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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1313 - (show annotations)
Wed Apr 24 12:07:09 2013 UTC (6 years, 4 months ago) by ph10
File size: 123739 byte(s)
Code (but not yet documentation) for *LIMIT_MATCH and *LIMIT_RECURSION.
1 .TH PCREAPI 3 "05 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 .sp
1197 The "magic number" is placed at the start of each compiled pattern as an simple
1198 check against passing an arbitrary memory pointer. The endianness error can
1199 occur if a compiled pattern is saved and reloaded on a different host. Here is
1200 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1201 pattern:
1202 .sp
1203 int rc;
1204 size_t length;
1205 rc = pcre_fullinfo(
1206 re, /* result of pcre_compile() */
1207 sd, /* result of pcre_study(), or NULL */
1208 PCRE_INFO_SIZE, /* what is required */
1209 &length); /* where to put the data */
1210 .sp
1211 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1212 as follows:
1213 .sp
1214 PCRE_INFO_BACKREFMAX
1215 .sp
1216 Return the number of the highest back reference in the pattern. The fourth
1217 argument should point to an \fBint\fP variable. Zero is returned if there are
1218 no back references.
1219 .sp
1220 PCRE_INFO_CAPTURECOUNT
1221 .sp
1222 Return the number of capturing subpatterns in the pattern. The fourth argument
1223 should point to an \fBint\fP variable.
1224 .sp
1225 PCRE_INFO_DEFAULT_TABLES
1226 .sp
1227 Return a pointer to the internal default character tables within PCRE. The
1228 fourth argument should point to an \fBunsigned char *\fP variable. This
1229 information call is provided for internal use by the \fBpcre_study()\fP
1230 function. External callers can cause PCRE to use its internal tables by passing
1231 a NULL table pointer.
1232 .sp
1233 PCRE_INFO_FIRSTBYTE
1234 .sp
1235 Return information about the first data unit of any matched string, for a
1236 non-anchored pattern. (The name of this option refers to the 8-bit library,
1237 where data units are bytes.) The fourth argument should point to an \fBint\fP
1238 variable.
1239 .P
1240 If there is a fixed first value, for example, the letter "c" from a pattern
1241 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1242 value is always less than 256. In the 16-bit library the value can be up to
1243 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1244 .P
1245 If there is no fixed first value, and if either
1246 .sp
1247 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1248 starts with "^", or
1249 .sp
1250 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1251 (if it were set, the pattern would be anchored),
1252 .sp
1253 -1 is returned, indicating that the pattern matches only at the start of a
1254 subject string or after any newline within the string. Otherwise -2 is
1255 returned. For anchored patterns, -2 is returned.
1256 .P
1257 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1258 to return the full 32-bit range of the character, this value is deprecated;
1259 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1260 should be used.
1261 .sp
1262 PCRE_INFO_FIRSTTABLE
1263 .sp
1264 If the pattern was studied, and this resulted in the construction of a 256-bit
1265 table indicating a fixed set of values for the first data unit in any matching
1266 string, a pointer to the table is returned. Otherwise NULL is returned. The
1267 fourth argument should point to an \fBunsigned char *\fP variable.
1268 .sp
1269 PCRE_INFO_HASCRORLF
1270 .sp
1271 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1272 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1273 explicit match is either a literal CR or LF character, or \er or \en.
1274 .sp
1275 PCRE_INFO_JCHANGED
1276 .sp
1277 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1278 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1279 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1280 .sp
1281 PCRE_INFO_JIT
1282 .sp
1283 Return 1 if the pattern was studied with one of the JIT options, and
1284 just-in-time compiling was successful. The fourth argument should point to an
1285 \fBint\fP variable. A return value of 0 means that JIT support is not available
1286 in this version of PCRE, or that the pattern was not studied with a JIT option,
1287 or that the JIT compiler could not handle this particular pattern. See the
1288 .\" HREF
1289 \fBpcrejit\fP
1290 .\"
1291 documentation for details of what can and cannot be handled.
1292 .sp
1293 PCRE_INFO_JITSIZE
1294 .sp
1295 If the pattern was successfully studied with a JIT option, return the size of
1296 the JIT compiled code, otherwise return zero. The fourth argument should point
1297 to a \fBsize_t\fP variable.
1298 .sp
1299 PCRE_INFO_LASTLITERAL
1300 .sp
1301 Return the value of the rightmost literal data unit that must exist in any
1302 matched string, other than at its start, if such a value has been recorded. The
1303 fourth argument should point to an \fBint\fP variable. If there is no such
1304 value, -1 is returned. For anchored patterns, a last literal value is recorded
1305 only if it follows something of variable length. For example, for the pattern
1306 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1307 is -1.
1308 .P
1309 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1310 to return the full 32-bit range of the character, this value is deprecated;
1311 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1312 be used.
1313 .sp
1314 PCRE_INFO_MAXLOOKBEHIND
1315 .sp
1316 Return the number of characters (NB not bytes) in the longest lookbehind
1317 assertion in the pattern. This information is useful when doing multi-segment
1318 matching using the partial matching facilities. Note that the simple assertions
1319 \eb and \eB require a one-character lookbehind. \eA also registers a
1320 one-character lookbehind, though it does not actually inspect the previous
1321 character. This is to ensure that at least one character from the old segment
1322 is retained when a new segment is processed. Otherwise, if there are no
1323 lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1324 segment.
1325 .sp
1326 PCRE_INFO_MINLENGTH
1327 .sp
1328 If the pattern was studied and a minimum length for matching subject strings
1329 was computed, its value is returned. Otherwise the returned value is -1. The
1330 value is a number of characters, which in UTF-8 mode may be different from the
1331 number of bytes. The fourth argument should point to an \fBint\fP variable. A
1332 non-negative value is a lower bound to the length of any matching string. There
1333 may not be any strings of that length that do actually match, but every string
1334 that does match is at least that long.
1335 .sp
1336 PCRE_INFO_NAMECOUNT
1337 PCRE_INFO_NAMEENTRYSIZE
1338 PCRE_INFO_NAMETABLE
1339 .sp
1340 PCRE supports the use of named as well as numbered capturing parentheses. The
1341 names are just an additional way of identifying the parentheses, which still
1342 acquire numbers. Several convenience functions such as
1343 \fBpcre_get_named_substring()\fP are provided for extracting captured
1344 substrings by name. It is also possible to extract the data directly, by first
1345 converting the name to a number in order to access the correct pointers in the
1346 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1347 you need to use the name-to-number map, which is described by these three
1348 values.
1349 .P
1350 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1351 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1352 entry; both of these return an \fBint\fP value. The entry size depends on the
1353 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1354 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1355 the first two bytes of each entry are the number of the capturing parenthesis,
1356 most significant byte first. In the 16-bit library, the pointer points to
1357 16-bit data units, the first of which contains the parenthesis number.
1358 In the 32-bit library, the pointer points to 32-bit data units, the first of
1359 which contains the parenthesis number. The rest
1360 of the entry is the corresponding name, zero terminated.
1361 .P
1362 The names are in alphabetical order. Duplicate names may appear if (?| is used
1363 to create multiple groups with the same number, as described in the
1364 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1365 .\" </a>
1366 section on duplicate subpattern numbers
1367 .\"
1368 in the
1369 .\" HREF
1370 \fBpcrepattern\fP
1371 .\"
1372 page. Duplicate names for subpatterns with different numbers are permitted only
1373 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1374 table in the order in which they were found in the pattern. In the absence of
1375 (?| this is the order of increasing number; when (?| is used this is not
1376 necessarily the case because later subpatterns may have lower numbers.
1377 .P
1378 As a simple example of the name/number table, consider the following pattern
1379 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1380 space - including newlines - is ignored):
1381 .sp
1382 .\" JOIN
1383 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1384 (?<month>\ed\ed) - (?<day>\ed\ed) )
1385 .sp
1386 There are four named subpatterns, so the table has four entries, and each entry
1387 in the table is eight bytes long. The table is as follows, with non-printing
1388 bytes shows in hexadecimal, and undefined bytes shown as ??:
1389 .sp
1390 00 01 d a t e 00 ??
1391 00 05 d a y 00 ?? ??
1392 00 04 m o n t h 00
1393 00 02 y e a r 00 ??
1394 .sp
1395 When writing code to extract data from named subpatterns using the
1396 name-to-number map, remember that the length of the entries is likely to be
1397 different for each compiled pattern.
1398 .sp
1399 PCRE_INFO_OKPARTIAL
1400 .sp
1401 Return 1 if the pattern can be used for partial matching with
1402 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1403 \fBint\fP variable. From release 8.00, this always returns 1, because the
1404 restrictions that previously applied to partial matching have been lifted. The
1405 .\" HREF
1406 \fBpcrepartial\fP
1407 .\"
1408 documentation gives details of partial matching.
1409 .sp
1410 PCRE_INFO_OPTIONS
1411 .sp
1412 Return a copy of the options with which the pattern was compiled. The fourth
1413 argument should point to an \fBunsigned long int\fP variable. These option bits
1414 are those specified in the call to \fBpcre_compile()\fP, modified by any
1415 top-level option settings at the start of the pattern itself. In other words,
1416 they are the options that will be in force when matching starts. For example,
1417 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1418 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1419 .P
1420 A pattern is automatically anchored by PCRE if all of its top-level
1421 alternatives begin with one of the following:
1422 .sp
1423 ^ unless PCRE_MULTILINE is set
1424 \eA always
1425 \eG always
1426 .\" JOIN
1427 .* if PCRE_DOTALL is set and there are no back
1428 references to the subpattern in which .* appears
1429 .sp
1430 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1431 \fBpcre_fullinfo()\fP.
1432 .sp
1433 PCRE_INFO_SIZE
1434 .sp
1435 Return the size of the compiled pattern in bytes (for both libraries). The
1436 fourth argument should point to a \fBsize_t\fP variable. This value does not
1437 include the size of the \fBpcre\fP structure that is returned by
1438 \fBpcre_compile()\fP. The value that is passed as the argument to
1439 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1440 place the compiled data is the value returned by this option plus the size of
1441 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1442 does not alter the value returned by this option.
1443 .sp
1444 PCRE_INFO_STUDYSIZE
1445 .sp
1446 Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1447 field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1448 study data, zero is returned. The fourth argument should point to a
1449 \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1450 to record information that will speed up matching (see the section entitled
1451 .\" HTML <a href="#studyingapattern">
1452 .\" </a>
1453 "Studying a pattern"
1454 .\"
1455 above). The format of the \fIstudy_data\fP block is private, but its length
1456 is made available via this option so that it can be saved and restored (see the
1457 .\" HREF
1458 \fBpcreprecompile\fP
1459 .\"
1460 documentation for details).
1461 .sp
1462 PCRE_INFO_FIRSTCHARACTERFLAGS
1463 .sp
1464 Return information about the first data unit of any matched string, for a
1465 non-anchored pattern. The fourth argument should point to an \fBint\fP
1466 variable.
1467 .P
1468 If there is a fixed first value, for example, the letter "c" from a pattern
1469 such as (cat|cow|coyote), 1 is returned, and the character value can be
1470 retrieved using PCRE_INFO_FIRSTCHARACTER.
1471 .P
1472 If there is no fixed first value, and if either
1473 .sp
1474 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1475 starts with "^", or
1476 .sp
1477 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1478 (if it were set, the pattern would be anchored),
1479 .sp
1480 2 is returned, indicating that the pattern matches only at the start of a
1481 subject string or after any newline within the string. Otherwise 0 is
1482 returned. For anchored patterns, 0 is returned.
1483 .sp
1484 PCRE_INFO_FIRSTCHARACTER
1485 .sp
1486 Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1487 returned 1; otherwise returns 0. The fourth argument should point to an
1488 \fBuint_t\fP variable.
1489 .P
1490 In the 8-bit library, the value is always less than 256. In the 16-bit library
1491 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1492 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1493 .P
1494 If there is no fixed first value, and if either
1495 .sp
1496 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1497 starts with "^", or
1498 .sp
1499 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1500 (if it were set, the pattern would be anchored),
1501 .sp
1502 -1 is returned, indicating that the pattern matches only at the start of a
1503 subject string or after any newline within the string. Otherwise -2 is
1504 returned. For anchored patterns, -2 is returned.
1505 .sp
1506 PCRE_INFO_REQUIREDCHARFLAGS
1507 .sp
1508 Returns 1 if there is a rightmost literal data unit that must exist in any
1509 matched string, other than at its start. The fourth argument should point to
1510 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1511 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1512 .P
1513 For anchored patterns, a last literal value is recorded only if it follows
1514 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1515 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1516 /^a\edz\ed/ the returned value is 0.
1517 .sp
1518 PCRE_INFO_REQUIREDCHAR
1519 .sp
1520 Return the value of the rightmost literal data unit that must exist in any
1521 matched string, other than at its start, if such a value has been recorded. The
1522 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1523 value, 0 is returned.
1524 .
1525 .
1526 .SH "REFERENCE COUNTS"
1527 .rs
1528 .sp
1529 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1530 .PP
1531 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1532 data block that contains a compiled pattern. It is provided for the benefit of
1533 applications that operate in an object-oriented manner, where different parts
1534 of the application may be using the same compiled pattern, but you want to free
1535 the block when they are all done.
1536 .P
1537 When a pattern is compiled, the reference count field is initialized to zero.
1538 It is changed only by calling this function, whose action is to add the
1539 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1540 function is the new value. However, the value of the count is constrained to
1541 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1542 it is forced to the appropriate limit value.
1543 .P
1544 Except when it is zero, the reference count is not correctly preserved if a
1545 pattern is compiled on one host and then transferred to a host whose byte-order
1546 is different. (This seems a highly unlikely scenario.)
1547 .
1548 .
1549 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1550 .rs
1551 .sp
1552 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1553 .ti +5n
1554 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1555 .ti +5n
1556 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1557 .P
1558 The function \fBpcre_exec()\fP is called to match a subject string against a
1559 compiled pattern, which is passed in the \fIcode\fP argument. If the
1560 pattern was studied, the result of the study should be passed in the
1561 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1562 and \fIextra\fP arguments as many times as you like, in order to match
1563 different subject strings with the same pattern.
1564 .P
1565 This function is the main matching facility of the library, and it operates in
1566 a Perl-like manner. For specialist use there is also an alternative matching
1567 function, which is described
1568 .\" HTML <a href="#dfamatch">
1569 .\" </a>
1570 below
1571 .\"
1572 in the section about the \fBpcre_dfa_exec()\fP function.
1573 .P
1574 In most applications, the pattern will have been compiled (and optionally
1575 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1576 possible to save compiled patterns and study data, and then use them later
1577 in different processes, possibly even on different hosts. For a discussion
1578 about this, see the
1579 .\" HREF
1580 \fBpcreprecompile\fP
1581 .\"
1582 documentation.
1583 .P
1584 Here is an example of a simple call to \fBpcre_exec()\fP:
1585 .sp
1586 int rc;
1587 int ovector[30];
1588 rc = pcre_exec(
1589 re, /* result of pcre_compile() */
1590 NULL, /* we didn't study the pattern */
1591 "some string", /* the subject string */
1592 11, /* the length of the subject string */
1593 0, /* start at offset 0 in the subject */
1594 0, /* default options */
1595 ovector, /* vector of integers for substring information */
1596 30); /* number of elements (NOT size in bytes) */
1597 .
1598 .
1599 .\" HTML <a name="extradata"></a>
1600 .SS "Extra data for \fBpcre_exec()\fR"
1601 .rs
1602 .sp
1603 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1604 data block. The \fBpcre_study()\fP function returns such a block (when it
1605 doesn't return NULL), but you can also create one for yourself, and pass
1606 additional information in it. The \fBpcre_extra\fP block contains the following
1607 fields (not necessarily in this order):
1608 .sp
1609 unsigned long int \fIflags\fP;
1610 void *\fIstudy_data\fP;
1611 void *\fIexecutable_jit\fP;
1612 unsigned long int \fImatch_limit\fP;
1613 unsigned long int \fImatch_limit_recursion\fP;
1614 void *\fIcallout_data\fP;
1615 const unsigned char *\fItables\fP;
1616 unsigned char **\fImark\fP;
1617 .sp
1618 In the 16-bit version of this structure, the \fImark\fP field has type
1619 "PCRE_UCHAR16 **".
1620 .sp
1621 In the 32-bit version of this structure, the \fImark\fP field has type
1622 "PCRE_UCHAR32 **".
1623 .P
1624 The \fIflags\fP field is used to specify which of the other fields are set. The
1625 flag bits are:
1626 .sp
1627 PCRE_EXTRA_CALLOUT_DATA
1628 PCRE_EXTRA_EXECUTABLE_JIT
1629 PCRE_EXTRA_MARK
1630 PCRE_EXTRA_MATCH_LIMIT
1631 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1632 PCRE_EXTRA_STUDY_DATA
1633 PCRE_EXTRA_TABLES
1634 .sp
1635 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1636 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1637 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1638 should not set these yourself, but you may add to the block by setting other
1639 fields and their corresponding flag bits.
1640 .P
1641 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1642 vast amount of resources when running patterns that are not going to match,
1643 but which have a very large number of possibilities in their search trees. The
1644 classic example is a pattern that uses nested unlimited repeats.
1645 .P
1646 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1647 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1648 imposed on the number of times this function is called during a match, which
1649 has the effect of limiting the amount of backtracking that can take place. For
1650 patterns that are not anchored, the count restarts from zero for each position
1651 in the subject string.
1652 .P
1653 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1654 with a JIT option, the way that the matching is executed is entirely different.
1655 However, there is still the possibility of runaway matching that goes on for a
1656 very long time, and so the \fImatch_limit\fP value is also used in this case
1657 (but in a different way) to limit how long the matching can continue.
1658 .P
1659 The default value for the limit can be set when PCRE is built; the default
1660 default is 10 million, which handles all but the most extreme cases. You can
1661 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1662 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1663 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1664 PCRE_ERROR_MATCHLIMIT.
1665 .P
1666 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1667 instead of limiting the total number of times that \fBmatch()\fP is called, it
1668 limits the depth of recursion. The recursion depth is a smaller number than the
1669 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1670 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1671 .P
1672 Limiting the recursion depth limits the amount of machine stack that can be
1673 used, or, when PCRE has been compiled to use memory on the heap instead of the
1674 stack, the amount of heap memory that can be used. This limit is not relevant,
1675 and is ignored, when matching is done using JIT compiled code.
1676 .P
1677 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1678 built; the default default is the same value as the default for
1679 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1680 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1681 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1682 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1683 .P
1684 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1685 and is described in the
1686 .\" HREF
1687 \fBpcrecallout\fP
1688 .\"
1689 documentation.
1690 .P
1691 The \fItables\fP field is used to pass a character tables pointer to
1692 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1693 pattern. A non-NULL value is stored with the compiled pattern only if custom
1694 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1695 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1696 internal tables to be used. This facility is helpful when re-using patterns
1697 that have been saved after compiling with an external set of tables, because
1698 the external tables might be at a different address when \fBpcre_exec()\fP is
1699 called. See the
1700 .\" HREF
1701 \fBpcreprecompile\fP
1702 .\"
1703 documentation for a discussion of saving compiled patterns for later use.
1704 .P
1705 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1706 be set to point to a suitable variable. If the pattern contains any
1707 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1708 a name to pass back, a pointer to the name string (zero terminated) is placed
1709 in the variable pointed to by the \fImark\fP field. The names are within the
1710 compiled pattern; if you wish to retain such a name you must copy it before
1711 freeing the memory of a compiled pattern. If there is no name to pass back, the
1712 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1713 backtracking control verbs, see the section entitled
1714 .\" HTML <a href="pcrepattern#backtrackcontrol">
1715 .\" </a>
1716 "Backtracking control"
1717 .\"
1718 in the
1719 .\" HREF
1720 \fBpcrepattern\fP
1721 .\"
1722 documentation.
1723 .
1724 .
1725 .\" HTML <a name="execoptions"></a>
1726 .SS "Option bits for \fBpcre_exec()\fP"
1727 .rs
1728 .sp
1729 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1730 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1731 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1732 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1733 PCRE_PARTIAL_SOFT.
1734 .P
1735 If the pattern was successfully studied with one of the just-in-time (JIT)
1736 compile options, the only supported options for JIT execution are
1737 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1738 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1739 unsupported option is used, JIT execution is disabled and the normal
1740 interpretive code in \fBpcre_exec()\fP is run.
1741 .sp
1742 PCRE_ANCHORED
1743 .sp
1744 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1745 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1746 to be anchored by virtue of its contents, it cannot be made unachored at
1747 matching time.
1748 .sp
1749 PCRE_BSR_ANYCRLF
1750 PCRE_BSR_UNICODE
1751 .sp
1752 These options (which are mutually exclusive) control what the \eR escape
1753 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1754 match any Unicode newline sequence. These options override the choice that was
1755 made or defaulted when the pattern was compiled.
1756 .sp
1757 PCRE_NEWLINE_CR
1758 PCRE_NEWLINE_LF
1759 PCRE_NEWLINE_CRLF
1760 PCRE_NEWLINE_ANYCRLF
1761 PCRE_NEWLINE_ANY
1762 .sp
1763 These options override the newline definition that was chosen or defaulted when
1764 the pattern was compiled. For details, see the description of
1765 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1766 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1767 the way the match position is advanced after a match failure for an unanchored
1768 pattern.
1769 .P
1770 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1771 match attempt for an unanchored pattern fails when the current position is at a
1772 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1773 characters, the match position is advanced by two characters instead of one, in
1774 other words, to after the CRLF.
1775 .P
1776 The above rule is a compromise that makes the most common cases work as
1777 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1778 set), it does not match the string "\er\enA" because, after failing at the
1779 start, it skips both the CR and the LF before retrying. However, the pattern
1780 [\er\en]A does match that string, because it contains an explicit CR or LF
1781 reference, and so advances only by one character after the first failure.
1782 .P
1783 An explicit match for CR of LF is either a literal appearance of one of those
1784 characters, or one of the \er or \en escape sequences. Implicit matches such as
1785 [^X] do not count, nor does \es (which includes CR and LF in the characters
1786 that it matches).
1787 .P
1788 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1789 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1790 .sp
1791 PCRE_NOTBOL
1792 .sp
1793 This option specifies that first character of the subject string is not the
1794 beginning of a line, so the circumflex metacharacter should not match before
1795 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1796 never to match. This option affects only the behaviour of the circumflex
1797 metacharacter. It does not affect \eA.
1798 .sp
1799 PCRE_NOTEOL
1800 .sp
1801 This option specifies that the end of the subject string is not the end of a
1802 line, so the dollar metacharacter should not match it nor (except in multiline
1803 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1804 compile time) causes dollar never to match. This option affects only the
1805 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1806 .sp
1807 PCRE_NOTEMPTY
1808 .sp
1809 An empty string is not considered to be a valid match if this option is set. If
1810 there are alternatives in the pattern, they are tried. If all the alternatives
1811 match the empty string, the entire match fails. For example, if the pattern
1812 .sp
1813 a?b?
1814 .sp
1815 is applied to a string not beginning with "a" or "b", it matches an empty
1816 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1817 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1818 .sp
1819 PCRE_NOTEMPTY_ATSTART
1820 .sp
1821 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1822 the start of the subject is permitted. If the pattern is anchored, such a match
1823 can occur only if the pattern contains \eK.
1824 .P
1825 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1826 does make a special case of a pattern match of the empty string within its
1827 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1828 emulate Perl's behaviour after matching a null string by first trying the match
1829 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1830 if that fails, by advancing the starting offset (see below) and trying an
1831 ordinary match again. There is some code that demonstrates how to do this in
1832 the
1833 .\" HREF
1834 \fBpcredemo\fP
1835 .\"
1836 sample program. In the most general case, you have to check to see if the
1837 newline convention recognizes CRLF as a newline, and if so, and the current
1838 character is CR followed by LF, advance the starting offset by two characters
1839 instead of one.
1840 .sp
1841 PCRE_NO_START_OPTIMIZE
1842 .sp
1843 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1844 a match, in order to speed up the process. For example, if it is known that an
1845 unanchored match must start with a specific character, it searches the subject
1846 for that character, and fails immediately if it cannot find it, without
1847 actually running the main matching function. This means that a special item
1848 such as (*COMMIT) at the start of a pattern is not considered until after a
1849 suitable starting point for the match has been found. Also, when callouts or
1850 (*MARK) items are in use, these "start-up" optimizations can cause them to be
1851 skipped if the pattern is never actually used. The start-up optimizations are
1852 in effect a pre-scan of the subject that takes place before the pattern is run.
1853 .P
1854 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1855 causing performance to suffer, but ensuring that in cases where the result is
1856 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1857 are considered at every possible starting position in the subject string. If
1858 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1859 time. The use of PCRE_NO_START_OPTIMIZE at matching time (that is, passing it
1860 to \fBpcre_exec()\fP) disables JIT execution; in this situation, matching is
1861 always done using interpretively.
1862 .P
1863 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1864 Consider the pattern
1865 .sp
1866 (*COMMIT)ABC
1867 .sp
1868 When this is compiled, PCRE records the fact that a match must start with the
1869 character "A". Suppose the subject string is "DEFABC". The start-up
1870 optimization scans along the subject, finds "A" and runs the first match
1871 attempt from there. The (*COMMIT) item means that the pattern must match the
1872 current starting position, which in this case, it does. However, if the same
1873 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1874 subject string does not happen. The first match attempt is run starting from
1875 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1876 the overall result is "no match". If the pattern is studied, more start-up
1877 optimizations may be used. For example, a minimum length for the subject may be
1878 recorded. Consider the pattern
1879 .sp
1880 (*MARK:A)(X|Y)
1881 .sp
1882 The minimum length for a match is one character. If the subject is "ABC", there
1883 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1884 If the pattern is studied, the final attempt does not take place, because PCRE
1885 knows that the subject is too short, and so the (*MARK) is never encountered.
1886 In this case, studying the pattern does not affect the overall match result,
1887 which is still "no match", but it does affect the auxiliary information that is
1888 returned.
1889 .sp
1890 PCRE_NO_UTF8_CHECK
1891 .sp
1892 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1893 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1894 The entire string is checked before any other processing takes place. The value
1895 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1896 UTF-8 character. There is a discussion about the
1897 .\" HTML <a href="pcreunicode.html#utf8strings">
1898 .\" </a>
1899 validity of UTF-8 strings
1900 .\"
1901 in the
1902 .\" HREF
1903 \fBpcreunicode\fP
1904 .\"
1905 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1906 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1907 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1908 cases, information about the precise nature of the error may also be returned
1909 (see the descriptions of these errors in the section entitled \fIError return
1910 values from\fP \fBpcre_exec()\fP
1911 .\" HTML <a href="#errorlist">
1912 .\" </a>
1913 below).
1914 .\"
1915 If \fIstartoffset\fP contains a value that does not point to the start of a
1916 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1917 returned.
1918 .P
1919 If you already know that your subject is valid, and you want to skip these
1920 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1921 calling \fBpcre_exec()\fP. You might want to do this for the second and
1922 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1923 all the matches in a single subject string. However, you should be sure that
1924 the value of \fIstartoffset\fP points to the start of a character (or the end
1925 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1926 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1927 undefined. Your program may crash.
1928 .sp
1929 PCRE_PARTIAL_HARD
1930 PCRE_PARTIAL_SOFT
1931 .sp
1932 These options turn on the partial matching feature. For backwards
1933 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1934 occurs if the end of the subject string is reached successfully, but there are
1935 not enough subject characters to complete the match. If this happens when
1936 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1937 testing any remaining alternatives. Only if no complete match can be found is
1938 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1939 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1940 but only if no complete match can be found.
1941 .P
1942 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1943 partial match is found, \fBpcre_exec()\fP immediately returns
1944 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1945 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1946 important that an alternative complete match.
1947 .P
1948 In both cases, the portion of the string that was inspected when the partial
1949 match was found is set as the first matching string. There is a more detailed
1950 discussion of partial and multi-segment matching, with examples, in the
1951 .\" HREF
1952 \fBpcrepartial\fP
1953 .\"
1954 documentation.
1955 .
1956 .
1957 .SS "The string to be matched by \fBpcre_exec()\fP"
1958 .rs
1959 .sp
1960 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1961 \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1962 in \fIstartoffset\fP. If this is negative or greater than the length of the
1963 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1964 offset is zero, the search for a match starts at the beginning of the subject,
1965 and this is by far the most common case. In UTF-8 mode, the byte offset must
1966 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1967 pattern string, the subject may contain binary zero bytes.
1968 .P
1969 A non-zero starting offset is useful when searching for another match in the
1970 same subject by calling \fBpcre_exec()\fP again after a previous success.
1971 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1972 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1973 lookbehind. For example, consider the pattern
1974 .sp
1975 \eBiss\eB
1976 .sp
1977 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1978 the current position in the subject is not a word boundary.) When applied to
1979 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1980 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1981 subject, namely "issipi", it does not match, because \eB is always false at the
1982 start of the subject, which is deemed to be a word boundary. However, if
1983 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1984 set to 4, it finds the second occurrence of "iss" because it is able to look
1985 behind the starting point to discover that it is preceded by a letter.
1986 .P
1987 Finding all the matches in a subject is tricky when the pattern can match an
1988 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1989 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1990 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1991 and trying an ordinary match again. There is some code that demonstrates how to
1992 do this in the
1993 .\" HREF
1994 \fBpcredemo\fP
1995 .\"
1996 sample program. In the most general case, you have to check to see if the
1997 newline convention recognizes CRLF as a newline, and if so, and the current
1998 character is CR followed by LF, advance the starting offset by two characters
1999 instead of one.
2000 .P
2001 If a non-zero starting offset is passed when the pattern is anchored, one
2002 attempt to match at the given offset is made. This can only succeed if the
2003 pattern does not require the match to be at the start of the subject.
2004 .
2005 .
2006 .SS "How \fBpcre_exec()\fP returns captured substrings"
2007 .rs
2008 .sp
2009 In general, a pattern matches a certain portion of the subject, and in
2010 addition, further substrings from the subject may be picked out by parts of the
2011 pattern. Following the usage in Jeffrey Friedl's book, this is called
2012 "capturing" in what follows, and the phrase "capturing subpattern" is used for
2013 a fragment of a pattern that picks out a substring. PCRE supports several other
2014 kinds of parenthesized subpattern that do not cause substrings to be captured.
2015 .P
2016 Captured substrings are returned to the caller via a vector of integers whose
2017 address is passed in \fIovector\fP. The number of elements in the vector is
2018 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2019 argument is NOT the size of \fIovector\fP in bytes.
2020 .P
2021 The first two-thirds of the vector is used to pass back captured substrings,
2022 each substring using a pair of integers. The remaining third of the vector is
2023 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2024 and is not available for passing back information. The number passed in
2025 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2026 rounded down.
2027 .P
2028 When a match is successful, information about captured substrings is returned
2029 in pairs of integers, starting at the beginning of \fIovector\fP, and
2030 continuing up to two-thirds of its length at the most. The first element of
2031 each pair is set to the byte offset of the first character in a substring, and
2032 the second is set to the byte offset of the first character after the end of a
2033 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
2034 mode. They are not character counts.
2035 .P
2036 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2037 portion of the subject string matched by the entire pattern. The next pair is
2038 used for the first capturing subpattern, and so on. The value returned by
2039 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2040 For example, if two substrings have been captured, the returned value is 3. If
2041 there are no capturing subpatterns, the return value from a successful match is
2042 1, indicating that just the first pair of offsets has been set.
2043 .P
2044 If a capturing subpattern is matched repeatedly, it is the last portion of the
2045 string that it matched that is returned.
2046 .P
2047 If the vector is too small to hold all the captured substring offsets, it is
2048 used as far as possible (up to two-thirds of its length), and the function
2049 returns a value of zero. If neither the actual string matched nor any captured
2050 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2051 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2052 back references and the \fIovector\fP is not big enough to remember the related
2053 substrings, PCRE has to get additional memory for use during matching. Thus it
2054 is usually advisable to supply an \fIovector\fP of reasonable size.
2055 .P
2056 There are some cases where zero is returned (indicating vector overflow) when
2057 in fact the vector is exactly the right size for the final match. For example,
2058 consider the pattern
2059 .sp
2060 (a)(?:(b)c|bd)
2061 .sp
2062 If a vector of 6 elements (allowing for only 1 captured substring) is given
2063 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2064 captured string, thereby recording a vector overflow, before failing to match
2065 "c" and backing up to try the second alternative. The zero return, however,
2066 does correctly indicate that the maximum number of slots (namely 2) have been
2067 filled. In similar cases where there is temporary overflow, but the final
2068 number of used slots is actually less than the maximum, a non-zero value is
2069 returned.
2070 .P
2071 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2072 subpatterns there are in a compiled pattern. The smallest size for
2073 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2074 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2075 .P
2076 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2077 the subject when subpattern \fIn\fP has not been used at all. For example, if
2078 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2079 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2080 happens, both values in the offset pairs corresponding to unused subpatterns
2081 are set to -1.
2082 .P
2083 Offset values that correspond to unused subpatterns at the end of the
2084 expression are also set to -1. For example, if the string "abc" is matched
2085 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2086 return from the function is 2, because the highest used capturing subpattern
2087 number is 1, and the offsets for for the second and third capturing subpatterns
2088 (assuming the vector is large enough, of course) are set to -1.
2089 .P
2090 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2091 correspond to capturing parentheses in the pattern are never changed. That is,
2092 if a pattern contains \fIn\fP capturing parentheses, no more than
2093 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2094 elements (in the first two-thirds) retain whatever values they previously had.
2095 .P
2096 Some convenience functions are provided for extracting the captured substrings
2097 as separate strings. These are described below.
2098 .
2099 .
2100 .\" HTML <a name="errorlist"></a>
2101 .SS "Error return values from \fBpcre_exec()\fP"
2102 .rs
2103 .sp
2104 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2105 defined in the header file:
2106 .sp
2107 PCRE_ERROR_NOMATCH (-1)
2108 .sp
2109 The subject string did not match the pattern.
2110 .sp
2111 PCRE_ERROR_NULL (-2)
2112 .sp
2113 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2114 NULL and \fIovecsize\fP was not zero.
2115 .sp
2116 PCRE_ERROR_BADOPTION (-3)
2117 .sp
2118 An unrecognized bit was set in the \fIoptions\fP argument.
2119 .sp
2120 PCRE_ERROR_BADMAGIC (-4)
2121 .sp
2122 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2123 the case when it is passed a junk pointer and to detect when a pattern that was
2124 compiled in an environment of one endianness is run in an environment with the
2125 other endianness. This is the error that PCRE gives when the magic number is
2126 not present.
2127 .sp
2128 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2129 .sp
2130 While running the pattern match, an unknown item was encountered in the
2131 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2132 of the compiled pattern.
2133 .sp
2134 PCRE_ERROR_NOMEMORY (-6)
2135 .sp
2136 If a pattern contains back references, but the \fIovector\fP that is passed to
2137 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2138 gets a block of memory at the start of matching to use for this purpose. If the
2139 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2140 automatically freed at the end of matching.
2141 .P
2142 This error is also given if \fBpcre_stack_malloc()\fP fails in
2143 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2144 \fB--disable-stack-for-recursion\fP.
2145 .sp
2146 PCRE_ERROR_NOSUBSTRING (-7)
2147 .sp
2148 This error is used by the \fBpcre_copy_substring()\fP,
2149 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2150 below). It is never returned by \fBpcre_exec()\fP.
2151 .sp
2152 PCRE_ERROR_MATCHLIMIT (-8)
2153 .sp
2154 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2155 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2156 above.
2157 .sp
2158 PCRE_ERROR_CALLOUT (-9)
2159 .sp
2160 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2161 use by callout functions that want to yield a distinctive error code. See the
2162 .\" HREF
2163 \fBpcrecallout\fP
2164 .\"
2165 documentation for details.
2166 .sp
2167 PCRE_ERROR_BADUTF8 (-10)
2168 .sp
2169 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2170 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2171 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2172 UTF-8 character is placed in the first element, and a reason code is placed in
2173 the second element. The reason codes are listed in the
2174 .\" HTML <a href="#badutf8reasons">
2175 .\" </a>
2176 following section.
2177 .\"
2178 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2179 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2180 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2181 .sp
2182 PCRE_ERROR_BADUTF8_OFFSET (-11)
2183 .sp
2184 The UTF-8 byte sequence that was passed as a subject was checked and found to
2185 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2186 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2187 end of the subject.
2188 .sp
2189 PCRE_ERROR_PARTIAL (-12)
2190 .sp
2191 The subject string did not match, but it did match partially. See the
2192 .\" HREF
2193 \fBpcrepartial\fP
2194 .\"
2195 documentation for details of partial matching.
2196 .sp
2197 PCRE_ERROR_BADPARTIAL (-13)
2198 .sp
2199 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2200 option was used with a compiled pattern containing items that were not
2201 supported for partial matching. From release 8.00 onwards, there are no
2202 restrictions on partial matching.
2203 .sp
2204 PCRE_ERROR_INTERNAL (-14)
2205 .sp
2206 An unexpected internal error has occurred. This error could be caused by a bug
2207 in PCRE or by overwriting of the compiled pattern.
2208 .sp
2209 PCRE_ERROR_BADCOUNT (-15)
2210 .sp
2211 This error is given if the value of the \fIovecsize\fP argument is negative.
2212 .sp
2213 PCRE_ERROR_RECURSIONLIMIT (-21)
2214 .sp
2215 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2216 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2217 description above.
2218 .sp
2219 PCRE_ERROR_BADNEWLINE (-23)
2220 .sp
2221 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2222 .sp
2223 PCRE_ERROR_BADOFFSET (-24)
2224 .sp
2225 The value of \fIstartoffset\fP was negative or greater than the length of the
2226 subject, that is, the value in \fIlength\fP.
2227 .sp
2228 PCRE_ERROR_SHORTUTF8 (-25)
2229 .sp
2230 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2231 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2232 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2233 fact sufficient to detect this case, but this special error code for
2234 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2235 retained for backwards compatibility.
2236 .sp
2237 PCRE_ERROR_RECURSELOOP (-26)
2238 .sp
2239 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2240 the pattern. Specifically, it means that either the whole pattern or a
2241 subpattern has been called recursively for the second time at the same position
2242 in the subject string. Some simple patterns that might do this are detected and
2243 faulted at compile time, but more complicated cases, in particular mutual
2244 recursions between two different subpatterns, cannot be detected until run
2245 time.
2246 .sp
2247 PCRE_ERROR_JIT_STACKLIMIT (-27)
2248 .sp
2249 This error is returned when a pattern that was successfully studied using a
2250 JIT compile option is being matched, but the memory available for the
2251 just-in-time processing stack is not large enough. See the
2252 .\" HREF
2253 \fBpcrejit\fP
2254 .\"
2255 documentation for more details.
2256 .sp
2257 PCRE_ERROR_BADMODE (-28)
2258 .sp
2259 This error is given if a pattern that was compiled by the 8-bit library is
2260 passed to a 16-bit or 32-bit library function, or vice versa.
2261 .sp
2262 PCRE_ERROR_BADENDIANNESS (-29)
2263 .sp
2264 This error is given if a pattern that was compiled and saved is reloaded on a
2265 host with different endianness. The utility function
2266 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2267 so that it runs on the new host.
2268 .sp
2269 PCRE_ERROR_JIT_BADOPTION
2270 .sp
2271 This error is returned when a pattern that was successfully studied using a JIT
2272 compile option is being matched, but the matching mode (partial or complete
2273 match) does not correspond to any JIT compilation mode. When the JIT fast path
2274 function is used, this error may be also given for invalid options. See the
2275 .\" HREF
2276 \fBpcrejit\fP
2277 .\"
2278 documentation for more details.
2279 .sp
2280 PCRE_ERROR_BADLENGTH (-32)
2281 .sp
2282 This error is given if \fBpcre_exec()\fP is called with a negative value for
2283 the \fIlength\fP argument.
2284 .P
2285 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2286 .
2287 .
2288 .\" HTML <a name="badutf8reasons"></a>
2289 .SS "Reason codes for invalid UTF-8 strings"
2290 .rs
2291 .sp
2292 This section applies only to the 8-bit library. The corresponding information
2293 for the 16-bit and 32-bit libraries is given in the
2294 .\" HREF
2295 \fBpcre16\fP
2296 .\"
2297 and
2298 .\" HREF
2299 \fBpcre32\fP
2300 .\"
2301 pages.
2302 .P
2303 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2304 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2305 least 2, the offset of the start of the invalid UTF-8 character is placed in
2306 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2307 in the second element (\fIovector[1]\fP). The reason codes are given names in
2308 the \fBpcre.h\fP header file:
2309 .sp
2310 PCRE_UTF8_ERR1
2311 PCRE_UTF8_ERR2
2312 PCRE_UTF8_ERR3
2313 PCRE_UTF8_ERR4
2314 PCRE_UTF8_ERR5
2315 .sp
2316 The string ends with a truncated UTF-8 character; the code specifies how many
2317 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2318 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2319 allows for up to 6 bytes, and this is checked first; hence the possibility of
2320 4 or 5 missing bytes.
2321 .sp
2322 PCRE_UTF8_ERR6
2323 PCRE_UTF8_ERR7
2324 PCRE_UTF8_ERR8
2325 PCRE_UTF8_ERR9
2326 PCRE_UTF8_ERR10
2327 .sp
2328 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2329 character do not have the binary value 0b10 (that is, either the most
2330 significant bit is 0, or the next bit is 1).
2331 .sp
2332 PCRE_UTF8_ERR11
2333 PCRE_UTF8_ERR12
2334 .sp
2335 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2336 these code points are excluded by RFC 3629.
2337 .sp
2338 PCRE_UTF8_ERR13
2339 .sp
2340 A 4-byte character has a value greater than 0x10fff; these code points are
2341 excluded by RFC 3629.
2342 .sp
2343 PCRE_UTF8_ERR14
2344 .sp
2345 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2346 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2347 from UTF-8.
2348 .sp
2349 PCRE_UTF8_ERR15
2350 PCRE_UTF8_ERR16
2351 PCRE_UTF8_ERR17
2352 PCRE_UTF8_ERR18
2353 PCRE_UTF8_ERR19
2354 .sp
2355 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2356 value that can be represented by fewer bytes, which is invalid. For example,
2357 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2358 one byte.
2359 .sp
2360 PCRE_UTF8_ERR20
2361 .sp
2362 The two most significant bits of the first byte of a character have the binary
2363 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2364 byte can only validly occur as the second or subsequent byte of a multi-byte
2365 character.
2366 .sp
2367 PCRE_UTF8_ERR21
2368 .sp
2369 The first byte of a character has the value 0xfe or 0xff. These values can
2370 never occur in a valid UTF-8 string.
2371 .sp
2372 PCRE_UTF8_ERR22
2373 .sp
2374 This error code was formerly used when the presence of a so-called
2375 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2376 such characters should not cause a string to be rejected, and so this code is
2377 no longer in use and is never returned.
2378 .
2379 .
2380 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2381 .rs
2382 .sp
2383 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2384 .ti +5n
2385 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2386 .ti +5n
2387 .B int \fIbuffersize\fP);
2388 .PP
2389 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2390 .ti +5n
2391 .B int \fIstringcount\fP, int \fIstringnumber\fP,
2392 .ti +5n
2393 .B const char **\fIstringptr\fP);
2394 .PP
2395 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2396 .ti +5n
2397 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2398 .PP
2399 Captured substrings can be accessed directly by using the offsets returned by
2400 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2401 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2402 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2403 as new, separate, zero-terminated strings. These functions identify substrings
2404 by number. The next section describes functions for extracting named
2405 substrings.
2406 .P
2407 A substring that contains a binary zero is correctly extracted and has a
2408 further zero added on the end, but the result is not, of course, a C string.
2409 However, you can process such a string by referring to the length that is
2410 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2411 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2412 for handling strings containing binary zeros, because the end of the final
2413 string is not independently indicated.
2414 .P
2415 The first three arguments are the same for all three of these functions:
2416 \fIsubject\fP is the subject string that has just been successfully matched,
2417 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2418 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2419 captured by the match, including the substring that matched the entire regular
2420 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2421 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2422 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2423 number of elements in the vector divided by three.
2424 .P
2425 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2426 extract a single substring, whose number is given as \fIstringnumber\fP. A
2427 value of zero extracts the substring that matched the entire pattern, whereas
2428 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2429 the string is placed in \fIbuffer\fP, whose length is given by
2430 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2431 obtained via \fBpcre_malloc\fP, and its address is returned via
2432 \fIstringptr\fP. The yield of the function is the length of the string, not
2433 including the terminating zero, or one of these error codes:
2434 .sp
2435 PCRE_ERROR_NOMEMORY (-6)
2436 .sp
2437 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2438 memory failed for \fBpcre_get_substring()\fP.
2439 .sp
2440 PCRE_ERROR_NOSUBSTRING (-7)
2441 .sp
2442 There is no substring whose number is \fIstringnumber\fP.
2443 .P
2444 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2445 and builds a list of pointers to them. All this is done in a single block of
2446 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2447 is returned via \fIlistptr\fP, which is also the start of the list of string
2448 pointers. The end of the list is marked by a NULL pointer. The yield of the
2449 function is zero if all went well, or the error code
2450 .sp
2451 PCRE_ERROR_NOMEMORY (-6)
2452 .sp
2453 if the attempt to get the memory block failed.
2454 .P
2455 When any of these functions encounter a substring that is unset, which can
2456 happen when capturing subpattern number \fIn+1\fP matches some part of the
2457 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2458 string. This can be distinguished from a genuine zero-length substring by
2459 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2460 substrings.
2461 .P
2462 The two convenience functions \fBpcre_free_substring()\fP and
2463 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2464 a previous call of \fBpcre_get_substring()\fP or
2465 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2466 the function pointed to by \fBpcre_free\fP, which of course could be called
2467 directly from a C program. However, PCRE is used in some situations where it is
2468 linked via a special interface to another programming language that cannot use
2469 \fBpcre_free\fP directly; it is for these cases that the functions are
2470 provided.
2471 .
2472 .
2473 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2474 .rs
2475 .sp
2476 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2477 .ti +5n
2478 .B const char *\fIname\fP);
2479 .PP
2480 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2481 .ti +5n
2482 .B const char *\fIsubject\fP, int *\fIovector\fP,
2483 .ti +5n
2484 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2485 .ti +5n
2486 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2487 .PP
2488 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2489 .ti +5n
2490 .B const char *\fIsubject\fP, int *\fIovector\fP,
2491 .ti +5n
2492 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2493 .ti +5n
2494 .B const char **\fIstringptr\fP);
2495 .PP
2496 To extract a substring by name, you first have to find associated number.
2497 For example, for this pattern
2498 .sp
2499 (a+)b(?<xxx>\ed+)...
2500 .sp
2501 the number of the subpattern called "xxx" is 2. If the name is known to be
2502 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2503 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2504 pattern, and the second is the name. The yield of the function is the
2505 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2506 that name.
2507 .P
2508 Given the number, you can extract the substring directly, or use one of the
2509 functions described in the previous section. For convenience, there are also
2510 two functions that do the whole job.
2511 .P
2512 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2513 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2514 functions that extract by number. As these are described in the previous
2515 section, they are not re-described here. There are just two differences:
2516 .P
2517 First, instead of a substring number, a substring name is given. Second, there
2518 is an extra argument, given at the start, which is a pointer to the compiled
2519 pattern. This is needed in order to gain access to the name-to-number
2520 translation table.
2521 .P
2522 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2523 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2524 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2525 the behaviour may not be what you want (see the next section).
2526 .P
2527 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2528 subpatterns with the same number, as described in the
2529 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2530 .\" </a>
2531 section on duplicate subpattern numbers
2532 .\"
2533 in the
2534 .\" HREF
2535 \fBpcrepattern\fP
2536 .\"
2537 page, you cannot use names to distinguish the different subpatterns, because
2538 names are not included in the compiled code. The matching process uses only
2539 numbers. For this reason, the use of different names for subpatterns of the
2540 same number causes an error at compile time.
2541 .
2542 .
2543 .SH "DUPLICATE SUBPATTERN NAMES"
2544 .rs
2545 .sp
2546 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2547 .ti +5n
2548 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2549 .PP
2550 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2551 are not required to be unique. (Duplicate names are always allowed for
2552 subpatterns with the same number, created by using the (?| feature. Indeed, if
2553 such subpatterns are named, they are required to use the same names.)
2554 .P
2555 Normally, patterns with duplicate names are such that in any one match, only
2556 one of the named subpatterns participates. An example is shown in the
2557 .\" HREF
2558 \fBpcrepattern\fP
2559 .\"
2560 documentation.
2561 .P
2562 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2563 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2564 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2565 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2566 returns one of the numbers that are associated with the name, but it is not
2567 defined which it is.
2568 .P
2569 If you want to get full details of all captured substrings for a given name,
2570 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2571 argument is the compiled pattern, and the second is the name. The third and
2572 fourth are pointers to variables which are updated by the function. After it
2573 has run, they point to the first and last entries in the name-to-number table
2574 for the given name. The function itself returns the length of each entry, or
2575 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2576 described above in the section entitled \fIInformation about a pattern\fP
2577 .\" HTML <a href="#infoaboutpattern">
2578 .\" </a>
2579 above.
2580 .\"
2581 Given all the relevant entries for the name, you can extract each of their
2582 numbers, and hence the captured data, if any.
2583 .
2584 .
2585 .SH "FINDING ALL POSSIBLE MATCHES"
2586 .rs
2587 .sp
2588 The traditional matching function uses a similar algorithm to Perl, which stops
2589 when it finds the first match, starting at a given point in the subject. If you
2590 want to find all possible matches, or the longest possible match, consider
2591 using the alternative matching function (see below) instead. If you cannot use
2592 the alternative function, but still need to find all possible matches, you
2593 can kludge it up by making use of the callout facility, which is described in
2594 the
2595 .\" HREF
2596 \fBpcrecallout\fP
2597 .\"
2598 documentation.
2599 .P
2600 What you have to do is to insert a callout right at the end of the pattern.
2601 When your callout function is called, extract and save the current matched
2602 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2603 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2604 will yield PCRE_ERROR_NOMATCH.
2605 .
2606 .
2607 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2608 .rs
2609 .sp
2610 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2611 stack, which in certain environments can be rather limited in size. Some users
2612 find it helpful to have an estimate of the amount of stack that is used by
2613 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2614 .\" HREF
2615 \fBpcrestack\fP
2616 .\"
2617 documentation. The estimate that is output by \fBpcretest\fP when called with
2618 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2619 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2620 .P
2621 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2622 the negative error code PCRE_ERROR_NULL, but with this special combination of
2623 arguments, it returns instead a negative number whose absolute value is the
2624 approximate stack frame size in bytes. (A negative number is used so that it is
2625 clear that no match has happened.) The value is approximate because in some
2626 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2627 additional variables on the stack.
2628 .P
2629 If PCRE has been compiled to use the heap instead of the stack for recursion,
2630 the value returned is the size of each block that is obtained from the heap.
2631 .
2632 .
2633 .\" HTML <a name="dfamatch"></a>
2634 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2635 .rs
2636 .sp
2637 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2638 .ti +5n
2639 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2640 .ti +5n
2641 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2642 .ti +5n
2643 .B int *\fIworkspace\fP, int \fIwscount\fP);
2644 .P
2645 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2646 a compiled pattern, using a matching algorithm that scans the subject string
2647 just once, and does not backtrack. This has different characteristics to the
2648 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2649 patterns are not supported. Nevertheless, there are times when this kind of
2650 matching can be useful. For a discussion of the two matching algorithms, and a
2651 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2652 .\" HREF
2653 \fBpcrematching\fP
2654 .\"
2655 documentation.
2656 .P
2657 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2658 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2659 different way, and this is described below. The other common arguments are used
2660 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2661 here.
2662 .P
2663 The two additional arguments provide workspace for the function. The workspace
2664 vector should contain at least 20 elements. It is used for keeping track of
2665 multiple paths through the pattern tree. More workspace will be needed for
2666 patterns and subjects where there are a lot of potential matches.
2667 .P
2668 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2669 .sp
2670 int rc;
2671 int ovector[10];
2672 int wspace[20];
2673 rc = pcre_dfa_exec(
2674 re, /* result of pcre_compile() */
2675 NULL, /* we didn't study the pattern */
2676 "some string", /* the subject string */
2677 11, /* the length of the subject string */
2678 0, /* start at offset 0 in the subject */
2679 0, /* default options */
2680 ovector, /* vector of integers for substring information */
2681 10, /* number of elements (NOT size in bytes) */
2682 wspace, /* working space vector */
2683 20); /* number of elements (NOT size in bytes) */
2684 .
2685 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2686 .rs
2687 .sp
2688 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2689 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2690 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2691 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2692 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2693 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2694 so their description is not repeated here.
2695 .sp
2696 PCRE_PARTIAL_HARD
2697 PCRE_PARTIAL_SOFT
2698 .sp
2699 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2700 details are slightly different. When PCRE_PARTIAL_HARD is set for
2701 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2702 is reached and there is still at least one matching possibility that requires
2703 additional characters. This happens even if some complete matches have also
2704 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2705 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2706 there have been no complete matches, but there is still at least one matching
2707 possibility. The portion of the string that was inspected when the longest
2708 partial match was found is set as the first matching string in both cases.
2709 There is a more detailed discussion of partial and multi-segment matching, with
2710 examples, in the
2711 .\" HREF
2712 \fBpcrepartial\fP
2713 .\"
2714 documentation.
2715 .sp
2716 PCRE_DFA_SHORTEST
2717 .sp
2718 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2719 soon as it has found one match. Because of the way the alternative algorithm
2720 works, this is necessarily the shortest possible match at the first possible
2721 matching point in the subject string.
2722 .sp
2723 PCRE_DFA_RESTART
2724 .sp
2725 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2726 again, with additional subject characters, and have it continue with the same
2727 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2728 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2729 before because data about the match so far is left in them after a partial
2730 match. There is more discussion of this facility in the
2731 .\" HREF
2732 \fBpcrepartial\fP
2733 .\"
2734 documentation.
2735 .
2736 .
2737 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2738 .rs
2739 .sp
2740 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2741 substring in the subject. Note, however, that all the matches from one run of
2742 the function start at the same point in the subject. The shorter matches are
2743 all initial substrings of the longer matches. For example, if the pattern
2744 .sp
2745 <.*>
2746 .sp
2747 is matched against the string
2748 .sp
2749 This is <something> <something else> <something further> no more
2750 .sp
2751 the three matched strings are
2752 .sp
2753 <something>
2754 <something> <something else>
2755 <something> <something else> <something further>
2756 .sp
2757 On success, the yield of the function is a number greater than zero, which is
2758 the number of matched substrings. The substrings themselves are returned in
2759 \fIovector\fP. Each string uses two elements; the first is the offset to the
2760 start, and the second is the offset to the end. In fact, all the strings have
2761 the same start offset. (Space could have been saved by giving this only once,
2762 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2763 returns data, even though the meaning of the strings is different.)
2764 .P
2765 The strings are returned in reverse order of length; that is, the longest
2766 matching string is given first. If there were too many matches to fit into
2767 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2768 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2769 the entire \fIovector\fP for returning matched strings.
2770 .
2771 .
2772 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2773 .rs
2774 .sp
2775 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2776 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2777 described
2778 .\" HTML <a href="#errorlist">
2779 .\" </a>
2780 above.
2781 .\"
2782 There are in addition the following errors that are specific to
2783 \fBpcre_dfa_exec()\fP:
2784 .sp
2785 PCRE_ERROR_DFA_UITEM (-16)
2786 .sp
2787 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2788 that it does not support, for instance, the use of \eC or a back reference.
2789 .sp
2790 PCRE_ERROR_DFA_UCOND (-17)
2791 .sp
2792 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2793 uses a back reference for the condition, or a test for recursion in a specific
2794 group. These are not supported.
2795 .sp
2796 PCRE_ERROR_DFA_UMLIMIT (-18)
2797 .sp
2798 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2799 block that contains a setting of the \fImatch_limit\fP or
2800 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2801 meaningless for DFA matching).
2802 .sp
2803 PCRE_ERROR_DFA_WSSIZE (-19)
2804 .sp
2805 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2806 \fIworkspace\fP vector.
2807 .sp
2808 PCRE_ERROR_DFA_RECURSE (-20)
2809 .sp
2810 When a recursive subpattern is processed, the matching function calls itself
2811 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2812 error is given if the output vector is not large enough. This should be
2813 extremely rare, as a vector of size 1000 is used.
2814 .sp
2815 PCRE_ERROR_DFA_BADRESTART (-30)
2816 .sp
2817 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2818 some plausibility checks are made on the contents of the workspace, which
2819 should contain data about the previous partial match. If any of these checks
2820 fail, this error is given.
2821 .
2822 .
2823 .SH "SEE ALSO"
2824 .rs
2825 .sp
2826 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2827 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2828 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2829 \fBpcrestack\fP(3).
2830 .
2831 .
2832 .SH AUTHOR
2833 .rs
2834 .sp
2835 .nf
2836 Philip Hazel
2837 University Computing Service
2838 Cambridge CB2 3QH, England.
2839 .fi
2840 .
2841 .
2842 .SH REVISION
2843 .rs
2844 .sp
2845 .nf
2846 Last updated: 05 April 2013
2847 Copyright (c) 1997-2013 University of Cambridge.
2848 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5