 NULL, /* we didn't study the pattern */
1464 "some string", /* the subject string */
1465 11, /* the length of the subject string */
1466 0, /* start at offset 0 in the subject */
1467 0, /* default options */
1468 ovector, /* vector of integers for substring information */
1469 30); /* number of elements (NOT size in bytes) */
1470 .
1471 .
1472 .\" HTML <a name="extradata"></a>
1473 .SS "Extra data for \fBpcre_exec()\fR"
1474 .rs
1475 .sp
1476 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1477 data block. The \fBpcre_study()\fP function returns such a block (when it
1478 doesn't return NULL), but you can also create one for yourself, and pass
1479 additional information in it. The \fBpcre_extra\fP block contains the following
1480 fields (not necessarily in this order):
1481 .sp
1482 unsigned long int \fIflags\fP;
1483 void *\fIstudy_data\fP;
1484 void *\fIexecutable_jit\fP;
1485 unsigned long int \fImatch_limit\fP;
1486 unsigned long int \fImatch_limit_recursion\fP;
1487 void *\fIcallout_data\fP;
1488 const unsigned char *\fItables\fP;
1489 unsigned char **\fImark\fP;
1490 .sp
1491 In the 16-bit version of this structure, the \fImark\fP field has type
1492 "PCRE_UCHAR16 **".
1493 .P
1494 The \fIflags\fP field is used to specify which of the other fields are set. The
1495 flag bits are:
1496 .sp
1504 .sp
1505 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1506 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1507 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1508 should not set these yourself, but you may add to the block by setting other
1509 fields and their corresponding flag bits.
1510 .P
1511 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1512 vast amount of resources when running patterns that are not going to match,
1513 but which have a very large number of possibilities in their search trees. The
1514 classic example is a pattern that uses nested unlimited repeats.
1515 .P
1516 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1517 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1518 imposed on the number of times this function is called during a match, which
1519 has the effect of limiting the amount of backtracking that can take place. For
1520 patterns that are not anchored, the count restarts from zero for each position
1521 in the subject string.
1522 .P
1523 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1524 with a JIT option, the way that the matching is executed is entirely different.
1525 However, there is still the possibility of runaway matching that goes on for a
1526 very long time, and so the \fImatch_limit\fP value is also used in this case
1527 (but in a different way) to limit how long the matching can continue.
1528 .P
1529 The default value for the limit can be set when PCRE is built; the default
1530 default is 10 million, which handles all but the most extreme cases. You can
1531 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1532 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1533 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1535 .P
1536 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1537 instead of limiting the total number of times that \fBmatch()\fP is called, it
1538 limits the depth of recursion. The recursion depth is a smaller number than the
1539 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1540 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1541 .P
1542 Limiting the recursion depth limits the amount of machine stack that can be
1543 used, or, when PCRE has been compiled to use memory on the heap instead of the
1544 stack, the amount of heap memory that can be used. This limit is not relevant,
1545 and is ignored, when matching is done using JIT compiled code.
1546 .P
1547 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1548 built; the default default is the same value as the default for
1549 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1550 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1551 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1552 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1553 .P
1554 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1555 and is described in the
1556 .\" HREF
1557 \fBpcrecallout\fP
1558 .\"
1559 documentation.
1560 .P
1561 The \fItables\fP field is used to pass a character tables pointer to
1562 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1563 pattern. A non-NULL value is stored with the compiled pattern only if custom
1564 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1565 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1566 internal tables to be used. This facility is helpful when re-using patterns
1567 that have been saved after compiling with an external set of tables, because
1568 the external tables might be at a different address when \fBpcre_exec()\fP is
1569 called. See the
1570 .\" HREF
1571 \fBpcreprecompile\fP
1572 .\"
1573 documentation for a discussion of saving compiled patterns for later use.
1574 .P
1575 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1576 be set to point to a suitable variable. If the pattern contains any
1577 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1578 a name to pass back, a pointer to the name string (zero terminated) is placed
1579 in the variable pointed to by the \fImark\fP field. The names are within the
1580 compiled pattern; if you wish to retain such a name you must copy it before
1581 freeing the memory of a compiled pattern. If there is no name to pass back, the
1582 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1583 backtracking control verbs, see the section entitled
1584 .\" HTML <a href="pcrepattern#backtrackcontrol">
1585 .\" </a>
1586 "Backtracking control"
1587 .\"
1588 in the
1589 .\" HREF
1590 \fBpcrepattern\fP
1591 .\"
1592 documentation.
1593 .
1594 .
1595 .\" HTML <a name="execoptions"></a>
1596 .SS "Option bits for \fBpcre_exec()\fP"
1597 .rs
1598 .sp
1599 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1600 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1604 .P
1605 If the pattern was successfully studied with one of the just-in-time (JIT)
1606 compile options, the only supported options for JIT execution are
1609 unsupported option is used, JIT execution is disabled and the normal
1610 interpretive code in \fBpcre_exec()\fP is run.
1611 .sp
1613 .sp
1614 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1615 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1616 to be anchored by virtue of its contents, it cannot be made unachored at
1617 matching time.
1618 .sp
1621 .sp
1622 These options (which are mutually exclusive) control what the \eR escape
1623 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1624 match any Unicode newline sequence. These options override the choice that was
1625 made or defaulted when the pattern was compiled.
1626 .sp
1632 .sp
1633 These options override the newline definition that was chosen or defaulted when
1634 the pattern was compiled. For details, see the description of
1635 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1636 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1637 the way the match position is advanced after a match failure for an unanchored
1638 pattern.
1639 .P
1641 match attempt for an unanchored pattern fails when the current position is at a
1642 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1643 characters, the match position is advanced by two characters instead of one, in
1644 other words, to after the CRLF.
1645 .P
1646 The above rule is a compromise that makes the most common cases work as
1647 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1648 set), it does not match the string "\er\enA" because, after failing at the
1649 start, it skips both the CR and the LF before retrying. However, the pattern
1650 [\er\en]A does match that string, because it contains an explicit CR or LF
1651 reference, and so advances only by one character after the first failure.
1652 .P
1653 An explicit match for CR of LF is either a literal appearance of one of those
1654 characters, or one of the \er or \en escape sequences. Implicit matches such as
1655 [^X] do not count, nor does \es (which includes CR and LF in the characters
1656 that it matches).
1657 .P
1658 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1659 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1660 .sp
1662 .sp
1663 This option specifies that first character of the subject string is not the
1664 beginning of a line, so the circumflex metacharacter should not match before
1665 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1666 never to match. This option affects only the behaviour of the circumflex
1667 metacharacter. It does not affect \eA.
1668 .sp
1670 .sp
1671 This option specifies that the end of the subject string is not the end of a
1672 line, so the dollar metacharacter should not match it nor (except in multiline
1673 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1674 compile time) causes dollar never to match. This option affects only the
1675 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1676 .sp
1678 .sp
1679 An empty string is not considered to be a valid match if this option is set. If
1680 there are alternatives in the pattern, they are tried. If all the alternatives
1681 match the empty string, the entire match fails. For example, if the pattern
1682 .sp
1683 a?b?
1684 .sp
1685 is applied to a string not beginning with "a" or "b", it matches an empty
1686 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1687 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1688 .sp
1690 .sp
1691 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1692 the start of the subject is permitted. If the pattern is anchored, such a match
1693 can occur only if the pattern contains \eK.
1694 .P
1695 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1696 does make a special case of a pattern match of the empty string within its
1697 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1698 emulate Perl's behaviour after matching a null string by first trying the match
1699 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1700 if that fails, by advancing the starting offset (see below) and trying an
1701 ordinary match again. There is some code that demonstrates how to do this in
1702 the
1703 .\" HREF
1704 \fBpcredemo\fP
1705 .\"
1706 sample program. In the most general case, you have to check to see if the
1707 newline convention recognizes CRLF as a newline, and if so, and the current
1708 character is CR followed by LF, advance the starting offset by two characters
1709 instead of one.
1710 .sp
1712 .sp
1713 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1714 a match, in order to speed up the process. For example, if it is known that an
1715 unanchored match must start with a specific character, it searches the subject
1716 for that character, and fails immediately if it cannot find it, without
1717 actually running the main matching function. This means that a special item
1718 such as (*COMMIT) at the start of a pattern is not considered until after a
1719 suitable starting point for the match has been found. When callouts or (*MARK)
1720 items are in use, these "start-up" optimizations can cause them to be skipped
1721 if the pattern is never actually used. The start-up optimizations are in effect
1722 a pre-scan of the subject that takes place before the pattern is run.
1723 .P
1724 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1725 causing performance to suffer, but ensuring that in cases where the result is
1726 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1727 are considered at every possible starting position in the subject string. If
1728 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1729 time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1730 matching is always done using interpretively.
1731 .P
1732 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1733 Consider the pattern
1734 .sp
1736 .sp
1737 When this is compiled, PCRE records the fact that a match must start with the
1738 character "A". Suppose the subject string is "DEFABC". The start-up
1739 optimization scans along the subject, finds "A" and runs the first match
1740 attempt from there. The (*COMMIT) item means that the pattern must match the
1741 current starting position, which in this case, it does. However, if the same
1742 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1743 subject string does not happen. The first match attempt is run starting from
1744 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1745 the overall result is "no match". If the pattern is studied, more start-up
1746 optimizations may be used. For example, a minimum length for the subject may be
1747 recorded. Consider the pattern
1748 .sp
1749 (*MARK:A)(X|Y)
1750 .sp
1751 The minimum length for a match is one character. If the subject is "ABC", there
1752 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1753 If the pattern is studied, the final attempt does not take place, because PCRE
1754 knows that the subject is too short, and so the (*MARK) is never encountered.
1755 In this case, studying the pattern does not affect the overall match result,
1756 which is still "no match", but it does affect the auxiliary information that is
1757 returned.
1758 .sp
1760 .sp
1761 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1762 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1763 The entire string is checked before any other processing takes place. The value
1764 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1765 UTF-8 character. There is a discussion about the
1766 .\" HTML <a href="pcreunicode.html#utf8strings">
1767 .\" </a>
1768 validity of UTF-8 strings
1769 .\"
1770 in the
1771 .\" HREF
1772 \fBpcreunicode\fP
1773 .\"
1774 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1775 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1776 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1777 cases, information about the precise nature of the error may also be returned
1778 (see the descriptions of these errors in the section entitled \fIError return
1779 values from\fP \fBpcre_exec()\fP
1780 .\" HTML <a href="#errorlist">
1781 .\" </a>
1782 below).
1783 .\"
1784 If \fIstartoffset\fP contains a value that does not point to the start of a
1785 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1786 returned.
1787 .P
1788 If you already know that your subject is valid, and you want to skip these
1789 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1790 calling \fBpcre_exec()\fP. You might want to do this for the second and
1791 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1792 all the matches in a single subject string. However, you should be sure that
1793 the value of \fIstartoffset\fP points to the start of a character (or the end
1794 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1795 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1796 undefined. Your program may crash.
1797 .sp
1800 .sp
1801 These options turn on the partial matching feature. For backwards
1802 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1803 occurs if the end of the subject string is reached successfully, but there are
1804 not enough subject characters to complete the match. If this happens when
1805 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1806 testing any remaining alternatives. Only if no complete match can be found is
1807 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1808 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1809 but only if no complete match can be found.
1810 .P
1811 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1812 partial match is found, \fBpcre_exec()\fP immediately returns
1813 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1814 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1815 important that an alternative complete match.
1816 .P
1817 In both cases, the portion of the string that was inspected when the partial
1818 match was found is set as the first matching string. There is a more detailed
1819 discussion of partial and multi-segment matching, with examples, in the
1820 .\" HREF
1821 \fBpcrepartial\fP
1822 .\"
1823 documentation.
1824 .
1825 .
1826 .SS "The string to be matched by \fBpcre_exec()\fP"
1827 .rs
1828 .sp
1829 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1830 \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1831 in \fIstartoffset\fP. If this is negative or greater than the length of the
1832 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1833 offset is zero, the search for a match starts at the beginning of the subject,
1834 and this is by far the most common case. In UTF-8 mode, the byte offset must
1835 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1836 pattern string, the subject may contain binary zero bytes.
1837 .P
1838 A non-zero starting offset is useful when searching for another match in the
1839 same subject by calling \fBpcre_exec()\fP again after a previous success.
1840 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1841 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1842 lookbehind. For example, consider the pattern
1843 .sp
1844 \eBiss\eB
1845 .sp
1846 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1847 the current position in the subject is not a word boundary.) When applied to
1848 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1849 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1850 subject, namely "issipi", it does not match, because \eB is always false at the
1851 start of the subject, which is deemed to be a word boundary. However, if
1852 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1853 set to 4, it finds the second occurrence of "iss" because it is able to look
1854 behind the starting point to discover that it is preceded by a letter.
1855 .P
1856 Finding all the matches in a subject is tricky when the pattern can match an
1857 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1858 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1859 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1860 and trying an ordinary match again. There is some code that demonstrates how to
1861 do this in the
1862 .\" HREF
1863 \fBpcredemo\fP
1864 .\"
1865 sample program. In the most general case, you have to check to see if the
1866 newline convention recognizes CRLF as a newline, and if so, and the current
1867 character is CR followed by LF, advance the starting offset by two characters
1868 instead of one.
1869 .P
1870 If a non-zero starting offset is passed when the pattern is anchored, one
1871 attempt to match at the given offset is made. This can only succeed if the
1872 pattern does not require the match to be at the start of the subject.
1873 .
1874 .
1875 .SS "How \fBpcre_exec()\fP returns captured substrings"
1876 .rs
1877 .sp
1878 In general, a pattern matches a certain portion of the subject, and in
1879 addition, further substrings from the subject may be picked out by parts of the
1880 pattern. Following the usage in Jeffrey Friedl's book, this is called
1881 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1882 a fragment of a pattern that picks out a substring. PCRE supports several other
1883 kinds of parenthesized subpattern that do not cause substrings to be captured.
1884 .P
1885 Captured substrings are returned to the caller via a vector of integers whose
1886 address is passed in \fIovector\fP. The number of elements in the vector is
1887 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
1888 argument is NOT the size of \fIovector\fP in bytes.
1889 .P
1890 The first two-thirds of the vector is used to pass back captured substrings,
1891 each substring using a pair of integers. The remaining third of the vector is
1892 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
1893 and is not available for passing back information. The number passed in
1894 \fIovecsize\fP should always be a multiple of three. If it is not, it is
1895 rounded down.
1896 .P
1897 When a match is successful, information about captured substrings is returned
1898 in pairs of integers, starting at the beginning of \fIovector\fP, and
1899 continuing up to two-thirds of its length at the most. The first element of
1900 each pair is set to the byte offset of the first character in a substring, and
1901 the second is set to the byte offset of the first character after the end of a
1902 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
1903 mode. They are not character counts.
1904 .P
1905 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
1906 portion of the subject string matched by the entire pattern. The next pair is
1907 used for the first capturing subpattern, and so on. The value returned by
1908 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
1909 For example, if two substrings have been captured, the returned value is 3. If
1910 there are no capturing subpatterns, the return value from a successful match is
1911 1, indicating that just the first pair of offsets has been set.
1912 .P
1913 If a capturing subpattern is matched repeatedly, it is the last portion of the
1914 string that it matched that is returned.
1915 .P
1916 If the vector is too small to hold all the captured substring offsets, it is
1917 used as far as possible (up to two-thirds of its length), and the function
1918 returns a value of zero. If neither the actual string matched nor any captured
1919 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
1920 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
1921 back references and the \fIovector\fP is not big enough to remember the related
1922 substrings, PCRE has to get additional memory for use during matching. Thus it
1923 is usually advisable to supply an \fIovector\fP of reasonable size.
1924 .P
1925 There are some cases where zero is returned (indicating vector overflow) when
1926 in fact the vector is exactly the right size for the final match. For example,
1927 consider the pattern
1928 .sp
1929 (a)(?:(b)c|bd)
1930 .sp
1931 If a vector of 6 elements (allowing for only 1 captured substring) is given
1932 with subject string "abd", \fBpcre_exec()\fP will try to set the second
1933 captured string, thereby recording a vector overflow, before failing to match
1934 "c" and backing up to try the second alternative. The zero return, however,
1935 does correctly indicate that the maximum number of slots (namely 2) have been
1936 filled. In similar cases where there is temporary overflow, but the final
1937 number of used slots is actually less than the maximum, a non-zero value is
1938 returned.
1939 .P
1940 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
1941 subpatterns there are in a compiled pattern. The smallest size for
1942 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
1943 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
1944 .P
1945 It is possible for capturing subpattern number \fIn+1\fP to match some part of
1946 the subject when subpattern \fIn\fP has not been used at all. For example, if
1947 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
1948 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
1949 happens, both values in the offset pairs corresponding to unused subpatterns
1950 are set to -1.
1951 .P
1952 Offset values that correspond to unused subpatterns at the end of the
1953 expression are also set to -1. For example, if the string "abc" is matched
1954 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
1955 return from the function is 2, because the highest used capturing subpattern
1956 number is 1, and the offsets for for the second and third capturing subpatterns
1957 (assuming the vector is large enough, of course) are set to -1.
1958 .P
1959 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
1960 correspond to capturing parentheses in the pattern are never changed. That is,
1961 if a pattern contains \fIn\fP capturing parentheses, no more than
1962 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
1963 elements (in the first two-thirds) retain whatever values they previously had.
1964 .P
1965 Some convenience functions are provided for extracting the captured substrings
1966 as separate strings. These are described below.
1967 .
1968 .
1969 .\" HTML <a name="errorlist"></a>
1970 .SS "Error return values from \fBpcre_exec()\fP"
1971 .rs
1972 .sp
1973 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
1974 defined in the header file:
1975 .sp
1977 .sp
1978 The subject string did not match the pattern.
1979 .sp
1981 .sp
1982 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
1983 NULL and \fIovecsize\fP was not zero.
1984 .sp
1986 .sp
1987 An unrecognized bit was set in the \fIoptions\fP argument.
1988 .sp
1990 .sp
1991 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
1992 the case when it is passed a junk pointer and to detect when a pattern that was
1993 compiled in an environment of one endianness is run in an environment with the
1994 other endianness. This is the error that PCRE gives when the magic number is
1995 not present.
1996 .sp
1998 .sp
1999 While running the pattern match, an unknown item was encountered in the
2000 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2001 of the compiled pattern.
2002 .sp
2004 .sp
2005 If a pattern contains back references, but the \fIovector\fP that is passed to
2006 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2007 gets a block of memory at the start of matching to use for this purpose. If the
2008 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2009 automatically freed at the end of matching.
2010 .P
2011 This error is also given if \fBpcre_stack_malloc()\fP fails in
2012 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2013 \fB--disable-stack-for-recursion\fP.
2014 .sp
2016 .sp
2017 This error is used by the \fBpcre_copy_substring()\fP,
2018 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2019 below). It is never returned by \fBpcre_exec()\fP.
2020 .sp
2022 .sp
2023 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2024 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2025 above.
2026 .sp
2028 .sp
2029 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2030 use by callout functions that want to yield a distinctive error code. See the
2031 .\" HREF
2032 \fBpcrecallout\fP
2033 .\"
2034 documentation for details.
2035 .sp
2037 .sp
2038 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2039 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2040 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2041 UTF-8 character is placed in the first element, and a reason code is placed in
2042 the second element. The reason codes are listed in the
2043 .\" HTML <a href="#badutf8reasons">
2044 .\" </a>
2045 following section.
2046 .\"
2047 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2048 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2049 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2050 .sp
2052 .sp
2053 The UTF-8 byte sequence that was passed as a subject was checked and found to
2054 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2055 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2056 end of the subject.
2057 .sp
2059 .sp
2060 The subject string did not match, but it did match partially. See the
2061 .\" HREF
2062 \fBpcrepartial\fP
2063 .\"
2064 documentation for details of partial matching.
2065 .sp
2067 .sp
2068 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2069 option was used with a compiled pattern containing items that were not
2070 supported for partial matching. From release 8.00 onwards, there are no
2071 restrictions on partial matching.
2072 .sp
2074 .sp
2075 An unexpected internal error has occurred. This error could be caused by a bug
2076 in PCRE or by overwriting of the compiled pattern.
2077 .sp
2079 .sp
2080 This error is given if the value of the \fIovecsize\fP argument is negative.
2081 .sp
2083 .sp
2084 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2085 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2086 description above.
2087 .sp
2089 .sp
2090 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2091 .sp
2093 .sp
2094 The value of \fIstartoffset\fP was negative or greater than the length of the
2095 subject, that is, the value in \fIlength\fP.
2096 .sp
2098 .sp
2099 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2100 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2101 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2102 fact sufficient to detect this case, but this special error code for
2103 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2104 retained for backwards compatibility.
2105 .sp
2107 .sp
2108 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2109 the pattern. Specifically, it means that either the whole pattern or a
2110 subpattern has been called recursively for the second time at the same position
2111 in the subject string. Some simple patterns that might do this are detected and
2112 faulted at compile time, but more complicated cases, in particular mutual
2113 recursions between two different subpatterns, cannot be detected until run
2114 time.
2115 .sp
2117 .sp
2118 This error is returned when a pattern that was successfully studied using a
2119 JIT compile option is being matched, but the memory available for the
2120 just-in-time processing stack is not large enough. See the
2121 .\" HREF
2122 \fBpcrejit\fP
2123 .\"
2124 documentation for more details.
2125 .sp
2127 .sp
2128 This error is given if a pattern that was compiled by the 8-bit library is
2129 passed to a 16-bit library function, or vice versa.
2130 .sp
2132 .sp
2133 This error is given if a pattern that was compiled and saved is reloaded on a
2134 host with different endianness. The utility function
2135 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2136 so that it runs on the new host.
2137 .P
2138 Error numbers -16 to -20, -22, and -30 are not used by \fBpcre_exec()\fP.
2139 .
2140 .
2141 .\" HTML <a name="badutf8reasons"></a>
2142 .SS "Reason codes for invalid UTF-8 strings"
2143 .rs
2144 .sp
2145 This section applies only to the 8-bit library. The corresponding information
2146 for the 16-bit library is given in the
2147 .\" HREF
2148 \fBpcre16\fP
2149 .\"
2150 page.
2151 .P
2152 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2153 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2154 least 2, the offset of the start of the invalid UTF-8 character is placed in
2155 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2156 in the second element (\fIovector[1]\fP). The reason codes are given names in
2157 the \fBpcre.h\fP header file:
2158 .sp
2164 .sp
2165 The string ends with a truncated UTF-8 character; the code specifies how many
2166 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2167 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2168 allows for up to 6 bytes, and this is checked first; hence the possibility of
2169 4 or 5 missing bytes.
2170 .sp
2175 PCRE_UTF8_ERR10
2176 .sp
2177 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2178 character do not have the binary value 0b10 (that is, either the most
2179 significant bit is 0, or the next bit is 1).
2180 .sp
2181 PCRE_UTF8_ERR11
2182 PCRE_UTF8_ERR12
2183 .sp
2184 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2185 these code points are excluded by RFC 3629.
2186 .sp
2187 PCRE_UTF8_ERR13
2188 .sp
2189 A 4-byte character has a value greater than 0x10fff; these code points are
2190 excluded by RFC 3629.
2191 .sp
2192 PCRE_UTF8_ERR14
2193 .sp
2194 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2195 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2196 from UTF-8.
2197 .sp
2198 PCRE_UTF8_ERR15
2199 PCRE_UTF8_ERR16
2200 PCRE_UTF8_ERR17
2201 PCRE_UTF8_ERR18
2202 PCRE_UTF8_ERR19
2203 .sp
2204 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2205 value that can be represented by fewer bytes, which is invalid. For example,
2206 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2207 one byte.
2208 .sp
2209 PCRE_UTF8_ERR20
2210 .sp
2211 The two most significant bits of the first byte of a character have the binary
2212 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2213 byte can only validly occur as the second or subsequent byte of a multi-byte
2214 character.
2215 .sp
2216 PCRE_UTF8_ERR21
2217 .sp
2218 The first byte of a character has the value 0xfe or 0xff. These values can
2219 never occur in a valid UTF-8 string.
2220 .
2221 .
2223 .rs
2224 .sp
2225 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2226 .ti +5n
2227 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2228 .ti +5n
2229 .B int \fIbuffersize\fP);
2230 .PP
2231 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2232 .ti +5n
2233 .B int \fIstringcount\fP, int \fIstringnumber\fP,
2234 .ti +5n
2235 .B const char **\fIstringptr\fP);
2236 .PP
2237 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2238 .ti +5n
2239 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2240 .PP
2241 Captured substrings can be accessed directly by using the offsets returned by
2242 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2243 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2244 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2245 as new, separate, zero-terminated strings. These functions identify substrings
2246 by number. The next section describes functions for extracting named
2247 substrings.
2248 .P
2249 A substring that contains a binary zero is correctly extracted and has a
2250 further zero added on the end, but the result is not, of course, a C string.
2251 However, you can process such a string by referring to the length that is
2252 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2253 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2254 for handling strings containing binary zeros, because the end of the final
2255 string is not independently indicated.
2256 .P
2257 The first three arguments are the same for all three of these functions:
2258 \fIsubject\fP is the subject string that has just been successfully matched,
2259 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2260 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2261 captured by the match, including the substring that matched the entire regular
2262 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2263 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2264 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2265 number of elements in the vector divided by three.
2266 .P
2267 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2268 extract a single substring, whose number is given as \fIstringnumber\fP. A
2269 value of zero extracts the substring that matched the entire pattern, whereas
2270 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2271 the string is placed in \fIbuffer\fP, whose length is given by
2272 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2273 obtained via \fBpcre_malloc\fP, and its address is returned via
2274 \fIstringptr\fP. The yield of the function is the length of the string, not
2275 including the terminating zero, or one of these error codes:
2276 .sp
2278 .sp
2279 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2280 memory failed for \fBpcre_get_substring()\fP.
2281 .sp
2283 .sp
2284 There is no substring whose number is \fIstringnumber\fP.
2285 .P
2286 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2287 and builds a list of pointers to them. All this is done in a single block of
2288 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2289 is returned via \fIlistptr\fP, which is also the start of the list of string
2290 pointers. The end of the list is marked by a NULL pointer. The yield of the
2291 function is zero if all went well, or the error code
2292 .sp
2294 .sp
2295 if the attempt to get the memory block failed.
2296 .P
2297 When any of these functions encounter a substring that is unset, which can
2298 happen when capturing subpattern number \fIn+1\fP matches some part of the
2299 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2300 string. This can be distinguished from a genuine zero-length substring by
2301 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2302 substrings.
2303 .P
2304 The two convenience functions \fBpcre_free_substring()\fP and
2305 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2306 a previous call of \fBpcre_get_substring()\fP or
2307 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2308 the function pointed to by \fBpcre_free\fP, which of course could be called
2309 directly from a C program. However, PCRE is used in some situations where it is
2310 linked via a special interface to another programming language that cannot use
2311 \fBpcre_free\fP directly; it is for these cases that the functions are
2312 provided.
2313 .
2314 .
2316 .rs
2317 .sp
2318 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2319 .ti +5n
2320 .B const char *\fIname\fP);
2321 .PP
2322 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2323 .ti +5n
2324 .B const char *\fIsubject\fP, int *\fIovector\fP,
2325 .ti +5n
2326 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2327 .ti +5n
2328 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2329 .PP
2330 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2331 .ti +5n
2332 .B const char *\fIsubject\fP, int *\fIovector\fP,
2333 .ti +5n
2334 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2335 .ti +5n
2336 .B const char **\fIstringptr\fP);
2337 .PP
2338 To extract a substring by name, you first have to find associated number.
2339 For example, for this pattern
2340 .sp
2341 (a+)b(?<xxx>\ed+)...
2342 .sp
2343 the number of the subpattern called "xxx" is 2. If the name is known to be
2344 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2345 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2346 pattern, and the second is the name. The yield of the function is the
2347 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2348 that name.
2349 .P
2350 Given the number, you can extract the substring directly, or use one of the
2351 functions described in the previous section. For convenience, there are also
2352 two functions that do the whole job.
2353 .P
2354 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2355 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2356 functions that extract by number. As these are described in the previous
2357 section, they are not re-described here. There are just two differences:
2358 .P
2359 First, instead of a substring number, a substring name is given. Second, there
2360 is an extra argument, given at the start, which is a pointer to the compiled
2361 pattern. This is needed in order to gain access to the name-to-number
2362 translation table.
2363 .P
2364 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2365 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2366 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2367 the behaviour may not be what you want (see the next section).
2368 .P
2369 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2370 subpatterns with the same number, as described in the
2371 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2372 .\" </a>
2373 section on duplicate subpattern numbers
2374 .\"
2375 in the
2376 .\" HREF
2377 \fBpcrepattern\fP
2378 .\"
2379 page, you cannot use names to distinguish the different subpatterns, because
2380 names are not included in the compiled code. The matching process uses only
2381 numbers. For this reason, the use of different names for subpatterns of the
2382 same number causes an error at compile time.
2383 .
2384 .
2386 .rs
2387 .sp
2388 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2389 .ti +5n
2390 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2391 .PP
2392 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2393 are not required to be unique. (Duplicate names are always allowed for
2394 subpatterns with the same number, created by using the (?| feature. Indeed, if
2395 such subpatterns are named, they are required to use the same names.)
2396 .P
2397 Normally, patterns with duplicate names are such that in any one match, only
2398 one of the named subpatterns participates. An example is shown in the
2399 .\" HREF
2400 \fBpcrepattern\fP
2401 .\"
2402 documentation.
2403 .P
2404 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2405 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2406 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2407 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2408 returns one of the numbers that are associated with the name, but it is not
2409 defined which it is.
2410 .P
2411 If you want to get full details of all captured substrings for a given name,
2412 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2413 argument is the compiled pattern, and the second is the name. The third and
2414 fourth are pointers to variables which are updated by the function. After it
2415 has run, they point to the first and last entries in the name-to-number table
2416 for the given name. The function itself returns the length of each entry, or
2417 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2418 described above in the section entitled \fIInformation about a pattern\fP
2419 .\" HTML <a href="#infoaboutpattern">
2420 .\" </a>
2421 above.
2422 .\"
2423 Given all the relevant entries for the name, you can extract each of their
2424 numbers, and hence the captured data, if any.
2425 .
2426 .
2428 .rs
2429 .sp
2430 The traditional matching function uses a similar algorithm to Perl, which stops
2431 when it finds the first match, starting at a given point in the subject. If you
2432 want to find all possible matches, or the longest possible match, consider
2433 using the alternative matching function (see below) instead. If you cannot use
2434 the alternative function, but still need to find all possible matches, you
2435 can kludge it up by making use of the callout facility, which is described in
2436 the
2437 .\" HREF
2438 \fBpcrecallout\fP
2439 .\"
2440 documentation.
2441 .P
2442 What you have to do is to insert a callout right at the end of the pattern.
2443 When your callout function is called, extract and save the current matched
2444 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2445 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2446 will yield PCRE_ERROR_NOMATCH.
2447 .
2448 .
2450 .rs
2451 .sp
2452 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2453 stack, which in certain environments can be rather limited in size. Some users
2454 find it helpful to have an estimate of the amount of stack that is used by
2455 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2456 .\" HREF
2457 \fBpcrestack\fP
2458 .\"
2459 documentation. The estimate that is output by \fBpcretest\fP when called with
2460 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2461 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2462 .P
2463 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2464 the negative error code PCRE_ERROR_NULL, but with this special combination of
2465 arguments, it returns instead a negative number whose absolute value is the
2466 approximate stack frame size in bytes. (A negative number is used so that it is
2467 clear that no match has happened.) The value is approximate because in some
2468 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2469 additional variables on the stack.
2470 .P
2471 If PCRE has been compiled to use the heap instead of the stack for recursion,
2472 the value returned is the size of each block that is obtained from the heap.
2473 .
2474 .
2475 .\" HTML <a name="dfamatch"></a>
2477 .rs
2478 .sp
2479 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2480 .ti +5n
2481 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2482 .ti +5n
2483 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2484 .ti +5n
2485 .B int *\fIworkspace\fP, int \fIwscount\fP);
2486 .P
2487 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2488 a compiled pattern, using a matching algorithm that scans the subject string
2489 just once, and does not backtrack. This has different characteristics to the
2490 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2491 patterns are not supported. Nevertheless, there are times when this kind of
2492 matching can be useful. For a discussion of the two matching algorithms, and a
2493 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2494 .\" HREF
2495 \fBpcrematching\fP
2496 .\"
2497 documentation.
2498 .P
2499 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2500 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2501 different way, and this is described below. The other common arguments are used
2502 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2503 here.
2504 .P
2505 The two additional arguments provide workspace for the function. The workspace
2506 vector should contain at least 20 elements. It is used for keeping track of
2507 multiple paths through the pattern tree. More workspace will be needed for
2508 patterns and subjects where there are a lot of potential matches.
2509 .P
2510 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2511 .sp
2512 int rc;
2513 int ovector[10];
2514 int wspace[20];
2515 rc = pcre_dfa_exec(
2516 re, /* result of pcre_compile() */
2517 NULL, /* we didn't study the pattern */
2518 "some string", /* the subject string */
2519 11, /* the length of the subject string */
2520 0, /* start at offset 0 in the subject */
2521 0, /* default options */
2522 ovector, /* vector of integers for substring information */
2523 10, /* number of elements (NOT size in bytes) */
2524 wspace, /* working space vector */
2525 20); /* number of elements (NOT size in bytes) */
2526 .
2527 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2528 .rs
2529 .sp
2530 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2531 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2535 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2536 so their description is not repeated here.
2537 .sp
2540 .sp
2541 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2542 details are slightly different. When PCRE_PARTIAL_HARD is set for
2543 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2544 is reached and there is still at least one matching possibility that requires
2545 additional characters. This happens even if some complete matches have also
2546 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2547 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2548 there have been no complete matches, but there is still at least one matching
2549 possibility. The portion of the string that was inspected when the longest
2550 partial match was found is set as the first matching string in both cases.
2551 There is a more detailed discussion of partial and multi-segment matching, with
2552 examples, in the
2553 .\" HREF
2554 \fBpcrepartial\fP
2555 .\"
2556 documentation.
2557 .sp
2559 .sp
2560 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2561 soon as it has found one match. Because of the way the alternative algorithm
2562 works, this is necessarily the shortest possible match at the first possible
2563 matching point in the subject string.
2564 .sp
2566 .sp
2567 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2568 again, with additional subject characters, and have it continue with the same
2569 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2570 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2571 before because data about the match so far is left in them after a partial
2572 match. There is more discussion of this facility in the
2573 .\" HREF
2574 \fBpcrepartial\fP
2575 .\"
2576 documentation.
2577 .
2578 .
2579 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2580 .rs
2581 .sp
2582 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2583 substring in the subject. Note, however, that all the matches from one run of
2584 the function start at the same point in the subject. The shorter matches are
2585 all initial substrings of the longer matches. For example, if the pattern
2586 .sp
2587 <.*>
2588 .sp
2589 is matched against the string
2590 .sp
2591 This is <something> <something else> <something further> no more
2592 .sp
2593 the three matched strings are
2594 .sp
2595 <something>
2596 <something> <something else>
2597 <something> <something else> <something further>
2598 .sp
2599 On success, the yield of the function is a number greater than zero, which is
2600 the number of matched substrings. The substrings themselves are returned in
2601 \fIovector\fP. Each string uses two elements; the first is the offset to the
2602 start, and the second is the offset to the end. In fact, all the strings have
2603 the same start offset. (Space could have been saved by giving this only once,
2604 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2605 returns data, even though the meaning of the strings is different.)
2606 .P
2607 The strings are returned in reverse order of length; that is, the longest
2608 matching string is given first. If there were too many matches to fit into
2609 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2610 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2611 the entire \fIovector\fP for returning matched strings.
2612 .
2613 .
2614 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2615 .rs
2616 .sp
2617 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2618 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2619 described
2620 .\" HTML <a href="#errorlist">
2621 .\" </a>
2622 above.
2623 .\"
2624 There are in addition the following errors that are specific to
2625 \fBpcre_dfa_exec()\fP:
2626 .sp
2628 .sp
2629 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2630 that it does not support, for instance, the use of \eC or a back reference.
2631 .sp
2633 .sp
2634 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2635 uses a back reference for the condition, or a test for recursion in a specific
2636 group. These are not supported.
2637 .sp
2639 .sp
2640 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2641 block that contains a setting of the \fImatch_limit\fP or
2642 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2643 meaningless for DFA matching).
2644 .sp
2646 .sp
2647 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2648 \fIworkspace\fP vector.
2649 .sp
2651 .sp
2652 When a recursive subpattern is processed, the matching function calls itself
2653 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2654 error is given if the output vector is not large enough. This should be
2655 extremely rare, as a vector of size 1000 is used.
2656 .sp
2658 .sp
2659 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2660 some plausibility checks are made on the contents of the workspace, which
2661 should contain data about the previous partial match. If any of these checks
2662 fail, this error is given.
2663 .
2664 .
2665 .SH "SEE ALSO"
2666 .rs
2667 .sp
2668 \fBpcre16\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3), \fBpcrecpp(3)\fP(3),
2669 \fBpcrematching\fP(3), \fBpcrepartial\fP(3), \fBpcreposix\fP(3),
2670 \fBpcreprecompile\fP(3), \fBpcresample\fP(3), \fBpcrestack\fP(3).
2671 .
2672 .
2674 .rs
2675 .sp
2676 .nf
2677 Philip Hazel
2678 University Computing Service
2679 Cambridge CB2 3QH, England.
2680 .fi
2681 .
2682 .
2684 .rs
2685 .sp
2686 .nf
2687 Last updated: 07 September 2012
2688 Copyright (c) 1997-2012 University of Cambridge.
2689 .fi


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

  ViewVC Help
Powered by ViewVC 1.1.5