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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1314 - (show annotations)
Fri Apr 26 10:44:13 2013 UTC (6 years, 6 months ago) by ph10
File size: 125078 byte(s)
Documentation updates.
1 .TH PCREAPI 3 "26 April 2013" "PCRE 8.33"
2 .SH NAME
3 PCRE - Perl-compatible regular expressions
4 .sp
5 .B #include <pcre.h>
6 .
7 .
8 .SH "PCRE NATIVE API BASIC FUNCTIONS"
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 .
47 .SH "PCRE NATIVE API STRING EXTRACTION FUNCTIONS"
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 .
95 .SH "PCRE NATIVE API AUXILIARY FUNCTIONS"
96 .rs
97 .sp
98 .B int pcre_jit_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
99 .ti +5n
100 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
101 .ti +5n
102 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
103 .ti +5n
104 .B pcre_jit_stack *\fIjstack\fP);
105 .PP
106 .B pcre_jit_stack *pcre_jit_stack_alloc(int \fIstartsize\fP, int \fImaxsize\fP);
107 .PP
108 .B void pcre_jit_stack_free(pcre_jit_stack *\fIstack\fP);
109 .PP
110 .B void pcre_assign_jit_stack(pcre_extra *\fIextra\fP,
111 .ti +5n
112 .B pcre_jit_callback \fIcallback\fP, void *\fIdata\fP);
113 .PP
114 .B const unsigned char *pcre_maketables(void);
115 .PP
116 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
117 .ti +5n
118 .B int \fIwhat\fP, void *\fIwhere\fP);
119 .PP
120 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
121 .PP
122 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
123 .PP
124 .B const char *pcre_version(void);
125 .PP
126 .B int pcre_pattern_to_host_byte_order(pcre *\fIcode\fP,
127 .ti +5n
128 .B pcre_extra *\fIextra\fP, const unsigned char *\fItables\fP);
129 .
130 .
131 .SH "PCRE NATIVE API INDIRECTED FUNCTIONS"
132 .rs
133 .sp
134 .B void *(*pcre_malloc)(size_t);
135 .PP
136 .B void (*pcre_free)(void *);
137 .PP
138 .B void *(*pcre_stack_malloc)(size_t);
139 .PP
140 .B void (*pcre_stack_free)(void *);
141 .PP
142 .B int (*pcre_callout)(pcre_callout_block *);
143 .
144 .
145 .SH "PCRE 8-BIT, 16-BIT, AND 32-BIT LIBRARIES"
146 .rs
147 .sp
148 As well as support for 8-bit character strings, PCRE also supports 16-bit
149 strings (from release 8.30) and 32-bit strings (from release 8.32), by means of
150 two additional libraries. They can be built as well as, or instead of, the
151 8-bit library. To avoid too much complication, this document describes the
152 8-bit versions of the functions, with only occasional references to the 16-bit
153 and 32-bit libraries.
154 .P
155 The 16-bit and 32-bit functions operate in the same way as their 8-bit
156 counterparts; they just use different data types for their arguments and
157 results, and their names start with \fBpcre16_\fP or \fBpcre32_\fP instead of
158 \fBpcre_\fP. For every option that has UTF8 in its name (for example,
159 PCRE_UTF8), there are corresponding 16-bit and 32-bit names with UTF8 replaced
160 by UTF16 or UTF32, respectively. This facility is in fact just cosmetic; the
161 16-bit and 32-bit option names define the same bit values.
162 .P
163 References to bytes and UTF-8 in this document should be read as references to
164 16-bit data quantities and UTF-16 when using the 16-bit library, or 32-bit data
165 quantities and UTF-32 when using the 32-bit library, unless specified
166 otherwise. More details of the specific differences for the 16-bit and 32-bit
167 libraries are given in the
168 .\" HREF
169 \fBpcre16\fP
170 .\"
171 and
172 .\" HREF
173 \fBpcre32\fP
174 .\"
175 pages.
176 .
177 .
178 .SH "PCRE API OVERVIEW"
179 .rs
180 .sp
181 PCRE has its own native API, which is described in this document. There are
182 also some wrapper functions (for the 8-bit library only) that correspond to the
183 POSIX regular expression API, but they do not give access to all the
184 functionality. They are described in the
185 .\" HREF
186 \fBpcreposix\fP
187 .\"
188 documentation. Both of these APIs define a set of C function calls. A C++
189 wrapper (again for the 8-bit library only) is also distributed with PCRE. It is
190 documented in the
191 .\" HREF
192 \fBpcrecpp\fP
193 .\"
194 page.
195 .P
196 The native API C function prototypes are defined in the header file
197 \fBpcre.h\fP, and on Unix-like systems the (8-bit) library itself is called
198 \fBlibpcre\fP. It can normally be accessed by adding \fB-lpcre\fP to the
199 command for linking an application that uses PCRE. The header file defines the
200 macros PCRE_MAJOR and PCRE_MINOR to contain the major and minor release numbers
201 for the library. Applications can use these to include support for different
202 releases of PCRE.
203 .P
204 In a Windows environment, if you want to statically link an application program
205 against a non-dll \fBpcre.a\fP file, you must define PCRE_STATIC before
206 including \fBpcre.h\fP or \fBpcrecpp.h\fP, because otherwise the
207 \fBpcre_malloc()\fP and \fBpcre_free()\fP exported functions will be declared
208 \fB__declspec(dllimport)\fP, with unwanted results.
209 .P
210 The functions \fBpcre_compile()\fP, \fBpcre_compile2()\fP, \fBpcre_study()\fP,
211 and \fBpcre_exec()\fP are used for compiling and matching regular expressions
212 in a Perl-compatible manner. A sample program that demonstrates the simplest
213 way of using them is provided in the file called \fIpcredemo.c\fP in the PCRE
214 source distribution. A listing of this program is given in the
215 .\" HREF
216 \fBpcredemo\fP
217 .\"
218 documentation, and the
219 .\" HREF
220 \fBpcresample\fP
221 .\"
222 documentation describes how to compile and run it.
223 .P
224 Just-in-time compiler support is an optional feature of PCRE that can be built
225 in appropriate hardware environments. It greatly speeds up the matching
226 performance of many patterns. Simple programs can easily request that it be
227 used if available, by setting an option that is ignored when it is not
228 relevant. More complicated programs might need to make use of the functions
229 \fBpcre_jit_stack_alloc()\fP, \fBpcre_jit_stack_free()\fP, and
230 \fBpcre_assign_jit_stack()\fP in order to control the JIT code's memory usage.
231 .P
232 From release 8.32 there is also a direct interface for JIT execution, which
233 gives improved performance. The JIT-specific functions are discussed in the
234 .\" HREF
235 \fBpcrejit\fP
236 .\"
237 documentation.
238 .P
239 A second matching function, \fBpcre_dfa_exec()\fP, which is not
240 Perl-compatible, is also provided. This uses a different algorithm for the
241 matching. The alternative algorithm finds all possible matches (at a given
242 point in the subject), and scans the subject just once (unless there are
243 lookbehind assertions). However, this algorithm does not return captured
244 substrings. A description of the two matching algorithms and their advantages
245 and disadvantages is given in the
246 .\" HREF
247 \fBpcrematching\fP
248 .\"
249 documentation.
250 .P
251 In addition to the main compiling and matching functions, there are convenience
252 functions for extracting captured substrings from a subject string that is
253 matched by \fBpcre_exec()\fP. They are:
254 .sp
255 \fBpcre_copy_substring()\fP
256 \fBpcre_copy_named_substring()\fP
257 \fBpcre_get_substring()\fP
258 \fBpcre_get_named_substring()\fP
259 \fBpcre_get_substring_list()\fP
260 \fBpcre_get_stringnumber()\fP
261 \fBpcre_get_stringtable_entries()\fP
262 .sp
263 \fBpcre_free_substring()\fP and \fBpcre_free_substring_list()\fP are also
264 provided, to free the memory used for extracted strings.
265 .P
266 The function \fBpcre_maketables()\fP is used to build a set of character tables
267 in the current locale for passing to \fBpcre_compile()\fP, \fBpcre_exec()\fP,
268 or \fBpcre_dfa_exec()\fP. This is an optional facility that is provided for
269 specialist use. Most commonly, no special tables are passed, in which case
270 internal tables that are generated when PCRE is built are used.
271 .P
272 The function \fBpcre_fullinfo()\fP is used to find out information about a
273 compiled pattern. The function \fBpcre_version()\fP returns a pointer to a
274 string containing the version of PCRE and its date of release.
275 .P
276 The function \fBpcre_refcount()\fP maintains a reference count in a data block
277 containing a compiled pattern. This is provided for the benefit of
278 object-oriented applications.
279 .P
280 The global variables \fBpcre_malloc\fP and \fBpcre_free\fP initially contain
281 the entry points of the standard \fBmalloc()\fP and \fBfree()\fP functions,
282 respectively. PCRE calls the memory management functions via these variables,
283 so a calling program can replace them if it wishes to intercept the calls. This
284 should be done before calling any PCRE functions.
285 .P
286 The global variables \fBpcre_stack_malloc\fP and \fBpcre_stack_free\fP are also
287 indirections to memory management functions. These special functions are used
288 only when PCRE is compiled to use the heap for remembering data, instead of
289 recursive function calls, when running the \fBpcre_exec()\fP function. See the
290 .\" HREF
291 \fBpcrebuild\fP
292 .\"
293 documentation for details of how to do this. It is a non-standard way of
294 building PCRE, for use in environments that have limited stacks. Because of the
295 greater use of memory management, it runs more slowly. Separate functions are
296 provided so that special-purpose external code can be used for this case. When
297 used, these functions are always called in a stack-like manner (last obtained,
298 first freed), and always for memory blocks of the same size. There is a
299 discussion about PCRE's stack usage in the
300 .\" HREF
301 \fBpcrestack\fP
302 .\"
303 documentation.
304 .P
305 The global variable \fBpcre_callout\fP initially contains NULL. It can be set
306 by the caller to a "callout" function, which PCRE will then call at specified
307 points during a matching operation. Details are given in the
308 .\" HREF
309 \fBpcrecallout\fP
310 .\"
311 documentation.
312 .
313 .
314 .\" HTML <a name="newlines"></a>
315 .SH NEWLINES
316 .rs
317 .sp
318 PCRE supports five different conventions for indicating line breaks in
319 strings: a single CR (carriage return) character, a single LF (linefeed)
320 character, the two-character sequence CRLF, any of the three preceding, or any
321 Unicode newline sequence. The Unicode newline sequences are the three just
322 mentioned, plus the single characters VT (vertical tab, U+000B), FF (form feed,
323 U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
324 (paragraph separator, U+2029).
325 .P
326 Each of the first three conventions is used by at least one operating system as
327 its standard newline sequence. When PCRE is built, a default can be specified.
328 The default default is LF, which is the Unix standard. When PCRE is run, the
329 default can be overridden, either when a pattern is compiled, or when it is
330 matched.
331 .P
332 At compile time, the newline convention can be specified by the \fIoptions\fP
333 argument of \fBpcre_compile()\fP, or it can be specified by special text at the
334 start of the pattern itself; this overrides any other settings. See the
335 .\" HREF
336 \fBpcrepattern\fP
337 .\"
338 page for details of the special character sequences.
339 .P
340 In the PCRE documentation the word "newline" is used to mean "the character or
341 pair of characters that indicate a line break". The choice of newline
342 convention affects the handling of the dot, circumflex, and dollar
343 metacharacters, the handling of #-comments in /x mode, and, when CRLF is a
344 recognized line ending sequence, the match position advancement for a
345 non-anchored pattern. There is more detail about this in the
346 .\" HTML <a href="#execoptions">
347 .\" </a>
348 section on \fBpcre_exec()\fP options
349 .\"
350 below.
351 .P
352 The choice of newline convention does not affect the interpretation of
353 the \en or \er escape sequences, nor does it affect what \eR matches, which is
354 controlled in a similar way, but by separate options.
355 .
356 .
357 .SH MULTITHREADING
358 .rs
359 .sp
360 The PCRE functions can be used in multi-threading applications, with the
361 proviso that the memory management functions pointed to by \fBpcre_malloc\fP,
362 \fBpcre_free\fP, \fBpcre_stack_malloc\fP, and \fBpcre_stack_free\fP, and the
363 callout function pointed to by \fBpcre_callout\fP, are shared by all threads.
364 .P
365 The compiled form of a regular expression is not altered during matching, so
366 the same compiled pattern can safely be used by several threads at once.
367 .P
368 If the just-in-time optimization feature is being used, it needs separate
369 memory stack areas for each thread. See the
370 .\" HREF
371 \fBpcrejit\fP
372 .\"
373 documentation for more details.
374 .
375 .
376 .SH "SAVING PRECOMPILED PATTERNS FOR LATER USE"
377 .rs
378 .sp
379 The compiled form of a regular expression can be saved and re-used at a later
380 time, possibly by a different program, and even on a host other than the one on
381 which it was compiled. Details are given in the
382 .\" HREF
383 \fBpcreprecompile\fP
384 .\"
385 documentation, which includes a description of the
386 \fBpcre_pattern_to_host_byte_order()\fP function. However, compiling a regular
387 expression with one version of PCRE for use with a different version is not
388 guaranteed to work and may cause crashes.
389 .
390 .
391 .SH "CHECKING BUILD-TIME OPTIONS"
392 .rs
393 .sp
394 .B int pcre_config(int \fIwhat\fP, void *\fIwhere\fP);
395 .PP
396 The function \fBpcre_config()\fP makes it possible for a PCRE client to
397 discover which optional features have been compiled into the PCRE library. The
398 .\" HREF
399 \fBpcrebuild\fP
400 .\"
401 documentation has more details about these optional features.
402 .P
403 The first argument for \fBpcre_config()\fP is an integer, specifying which
404 information is required; the second argument is a pointer to a variable into
405 which the information is placed. The returned value is zero on success, or the
406 negative error code PCRE_ERROR_BADOPTION if the value in the first argument is
407 not recognized. The following information is available:
408 .sp
409 PCRE_CONFIG_UTF8
410 .sp
411 The output is an integer that is set to one if UTF-8 support is available;
412 otherwise it is set to zero. This value should normally be given to the 8-bit
413 version of this function, \fBpcre_config()\fP. If it is given to the 16-bit
414 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
415 .sp
416 PCRE_CONFIG_UTF16
417 .sp
418 The output is an integer that is set to one if UTF-16 support is available;
419 otherwise it is set to zero. This value should normally be given to the 16-bit
420 version of this function, \fBpcre16_config()\fP. If it is given to the 8-bit
421 or 32-bit version of this function, the result is PCRE_ERROR_BADOPTION.
422 .sp
423 PCRE_CONFIG_UTF32
424 .sp
425 The output is an integer that is set to one if UTF-32 support is available;
426 otherwise it is set to zero. This value should normally be given to the 32-bit
427 version of this function, \fBpcre32_config()\fP. If it is given to the 8-bit
428 or 16-bit version of this function, the result is PCRE_ERROR_BADOPTION.
429 .sp
430 PCRE_CONFIG_UNICODE_PROPERTIES
431 .sp
432 The output is an integer that is set to one if support for Unicode character
433 properties is available; otherwise it is set to zero.
434 .sp
435 PCRE_CONFIG_JIT
436 .sp
437 The output is an integer that is set to one if support for just-in-time
438 compiling is available; otherwise it is set to zero.
439 .sp
440 PCRE_CONFIG_JITTARGET
441 .sp
442 The output is a pointer to a zero-terminated "const char *" string. If JIT
443 support is available, the string contains the name of the architecture for
444 which the JIT compiler is configured, for example "x86 32bit (little endian +
445 unaligned)". If JIT support is not available, the result is NULL.
446 .sp
447 PCRE_CONFIG_NEWLINE
448 .sp
449 The output is an integer whose value specifies the default character sequence
450 that is recognized as meaning "newline". The values that are supported in
451 ASCII/Unicode environments are: 10 for LF, 13 for CR, 3338 for CRLF, -2 for
452 ANYCRLF, and -1 for ANY. In EBCDIC environments, CR, ANYCRLF, and ANY yield the
453 same values. However, the value for LF is normally 21, though some EBCDIC
454 environments use 37. The corresponding values for CRLF are 3349 and 3365. The
455 default should normally correspond to the standard sequence for your operating
456 system.
457 .sp
458 PCRE_CONFIG_BSR
459 .sp
460 The output is an integer whose value indicates what character sequences the \eR
461 escape sequence matches by default. A value of 0 means that \eR matches any
462 Unicode line ending sequence; a value of 1 means that \eR matches only CR, LF,
463 or CRLF. The default can be overridden when a pattern is compiled or matched.
464 .sp
465 PCRE_CONFIG_LINK_SIZE
466 .sp
467 The output is an integer that contains the number of bytes used for internal
468 linkage in compiled regular expressions. For the 8-bit library, the value can
469 be 2, 3, or 4. For the 16-bit library, the value is either 2 or 4 and is still
470 a number of bytes. For the 32-bit library, the value is either 2 or 4 and is
471 still a number of bytes. The default value of 2 is sufficient for all but the
472 most massive patterns, since it allows the compiled pattern to be up to 64K in
473 size. Larger values allow larger regular expressions to be compiled, at the
474 expense of slower matching.
475 .sp
476 PCRE_CONFIG_POSIX_MALLOC_THRESHOLD
477 .sp
478 The output is an integer that contains the threshold above which the POSIX
479 interface uses \fBmalloc()\fP for output vectors. Further details are given in
480 the
481 .\" HREF
482 \fBpcreposix\fP
483 .\"
484 documentation.
485 .sp
486 PCRE_CONFIG_MATCH_LIMIT
487 .sp
488 The output is a long integer that gives the default limit for the number of
489 internal matching function calls in a \fBpcre_exec()\fP execution. Further
490 details are given with \fBpcre_exec()\fP below.
491 .sp
492 PCRE_CONFIG_MATCH_LIMIT_RECURSION
493 .sp
494 The output is a long integer that gives the default limit for the depth of
495 recursion when calling the internal matching function in a \fBpcre_exec()\fP
496 execution. Further details are given with \fBpcre_exec()\fP below.
497 .sp
498 PCRE_CONFIG_STACKRECURSE
499 .sp
500 The output is an integer that is set to one if internal recursion when running
501 \fBpcre_exec()\fP is implemented by recursive function calls that use the stack
502 to remember their state. This is the usual way that PCRE is compiled. The
503 output is zero if PCRE was compiled to use blocks of data on the heap instead
504 of recursive function calls. In this case, \fBpcre_stack_malloc\fP and
505 \fBpcre_stack_free\fP are called to manage memory blocks on the heap, thus
506 avoiding the use of the stack.
507 .
508 .
509 .SH "COMPILING A PATTERN"
510 .rs
511 .sp
512 .B pcre *pcre_compile(const char *\fIpattern\fP, int \fIoptions\fP,
513 .ti +5n
514 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
515 .ti +5n
516 .B const unsigned char *\fItableptr\fP);
517 .sp
518 .B pcre *pcre_compile2(const char *\fIpattern\fP, int \fIoptions\fP,
519 .ti +5n
520 .B int *\fIerrorcodeptr\fP,
521 .ti +5n
522 .B const char **\fIerrptr\fP, int *\fIerroffset\fP,
523 .ti +5n
524 .B const unsigned char *\fItableptr\fP);
525 .P
526 Either of the functions \fBpcre_compile()\fP or \fBpcre_compile2()\fP can be
527 called to compile a pattern into an internal form. The only difference between
528 the two interfaces is that \fBpcre_compile2()\fP has an additional argument,
529 \fIerrorcodeptr\fP, via which a numerical error code can be returned. To avoid
530 too much repetition, we refer just to \fBpcre_compile()\fP below, but the
531 information applies equally to \fBpcre_compile2()\fP.
532 .P
533 The pattern is a C string terminated by a binary zero, and is passed in the
534 \fIpattern\fP argument. A pointer to a single block of memory that is obtained
535 via \fBpcre_malloc\fP is returned. This contains the compiled code and related
536 data. The \fBpcre\fP type is defined for the returned block; this is a typedef
537 for a structure whose contents are not externally defined. It is up to the
538 caller to free the memory (via \fBpcre_free\fP) when it is no longer required.
539 .P
540 Although the compiled code of a PCRE regex is relocatable, that is, it does not
541 depend on memory location, the complete \fBpcre\fP data block is not
542 fully relocatable, because it may contain a copy of the \fItableptr\fP
543 argument, which is an address (see below).
544 .P
545 The \fIoptions\fP argument contains various bit settings that affect the
546 compilation. It should be zero if no options are required. The available
547 options are described below. Some of them (in particular, those that are
548 compatible with Perl, but some others as well) can also be set and unset from
549 within the pattern (see the detailed description in the
550 .\" HREF
551 \fBpcrepattern\fP
552 .\"
553 documentation). For those options that can be different in different parts of
554 the pattern, the contents of the \fIoptions\fP argument specifies their
555 settings at the start of compilation and execution. The PCRE_ANCHORED,
556 PCRE_BSR_\fIxxx\fP, PCRE_NEWLINE_\fIxxx\fP, PCRE_NO_UTF8_CHECK, and
557 PCRE_NO_START_OPTIMIZE options can be set at the time of matching as well as at
558 compile time.
559 .P
560 If \fIerrptr\fP is NULL, \fBpcre_compile()\fP returns NULL immediately.
561 Otherwise, if compilation of a pattern fails, \fBpcre_compile()\fP returns
562 NULL, and sets the variable pointed to by \fIerrptr\fP to point to a textual
563 error message. This is a static string that is part of the library. You must
564 not try to free it. Normally, the offset from the start of the pattern to the
565 byte that was being processed when the error was discovered is placed in the
566 variable pointed to by \fIerroffset\fP, which must not be NULL (if it is, an
567 immediate error is given). However, for an invalid UTF-8 string, the offset is
568 that of the first byte of the failing character.
569 .P
570 Some errors are not detected until the whole pattern has been scanned; in these
571 cases, the offset passed back is the length of the pattern. Note that the
572 offset is in bytes, not characters, even in UTF-8 mode. It may sometimes point
573 into the middle of a UTF-8 character.
574 .P
575 If \fBpcre_compile2()\fP is used instead of \fBpcre_compile()\fP, and the
576 \fIerrorcodeptr\fP argument is not NULL, a non-zero error code number is
577 returned via this argument in the event of an error. This is in addition to the
578 textual error message. Error codes and messages are listed below.
579 .P
580 If the final argument, \fItableptr\fP, is NULL, PCRE uses a default set of
581 character tables that are built when PCRE is compiled, using the default C
582 locale. Otherwise, \fItableptr\fP must be an address that is the result of a
583 call to \fBpcre_maketables()\fP. This value is stored with the compiled
584 pattern, and used again by \fBpcre_exec()\fP, unless another table pointer is
585 passed to it. For more discussion, see the section on locale support below.
586 .P
587 This code fragment shows a typical straightforward call to \fBpcre_compile()\fP:
588 .sp
589 pcre *re;
590 const char *error;
591 int erroffset;
592 re = pcre_compile(
593 "^A.*Z", /* the pattern */
594 0, /* default options */
595 &error, /* for error message */
596 &erroffset, /* for error offset */
597 NULL); /* use default character tables */
598 .sp
599 The following names for option bits are defined in the \fBpcre.h\fP header
600 file:
601 .sp
602 PCRE_ANCHORED
603 .sp
604 If this bit is set, the pattern is forced to be "anchored", that is, it is
605 constrained to match only at the first matching point in the string that is
606 being searched (the "subject string"). This effect can also be achieved by
607 appropriate constructs in the pattern itself, which is the only way to do it in
608 Perl.
609 .sp
610 PCRE_AUTO_CALLOUT
611 .sp
612 If this bit is set, \fBpcre_compile()\fP automatically inserts callout items,
613 all with number 255, before each pattern item. For discussion of the callout
614 facility, see the
615 .\" HREF
616 \fBpcrecallout\fP
617 .\"
618 documentation.
619 .sp
620 PCRE_BSR_ANYCRLF
621 PCRE_BSR_UNICODE
622 .sp
623 These options (which are mutually exclusive) control what the \eR escape
624 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
625 match any Unicode newline sequence. The default is specified when PCRE is
626 built. It can be overridden from within the pattern, or by setting an option
627 when a compiled pattern is matched.
628 .sp
629 PCRE_CASELESS
630 .sp
631 If this bit is set, letters in the pattern match both upper and lower case
632 letters. It is equivalent to Perl's /i option, and it can be changed within a
633 pattern by a (?i) option setting. In UTF-8 mode, PCRE always understands the
634 concept of case for characters whose values are less than 128, so caseless
635 matching is always possible. For characters with higher values, the concept of
636 case is supported if PCRE is compiled with Unicode property support, but not
637 otherwise. If you want to use caseless matching for characters 128 and above,
638 you must ensure that PCRE is compiled with Unicode property support as well as
639 with UTF-8 support.
640 .sp
641 PCRE_DOLLAR_ENDONLY
642 .sp
643 If this bit is set, a dollar metacharacter in the pattern matches only at the
644 end of the subject string. Without this option, a dollar also matches
645 immediately before a newline at the end of the string (but not before any other
646 newlines). The PCRE_DOLLAR_ENDONLY option is ignored if PCRE_MULTILINE is set.
647 There is no equivalent to this option in Perl, and no way to set it within a
648 pattern.
649 .sp
650 PCRE_DOTALL
651 .sp
652 If this bit is set, a dot metacharacter in the pattern matches a character of
653 any value, including one that indicates a newline. However, it only ever
654 matches one character, even if newlines are coded as CRLF. Without this option,
655 a dot does not match when the current position is at a newline. This option is
656 equivalent to Perl's /s option, and it can be changed within a pattern by a
657 (?s) option setting. A negative class such as [^a] always matches newline
658 characters, independent of the setting of this option.
659 .sp
660 PCRE_DUPNAMES
661 .sp
662 If this bit is set, names used to identify capturing subpatterns need not be
663 unique. This can be helpful for certain types of pattern when it is known that
664 only one instance of the named subpattern can ever be matched. There are more
665 details of named subpatterns below; see also the
666 .\" HREF
667 \fBpcrepattern\fP
668 .\"
669 documentation.
670 .sp
671 PCRE_EXTENDED
672 .sp
673 If this bit is set, white space data characters in the pattern are totally
674 ignored except when escaped or inside a character class. White space does not
675 include the VT character (code 11). In addition, characters between an
676 unescaped # outside a character class and the next newline, inclusive, are also
677 ignored. This is equivalent to Perl's /x option, and it can be changed within a
678 pattern by a (?x) option setting.
679 .P
680 Which characters are interpreted as newlines is controlled by the options
681 passed to \fBpcre_compile()\fP or by a special sequence at the start of the
682 pattern, as described in the section entitled
683 .\" HTML <a href="pcrepattern.html#newlines">
684 .\" </a>
685 "Newline conventions"
686 .\"
687 in the \fBpcrepattern\fP documentation. Note that the end of this type of
688 comment is a literal newline sequence in the pattern; escape sequences that
689 happen to represent a newline do not count.
690 .P
691 This option makes it possible to include comments inside complicated patterns.
692 Note, however, that this applies only to data characters. White space characters
693 may never appear within special character sequences in a pattern, for example
694 within the sequence (?( that introduces a conditional subpattern.
695 .sp
696 PCRE_EXTRA
697 .sp
698 This option was invented in order to turn on additional functionality of PCRE
699 that is incompatible with Perl, but it is currently of very little use. When
700 set, any backslash in a pattern that is followed by a letter that has no
701 special meaning causes an error, thus reserving these combinations for future
702 expansion. By default, as in Perl, a backslash followed by a letter with no
703 special meaning is treated as a literal. (Perl can, however, be persuaded to
704 give an error for this, by running it with the -w option.) There are at present
705 no other features controlled by this option. It can also be set by a (?X)
706 option setting within a pattern.
707 .sp
708 PCRE_FIRSTLINE
709 .sp
710 If this option is set, an unanchored pattern is required to match before or at
711 the first newline in the subject string, though the matched text may continue
712 over the newline.
713 .sp
714 PCRE_JAVASCRIPT_COMPAT
715 .sp
716 If this option is set, PCRE's behaviour is changed in some ways so that it is
717 compatible with JavaScript rather than Perl. The changes are as follows:
718 .P
719 (1) A lone closing square bracket in a pattern causes a compile-time error,
720 because this is illegal in JavaScript (by default it is treated as a data
721 character). Thus, the pattern AB]CD becomes illegal when this option is set.
722 .P
723 (2) At run time, a back reference to an unset subpattern group matches an empty
724 string (by default this causes the current matching alternative to fail). A
725 pattern such as (\e1)(a) succeeds when this option is set (assuming it can find
726 an "a" in the subject), whereas it fails by default, for Perl compatibility.
727 .P
728 (3) \eU matches an upper case "U" character; by default \eU causes a compile
729 time error (Perl uses \eU to upper case subsequent characters).
730 .P
731 (4) \eu matches a lower case "u" character unless it is followed by four
732 hexadecimal digits, in which case the hexadecimal number defines the code point
733 to match. By default, \eu causes a compile time error (Perl uses it to upper
734 case the following character).
735 .P
736 (5) \ex matches a lower case "x" character unless it is followed by two
737 hexadecimal digits, in which case the hexadecimal number defines the code point
738 to match. By default, as in Perl, a hexadecimal number is always expected after
739 \ex, but it may have zero, one, or two digits (so, for example, \exz matches a
740 binary zero character followed by z).
741 .sp
742 PCRE_MULTILINE
743 .sp
744 By default, PCRE treats the subject string as consisting of a single line of
745 characters (even if it actually contains newlines). The "start of line"
746 metacharacter (^) matches only at the start of the string, while the "end of
747 line" metacharacter ($) matches only at the end of the string, or before a
748 terminating newline (unless PCRE_DOLLAR_ENDONLY is set). This is the same as
749 Perl.
750 .P
751 When PCRE_MULTILINE it is set, the "start of line" and "end of line" constructs
752 match immediately following or immediately before internal newlines in the
753 subject string, respectively, as well as at the very start and end. This is
754 equivalent to Perl's /m option, and it can be changed within a pattern by a
755 (?m) option setting. If there are no newlines in a subject string, or no
756 occurrences of ^ or $ in a pattern, setting PCRE_MULTILINE has no effect.
757 .sp
758 PCRE_NEVER_UTF
759 .sp
760 This option locks out interpretation of the pattern as UTF-8 (or UTF-16 or
761 UTF-32 in the 16-bit and 32-bit libraries). In particular, it prevents the
762 creator of the pattern from switching to UTF interpretation by starting the
763 pattern with (*UTF). This may be useful in applications that process patterns
764 from external sources. The combination of PCRE_UTF8 and PCRE_NEVER_UTF also
765 causes an error.
766 .sp
767 PCRE_NEWLINE_CR
768 PCRE_NEWLINE_LF
769 PCRE_NEWLINE_CRLF
770 PCRE_NEWLINE_ANYCRLF
771 PCRE_NEWLINE_ANY
772 .sp
773 These options override the default newline definition that was chosen when PCRE
774 was built. Setting the first or the second specifies that a newline is
775 indicated by a single character (CR or LF, respectively). Setting
776 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
777 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
778 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
779 that any Unicode newline sequence should be recognized.
780 .P
781 In an ASCII/Unicode environment, the Unicode newline sequences are the three
782 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
783 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
784 (paragraph separator, U+2029). For the 8-bit library, the last two are
785 recognized only in UTF-8 mode.
786 .P
787 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
788 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
789 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
790 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
791 less than 256. For more details, see the
792 .\" HREF
793 \fBpcrebuild\fP
794 .\"
795 documentation.
796 .P
797 The newline setting in the options word uses three bits that are treated
798 as a number, giving eight possibilities. Currently only six are used (default
799 plus the five values above). This means that if you set more than one newline
800 option, the combination may or may not be sensible. For example,
801 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
802 other combinations may yield unused numbers and cause an error.
803 .P
804 The only time that a line break in a pattern is specially recognized when
805 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
806 and so are ignored in this mode. Also, an unescaped # outside a character class
807 indicates a comment that lasts until after the next line break sequence. In
808 other circumstances, line break sequences in patterns are treated as literal
809 data.
810 .P
811 The newline option that is set at compile time becomes the default that is used
812 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
813 .sp
814 PCRE_NO_AUTO_CAPTURE
815 .sp
816 If this option is set, it disables the use of numbered capturing parentheses in
817 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
818 were followed by ?: but named parentheses can still be used for capturing (and
819 they acquire numbers in the usual way). There is no equivalent of this option
820 in Perl.
821 .sp
822 PCRE_NO_START_OPTIMIZE
823 .sp
824 This is an option that acts at matching time; that is, it is really an option
825 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
826 it is remembered with the compiled pattern and assumed at matching time. This
827 is necessary if you want to use JIT execution, because the JIT compiler needs
828 to know whether or not this option is set. For details see the discussion of
829 PCRE_NO_START_OPTIMIZE
830 .\" HTML <a href="#execoptions">
831 .\" </a>
832 below.
833 .\"
834 .sp
835 PCRE_UCP
836 .sp
837 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
838 \ew, and some of the POSIX character classes. By default, only ASCII characters
839 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
840 classify characters. More details are given in the section on
841 .\" HTML <a href="pcre.html#genericchartypes">
842 .\" </a>
843 generic character types
844 .\"
845 in the
846 .\" HREF
847 \fBpcrepattern\fP
848 .\"
849 page. If you set PCRE_UCP, matching one of the items it affects takes much
850 longer. The option is available only if PCRE has been compiled with Unicode
851 property support.
852 .sp
853 PCRE_UNGREEDY
854 .sp
855 This option inverts the "greediness" of the quantifiers so that they are not
856 greedy by default, but become greedy if followed by "?". It is not compatible
857 with Perl. It can also be set by a (?U) option setting within the pattern.
858 .sp
859 PCRE_UTF8
860 .sp
861 This option causes PCRE to regard both the pattern and the subject as strings
862 of UTF-8 characters instead of single-byte strings. However, it is available
863 only when PCRE is built to include UTF support. If not, the use of this option
864 provokes an error. Details of how this option changes the behaviour of PCRE are
865 given in the
866 .\" HREF
867 \fBpcreunicode\fP
868 .\"
869 page.
870 .sp
871 PCRE_NO_UTF8_CHECK
872 .sp
873 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
874 automatically checked. There is a discussion about the
875 .\" HTML <a href="pcreunicode.html#utf8strings">
876 .\" </a>
877 validity of UTF-8 strings
878 .\"
879 in the
880 .\" HREF
881 \fBpcreunicode\fP
882 .\"
883 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
884 error. If you already know that your pattern is valid, and you want to skip
885 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
886 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
887 undefined. It may cause your program to crash. Note that this option can also
888 be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
889 validity checking of subject strings only. If the same string is being matched
890 many times, the option can be safely set for the second and subsequent
891 matchings to improve performance.
892 .
893 .
894 .SH "COMPILATION ERROR CODES"
895 .rs
896 .sp
897 The following table lists the error codes than may be returned by
898 \fBpcre_compile2()\fP, along with the error messages that may be returned by
899 both compiling functions. Note that error messages are always 8-bit ASCII
900 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
901 have fallen out of use. To avoid confusion, they have not been re-used.
902 .sp
903 0 no error
904 1 \e at end of pattern
905 2 \ec at end of pattern
906 3 unrecognized character follows \e
907 4 numbers out of order in {} quantifier
908 5 number too big in {} quantifier
909 6 missing terminating ] for character class
910 7 invalid escape sequence in character class
911 8 range out of order in character class
912 9 nothing to repeat
913 10 [this code is not in use]
914 11 internal error: unexpected repeat
915 12 unrecognized character after (? or (?-
916 13 POSIX named classes are supported only within a class
917 14 missing )
918 15 reference to non-existent subpattern
919 16 erroffset passed as NULL
920 17 unknown option bit(s) set
921 18 missing ) after comment
922 19 [this code is not in use]
923 20 regular expression is too large
924 21 failed to get memory
925 22 unmatched parentheses
926 23 internal error: code overflow
927 24 unrecognized character after (?<
928 25 lookbehind assertion is not fixed length
929 26 malformed number or name after (?(
930 27 conditional group contains more than two branches
931 28 assertion expected after (?(
932 29 (?R or (?[+-]digits must be followed by )
933 30 unknown POSIX class name
934 31 POSIX collating elements are not supported
935 32 this version of PCRE is compiled without UTF support
936 33 [this code is not in use]
937 34 character value in \ex{...} sequence is too large
938 35 invalid condition (?(0)
939 36 \eC not allowed in lookbehind assertion
940 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
941 38 number after (?C is > 255
942 39 closing ) for (?C expected
943 40 recursive call could loop indefinitely
944 41 unrecognized character after (?P
945 42 syntax error in subpattern name (missing terminator)
946 43 two named subpatterns have the same name
947 44 invalid UTF-8 string (specifically UTF-8)
948 45 support for \eP, \ep, and \eX has not been compiled
949 46 malformed \eP or \ep sequence
950 47 unknown property name after \eP or \ep
951 48 subpattern name is too long (maximum 32 characters)
952 49 too many named subpatterns (maximum 10000)
953 50 [this code is not in use]
954 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
955 52 internal error: overran compiling workspace
956 53 internal error: previously-checked referenced subpattern
957 not found
958 54 DEFINE group contains more than one branch
959 55 repeating a DEFINE group is not allowed
960 56 inconsistent NEWLINE options
961 57 \eg is not followed by a braced, angle-bracketed, or quoted
962 name/number or by a plain number
963 58 a numbered reference must not be zero
964 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
965 60 (*VERB) not recognized or malformed
966 61 number is too big
967 62 subpattern name expected
968 63 digit expected after (?+
969 64 ] is an invalid data character in JavaScript compatibility mode
970 65 different names for subpatterns of the same number are
971 not allowed
972 66 (*MARK) must have an argument
973 67 this version of PCRE is not compiled with Unicode property
974 support
975 68 \ec must be followed by an ASCII character
976 69 \ek is not followed by a braced, angle-bracketed, or quoted name
977 70 internal error: unknown opcode in find_fixedlength()
978 71 \eN is not supported in a class
979 72 too many forward references
980 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
981 74 invalid UTF-16 string (specifically UTF-16)
982 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
983 76 character value in \eu.... sequence is too large
984 77 invalid UTF-32 string (specifically UTF-32)
985 .sp
986 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
987 be used if the limits were changed when PCRE was built.
988 .
989 .
990 .\" HTML <a name="studyingapattern"></a>
991 .SH "STUDYING A PATTERN"
992 .rs
993 .sp
994 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
995 .ti +5n
996 .B const char **\fIerrptr\fP);
997 .PP
998 If a compiled pattern is going to be used several times, it is worth spending
999 more time analyzing it in order to speed up the time taken for matching. The
1000 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
1001 argument. If studying the pattern produces additional information that will
1002 help speed up matching, \fBpcre_study()\fP returns a pointer to a
1003 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
1004 results of the study.
1005 .P
1006 The returned value from \fBpcre_study()\fP can be passed directly to
1007 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
1008 also contains other fields that can be set by the caller before the block is
1009 passed; these are described
1010 .\" HTML <a href="#extradata">
1011 .\" </a>
1012 below
1013 .\"
1014 in the section on matching a pattern.
1015 .P
1016 If studying the pattern does not produce any useful information,
1017 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1018 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1019 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1020 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1021 returns a \fBpcre_extra\fP block even if studying did not find any additional
1022 information. It may still return NULL, however, if an error occurs in
1023 \fBpcre_study()\fP.
1024 .P
1025 The second argument of \fBpcre_study()\fP contains option bits. There are three
1026 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1027 .sp
1028 PCRE_STUDY_JIT_COMPILE
1029 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1030 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1031 .sp
1032 If any of these are set, and the just-in-time compiler is available, the
1033 pattern is further compiled into machine code that executes much faster than
1034 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1035 compiler is not available, these options are ignored. All undefined bits in the
1036 \fIoptions\fP argument must be zero.
1037 .P
1038 JIT compilation is a heavyweight optimization. It can take some time for
1039 patterns to be analyzed, and for one-off matches and simple patterns the
1040 benefit of faster execution might be offset by a much slower study time.
1041 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1042 handled, matching automatically falls back to the \fBpcre_exec()\fP
1043 interpreter. For more details, see the
1044 .\" HREF
1045 \fBpcrejit\fP
1046 .\"
1047 documentation.
1048 .P
1049 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1050 studying succeeds (even if no data is returned), the variable it points to is
1051 set to NULL. Otherwise it is set to point to a textual error message. This is a
1052 static string that is part of the library. You must not try to free it. You
1053 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1054 sure that it has run successfully.
1055 .P
1056 When you are finished with a pattern, you can free the memory used for the
1057 study data by calling \fBpcre_free_study()\fP. This function was added to the
1058 API for release 8.20. For earlier versions, the memory could be freed with
1059 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1060 where JIT optimization is not used, but it is advisable to change to the new
1061 function when convenient.
1062 .P
1063 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1064 real application there should be tests for errors):
1065 .sp
1066 int rc;
1067 pcre *re;
1068 pcre_extra *sd;
1069 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1070 sd = pcre_study(
1071 re, /* result of pcre_compile() */
1072 0, /* no options */
1073 &error); /* set to NULL or points to a message */
1074 rc = pcre_exec( /* see below for details of pcre_exec() options */
1075 re, sd, "subject", 7, 0, 0, ovector, 30);
1076 ...
1077 pcre_free_study(sd);
1078 pcre_free(re);
1079 .sp
1080 Studying a pattern does two things: first, a lower bound for the length of
1081 subject string that is needed to match the pattern is computed. This does not
1082 mean that there are any strings of that length that match, but it does
1083 guarantee that no shorter strings match. The value is used to avoid wasting
1084 time by trying to match strings that are shorter than the lower bound. You can
1085 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1086 .P
1087 Studying a pattern is also useful for non-anchored patterns that do not have a
1088 single fixed starting character. A bitmap of possible starting bytes is
1089 created. This speeds up finding a position in the subject at which to start
1090 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1091 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1092 .P
1093 These two optimizations apply to both \fBpcre_exec()\fP and
1094 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1095 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option.
1096 You might want to do this if your pattern contains callouts or (*MARK) and you
1097 want to make use of these facilities in cases where matching fails.
1098 .P
1099 PCRE_NO_START_OPTIMIZE can be specified at either compile time or execution
1100 time. However, if PCRE_NO_START_OPTIMIZE is passed to \fBpcre_exec()\fP, (that
1101 is, after any JIT compilation has happened) JIT execution is disabled. For JIT
1102 execution to work with PCRE_NO_START_OPTIMIZE, the option must be set at
1103 compile time.
1104 .P
1105 There is a longer discussion of PCRE_NO_START_OPTIMIZE
1106 .\" HTML <a href="#execoptions">
1107 .\" </a>
1108 below.
1109 .\"
1110 .
1111 .
1112 .\" HTML <a name="localesupport"></a>
1113 .SH "LOCALE SUPPORT"
1114 .rs
1115 .sp
1116 PCRE handles caseless matching, and determines whether characters are letters,
1117 digits, or whatever, by reference to a set of tables, indexed by character
1118 value. When running in UTF-8 mode, this applies only to characters
1119 with codes less than 128. By default, higher-valued codes never match escapes
1120 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1121 Unicode character property support. Alternatively, the PCRE_UCP option can be
1122 set at compile time; this causes \ew and friends to use Unicode property
1123 support instead of built-in tables. The use of locales with Unicode is
1124 discouraged. If you are handling characters with codes greater than 128, you
1125 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1126 two.
1127 .P
1128 PCRE contains an internal set of tables that are used when the final argument
1129 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1130 Normally, the internal tables recognize only ASCII characters. However, when
1131 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1132 default "C" locale of the local system, which may cause them to be different.
1133 .P
1134 The internal tables can always be overridden by tables supplied by the
1135 application that calls PCRE. These may be created in a different locale from
1136 the default. As more and more applications change to using Unicode, the need
1137 for this locale support is expected to die away.
1138 .P
1139 External tables are built by calling the \fBpcre_maketables()\fP function,
1140 which has no arguments, in the relevant locale. The result can then be passed
1141 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1142 example, to build and use tables that are appropriate for the French locale
1143 (where accented characters with values greater than 128 are treated as letters),
1144 the following code could be used:
1145 .sp
1146 setlocale(LC_CTYPE, "fr_FR");
1147 tables = pcre_maketables();
1148 re = pcre_compile(..., tables);
1149 .sp
1150 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1151 are using Windows, the name for the French locale is "french".
1152 .P
1153 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1154 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1155 that the memory containing the tables remains available for as long as it is
1156 needed.
1157 .P
1158 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1159 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1160 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1161 pattern, compilation, studying and matching all happen in the same locale, but
1162 different patterns can be compiled in different locales.
1163 .P
1164 It is possible to pass a table pointer or NULL (indicating the use of the
1165 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1166 this facility could be used to match a pattern in a different locale from the
1167 one in which it was compiled. Passing table pointers at run time is discussed
1168 below in the section on matching a pattern.
1169 .
1170 .
1171 .\" HTML <a name="infoaboutpattern"></a>
1172 .SH "INFORMATION ABOUT A PATTERN"
1173 .rs
1174 .sp
1175 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1176 .ti +5n
1177 .B int \fIwhat\fP, void *\fIwhere\fP);
1178 .PP
1179 The \fBpcre_fullinfo()\fP function returns information about a compiled
1180 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1181 library at version 8.30, after more than 10 years of obsolescence.
1182 .P
1183 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1184 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1185 the pattern was not studied. The third argument specifies which piece of
1186 information is required, and the fourth argument is a pointer to a variable
1187 to receive the data. The yield of the function is zero for success, or one of
1188 the following negative numbers:
1189 .sp
1190 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1191 the argument \fIwhere\fP was NULL
1192 PCRE_ERROR_BADMAGIC the "magic number" was not found
1193 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1194 endianness
1195 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1196 PCRE_ERROR_UNSET the requested field is not set
1197 .sp
1198 The "magic number" is placed at the start of each compiled pattern as an simple
1199 check against passing an arbitrary memory pointer. The endianness error can
1200 occur if a compiled pattern is saved and reloaded on a different host. Here is
1201 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1202 pattern:
1203 .sp
1204 int rc;
1205 size_t length;
1206 rc = pcre_fullinfo(
1207 re, /* result of pcre_compile() */
1208 sd, /* result of pcre_study(), or NULL */
1209 PCRE_INFO_SIZE, /* what is required */
1210 &length); /* where to put the data */
1211 .sp
1212 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1213 as follows:
1214 .sp
1215 PCRE_INFO_BACKREFMAX
1216 .sp
1217 Return the number of the highest back reference in the pattern. The fourth
1218 argument should point to an \fBint\fP variable. Zero is returned if there are
1219 no back references.
1220 .sp
1221 PCRE_INFO_CAPTURECOUNT
1222 .sp
1223 Return the number of capturing subpatterns in the pattern. The fourth argument
1224 should point to an \fBint\fP variable.
1225 .sp
1226 PCRE_INFO_DEFAULT_TABLES
1227 .sp
1228 Return a pointer to the internal default character tables within PCRE. The
1229 fourth argument should point to an \fBunsigned char *\fP variable. This
1230 information call is provided for internal use by the \fBpcre_study()\fP
1231 function. External callers can cause PCRE to use its internal tables by passing
1232 a NULL table pointer.
1233 .sp
1234 PCRE_INFO_FIRSTBYTE
1235 .sp
1236 Return information about the first data unit of any matched string, for a
1237 non-anchored pattern. (The name of this option refers to the 8-bit library,
1238 where data units are bytes.) The fourth argument should point to an \fBint\fP
1239 variable.
1240 .P
1241 If there is a fixed first value, for example, the letter "c" from a pattern
1242 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1243 value is always less than 256. In the 16-bit library the value can be up to
1244 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1245 .P
1246 If there is no fixed first value, and if either
1247 .sp
1248 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1249 starts with "^", or
1250 .sp
1251 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1252 (if it were set, the pattern would be anchored),
1253 .sp
1254 -1 is returned, indicating that the pattern matches only at the start of a
1255 subject string or after any newline within the string. Otherwise -2 is
1256 returned. For anchored patterns, -2 is returned.
1257 .P
1258 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1259 to return the full 32-bit range of the character, this value is deprecated;
1260 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1261 should be used.
1262 .sp
1263 PCRE_INFO_FIRSTTABLE
1264 .sp
1265 If the pattern was studied, and this resulted in the construction of a 256-bit
1266 table indicating a fixed set of values for the first data unit in any matching
1267 string, a pointer to the table is returned. Otherwise NULL is returned. The
1268 fourth argument should point to an \fBunsigned char *\fP variable.
1269 .sp
1270 PCRE_INFO_HASCRORLF
1271 .sp
1272 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1273 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1274 explicit match is either a literal CR or LF character, or \er or \en.
1275 .sp
1276 PCRE_INFO_JCHANGED
1277 .sp
1278 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1279 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1280 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1281 .sp
1282 PCRE_INFO_JIT
1283 .sp
1284 Return 1 if the pattern was studied with one of the JIT options, and
1285 just-in-time compiling was successful. The fourth argument should point to an
1286 \fBint\fP variable. A return value of 0 means that JIT support is not available
1287 in this version of PCRE, or that the pattern was not studied with a JIT option,
1288 or that the JIT compiler could not handle this particular pattern. See the
1289 .\" HREF
1290 \fBpcrejit\fP
1291 .\"
1292 documentation for details of what can and cannot be handled.
1293 .sp
1294 PCRE_INFO_JITSIZE
1295 .sp
1296 If the pattern was successfully studied with a JIT option, return the size of
1297 the JIT compiled code, otherwise return zero. The fourth argument should point
1298 to a \fBsize_t\fP variable.
1299 .sp
1300 PCRE_INFO_LASTLITERAL
1301 .sp
1302 Return the value of the rightmost literal data unit that must exist in any
1303 matched string, other than at its start, if such a value has been recorded. The
1304 fourth argument should point to an \fBint\fP variable. If there is no such
1305 value, -1 is returned. For anchored patterns, a last literal value is recorded
1306 only if it follows something of variable length. For example, for the pattern
1307 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1308 is -1.
1309 .P
1310 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1311 to return the full 32-bit range of the character, this value is deprecated;
1312 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1313 be used.
1314 .sp
1315 PCRE_INFO_MATCHLIMIT
1316 .sp
1317 If the pattern set a match limit by including an item of the form
1318 (*LIMIT_MATCH=nnnn) at the start, the value is returned. The fourth argument
1319 should point to an unsigned 32-bit integer. If no such value has been set, the
1320 call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1321 .sp
1322 PCRE_INFO_MAXLOOKBEHIND
1323 .sp
1324 Return the number of characters (NB not bytes) in the longest lookbehind
1325 assertion in the pattern. This information is useful when doing multi-segment
1326 matching using the partial matching facilities. Note that the simple assertions
1327 \eb and \eB require a one-character lookbehind. \eA also registers a
1328 one-character lookbehind, though it does not actually inspect the previous
1329 character. This is to ensure that at least one character from the old segment
1330 is retained when a new segment is processed. Otherwise, if there are no
1331 lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1332 segment.
1333 .sp
1334 PCRE_INFO_MINLENGTH
1335 .sp
1336 If the pattern was studied and a minimum length for matching subject strings
1337 was computed, its value is returned. Otherwise the returned value is -1. The
1338 value is a number of characters, which in UTF-8 mode may be different from the
1339 number of bytes. The fourth argument should point to an \fBint\fP variable. A
1340 non-negative value is a lower bound to the length of any matching string. There
1341 may not be any strings of that length that do actually match, but every string
1342 that does match is at least that long.
1343 .sp
1344 PCRE_INFO_NAMECOUNT
1345 PCRE_INFO_NAMEENTRYSIZE
1346 PCRE_INFO_NAMETABLE
1347 .sp
1348 PCRE supports the use of named as well as numbered capturing parentheses. The
1349 names are just an additional way of identifying the parentheses, which still
1350 acquire numbers. Several convenience functions such as
1351 \fBpcre_get_named_substring()\fP are provided for extracting captured
1352 substrings by name. It is also possible to extract the data directly, by first
1353 converting the name to a number in order to access the correct pointers in the
1354 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1355 you need to use the name-to-number map, which is described by these three
1356 values.
1357 .P
1358 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1359 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1360 entry; both of these return an \fBint\fP value. The entry size depends on the
1361 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1362 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1363 the first two bytes of each entry are the number of the capturing parenthesis,
1364 most significant byte first. In the 16-bit library, the pointer points to
1365 16-bit data units, the first of which contains the parenthesis number.
1366 In the 32-bit library, the pointer points to 32-bit data units, the first of
1367 which contains the parenthesis number. The rest
1368 of the entry is the corresponding name, zero terminated.
1369 .P
1370 The names are in alphabetical order. Duplicate names may appear if (?| is used
1371 to create multiple groups with the same number, as described in the
1372 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1373 .\" </a>
1374 section on duplicate subpattern numbers
1375 .\"
1376 in the
1377 .\" HREF
1378 \fBpcrepattern\fP
1379 .\"
1380 page. Duplicate names for subpatterns with different numbers are permitted only
1381 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1382 table in the order in which they were found in the pattern. In the absence of
1383 (?| this is the order of increasing number; when (?| is used this is not
1384 necessarily the case because later subpatterns may have lower numbers.
1385 .P
1386 As a simple example of the name/number table, consider the following pattern
1387 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1388 space - including newlines - is ignored):
1389 .sp
1390 .\" JOIN
1391 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1392 (?<month>\ed\ed) - (?<day>\ed\ed) )
1393 .sp
1394 There are four named subpatterns, so the table has four entries, and each entry
1395 in the table is eight bytes long. The table is as follows, with non-printing
1396 bytes shows in hexadecimal, and undefined bytes shown as ??:
1397 .sp
1398 00 01 d a t e 00 ??
1399 00 05 d a y 00 ?? ??
1400 00 04 m o n t h 00
1401 00 02 y e a r 00 ??
1402 .sp
1403 When writing code to extract data from named subpatterns using the
1404 name-to-number map, remember that the length of the entries is likely to be
1405 different for each compiled pattern.
1406 .sp
1407 PCRE_INFO_OKPARTIAL
1408 .sp
1409 Return 1 if the pattern can be used for partial matching with
1410 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1411 \fBint\fP variable. From release 8.00, this always returns 1, because the
1412 restrictions that previously applied to partial matching have been lifted. The
1413 .\" HREF
1414 \fBpcrepartial\fP
1415 .\"
1416 documentation gives details of partial matching.
1417 .sp
1418 PCRE_INFO_OPTIONS
1419 .sp
1420 Return a copy of the options with which the pattern was compiled. The fourth
1421 argument should point to an \fBunsigned long int\fP variable. These option bits
1422 are those specified in the call to \fBpcre_compile()\fP, modified by any
1423 top-level option settings at the start of the pattern itself. In other words,
1424 they are the options that will be in force when matching starts. For example,
1425 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1426 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1427 .P
1428 A pattern is automatically anchored by PCRE if all of its top-level
1429 alternatives begin with one of the following:
1430 .sp
1431 ^ unless PCRE_MULTILINE is set
1432 \eA always
1433 \eG always
1434 .\" JOIN
1435 .* if PCRE_DOTALL is set and there are no back
1436 references to the subpattern in which .* appears
1437 .sp
1438 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1439 \fBpcre_fullinfo()\fP.
1440 .sp
1441 PCRE_INFO_RECURSIONLIMIT
1442 .sp
1443 If the pattern set a recursion limit by including an item of the form
1444 (*LIMIT_RECURSION=nnnn) at the start, the value is returned. The fourth
1445 argument should point to an unsigned 32-bit integer. If no such value has been
1446 set, the call to \fBpcre_fullinfo()\fP returns the error PCRE_ERROR_UNSET.
1447 .sp
1448 PCRE_INFO_SIZE
1449 .sp
1450 Return the size of the compiled pattern in bytes (for both libraries). The
1451 fourth argument should point to a \fBsize_t\fP variable. This value does not
1452 include the size of the \fBpcre\fP structure that is returned by
1453 \fBpcre_compile()\fP. The value that is passed as the argument to
1454 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1455 place the compiled data is the value returned by this option plus the size of
1456 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1457 does not alter the value returned by this option.
1458 .sp
1459 PCRE_INFO_STUDYSIZE
1460 .sp
1461 Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1462 field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1463 study data, zero is returned. The fourth argument should point to a
1464 \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1465 to record information that will speed up matching (see the section entitled
1466 .\" HTML <a href="#studyingapattern">
1467 .\" </a>
1468 "Studying a pattern"
1469 .\"
1470 above). The format of the \fIstudy_data\fP block is private, but its length
1471 is made available via this option so that it can be saved and restored (see the
1472 .\" HREF
1473 \fBpcreprecompile\fP
1474 .\"
1475 documentation for details).
1476 .sp
1477 PCRE_INFO_FIRSTCHARACTERFLAGS
1478 .sp
1479 Return information about the first data unit of any matched string, for a
1480 non-anchored pattern. The fourth argument should point to an \fBint\fP
1481 variable.
1482 .P
1483 If there is a fixed first value, for example, the letter "c" from a pattern
1484 such as (cat|cow|coyote), 1 is returned, and the character value can be
1485 retrieved using PCRE_INFO_FIRSTCHARACTER.
1486 .P
1487 If there is no fixed first value, and if either
1488 .sp
1489 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1490 starts with "^", or
1491 .sp
1492 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1493 (if it were set, the pattern would be anchored),
1494 .sp
1495 2 is returned, indicating that the pattern matches only at the start of a
1496 subject string or after any newline within the string. Otherwise 0 is
1497 returned. For anchored patterns, 0 is returned.
1498 .sp
1499 PCRE_INFO_FIRSTCHARACTER
1500 .sp
1501 Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1502 returned 1; otherwise returns 0. The fourth argument should point to an
1503 \fBuint_t\fP variable.
1504 .P
1505 In the 8-bit library, the value is always less than 256. In the 16-bit library
1506 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1507 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1508 .P
1509 If there is no fixed first value, and if either
1510 .sp
1511 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1512 starts with "^", or
1513 .sp
1514 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1515 (if it were set, the pattern would be anchored),
1516 .sp
1517 -1 is returned, indicating that the pattern matches only at the start of a
1518 subject string or after any newline within the string. Otherwise -2 is
1519 returned. For anchored patterns, -2 is returned.
1520 .sp
1521 PCRE_INFO_REQUIREDCHARFLAGS
1522 .sp
1523 Returns 1 if there is a rightmost literal data unit that must exist in any
1524 matched string, other than at its start. The fourth argument should point to
1525 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1526 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1527 .P
1528 For anchored patterns, a last literal value is recorded only if it follows
1529 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1530 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1531 /^a\edz\ed/ the returned value is 0.
1532 .sp
1533 PCRE_INFO_REQUIREDCHAR
1534 .sp
1535 Return the value of the rightmost literal data unit that must exist in any
1536 matched string, other than at its start, if such a value has been recorded. The
1537 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1538 value, 0 is returned.
1539 .
1540 .
1541 .SH "REFERENCE COUNTS"
1542 .rs
1543 .sp
1544 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1545 .PP
1546 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1547 data block that contains a compiled pattern. It is provided for the benefit of
1548 applications that operate in an object-oriented manner, where different parts
1549 of the application may be using the same compiled pattern, but you want to free
1550 the block when they are all done.
1551 .P
1552 When a pattern is compiled, the reference count field is initialized to zero.
1553 It is changed only by calling this function, whose action is to add the
1554 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1555 function is the new value. However, the value of the count is constrained to
1556 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1557 it is forced to the appropriate limit value.
1558 .P
1559 Except when it is zero, the reference count is not correctly preserved if a
1560 pattern is compiled on one host and then transferred to a host whose byte-order
1561 is different. (This seems a highly unlikely scenario.)
1562 .
1563 .
1564 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1565 .rs
1566 .sp
1567 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1568 .ti +5n
1569 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1570 .ti +5n
1571 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1572 .P
1573 The function \fBpcre_exec()\fP is called to match a subject string against a
1574 compiled pattern, which is passed in the \fIcode\fP argument. If the
1575 pattern was studied, the result of the study should be passed in the
1576 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1577 and \fIextra\fP arguments as many times as you like, in order to match
1578 different subject strings with the same pattern.
1579 .P
1580 This function is the main matching facility of the library, and it operates in
1581 a Perl-like manner. For specialist use there is also an alternative matching
1582 function, which is described
1583 .\" HTML <a href="#dfamatch">
1584 .\" </a>
1585 below
1586 .\"
1587 in the section about the \fBpcre_dfa_exec()\fP function.
1588 .P
1589 In most applications, the pattern will have been compiled (and optionally
1590 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1591 possible to save compiled patterns and study data, and then use them later
1592 in different processes, possibly even on different hosts. For a discussion
1593 about this, see the
1594 .\" HREF
1595 \fBpcreprecompile\fP
1596 .\"
1597 documentation.
1598 .P
1599 Here is an example of a simple call to \fBpcre_exec()\fP:
1600 .sp
1601 int rc;
1602 int ovector[30];
1603 rc = pcre_exec(
1604 re, /* result of pcre_compile() */
1605 NULL, /* we didn't study the pattern */
1606 "some string", /* the subject string */
1607 11, /* the length of the subject string */
1608 0, /* start at offset 0 in the subject */
1609 0, /* default options */
1610 ovector, /* vector of integers for substring information */
1611 30); /* number of elements (NOT size in bytes) */
1612 .
1613 .
1614 .\" HTML <a name="extradata"></a>
1615 .SS "Extra data for \fBpcre_exec()\fR"
1616 .rs
1617 .sp
1618 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1619 data block. The \fBpcre_study()\fP function returns such a block (when it
1620 doesn't return NULL), but you can also create one for yourself, and pass
1621 additional information in it. The \fBpcre_extra\fP block contains the following
1622 fields (not necessarily in this order):
1623 .sp
1624 unsigned long int \fIflags\fP;
1625 void *\fIstudy_data\fP;
1626 void *\fIexecutable_jit\fP;
1627 unsigned long int \fImatch_limit\fP;
1628 unsigned long int \fImatch_limit_recursion\fP;
1629 void *\fIcallout_data\fP;
1630 const unsigned char *\fItables\fP;
1631 unsigned char **\fImark\fP;
1632 .sp
1633 In the 16-bit version of this structure, the \fImark\fP field has type
1634 "PCRE_UCHAR16 **".
1635 .sp
1636 In the 32-bit version of this structure, the \fImark\fP field has type
1637 "PCRE_UCHAR32 **".
1638 .P
1639 The \fIflags\fP field is used to specify which of the other fields are set. The
1640 flag bits are:
1641 .sp
1642 PCRE_EXTRA_CALLOUT_DATA
1643 PCRE_EXTRA_EXECUTABLE_JIT
1644 PCRE_EXTRA_MARK
1645 PCRE_EXTRA_MATCH_LIMIT
1646 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1647 PCRE_EXTRA_STUDY_DATA
1648 PCRE_EXTRA_TABLES
1649 .sp
1650 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1651 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1652 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1653 should not set these yourself, but you may add to the block by setting other
1654 fields and their corresponding flag bits.
1655 .P
1656 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1657 vast amount of resources when running patterns that are not going to match,
1658 but which have a very large number of possibilities in their search trees. The
1659 classic example is a pattern that uses nested unlimited repeats.
1660 .P
1661 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1662 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1663 imposed on the number of times this function is called during a match, which
1664 has the effect of limiting the amount of backtracking that can take place. For
1665 patterns that are not anchored, the count restarts from zero for each position
1666 in the subject string.
1667 .P
1668 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1669 with a JIT option, the way that the matching is executed is entirely different.
1670 However, there is still the possibility of runaway matching that goes on for a
1671 very long time, and so the \fImatch_limit\fP value is also used in this case
1672 (but in a different way) to limit how long the matching can continue.
1673 .P
1674 The default value for the limit can be set when PCRE is built; the default
1675 default is 10 million, which handles all but the most extreme cases. You can
1676 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1677 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1678 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1679 PCRE_ERROR_MATCHLIMIT.
1680 .P
1681 A value for the match limit may also be supplied by an item at the start of a
1682 pattern of the form
1683 .sp
1684 (*LIMIT_MATCH=d)
1685 .sp
1686 where d is a decimal number. However, such a setting is ignored unless d is
1687 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1688 is set, less than the default.
1689 .P
1690 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1691 instead of limiting the total number of times that \fBmatch()\fP is called, it
1692 limits the depth of recursion. The recursion depth is a smaller number than the
1693 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1694 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1695 .P
1696 Limiting the recursion depth limits the amount of machine stack that can be
1697 used, or, when PCRE has been compiled to use memory on the heap instead of the
1698 stack, the amount of heap memory that can be used. This limit is not relevant,
1699 and is ignored, when matching is done using JIT compiled code.
1700 .P
1701 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1702 built; the default default is the same value as the default for
1703 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1704 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1705 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1706 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1707 .P
1708 A value for the recursion limit may also be supplied by an item at the start of
1709 a pattern of the form
1710 .sp
1711 (*LIMIT_RECURSION=d)
1712 .sp
1713 where d is a decimal number. However, such a setting is ignored unless d is
1714 less than the limit set by the caller of \fBpcre_exec()\fP or, if no such limit
1715 is set, less than the default.
1716 .P
1717 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1718 and is described in the
1719 .\" HREF
1720 \fBpcrecallout\fP
1721 .\"
1722 documentation.
1723 .P
1724 The \fItables\fP field is used to pass a character tables pointer to
1725 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1726 pattern. A non-NULL value is stored with the compiled pattern only if custom
1727 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1728 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1729 internal tables to be used. This facility is helpful when re-using patterns
1730 that have been saved after compiling with an external set of tables, because
1731 the external tables might be at a different address when \fBpcre_exec()\fP is
1732 called. See the
1733 .\" HREF
1734 \fBpcreprecompile\fP
1735 .\"
1736 documentation for a discussion of saving compiled patterns for later use.
1737 .P
1738 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1739 be set to point to a suitable variable. If the pattern contains any
1740 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1741 a name to pass back, a pointer to the name string (zero terminated) is placed
1742 in the variable pointed to by the \fImark\fP field. The names are within the
1743 compiled pattern; if you wish to retain such a name you must copy it before
1744 freeing the memory of a compiled pattern. If there is no name to pass back, the
1745 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1746 backtracking control verbs, see the section entitled
1747 .\" HTML <a href="pcrepattern#backtrackcontrol">
1748 .\" </a>
1749 "Backtracking control"
1750 .\"
1751 in the
1752 .\" HREF
1753 \fBpcrepattern\fP
1754 .\"
1755 documentation.
1756 .
1757 .
1758 .\" HTML <a name="execoptions"></a>
1759 .SS "Option bits for \fBpcre_exec()\fP"
1760 .rs
1761 .sp
1762 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1763 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1764 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1765 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1766 PCRE_PARTIAL_SOFT.
1767 .P
1768 If the pattern was successfully studied with one of the just-in-time (JIT)
1769 compile options, the only supported options for JIT execution are
1770 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1771 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1772 unsupported option is used, JIT execution is disabled and the normal
1773 interpretive code in \fBpcre_exec()\fP is run.
1774 .sp
1775 PCRE_ANCHORED
1776 .sp
1777 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1778 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1779 to be anchored by virtue of its contents, it cannot be made unachored at
1780 matching time.
1781 .sp
1782 PCRE_BSR_ANYCRLF
1783 PCRE_BSR_UNICODE
1784 .sp
1785 These options (which are mutually exclusive) control what the \eR escape
1786 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1787 match any Unicode newline sequence. These options override the choice that was
1788 made or defaulted when the pattern was compiled.
1789 .sp
1790 PCRE_NEWLINE_CR
1791 PCRE_NEWLINE_LF
1792 PCRE_NEWLINE_CRLF
1793 PCRE_NEWLINE_ANYCRLF
1794 PCRE_NEWLINE_ANY
1795 .sp
1796 These options override the newline definition that was chosen or defaulted when
1797 the pattern was compiled. For details, see the description of
1798 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1799 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1800 the way the match position is advanced after a match failure for an unanchored
1801 pattern.
1802 .P
1803 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1804 match attempt for an unanchored pattern fails when the current position is at a
1805 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1806 characters, the match position is advanced by two characters instead of one, in
1807 other words, to after the CRLF.
1808 .P
1809 The above rule is a compromise that makes the most common cases work as
1810 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1811 set), it does not match the string "\er\enA" because, after failing at the
1812 start, it skips both the CR and the LF before retrying. However, the pattern
1813 [\er\en]A does match that string, because it contains an explicit CR or LF
1814 reference, and so advances only by one character after the first failure.
1815 .P
1816 An explicit match for CR of LF is either a literal appearance of one of those
1817 characters, or one of the \er or \en escape sequences. Implicit matches such as
1818 [^X] do not count, nor does \es (which includes CR and LF in the characters
1819 that it matches).
1820 .P
1821 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1822 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1823 .sp
1824 PCRE_NOTBOL
1825 .sp
1826 This option specifies that first character of the subject string is not the
1827 beginning of a line, so the circumflex metacharacter should not match before
1828 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1829 never to match. This option affects only the behaviour of the circumflex
1830 metacharacter. It does not affect \eA.
1831 .sp
1832 PCRE_NOTEOL
1833 .sp
1834 This option specifies that the end of the subject string is not the end of a
1835 line, so the dollar metacharacter should not match it nor (except in multiline
1836 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1837 compile time) causes dollar never to match. This option affects only the
1838 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1839 .sp
1840 PCRE_NOTEMPTY
1841 .sp
1842 An empty string is not considered to be a valid match if this option is set. If
1843 there are alternatives in the pattern, they are tried. If all the alternatives
1844 match the empty string, the entire match fails. For example, if the pattern
1845 .sp
1846 a?b?
1847 .sp
1848 is applied to a string not beginning with "a" or "b", it matches an empty
1849 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1850 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1851 .sp
1852 PCRE_NOTEMPTY_ATSTART
1853 .sp
1854 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1855 the start of the subject is permitted. If the pattern is anchored, such a match
1856 can occur only if the pattern contains \eK.
1857 .P
1858 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1859 does make a special case of a pattern match of the empty string within its
1860 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1861 emulate Perl's behaviour after matching a null string by first trying the match
1862 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1863 if that fails, by advancing the starting offset (see below) and trying an
1864 ordinary match again. There is some code that demonstrates how to do this in
1865 the
1866 .\" HREF
1867 \fBpcredemo\fP
1868 .\"
1869 sample program. In the most general case, you have to check to see if the
1870 newline convention recognizes CRLF as a newline, and if so, and the current
1871 character is CR followed by LF, advance the starting offset by two characters
1872 instead of one.
1873 .sp
1874 PCRE_NO_START_OPTIMIZE
1875 .sp
1876 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1877 a match, in order to speed up the process. For example, if it is known that an
1878 unanchored match must start with a specific character, it searches the subject
1879 for that character, and fails immediately if it cannot find it, without
1880 actually running the main matching function. This means that a special item
1881 such as (*COMMIT) at the start of a pattern is not considered until after a
1882 suitable starting point for the match has been found. Also, when callouts or
1883 (*MARK) items are in use, these "start-up" optimizations can cause them to be
1884 skipped if the pattern is never actually used. The start-up optimizations are
1885 in effect a pre-scan of the subject that takes place before the pattern is run.
1886 .P
1887 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1888 causing performance to suffer, but ensuring that in cases where the result is
1889 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1890 are considered at every possible starting position in the subject string. If
1891 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1892 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1893 to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1894 always done using interpretively.
1895 .P
1896 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1897 Consider the pattern
1898 .sp
1899 (*COMMIT)ABC
1900 .sp
1901 When this is compiled, PCRE records the fact that a match must start with the
1902 character "A". Suppose the subject string is "DEFABC". The start-up
1903 optimization scans along the subject, finds "A" and runs the first match
1904 attempt from there. The (*COMMIT) item means that the pattern must match the
1905 current starting position, which in this case, it does. However, if the same
1906 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1907 subject string does not happen. The first match attempt is run starting from
1908 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1909 the overall result is "no match". If the pattern is studied, more start-up
1910 optimizations may be used. For example, a minimum length for the subject may be
1911 recorded. Consider the pattern
1912 .sp
1913 (*MARK:A)(X|Y)
1914 .sp
1915 The minimum length for a match is one character. If the subject is "ABC", there
1916 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1917 If the pattern is studied, the final attempt does not take place, because PCRE
1918 knows that the subject is too short, and so the (*MARK) is never encountered.
1919 In this case, studying the pattern does not affect the overall match result,
1920 which is still "no match", but it does affect the auxiliary information that is
1921 returned.
1922 .sp
1923 PCRE_NO_UTF8_CHECK
1924 .sp
1925 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1926 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1927 The entire string is checked before any other processing takes place. The value
1928 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1929 UTF-8 character. There is a discussion about the
1930 .\" HTML <a href="pcreunicode.html#utf8strings">
1931 .\" </a>
1932 validity of UTF-8 strings
1933 .\"
1934 in the
1935 .\" HREF
1936 \fBpcreunicode\fP
1937 .\"
1938 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1939 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1940 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1941 cases, information about the precise nature of the error may also be returned
1942 (see the descriptions of these errors in the section entitled \fIError return
1943 values from\fP \fBpcre_exec()\fP
1944 .\" HTML <a href="#errorlist">
1945 .\" </a>
1946 below).
1947 .\"
1948 If \fIstartoffset\fP contains a value that does not point to the start of a
1949 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1950 returned.
1951 .P
1952 If you already know that your subject is valid, and you want to skip these
1953 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1954 calling \fBpcre_exec()\fP. You might want to do this for the second and
1955 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1956 all the matches in a single subject string. However, you should be sure that
1957 the value of \fIstartoffset\fP points to the start of a character (or the end
1958 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1959 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1960 undefined. Your program may crash.
1961 .sp
1962 PCRE_PARTIAL_HARD
1963 PCRE_PARTIAL_SOFT
1964 .sp
1965 These options turn on the partial matching feature. For backwards
1966 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1967 occurs if the end of the subject string is reached successfully, but there are
1968 not enough subject characters to complete the match. If this happens when
1969 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1970 testing any remaining alternatives. Only if no complete match can be found is
1971 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1972 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1973 but only if no complete match can be found.
1974 .P
1975 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1976 partial match is found, \fBpcre_exec()\fP immediately returns
1977 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1978 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1979 important that an alternative complete match.
1980 .P
1981 In both cases, the portion of the string that was inspected when the partial
1982 match was found is set as the first matching string. There is a more detailed
1983 discussion of partial and multi-segment matching, with examples, in the
1984 .\" HREF
1985 \fBpcrepartial\fP
1986 .\"
1987 documentation.
1988 .
1989 .
1990 .SS "The string to be matched by \fBpcre_exec()\fP"
1991 .rs
1992 .sp
1993 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1994 \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1995 in \fIstartoffset\fP. If this is negative or greater than the length of the
1996 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1997 offset is zero, the search for a match starts at the beginning of the subject,
1998 and this is by far the most common case. In UTF-8 mode, the byte offset must
1999 point to the start of a UTF-8 character (or the end of the subject). Unlike the
2000 pattern string, the subject may contain binary zero bytes.
2001 .P
2002 A non-zero starting offset is useful when searching for another match in the
2003 same subject by calling \fBpcre_exec()\fP again after a previous success.
2004 Setting \fIstartoffset\fP differs from just passing over a shortened string and
2005 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
2006 lookbehind. For example, consider the pattern
2007 .sp
2008 \eBiss\eB
2009 .sp
2010 which finds occurrences of "iss" in the middle of words. (\eB matches only if
2011 the current position in the subject is not a word boundary.) When applied to
2012 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
2013 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
2014 subject, namely "issipi", it does not match, because \eB is always false at the
2015 start of the subject, which is deemed to be a word boundary. However, if
2016 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
2017 set to 4, it finds the second occurrence of "iss" because it is able to look
2018 behind the starting point to discover that it is preceded by a letter.
2019 .P
2020 Finding all the matches in a subject is tricky when the pattern can match an
2021 empty string. It is possible to emulate Perl's /g behaviour by first trying the
2022 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
2023 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
2024 and trying an ordinary match again. There is some code that demonstrates how to
2025 do this in the
2026 .\" HREF
2027 \fBpcredemo\fP
2028 .\"
2029 sample program. In the most general case, you have to check to see if the
2030 newline convention recognizes CRLF as a newline, and if so, and the current
2031 character is CR followed by LF, advance the starting offset by two characters
2032 instead of one.
2033 .P
2034 If a non-zero starting offset is passed when the pattern is anchored, one
2035 attempt to match at the given offset is made. This can only succeed if the
2036 pattern does not require the match to be at the start of the subject.
2037 .
2038 .
2039 .SS "How \fBpcre_exec()\fP returns captured substrings"
2040 .rs
2041 .sp
2042 In general, a pattern matches a certain portion of the subject, and in
2043 addition, further substrings from the subject may be picked out by parts of the
2044 pattern. Following the usage in Jeffrey Friedl's book, this is called
2045 "capturing" in what follows, and the phrase "capturing subpattern" is used for
2046 a fragment of a pattern that picks out a substring. PCRE supports several other
2047 kinds of parenthesized subpattern that do not cause substrings to be captured.
2048 .P
2049 Captured substrings are returned to the caller via a vector of integers whose
2050 address is passed in \fIovector\fP. The number of elements in the vector is
2051 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2052 argument is NOT the size of \fIovector\fP in bytes.
2053 .P
2054 The first two-thirds of the vector is used to pass back captured substrings,
2055 each substring using a pair of integers. The remaining third of the vector is
2056 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2057 and is not available for passing back information. The number passed in
2058 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2059 rounded down.
2060 .P
2061 When a match is successful, information about captured substrings is returned
2062 in pairs of integers, starting at the beginning of \fIovector\fP, and
2063 continuing up to two-thirds of its length at the most. The first element of
2064 each pair is set to the byte offset of the first character in a substring, and
2065 the second is set to the byte offset of the first character after the end of a
2066 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
2067 mode. They are not character counts.
2068 .P
2069 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2070 portion of the subject string matched by the entire pattern. The next pair is
2071 used for the first capturing subpattern, and so on. The value returned by
2072 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2073 For example, if two substrings have been captured, the returned value is 3. If
2074 there are no capturing subpatterns, the return value from a successful match is
2075 1, indicating that just the first pair of offsets has been set.
2076 .P
2077 If a capturing subpattern is matched repeatedly, it is the last portion of the
2078 string that it matched that is returned.
2079 .P
2080 If the vector is too small to hold all the captured substring offsets, it is
2081 used as far as possible (up to two-thirds of its length), and the function
2082 returns a value of zero. If neither the actual string matched nor any captured
2083 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2084 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2085 back references and the \fIovector\fP is not big enough to remember the related
2086 substrings, PCRE has to get additional memory for use during matching. Thus it
2087 is usually advisable to supply an \fIovector\fP of reasonable size.
2088 .P
2089 There are some cases where zero is returned (indicating vector overflow) when
2090 in fact the vector is exactly the right size for the final match. For example,
2091 consider the pattern
2092 .sp
2093 (a)(?:(b)c|bd)
2094 .sp
2095 If a vector of 6 elements (allowing for only 1 captured substring) is given
2096 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2097 captured string, thereby recording a vector overflow, before failing to match
2098 "c" and backing up to try the second alternative. The zero return, however,
2099 does correctly indicate that the maximum number of slots (namely 2) have been
2100 filled. In similar cases where there is temporary overflow, but the final
2101 number of used slots is actually less than the maximum, a non-zero value is
2102 returned.
2103 .P
2104 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2105 subpatterns there are in a compiled pattern. The smallest size for
2106 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2107 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2108 .P
2109 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2110 the subject when subpattern \fIn\fP has not been used at all. For example, if
2111 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2112 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2113 happens, both values in the offset pairs corresponding to unused subpatterns
2114 are set to -1.
2115 .P
2116 Offset values that correspond to unused subpatterns at the end of the
2117 expression are also set to -1. For example, if the string "abc" is matched
2118 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2119 return from the function is 2, because the highest used capturing subpattern
2120 number is 1, and the offsets for for the second and third capturing subpatterns
2121 (assuming the vector is large enough, of course) are set to -1.
2122 .P
2123 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2124 correspond to capturing parentheses in the pattern are never changed. That is,
2125 if a pattern contains \fIn\fP capturing parentheses, no more than
2126 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2127 elements (in the first two-thirds) retain whatever values they previously had.
2128 .P
2129 Some convenience functions are provided for extracting the captured substrings
2130 as separate strings. These are described below.
2131 .
2132 .
2133 .\" HTML <a name="errorlist"></a>
2134 .SS "Error return values from \fBpcre_exec()\fP"
2135 .rs
2136 .sp
2137 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2138 defined in the header file:
2139 .sp
2140 PCRE_ERROR_NOMATCH (-1)
2141 .sp
2142 The subject string did not match the pattern.
2143 .sp
2144 PCRE_ERROR_NULL (-2)
2145 .sp
2146 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2147 NULL and \fIovecsize\fP was not zero.
2148 .sp
2149 PCRE_ERROR_BADOPTION (-3)
2150 .sp
2151 An unrecognized bit was set in the \fIoptions\fP argument.
2152 .sp
2153 PCRE_ERROR_BADMAGIC (-4)
2154 .sp
2155 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2156 the case when it is passed a junk pointer and to detect when a pattern that was
2157 compiled in an environment of one endianness is run in an environment with the
2158 other endianness. This is the error that PCRE gives when the magic number is
2159 not present.
2160 .sp
2161 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2162 .sp
2163 While running the pattern match, an unknown item was encountered in the
2164 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2165 of the compiled pattern.
2166 .sp
2167 PCRE_ERROR_NOMEMORY (-6)
2168 .sp
2169 If a pattern contains back references, but the \fIovector\fP that is passed to
2170 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2171 gets a block of memory at the start of matching to use for this purpose. If the
2172 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2173 automatically freed at the end of matching.
2174 .P
2175 This error is also given if \fBpcre_stack_malloc()\fP fails in
2176 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2177 \fB--disable-stack-for-recursion\fP.
2178 .sp
2179 PCRE_ERROR_NOSUBSTRING (-7)
2180 .sp
2181 This error is used by the \fBpcre_copy_substring()\fP,
2182 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2183 below). It is never returned by \fBpcre_exec()\fP.
2184 .sp
2185 PCRE_ERROR_MATCHLIMIT (-8)
2186 .sp
2187 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2188 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2189 above.
2190 .sp
2191 PCRE_ERROR_CALLOUT (-9)
2192 .sp
2193 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2194 use by callout functions that want to yield a distinctive error code. See the
2195 .\" HREF
2196 \fBpcrecallout\fP
2197 .\"
2198 documentation for details.
2199 .sp
2200 PCRE_ERROR_BADUTF8 (-10)
2201 .sp
2202 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2203 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2204 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2205 UTF-8 character is placed in the first element, and a reason code is placed in
2206 the second element. The reason codes are listed in the
2207 .\" HTML <a href="#badutf8reasons">
2208 .\" </a>
2209 following section.
2210 .\"
2211 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2212 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2213 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2214 .sp
2215 PCRE_ERROR_BADUTF8_OFFSET (-11)
2216 .sp
2217 The UTF-8 byte sequence that was passed as a subject was checked and found to
2218 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2219 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2220 end of the subject.
2221 .sp
2222 PCRE_ERROR_PARTIAL (-12)
2223 .sp
2224 The subject string did not match, but it did match partially. See the
2225 .\" HREF
2226 \fBpcrepartial\fP
2227 .\"
2228 documentation for details of partial matching.
2229 .sp
2230 PCRE_ERROR_BADPARTIAL (-13)
2231 .sp
2232 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2233 option was used with a compiled pattern containing items that were not
2234 supported for partial matching. From release 8.00 onwards, there are no
2235 restrictions on partial matching.
2236 .sp
2237 PCRE_ERROR_INTERNAL (-14)
2238 .sp
2239 An unexpected internal error has occurred. This error could be caused by a bug
2240 in PCRE or by overwriting of the compiled pattern.
2241 .sp
2242 PCRE_ERROR_BADCOUNT (-15)
2243 .sp
2244 This error is given if the value of the \fIovecsize\fP argument is negative.
2245 .sp
2246 PCRE_ERROR_RECURSIONLIMIT (-21)
2247 .sp
2248 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2249 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2250 description above.
2251 .sp
2252 PCRE_ERROR_BADNEWLINE (-23)
2253 .sp
2254 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2255 .sp
2256 PCRE_ERROR_BADOFFSET (-24)
2257 .sp
2258 The value of \fIstartoffset\fP was negative or greater than the length of the
2259 subject, that is, the value in \fIlength\fP.
2260 .sp
2261 PCRE_ERROR_SHORTUTF8 (-25)
2262 .sp
2263 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2264 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2265 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2266 fact sufficient to detect this case, but this special error code for
2267 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2268 retained for backwards compatibility.
2269 .sp
2270 PCRE_ERROR_RECURSELOOP (-26)
2271 .sp
2272 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2273 the pattern. Specifically, it means that either the whole pattern or a
2274 subpattern has been called recursively for the second time at the same position
2275 in the subject string. Some simple patterns that might do this are detected and
2276 faulted at compile time, but more complicated cases, in particular mutual
2277 recursions between two different subpatterns, cannot be detected until run
2278 time.
2279 .sp
2280 PCRE_ERROR_JIT_STACKLIMIT (-27)
2281 .sp
2282 This error is returned when a pattern that was successfully studied using a
2283 JIT compile option is being matched, but the memory available for the
2284 just-in-time processing stack is not large enough. See the
2285 .\" HREF
2286 \fBpcrejit\fP
2287 .\"
2288 documentation for more details.
2289 .sp
2290 PCRE_ERROR_BADMODE (-28)
2291 .sp
2292 This error is given if a pattern that was compiled by the 8-bit library is
2293 passed to a 16-bit or 32-bit library function, or vice versa.
2294 .sp
2295 PCRE_ERROR_BADENDIANNESS (-29)
2296 .sp
2297 This error is given if a pattern that was compiled and saved is reloaded on a
2298 host with different endianness. The utility function
2299 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2300 so that it runs on the new host.
2301 .sp
2302 PCRE_ERROR_JIT_BADOPTION
2303 .sp
2304 This error is returned when a pattern that was successfully studied using a JIT
2305 compile option is being matched, but the matching mode (partial or complete
2306 match) does not correspond to any JIT compilation mode. When the JIT fast path
2307 function is used, this error may be also given for invalid options. See the
2308 .\" HREF
2309 \fBpcrejit\fP
2310 .\"
2311 documentation for more details.
2312 .sp
2313 PCRE_ERROR_BADLENGTH (-32)
2314 .sp
2315 This error is given if \fBpcre_exec()\fP is called with a negative value for
2316 the \fIlength\fP argument.
2317 .P
2318 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2319 .
2320 .
2321 .\" HTML <a name="badutf8reasons"></a>
2322 .SS "Reason codes for invalid UTF-8 strings"
2323 .rs
2324 .sp
2325 This section applies only to the 8-bit library. The corresponding information
2326 for the 16-bit and 32-bit libraries is given in the
2327 .\" HREF
2328 \fBpcre16\fP
2329 .\"
2330 and
2331 .\" HREF
2332 \fBpcre32\fP
2333 .\"
2334 pages.
2335 .P
2336 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2337 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2338 least 2, the offset of the start of the invalid UTF-8 character is placed in
2339 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2340 in the second element (\fIovector[1]\fP). The reason codes are given names in
2341 the \fBpcre.h\fP header file:
2342 .sp
2343 PCRE_UTF8_ERR1
2344 PCRE_UTF8_ERR2
2345 PCRE_UTF8_ERR3
2346 PCRE_UTF8_ERR4
2347 PCRE_UTF8_ERR5
2348 .sp
2349 The string ends with a truncated UTF-8 character; the code specifies how many
2350 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2351 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2352 allows for up to 6 bytes, and this is checked first; hence the possibility of
2353 4 or 5 missing bytes.
2354 .sp
2355 PCRE_UTF8_ERR6
2356 PCRE_UTF8_ERR7
2357 PCRE_UTF8_ERR8
2358 PCRE_UTF8_ERR9
2359 PCRE_UTF8_ERR10
2360 .sp
2361 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2362 character do not have the binary value 0b10 (that is, either the most
2363 significant bit is 0, or the next bit is 1).
2364 .sp
2365 PCRE_UTF8_ERR11
2366 PCRE_UTF8_ERR12
2367 .sp
2368 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2369 these code points are excluded by RFC 3629.
2370 .sp
2371 PCRE_UTF8_ERR13
2372 .sp
2373 A 4-byte character has a value greater than 0x10fff; these code points are
2374 excluded by RFC 3629.
2375 .sp
2376 PCRE_UTF8_ERR14
2377 .sp
2378 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2379 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2380 from UTF-8.
2381 .sp
2382 PCRE_UTF8_ERR15
2383 PCRE_UTF8_ERR16
2384 PCRE_UTF8_ERR17
2385 PCRE_UTF8_ERR18
2386 PCRE_UTF8_ERR19
2387 .sp
2388 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2389 value that can be represented by fewer bytes, which is invalid. For example,
2390 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2391 one byte.
2392 .sp
2393 PCRE_UTF8_ERR20
2394 .sp
2395 The two most significant bits of the first byte of a character have the binary
2396 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2397 byte can only validly occur as the second or subsequent byte of a multi-byte
2398 character.
2399 .sp
2400 PCRE_UTF8_ERR21
2401 .sp
2402 The first byte of a character has the value 0xfe or 0xff. These values can
2403 never occur in a valid UTF-8 string.
2404 .sp
2405 PCRE_UTF8_ERR22
2406 .sp
2407 This error code was formerly used when the presence of a so-called
2408 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2409 such characters should not cause a string to be rejected, and so this code is
2410 no longer in use and is never returned.
2411 .
2412 .
2413 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2414 .rs
2415 .sp
2416 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2417 .ti +5n
2418 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2419 .ti +5n
2420 .B int \fIbuffersize\fP);
2421 .PP
2422 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2423 .ti +5n
2424 .B int \fIstringcount\fP, int \fIstringnumber\fP,
2425 .ti +5n
2426 .B const char **\fIstringptr\fP);
2427 .PP
2428 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2429 .ti +5n
2430 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2431 .PP
2432 Captured substrings can be accessed directly by using the offsets returned by
2433 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2434 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2435 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2436 as new, separate, zero-terminated strings. These functions identify substrings
2437 by number. The next section describes functions for extracting named
2438 substrings.
2439 .P
2440 A substring that contains a binary zero is correctly extracted and has a
2441 further zero added on the end, but the result is not, of course, a C string.
2442 However, you can process such a string by referring to the length that is
2443 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2444 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2445 for handling strings containing binary zeros, because the end of the final
2446 string is not independently indicated.
2447 .P
2448 The first three arguments are the same for all three of these functions:
2449 \fIsubject\fP is the subject string that has just been successfully matched,
2450 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2451 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2452 captured by the match, including the substring that matched the entire regular
2453 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2454 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2455 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2456 number of elements in the vector divided by three.
2457 .P
2458 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2459 extract a single substring, whose number is given as \fIstringnumber\fP. A
2460 value of zero extracts the substring that matched the entire pattern, whereas
2461 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2462 the string is placed in \fIbuffer\fP, whose length is given by
2463 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2464 obtained via \fBpcre_malloc\fP, and its address is returned via
2465 \fIstringptr\fP. The yield of the function is the length of the string, not
2466 including the terminating zero, or one of these error codes:
2467 .sp
2468 PCRE_ERROR_NOMEMORY (-6)
2469 .sp
2470 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2471 memory failed for \fBpcre_get_substring()\fP.
2472 .sp
2473 PCRE_ERROR_NOSUBSTRING (-7)
2474 .sp
2475 There is no substring whose number is \fIstringnumber\fP.
2476 .P
2477 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2478 and builds a list of pointers to them. All this is done in a single block of
2479 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2480 is returned via \fIlistptr\fP, which is also the start of the list of string
2481 pointers. The end of the list is marked by a NULL pointer. The yield of the
2482 function is zero if all went well, or the error code
2483 .sp
2484 PCRE_ERROR_NOMEMORY (-6)
2485 .sp
2486 if the attempt to get the memory block failed.
2487 .P
2488 When any of these functions encounter a substring that is unset, which can
2489 happen when capturing subpattern number \fIn+1\fP matches some part of the
2490 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2491 string. This can be distinguished from a genuine zero-length substring by
2492 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2493 substrings.
2494 .P
2495 The two convenience functions \fBpcre_free_substring()\fP and
2496 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2497 a previous call of \fBpcre_get_substring()\fP or
2498 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2499 the function pointed to by \fBpcre_free\fP, which of course could be called
2500 directly from a C program. However, PCRE is used in some situations where it is
2501 linked via a special interface to another programming language that cannot use
2502 \fBpcre_free\fP directly; it is for these cases that the functions are
2503 provided.
2504 .
2505 .
2506 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2507 .rs
2508 .sp
2509 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2510 .ti +5n
2511 .B const char *\fIname\fP);
2512 .PP
2513 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2514 .ti +5n
2515 .B const char *\fIsubject\fP, int *\fIovector\fP,
2516 .ti +5n
2517 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2518 .ti +5n
2519 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2520 .PP
2521 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2522 .ti +5n
2523 .B const char *\fIsubject\fP, int *\fIovector\fP,
2524 .ti +5n
2525 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2526 .ti +5n
2527 .B const char **\fIstringptr\fP);
2528 .PP
2529 To extract a substring by name, you first have to find associated number.
2530 For example, for this pattern
2531 .sp
2532 (a+)b(?<xxx>\ed+)...
2533 .sp
2534 the number of the subpattern called "xxx" is 2. If the name is known to be
2535 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2536 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2537 pattern, and the second is the name. The yield of the function is the
2538 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2539 that name.
2540 .P
2541 Given the number, you can extract the substring directly, or use one of the
2542 functions described in the previous section. For convenience, there are also
2543 two functions that do the whole job.
2544 .P
2545 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2546 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2547 functions that extract by number. As these are described in the previous
2548 section, they are not re-described here. There are just two differences:
2549 .P
2550 First, instead of a substring number, a substring name is given. Second, there
2551 is an extra argument, given at the start, which is a pointer to the compiled
2552 pattern. This is needed in order to gain access to the name-to-number
2553 translation table.
2554 .P
2555 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2556 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2557 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2558 the behaviour may not be what you want (see the next section).
2559 .P
2560 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2561 subpatterns with the same number, as described in the
2562 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2563 .\" </a>
2564 section on duplicate subpattern numbers
2565 .\"
2566 in the
2567 .\" HREF
2568 \fBpcrepattern\fP
2569 .\"
2570 page, you cannot use names to distinguish the different subpatterns, because
2571 names are not included in the compiled code. The matching process uses only
2572 numbers. For this reason, the use of different names for subpatterns of the
2573 same number causes an error at compile time.
2574 .
2575 .
2576 .SH "DUPLICATE SUBPATTERN NAMES"
2577 .rs
2578 .sp
2579 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2580 .ti +5n
2581 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2582 .PP
2583 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2584 are not required to be unique. (Duplicate names are always allowed for
2585 subpatterns with the same number, created by using the (?| feature. Indeed, if
2586 such subpatterns are named, they are required to use the same names.)
2587 .P
2588 Normally, patterns with duplicate names are such that in any one match, only
2589 one of the named subpatterns participates. An example is shown in the
2590 .\" HREF
2591 \fBpcrepattern\fP
2592 .\"
2593 documentation.
2594 .P
2595 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2596 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2597 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2598 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2599 returns one of the numbers that are associated with the name, but it is not
2600 defined which it is.
2601 .P
2602 If you want to get full details of all captured substrings for a given name,
2603 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2604 argument is the compiled pattern, and the second is the name. The third and
2605 fourth are pointers to variables which are updated by the function. After it
2606 has run, they point to the first and last entries in the name-to-number table
2607 for the given name. The function itself returns the length of each entry, or
2608 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2609 described above in the section entitled \fIInformation about a pattern\fP
2610 .\" HTML <a href="#infoaboutpattern">
2611 .\" </a>
2612 above.
2613 .\"
2614 Given all the relevant entries for the name, you can extract each of their
2615 numbers, and hence the captured data, if any.
2616 .
2617 .
2618 .SH "FINDING ALL POSSIBLE MATCHES"
2619 .rs
2620 .sp
2621 The traditional matching function uses a similar algorithm to Perl, which stops
2622 when it finds the first match, starting at a given point in the subject. If you
2623 want to find all possible matches, or the longest possible match, consider
2624 using the alternative matching function (see below) instead. If you cannot use
2625 the alternative function, but still need to find all possible matches, you
2626 can kludge it up by making use of the callout facility, which is described in
2627 the
2628 .\" HREF
2629 \fBpcrecallout\fP
2630 .\"
2631 documentation.
2632 .P
2633 What you have to do is to insert a callout right at the end of the pattern.
2634 When your callout function is called, extract and save the current matched
2635 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2636 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2637 will yield PCRE_ERROR_NOMATCH.
2638 .
2639 .
2640 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2641 .rs
2642 .sp
2643 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2644 stack, which in certain environments can be rather limited in size. Some users
2645 find it helpful to have an estimate of the amount of stack that is used by
2646 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2647 .\" HREF
2648 \fBpcrestack\fP
2649 .\"
2650 documentation. The estimate that is output by \fBpcretest\fP when called with
2651 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2652 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2653 .P
2654 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2655 the negative error code PCRE_ERROR_NULL, but with this special combination of
2656 arguments, it returns instead a negative number whose absolute value is the
2657 approximate stack frame size in bytes. (A negative number is used so that it is
2658 clear that no match has happened.) The value is approximate because in some
2659 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2660 additional variables on the stack.
2661 .P
2662 If PCRE has been compiled to use the heap instead of the stack for recursion,
2663 the value returned is the size of each block that is obtained from the heap.
2664 .
2665 .
2666 .\" HTML <a name="dfamatch"></a>
2667 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2668 .rs
2669 .sp
2670 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2671 .ti +5n
2672 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2673 .ti +5n
2674 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2675 .ti +5n
2676 .B int *\fIworkspace\fP, int \fIwscount\fP);
2677 .P
2678 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2679 a compiled pattern, using a matching algorithm that scans the subject string
2680 just once, and does not backtrack. This has different characteristics to the
2681 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2682 patterns are not supported. Nevertheless, there are times when this kind of
2683 matching can be useful. For a discussion of the two matching algorithms, and a
2684 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2685 .\" HREF
2686 \fBpcrematching\fP
2687 .\"
2688 documentation.
2689 .P
2690 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2691 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2692 different way, and this is described below. The other common arguments are used
2693 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2694 here.
2695 .P
2696 The two additional arguments provide workspace for the function. The workspace
2697 vector should contain at least 20 elements. It is used for keeping track of
2698 multiple paths through the pattern tree. More workspace will be needed for
2699 patterns and subjects where there are a lot of potential matches.
2700 .P
2701 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2702 .sp
2703 int rc;
2704 int ovector[10];
2705 int wspace[20];
2706 rc = pcre_dfa_exec(
2707 re, /* result of pcre_compile() */
2708 NULL, /* we didn't study the pattern */
2709 "some string", /* the subject string */
2710 11, /* the length of the subject string */
2711 0, /* start at offset 0 in the subject */
2712 0, /* default options */
2713 ovector, /* vector of integers for substring information */
2714 10, /* number of elements (NOT size in bytes) */
2715 wspace, /* working space vector */
2716 20); /* number of elements (NOT size in bytes) */
2717 .
2718 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2719 .rs
2720 .sp
2721 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2722 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2723 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2724 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2725 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2726 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2727 so their description is not repeated here.
2728 .sp
2729 PCRE_PARTIAL_HARD
2730 PCRE_PARTIAL_SOFT
2731 .sp
2732 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2733 details are slightly different. When PCRE_PARTIAL_HARD is set for
2734 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2735 is reached and there is still at least one matching possibility that requires
2736 additional characters. This happens even if some complete matches have also
2737 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2738 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2739 there have been no complete matches, but there is still at least one matching
2740 possibility. The portion of the string that was inspected when the longest
2741 partial match was found is set as the first matching string in both cases.
2742 There is a more detailed discussion of partial and multi-segment matching, with
2743 examples, in the
2744 .\" HREF
2745 \fBpcrepartial\fP
2746 .\"
2747 documentation.
2748 .sp
2749 PCRE_DFA_SHORTEST
2750 .sp
2751 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2752 soon as it has found one match. Because of the way the alternative algorithm
2753 works, this is necessarily the shortest possible match at the first possible
2754 matching point in the subject string.
2755 .sp
2756 PCRE_DFA_RESTART
2757 .sp
2758 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2759 again, with additional subject characters, and have it continue with the same
2760 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2761 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2762 before because data about the match so far is left in them after a partial
2763 match. There is more discussion of this facility in the
2764 .\" HREF
2765 \fBpcrepartial\fP
2766 .\"
2767 documentation.
2768 .
2769 .
2770 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2771 .rs
2772 .sp
2773 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2774 substring in the subject. Note, however, that all the matches from one run of
2775 the function start at the same point in the subject. The shorter matches are
2776 all initial substrings of the longer matches. For example, if the pattern
2777 .sp
2778 <.*>
2779 .sp
2780 is matched against the string
2781 .sp
2782 This is <something> <something else> <something further> no more
2783 .sp
2784 the three matched strings are
2785 .sp
2786 <something>
2787 <something> <something else>
2788 <something> <something else> <something further>
2789 .sp
2790 On success, the yield of the function is a number greater than zero, which is
2791 the number of matched substrings. The substrings themselves are returned in
2792 \fIovector\fP. Each string uses two elements; the first is the offset to the
2793 start, and the second is the offset to the end. In fact, all the strings have
2794 the same start offset. (Space could have been saved by giving this only once,
2795 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2796 returns data, even though the meaning of the strings is different.)
2797 .P
2798 The strings are returned in reverse order of length; that is, the longest
2799 matching string is given first. If there were too many matches to fit into
2800 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2801 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2802 the entire \fIovector\fP for returning matched strings.
2803 .
2804 .
2805 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2806 .rs
2807 .sp
2808 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2809 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2810 described
2811 .\" HTML <a href="#errorlist">
2812 .\" </a>
2813 above.
2814 .\"
2815 There are in addition the following errors that are specific to
2816 \fBpcre_dfa_exec()\fP:
2817 .sp
2818 PCRE_ERROR_DFA_UITEM (-16)
2819 .sp
2820 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2821 that it does not support, for instance, the use of \eC or a back reference.
2822 .sp
2823 PCRE_ERROR_DFA_UCOND (-17)
2824 .sp
2825 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2826 uses a back reference for the condition, or a test for recursion in a specific
2827 group. These are not supported.
2828 .sp
2829 PCRE_ERROR_DFA_UMLIMIT (-18)
2830 .sp
2831 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2832 block that contains a setting of the \fImatch_limit\fP or
2833 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2834 meaningless for DFA matching).
2835 .sp
2836 PCRE_ERROR_DFA_WSSIZE (-19)
2837 .sp
2838 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2839 \fIworkspace\fP vector.
2840 .sp
2841 PCRE_ERROR_DFA_RECURSE (-20)
2842 .sp
2843 When a recursive subpattern is processed, the matching function calls itself
2844 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2845 error is given if the output vector is not large enough. This should be
2846 extremely rare, as a vector of size 1000 is used.
2847 .sp
2848 PCRE_ERROR_DFA_BADRESTART (-30)
2849 .sp
2850 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2851 some plausibility checks are made on the contents of the workspace, which
2852 should contain data about the previous partial match. If any of these checks
2853 fail, this error is given.
2854 .
2855 .
2856 .SH "SEE ALSO"
2857 .rs
2858 .sp
2859 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2860 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2861 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2862 \fBpcrestack\fP(3).
2863 .
2864 .
2865 .SH AUTHOR
2866 .rs
2867 .sp
2868 .nf
2869 Philip Hazel
2870 University Computing Service
2871 Cambridge CB2 3QH, England.
2872 .fi
2873 .
2874 .
2875 .SH REVISION
2876 .rs
2877 .sp
2878 .nf
2879 Last updated: 26 April 2013
2880 Copyright (c) 1997-2013 University of Cambridge.
2881 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5