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

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

Parent Directory Parent Directory | Revision Log Revision Log


Revision 1261 - (show annotations)
Wed Feb 27 16:27:01 2013 UTC (6 years, 5 months ago) by ph10
File size: 122877 byte(s)
Correct Unicode string checking in the light of corrigendum #9.
1 .TH PCREAPI 3 "27 February 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_NEWLINE_CR
759 PCRE_NEWLINE_LF
760 PCRE_NEWLINE_CRLF
761 PCRE_NEWLINE_ANYCRLF
762 PCRE_NEWLINE_ANY
763 .sp
764 These options override the default newline definition that was chosen when PCRE
765 was built. Setting the first or the second specifies that a newline is
766 indicated by a single character (CR or LF, respectively). Setting
767 PCRE_NEWLINE_CRLF specifies that a newline is indicated by the two-character
768 CRLF sequence. Setting PCRE_NEWLINE_ANYCRLF specifies that any of the three
769 preceding sequences should be recognized. Setting PCRE_NEWLINE_ANY specifies
770 that any Unicode newline sequence should be recognized.
771 .P
772 In an ASCII/Unicode environment, the Unicode newline sequences are the three
773 just mentioned, plus the single characters VT (vertical tab, U+000B), FF (form
774 feed, U+000C), NEL (next line, U+0085), LS (line separator, U+2028), and PS
775 (paragraph separator, U+2029). For the 8-bit library, the last two are
776 recognized only in UTF-8 mode.
777 .P
778 When PCRE is compiled to run in an EBCDIC (mainframe) environment, the code for
779 CR is 0x0d, the same as ASCII. However, the character code for LF is normally
780 0x15, though in some EBCDIC environments 0x25 is used. Whichever of these is
781 not LF is made to correspond to Unicode's NEL character. EBCDIC codes are all
782 less than 256. For more details, see the
783 .\" HREF
784 \fBpcrebuild\fP
785 .\"
786 documentation.
787 .P
788 The newline setting in the options word uses three bits that are treated
789 as a number, giving eight possibilities. Currently only six are used (default
790 plus the five values above). This means that if you set more than one newline
791 option, the combination may or may not be sensible. For example,
792 PCRE_NEWLINE_CR with PCRE_NEWLINE_LF is equivalent to PCRE_NEWLINE_CRLF, but
793 other combinations may yield unused numbers and cause an error.
794 .P
795 The only time that a line break in a pattern is specially recognized when
796 compiling is when PCRE_EXTENDED is set. CR and LF are white space characters,
797 and so are ignored in this mode. Also, an unescaped # outside a character class
798 indicates a comment that lasts until after the next line break sequence. In
799 other circumstances, line break sequences in patterns are treated as literal
800 data.
801 .P
802 The newline option that is set at compile time becomes the default that is used
803 for \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, but it can be overridden.
804 .sp
805 PCRE_NO_AUTO_CAPTURE
806 .sp
807 If this option is set, it disables the use of numbered capturing parentheses in
808 the pattern. Any opening parenthesis that is not followed by ? behaves as if it
809 were followed by ?: but named parentheses can still be used for capturing (and
810 they acquire numbers in the usual way). There is no equivalent of this option
811 in Perl.
812 .sp
813 NO_START_OPTIMIZE
814 .sp
815 This is an option that acts at matching time; that is, it is really an option
816 for \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. If it is set at compile time,
817 it is remembered with the compiled pattern and assumed at matching time. For
818 details see the discussion of PCRE_NO_START_OPTIMIZE
819 .\" HTML <a href="#execoptions">
820 .\" </a>
821 below.
822 .\"
823 .sp
824 PCRE_UCP
825 .sp
826 This option changes the way PCRE processes \eB, \eb, \eD, \ed, \eS, \es, \eW,
827 \ew, and some of the POSIX character classes. By default, only ASCII characters
828 are recognized, but if PCRE_UCP is set, Unicode properties are used instead to
829 classify characters. More details are given in the section on
830 .\" HTML <a href="pcre.html#genericchartypes">
831 .\" </a>
832 generic character types
833 .\"
834 in the
835 .\" HREF
836 \fBpcrepattern\fP
837 .\"
838 page. If you set PCRE_UCP, matching one of the items it affects takes much
839 longer. The option is available only if PCRE has been compiled with Unicode
840 property support.
841 .sp
842 PCRE_UNGREEDY
843 .sp
844 This option inverts the "greediness" of the quantifiers so that they are not
845 greedy by default, but become greedy if followed by "?". It is not compatible
846 with Perl. It can also be set by a (?U) option setting within the pattern.
847 .sp
848 PCRE_UTF8
849 .sp
850 This option causes PCRE to regard both the pattern and the subject as strings
851 of UTF-8 characters instead of single-byte strings. However, it is available
852 only when PCRE is built to include UTF support. If not, the use of this option
853 provokes an error. Details of how this option changes the behaviour of PCRE are
854 given in the
855 .\" HREF
856 \fBpcreunicode\fP
857 .\"
858 page.
859 .sp
860 PCRE_NO_UTF8_CHECK
861 .sp
862 When PCRE_UTF8 is set, the validity of the pattern as a UTF-8 string is
863 automatically checked. There is a discussion about the
864 .\" HTML <a href="pcreunicode.html#utf8strings">
865 .\" </a>
866 validity of UTF-8 strings
867 .\"
868 in the
869 .\" HREF
870 \fBpcreunicode\fP
871 .\"
872 page. If an invalid UTF-8 sequence is found, \fBpcre_compile()\fP returns an
873 error. If you already know that your pattern is valid, and you want to skip
874 this check for performance reasons, you can set the PCRE_NO_UTF8_CHECK option.
875 When it is set, the effect of passing an invalid UTF-8 string as a pattern is
876 undefined. It may cause your program to crash. Note that this option can also
877 be passed to \fBpcre_exec()\fP and \fBpcre_dfa_exec()\fP, to suppress the
878 validity checking of subject strings only. If the same string is being matched
879 many times, the option can be safely set for the second and subsequent
880 matchings to improve performance.
881 .
882 .
883 .SH "COMPILATION ERROR CODES"
884 .rs
885 .sp
886 The following table lists the error codes than may be returned by
887 \fBpcre_compile2()\fP, along with the error messages that may be returned by
888 both compiling functions. Note that error messages are always 8-bit ASCII
889 strings, even in 16-bit or 32-bit mode. As PCRE has developed, some error codes
890 have fallen out of use. To avoid confusion, they have not been re-used.
891 .sp
892 0 no error
893 1 \e at end of pattern
894 2 \ec at end of pattern
895 3 unrecognized character follows \e
896 4 numbers out of order in {} quantifier
897 5 number too big in {} quantifier
898 6 missing terminating ] for character class
899 7 invalid escape sequence in character class
900 8 range out of order in character class
901 9 nothing to repeat
902 10 [this code is not in use]
903 11 internal error: unexpected repeat
904 12 unrecognized character after (? or (?-
905 13 POSIX named classes are supported only within a class
906 14 missing )
907 15 reference to non-existent subpattern
908 16 erroffset passed as NULL
909 17 unknown option bit(s) set
910 18 missing ) after comment
911 19 [this code is not in use]
912 20 regular expression is too large
913 21 failed to get memory
914 22 unmatched parentheses
915 23 internal error: code overflow
916 24 unrecognized character after (?<
917 25 lookbehind assertion is not fixed length
918 26 malformed number or name after (?(
919 27 conditional group contains more than two branches
920 28 assertion expected after (?(
921 29 (?R or (?[+-]digits must be followed by )
922 30 unknown POSIX class name
923 31 POSIX collating elements are not supported
924 32 this version of PCRE is compiled without UTF support
925 33 [this code is not in use]
926 34 character value in \ex{...} sequence is too large
927 35 invalid condition (?(0)
928 36 \eC not allowed in lookbehind assertion
929 37 PCRE does not support \eL, \el, \eN{name}, \eU, or \eu
930 38 number after (?C is > 255
931 39 closing ) for (?C expected
932 40 recursive call could loop indefinitely
933 41 unrecognized character after (?P
934 42 syntax error in subpattern name (missing terminator)
935 43 two named subpatterns have the same name
936 44 invalid UTF-8 string (specifically UTF-8)
937 45 support for \eP, \ep, and \eX has not been compiled
938 46 malformed \eP or \ep sequence
939 47 unknown property name after \eP or \ep
940 48 subpattern name is too long (maximum 32 characters)
941 49 too many named subpatterns (maximum 10000)
942 50 [this code is not in use]
943 51 octal value is greater than \e377 in 8-bit non-UTF-8 mode
944 52 internal error: overran compiling workspace
945 53 internal error: previously-checked referenced subpattern
946 not found
947 54 DEFINE group contains more than one branch
948 55 repeating a DEFINE group is not allowed
949 56 inconsistent NEWLINE options
950 57 \eg is not followed by a braced, angle-bracketed, or quoted
951 name/number or by a plain number
952 58 a numbered reference must not be zero
953 59 an argument is not allowed for (*ACCEPT), (*FAIL), or (*COMMIT)
954 60 (*VERB) not recognized
955 61 number is too big
956 62 subpattern name expected
957 63 digit expected after (?+
958 64 ] is an invalid data character in JavaScript compatibility mode
959 65 different names for subpatterns of the same number are
960 not allowed
961 66 (*MARK) must have an argument
962 67 this version of PCRE is not compiled with Unicode property
963 support
964 68 \ec must be followed by an ASCII character
965 69 \ek is not followed by a braced, angle-bracketed, or quoted name
966 70 internal error: unknown opcode in find_fixedlength()
967 71 \eN is not supported in a class
968 72 too many forward references
969 73 disallowed Unicode code point (>= 0xd800 && <= 0xdfff)
970 74 invalid UTF-16 string (specifically UTF-16)
971 75 name is too long in (*MARK), (*PRUNE), (*SKIP), or (*THEN)
972 76 character value in \eu.... sequence is too large
973 77 invalid UTF-32 string (specifically UTF-32)
974 .sp
975 The numbers 32 and 10000 in errors 48 and 49 are defaults; different values may
976 be used if the limits were changed when PCRE was built.
977 .
978 .
979 .\" HTML <a name="studyingapattern"></a>
980 .SH "STUDYING A PATTERN"
981 .rs
982 .sp
983 .B pcre_extra *pcre_study(const pcre *\fIcode\fP, int \fIoptions\fP
984 .ti +5n
985 .B const char **\fIerrptr\fP);
986 .PP
987 If a compiled pattern is going to be used several times, it is worth spending
988 more time analyzing it in order to speed up the time taken for matching. The
989 function \fBpcre_study()\fP takes a pointer to a compiled pattern as its first
990 argument. If studying the pattern produces additional information that will
991 help speed up matching, \fBpcre_study()\fP returns a pointer to a
992 \fBpcre_extra\fP block, in which the \fIstudy_data\fP field points to the
993 results of the study.
994 .P
995 The returned value from \fBpcre_study()\fP can be passed directly to
996 \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP. However, a \fBpcre_extra\fP block
997 also contains other fields that can be set by the caller before the block is
998 passed; these are described
999 .\" HTML <a href="#extradata">
1000 .\" </a>
1001 below
1002 .\"
1003 in the section on matching a pattern.
1004 .P
1005 If studying the pattern does not produce any useful information,
1006 \fBpcre_study()\fP returns NULL by default. In that circumstance, if the
1007 calling program wants to pass any of the other fields to \fBpcre_exec()\fP or
1008 \fBpcre_dfa_exec()\fP, it must set up its own \fBpcre_extra\fP block. However,
1009 if \fBpcre_study()\fP is called with the PCRE_STUDY_EXTRA_NEEDED option, it
1010 returns a \fBpcre_extra\fP block even if studying did not find any additional
1011 information. It may still return NULL, however, if an error occurs in
1012 \fBpcre_study()\fP.
1013 .P
1014 The second argument of \fBpcre_study()\fP contains option bits. There are three
1015 further options in addition to PCRE_STUDY_EXTRA_NEEDED:
1016 .sp
1017 PCRE_STUDY_JIT_COMPILE
1018 PCRE_STUDY_JIT_PARTIAL_HARD_COMPILE
1019 PCRE_STUDY_JIT_PARTIAL_SOFT_COMPILE
1020 .sp
1021 If any of these are set, and the just-in-time compiler is available, the
1022 pattern is further compiled into machine code that executes much faster than
1023 the \fBpcre_exec()\fP interpretive matching function. If the just-in-time
1024 compiler is not available, these options are ignored. All undefined bits in the
1025 \fIoptions\fP argument must be zero.
1026 .P
1027 JIT compilation is a heavyweight optimization. It can take some time for
1028 patterns to be analyzed, and for one-off matches and simple patterns the
1029 benefit of faster execution might be offset by a much slower study time.
1030 Not all patterns can be optimized by the JIT compiler. For those that cannot be
1031 handled, matching automatically falls back to the \fBpcre_exec()\fP
1032 interpreter. For more details, see the
1033 .\" HREF
1034 \fBpcrejit\fP
1035 .\"
1036 documentation.
1037 .P
1038 The third argument for \fBpcre_study()\fP is a pointer for an error message. If
1039 studying succeeds (even if no data is returned), the variable it points to is
1040 set to NULL. Otherwise it is set to point to a textual error message. This is a
1041 static string that is part of the library. You must not try to free it. You
1042 should test the error pointer for NULL after calling \fBpcre_study()\fP, to be
1043 sure that it has run successfully.
1044 .P
1045 When you are finished with a pattern, you can free the memory used for the
1046 study data by calling \fBpcre_free_study()\fP. This function was added to the
1047 API for release 8.20. For earlier versions, the memory could be freed with
1048 \fBpcre_free()\fP, just like the pattern itself. This will still work in cases
1049 where JIT optimization is not used, but it is advisable to change to the new
1050 function when convenient.
1051 .P
1052 This is a typical way in which \fBpcre_study\fP() is used (except that in a
1053 real application there should be tests for errors):
1054 .sp
1055 int rc;
1056 pcre *re;
1057 pcre_extra *sd;
1058 re = pcre_compile("pattern", 0, &error, &erroroffset, NULL);
1059 sd = pcre_study(
1060 re, /* result of pcre_compile() */
1061 0, /* no options */
1062 &error); /* set to NULL or points to a message */
1063 rc = pcre_exec( /* see below for details of pcre_exec() options */
1064 re, sd, "subject", 7, 0, 0, ovector, 30);
1065 ...
1066 pcre_free_study(sd);
1067 pcre_free(re);
1068 .sp
1069 Studying a pattern does two things: first, a lower bound for the length of
1070 subject string that is needed to match the pattern is computed. This does not
1071 mean that there are any strings of that length that match, but it does
1072 guarantee that no shorter strings match. The value is used to avoid wasting
1073 time by trying to match strings that are shorter than the lower bound. You can
1074 find out the value in a calling program via the \fBpcre_fullinfo()\fP function.
1075 .P
1076 Studying a pattern is also useful for non-anchored patterns that do not have a
1077 single fixed starting character. A bitmap of possible starting bytes is
1078 created. This speeds up finding a position in the subject at which to start
1079 matching. (In 16-bit mode, the bitmap is used for 16-bit values less than 256.
1080 In 32-bit mode, the bitmap is used for 32-bit values less than 256.)
1081 .P
1082 These two optimizations apply to both \fBpcre_exec()\fP and
1083 \fBpcre_dfa_exec()\fP, and the information is also used by the JIT compiler.
1084 The optimizations can be disabled by setting the PCRE_NO_START_OPTIMIZE option
1085 when calling \fBpcre_exec()\fP or \fBpcre_dfa_exec()\fP, but if this is done,
1086 JIT execution is also disabled. You might want to do this if your pattern
1087 contains callouts or (*MARK) and you want to make use of these facilities in
1088 cases where matching fails. See the discussion of PCRE_NO_START_OPTIMIZE
1089 .\" HTML <a href="#execoptions">
1090 .\" </a>
1091 below.
1092 .\"
1093 .
1094 .
1095 .\" HTML <a name="localesupport"></a>
1096 .SH "LOCALE SUPPORT"
1097 .rs
1098 .sp
1099 PCRE handles caseless matching, and determines whether characters are letters,
1100 digits, or whatever, by reference to a set of tables, indexed by character
1101 value. When running in UTF-8 mode, this applies only to characters
1102 with codes less than 128. By default, higher-valued codes never match escapes
1103 such as \ew or \ed, but they can be tested with \ep if PCRE is built with
1104 Unicode character property support. Alternatively, the PCRE_UCP option can be
1105 set at compile time; this causes \ew and friends to use Unicode property
1106 support instead of built-in tables. The use of locales with Unicode is
1107 discouraged. If you are handling characters with codes greater than 128, you
1108 should either use UTF-8 and Unicode, or use locales, but not try to mix the
1109 two.
1110 .P
1111 PCRE contains an internal set of tables that are used when the final argument
1112 of \fBpcre_compile()\fP is NULL. These are sufficient for many applications.
1113 Normally, the internal tables recognize only ASCII characters. However, when
1114 PCRE is built, it is possible to cause the internal tables to be rebuilt in the
1115 default "C" locale of the local system, which may cause them to be different.
1116 .P
1117 The internal tables can always be overridden by tables supplied by the
1118 application that calls PCRE. These may be created in a different locale from
1119 the default. As more and more applications change to using Unicode, the need
1120 for this locale support is expected to die away.
1121 .P
1122 External tables are built by calling the \fBpcre_maketables()\fP function,
1123 which has no arguments, in the relevant locale. The result can then be passed
1124 to \fBpcre_compile()\fP or \fBpcre_exec()\fP as often as necessary. For
1125 example, to build and use tables that are appropriate for the French locale
1126 (where accented characters with values greater than 128 are treated as letters),
1127 the following code could be used:
1128 .sp
1129 setlocale(LC_CTYPE, "fr_FR");
1130 tables = pcre_maketables();
1131 re = pcre_compile(..., tables);
1132 .sp
1133 The locale name "fr_FR" is used on Linux and other Unix-like systems; if you
1134 are using Windows, the name for the French locale is "french".
1135 .P
1136 When \fBpcre_maketables()\fP runs, the tables are built in memory that is
1137 obtained via \fBpcre_malloc\fP. It is the caller's responsibility to ensure
1138 that the memory containing the tables remains available for as long as it is
1139 needed.
1140 .P
1141 The pointer that is passed to \fBpcre_compile()\fP is saved with the compiled
1142 pattern, and the same tables are used via this pointer by \fBpcre_study()\fP
1143 and normally also by \fBpcre_exec()\fP. Thus, by default, for any single
1144 pattern, compilation, studying and matching all happen in the same locale, but
1145 different patterns can be compiled in different locales.
1146 .P
1147 It is possible to pass a table pointer or NULL (indicating the use of the
1148 internal tables) to \fBpcre_exec()\fP. Although not intended for this purpose,
1149 this facility could be used to match a pattern in a different locale from the
1150 one in which it was compiled. Passing table pointers at run time is discussed
1151 below in the section on matching a pattern.
1152 .
1153 .
1154 .\" HTML <a name="infoaboutpattern"></a>
1155 .SH "INFORMATION ABOUT A PATTERN"
1156 .rs
1157 .sp
1158 .B int pcre_fullinfo(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1159 .ti +5n
1160 .B int \fIwhat\fP, void *\fIwhere\fP);
1161 .PP
1162 The \fBpcre_fullinfo()\fP function returns information about a compiled
1163 pattern. It replaces the \fBpcre_info()\fP function, which was removed from the
1164 library at version 8.30, after more than 10 years of obsolescence.
1165 .P
1166 The first argument for \fBpcre_fullinfo()\fP is a pointer to the compiled
1167 pattern. The second argument is the result of \fBpcre_study()\fP, or NULL if
1168 the pattern was not studied. The third argument specifies which piece of
1169 information is required, and the fourth argument is a pointer to a variable
1170 to receive the data. The yield of the function is zero for success, or one of
1171 the following negative numbers:
1172 .sp
1173 PCRE_ERROR_NULL the argument \fIcode\fP was NULL
1174 the argument \fIwhere\fP was NULL
1175 PCRE_ERROR_BADMAGIC the "magic number" was not found
1176 PCRE_ERROR_BADENDIANNESS the pattern was compiled with different
1177 endianness
1178 PCRE_ERROR_BADOPTION the value of \fIwhat\fP was invalid
1179 .sp
1180 The "magic number" is placed at the start of each compiled pattern as an simple
1181 check against passing an arbitrary memory pointer. The endianness error can
1182 occur if a compiled pattern is saved and reloaded on a different host. Here is
1183 a typical call of \fBpcre_fullinfo()\fP, to obtain the length of the compiled
1184 pattern:
1185 .sp
1186 int rc;
1187 size_t length;
1188 rc = pcre_fullinfo(
1189 re, /* result of pcre_compile() */
1190 sd, /* result of pcre_study(), or NULL */
1191 PCRE_INFO_SIZE, /* what is required */
1192 &length); /* where to put the data */
1193 .sp
1194 The possible values for the third argument are defined in \fBpcre.h\fP, and are
1195 as follows:
1196 .sp
1197 PCRE_INFO_BACKREFMAX
1198 .sp
1199 Return the number of the highest back reference in the pattern. The fourth
1200 argument should point to an \fBint\fP variable. Zero is returned if there are
1201 no back references.
1202 .sp
1203 PCRE_INFO_CAPTURECOUNT
1204 .sp
1205 Return the number of capturing subpatterns in the pattern. The fourth argument
1206 should point to an \fBint\fP variable.
1207 .sp
1208 PCRE_INFO_DEFAULT_TABLES
1209 .sp
1210 Return a pointer to the internal default character tables within PCRE. The
1211 fourth argument should point to an \fBunsigned char *\fP variable. This
1212 information call is provided for internal use by the \fBpcre_study()\fP
1213 function. External callers can cause PCRE to use its internal tables by passing
1214 a NULL table pointer.
1215 .sp
1216 PCRE_INFO_FIRSTBYTE
1217 .sp
1218 Return information about the first data unit of any matched string, for a
1219 non-anchored pattern. (The name of this option refers to the 8-bit library,
1220 where data units are bytes.) The fourth argument should point to an \fBint\fP
1221 variable.
1222 .P
1223 If there is a fixed first value, for example, the letter "c" from a pattern
1224 such as (cat|cow|coyote), its value is returned. In the 8-bit library, the
1225 value is always less than 256. In the 16-bit library the value can be up to
1226 0xffff. In the 32-bit library the value can be up to 0x10ffff.
1227 .P
1228 If there is no fixed first value, and if either
1229 .sp
1230 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1231 starts with "^", or
1232 .sp
1233 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1234 (if it were set, the pattern would be anchored),
1235 .sp
1236 -1 is returned, indicating that the pattern matches only at the start of a
1237 subject string or after any newline within the string. Otherwise -2 is
1238 returned. For anchored patterns, -2 is returned.
1239 .P
1240 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1241 to return the full 32-bit range of the character, this value is deprecated;
1242 instead the PCRE_INFO_FIRSTCHARACTERFLAGS and PCRE_INFO_FIRSTCHARACTER values
1243 should be used.
1244 .sp
1245 PCRE_INFO_FIRSTTABLE
1246 .sp
1247 If the pattern was studied, and this resulted in the construction of a 256-bit
1248 table indicating a fixed set of values for the first data unit in any matching
1249 string, a pointer to the table is returned. Otherwise NULL is returned. The
1250 fourth argument should point to an \fBunsigned char *\fP variable.
1251 .sp
1252 PCRE_INFO_HASCRORLF
1253 .sp
1254 Return 1 if the pattern contains any explicit matches for CR or LF characters,
1255 otherwise 0. The fourth argument should point to an \fBint\fP variable. An
1256 explicit match is either a literal CR or LF character, or \er or \en.
1257 .sp
1258 PCRE_INFO_JCHANGED
1259 .sp
1260 Return 1 if the (?J) or (?-J) option setting is used in the pattern, otherwise
1261 0. The fourth argument should point to an \fBint\fP variable. (?J) and
1262 (?-J) set and unset the local PCRE_DUPNAMES option, respectively.
1263 .sp
1264 PCRE_INFO_JIT
1265 .sp
1266 Return 1 if the pattern was studied with one of the JIT options, and
1267 just-in-time compiling was successful. The fourth argument should point to an
1268 \fBint\fP variable. A return value of 0 means that JIT support is not available
1269 in this version of PCRE, or that the pattern was not studied with a JIT option,
1270 or that the JIT compiler could not handle this particular pattern. See the
1271 .\" HREF
1272 \fBpcrejit\fP
1273 .\"
1274 documentation for details of what can and cannot be handled.
1275 .sp
1276 PCRE_INFO_JITSIZE
1277 .sp
1278 If the pattern was successfully studied with a JIT option, return the size of
1279 the JIT compiled code, otherwise return zero. The fourth argument should point
1280 to a \fBsize_t\fP variable.
1281 .sp
1282 PCRE_INFO_LASTLITERAL
1283 .sp
1284 Return the value of the rightmost literal data unit that must exist in any
1285 matched string, other than at its start, if such a value has been recorded. The
1286 fourth argument should point to an \fBint\fP variable. If there is no such
1287 value, -1 is returned. For anchored patterns, a last literal value is recorded
1288 only if it follows something of variable length. For example, for the pattern
1289 /^a\ed+z\ed+/ the returned value is "z", but for /^a\edz\ed/ the returned value
1290 is -1.
1291 .P
1292 Since for the 32-bit library using the non-UTF-32 mode, this function is unable
1293 to return the full 32-bit range of the character, this value is deprecated;
1294 instead the PCRE_INFO_REQUIREDCHARFLAGS and PCRE_INFO_REQUIREDCHAR values should
1295 be used.
1296 .sp
1297 PCRE_INFO_MAXLOOKBEHIND
1298 .sp
1299 Return the number of characters (NB not bytes) in the longest lookbehind
1300 assertion in the pattern. This information is useful when doing multi-segment
1301 matching using the partial matching facilities. Note that the simple assertions
1302 \eb and \eB require a one-character lookbehind. \eA also registers a
1303 one-character lookbehind, though it does not actually inspect the previous
1304 character. This is to ensure that at least one character from the old segment
1305 is retained when a new segment is processed. Otherwise, if there are no
1306 lookbehinds in the pattern, \eA might match incorrectly at the start of a new
1307 segment.
1308 .sp
1309 PCRE_INFO_MINLENGTH
1310 .sp
1311 If the pattern was studied and a minimum length for matching subject strings
1312 was computed, its value is returned. Otherwise the returned value is -1. The
1313 value is a number of characters, which in UTF-8 mode may be different from the
1314 number of bytes. The fourth argument should point to an \fBint\fP variable. A
1315 non-negative value is a lower bound to the length of any matching string. There
1316 may not be any strings of that length that do actually match, but every string
1317 that does match is at least that long.
1318 .sp
1319 PCRE_INFO_NAMECOUNT
1320 PCRE_INFO_NAMEENTRYSIZE
1321 PCRE_INFO_NAMETABLE
1322 .sp
1323 PCRE supports the use of named as well as numbered capturing parentheses. The
1324 names are just an additional way of identifying the parentheses, which still
1325 acquire numbers. Several convenience functions such as
1326 \fBpcre_get_named_substring()\fP are provided for extracting captured
1327 substrings by name. It is also possible to extract the data directly, by first
1328 converting the name to a number in order to access the correct pointers in the
1329 output vector (described with \fBpcre_exec()\fP below). To do the conversion,
1330 you need to use the name-to-number map, which is described by these three
1331 values.
1332 .P
1333 The map consists of a number of fixed-size entries. PCRE_INFO_NAMECOUNT gives
1334 the number of entries, and PCRE_INFO_NAMEENTRYSIZE gives the size of each
1335 entry; both of these return an \fBint\fP value. The entry size depends on the
1336 length of the longest name. PCRE_INFO_NAMETABLE returns a pointer to the first
1337 entry of the table. This is a pointer to \fBchar\fP in the 8-bit library, where
1338 the first two bytes of each entry are the number of the capturing parenthesis,
1339 most significant byte first. In the 16-bit library, the pointer points to
1340 16-bit data units, the first of which contains the parenthesis number.
1341 In the 32-bit library, the pointer points to 32-bit data units, the first of
1342 which contains the parenthesis number. The rest
1343 of the entry is the corresponding name, zero terminated.
1344 .P
1345 The names are in alphabetical order. Duplicate names may appear if (?| is used
1346 to create multiple groups with the same number, as described in the
1347 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
1348 .\" </a>
1349 section on duplicate subpattern numbers
1350 .\"
1351 in the
1352 .\" HREF
1353 \fBpcrepattern\fP
1354 .\"
1355 page. Duplicate names for subpatterns with different numbers are permitted only
1356 if PCRE_DUPNAMES is set. In all cases of duplicate names, they appear in the
1357 table in the order in which they were found in the pattern. In the absence of
1358 (?| this is the order of increasing number; when (?| is used this is not
1359 necessarily the case because later subpatterns may have lower numbers.
1360 .P
1361 As a simple example of the name/number table, consider the following pattern
1362 after compilation by the 8-bit library (assume PCRE_EXTENDED is set, so white
1363 space - including newlines - is ignored):
1364 .sp
1365 .\" JOIN
1366 (?<date> (?<year>(\ed\ed)?\ed\ed) -
1367 (?<month>\ed\ed) - (?<day>\ed\ed) )
1368 .sp
1369 There are four named subpatterns, so the table has four entries, and each entry
1370 in the table is eight bytes long. The table is as follows, with non-printing
1371 bytes shows in hexadecimal, and undefined bytes shown as ??:
1372 .sp
1373 00 01 d a t e 00 ??
1374 00 05 d a y 00 ?? ??
1375 00 04 m o n t h 00
1376 00 02 y e a r 00 ??
1377 .sp
1378 When writing code to extract data from named subpatterns using the
1379 name-to-number map, remember that the length of the entries is likely to be
1380 different for each compiled pattern.
1381 .sp
1382 PCRE_INFO_OKPARTIAL
1383 .sp
1384 Return 1 if the pattern can be used for partial matching with
1385 \fBpcre_exec()\fP, otherwise 0. The fourth argument should point to an
1386 \fBint\fP variable. From release 8.00, this always returns 1, because the
1387 restrictions that previously applied to partial matching have been lifted. The
1388 .\" HREF
1389 \fBpcrepartial\fP
1390 .\"
1391 documentation gives details of partial matching.
1392 .sp
1393 PCRE_INFO_OPTIONS
1394 .sp
1395 Return a copy of the options with which the pattern was compiled. The fourth
1396 argument should point to an \fBunsigned long int\fP variable. These option bits
1397 are those specified in the call to \fBpcre_compile()\fP, modified by any
1398 top-level option settings at the start of the pattern itself. In other words,
1399 they are the options that will be in force when matching starts. For example,
1400 if the pattern /(?im)abc(?-i)d/ is compiled with the PCRE_EXTENDED option, the
1401 result is PCRE_CASELESS, PCRE_MULTILINE, and PCRE_EXTENDED.
1402 .P
1403 A pattern is automatically anchored by PCRE if all of its top-level
1404 alternatives begin with one of the following:
1405 .sp
1406 ^ unless PCRE_MULTILINE is set
1407 \eA always
1408 \eG always
1409 .\" JOIN
1410 .* if PCRE_DOTALL is set and there are no back
1411 references to the subpattern in which .* appears
1412 .sp
1413 For such patterns, the PCRE_ANCHORED bit is set in the options returned by
1414 \fBpcre_fullinfo()\fP.
1415 .sp
1416 PCRE_INFO_SIZE
1417 .sp
1418 Return the size of the compiled pattern in bytes (for both libraries). The
1419 fourth argument should point to a \fBsize_t\fP variable. This value does not
1420 include the size of the \fBpcre\fP structure that is returned by
1421 \fBpcre_compile()\fP. The value that is passed as the argument to
1422 \fBpcre_malloc()\fP when \fBpcre_compile()\fP is getting memory in which to
1423 place the compiled data is the value returned by this option plus the size of
1424 the \fBpcre\fP structure. Studying a compiled pattern, with or without JIT,
1425 does not alter the value returned by this option.
1426 .sp
1427 PCRE_INFO_STUDYSIZE
1428 .sp
1429 Return the size in bytes of the data block pointed to by the \fIstudy_data\fP
1430 field in a \fBpcre_extra\fP block. If \fBpcre_extra\fP is NULL, or there is no
1431 study data, zero is returned. The fourth argument should point to a
1432 \fBsize_t\fP variable. The \fIstudy_data\fP field is set by \fBpcre_study()\fP
1433 to record information that will speed up matching (see the section entitled
1434 .\" HTML <a href="#studyingapattern">
1435 .\" </a>
1436 "Studying a pattern"
1437 .\"
1438 above). The format of the \fIstudy_data\fP block is private, but its length
1439 is made available via this option so that it can be saved and restored (see the
1440 .\" HREF
1441 \fBpcreprecompile\fP
1442 .\"
1443 documentation for details).
1444 .sp
1445 PCRE_INFO_FIRSTCHARACTERFLAGS
1446 .sp
1447 Return information about the first data unit of any matched string, for a
1448 non-anchored pattern. The fourth argument should point to an \fBint\fP
1449 variable.
1450 .P
1451 If there is a fixed first value, for example, the letter "c" from a pattern
1452 such as (cat|cow|coyote), 1 is returned, and the character value can be
1453 retrieved using PCRE_INFO_FIRSTCHARACTER.
1454 .P
1455 If there is no fixed first value, and if either
1456 .sp
1457 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1458 starts with "^", or
1459 .sp
1460 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1461 (if it were set, the pattern would be anchored),
1462 .sp
1463 2 is returned, indicating that the pattern matches only at the start of a
1464 subject string or after any newline within the string. Otherwise 0 is
1465 returned. For anchored patterns, 0 is returned.
1466 .sp
1467 PCRE_INFO_FIRSTCHARACTER
1468 .sp
1469 Return the fixed first character value, if PCRE_INFO_FIRSTCHARACTERFLAGS
1470 returned 1; otherwise returns 0. The fourth argument should point to an
1471 \fBuint_t\fP variable.
1472 .P
1473 In the 8-bit library, the value is always less than 256. In the 16-bit library
1474 the value can be up to 0xffff. In the 32-bit library in UTF-32 mode the value
1475 can be up to 0x10ffff, and up to 0xffffffff when not using UTF-32 mode.
1476 .P
1477 If there is no fixed first value, and if either
1478 .sp
1479 (a) the pattern was compiled with the PCRE_MULTILINE option, and every branch
1480 starts with "^", or
1481 .sp
1482 (b) every branch of the pattern starts with ".*" and PCRE_DOTALL is not set
1483 (if it were set, the pattern would be anchored),
1484 .sp
1485 -1 is returned, indicating that the pattern matches only at the start of a
1486 subject string or after any newline within the string. Otherwise -2 is
1487 returned. For anchored patterns, -2 is returned.
1488 .sp
1489 PCRE_INFO_REQUIREDCHARFLAGS
1490 .sp
1491 Returns 1 if there is a rightmost literal data unit that must exist in any
1492 matched string, other than at its start. The fourth argument should point to
1493 an \fBint\fP variable. If there is no such value, 0 is returned. If returning
1494 1, the character value itself can be retrieved using PCRE_INFO_REQUIREDCHAR.
1495 .P
1496 For anchored patterns, a last literal value is recorded only if it follows
1497 something of variable length. For example, for the pattern /^a\ed+z\ed+/ the
1498 returned value 1 (with "z" returned from PCRE_INFO_REQUIREDCHAR), but for
1499 /^a\edz\ed/ the returned value is 0.
1500 .sp
1501 PCRE_INFO_REQUIREDCHAR
1502 .sp
1503 Return the value of the rightmost literal data unit that must exist in any
1504 matched string, other than at its start, if such a value has been recorded. The
1505 fourth argument should point to an \fBuint32_t\fP variable. If there is no such
1506 value, 0 is returned.
1507 .
1508 .
1509 .SH "REFERENCE COUNTS"
1510 .rs
1511 .sp
1512 .B int pcre_refcount(pcre *\fIcode\fP, int \fIadjust\fP);
1513 .PP
1514 The \fBpcre_refcount()\fP function is used to maintain a reference count in the
1515 data block that contains a compiled pattern. It is provided for the benefit of
1516 applications that operate in an object-oriented manner, where different parts
1517 of the application may be using the same compiled pattern, but you want to free
1518 the block when they are all done.
1519 .P
1520 When a pattern is compiled, the reference count field is initialized to zero.
1521 It is changed only by calling this function, whose action is to add the
1522 \fIadjust\fP value (which may be positive or negative) to it. The yield of the
1523 function is the new value. However, the value of the count is constrained to
1524 lie between 0 and 65535, inclusive. If the new value is outside these limits,
1525 it is forced to the appropriate limit value.
1526 .P
1527 Except when it is zero, the reference count is not correctly preserved if a
1528 pattern is compiled on one host and then transferred to a host whose byte-order
1529 is different. (This seems a highly unlikely scenario.)
1530 .
1531 .
1532 .SH "MATCHING A PATTERN: THE TRADITIONAL FUNCTION"
1533 .rs
1534 .sp
1535 .B int pcre_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
1536 .ti +5n
1537 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
1538 .ti +5n
1539 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP);
1540 .P
1541 The function \fBpcre_exec()\fP is called to match a subject string against a
1542 compiled pattern, which is passed in the \fIcode\fP argument. If the
1543 pattern was studied, the result of the study should be passed in the
1544 \fIextra\fP argument. You can call \fBpcre_exec()\fP with the same \fIcode\fP
1545 and \fIextra\fP arguments as many times as you like, in order to match
1546 different subject strings with the same pattern.
1547 .P
1548 This function is the main matching facility of the library, and it operates in
1549 a Perl-like manner. For specialist use there is also an alternative matching
1550 function, which is described
1551 .\" HTML <a href="#dfamatch">
1552 .\" </a>
1553 below
1554 .\"
1555 in the section about the \fBpcre_dfa_exec()\fP function.
1556 .P
1557 In most applications, the pattern will have been compiled (and optionally
1558 studied) in the same process that calls \fBpcre_exec()\fP. However, it is
1559 possible to save compiled patterns and study data, and then use them later
1560 in different processes, possibly even on different hosts. For a discussion
1561 about this, see the
1562 .\" HREF
1563 \fBpcreprecompile\fP
1564 .\"
1565 documentation.
1566 .P
1567 Here is an example of a simple call to \fBpcre_exec()\fP:
1568 .sp
1569 int rc;
1570 int ovector[30];
1571 rc = pcre_exec(
1572 re, /* result of pcre_compile() */
1573 NULL, /* we didn't study the pattern */
1574 "some string", /* the subject string */
1575 11, /* the length of the subject string */
1576 0, /* start at offset 0 in the subject */
1577 0, /* default options */
1578 ovector, /* vector of integers for substring information */
1579 30); /* number of elements (NOT size in bytes) */
1580 .
1581 .
1582 .\" HTML <a name="extradata"></a>
1583 .SS "Extra data for \fBpcre_exec()\fR"
1584 .rs
1585 .sp
1586 If the \fIextra\fP argument is not NULL, it must point to a \fBpcre_extra\fP
1587 data block. The \fBpcre_study()\fP function returns such a block (when it
1588 doesn't return NULL), but you can also create one for yourself, and pass
1589 additional information in it. The \fBpcre_extra\fP block contains the following
1590 fields (not necessarily in this order):
1591 .sp
1592 unsigned long int \fIflags\fP;
1593 void *\fIstudy_data\fP;
1594 void *\fIexecutable_jit\fP;
1595 unsigned long int \fImatch_limit\fP;
1596 unsigned long int \fImatch_limit_recursion\fP;
1597 void *\fIcallout_data\fP;
1598 const unsigned char *\fItables\fP;
1599 unsigned char **\fImark\fP;
1600 .sp
1601 In the 16-bit version of this structure, the \fImark\fP field has type
1602 "PCRE_UCHAR16 **".
1603 .sp
1604 In the 32-bit version of this structure, the \fImark\fP field has type
1605 "PCRE_UCHAR32 **".
1606 .P
1607 The \fIflags\fP field is used to specify which of the other fields are set. The
1608 flag bits are:
1609 .sp
1610 PCRE_EXTRA_CALLOUT_DATA
1611 PCRE_EXTRA_EXECUTABLE_JIT
1612 PCRE_EXTRA_MARK
1613 PCRE_EXTRA_MATCH_LIMIT
1614 PCRE_EXTRA_MATCH_LIMIT_RECURSION
1615 PCRE_EXTRA_STUDY_DATA
1616 PCRE_EXTRA_TABLES
1617 .sp
1618 Other flag bits should be set to zero. The \fIstudy_data\fP field and sometimes
1619 the \fIexecutable_jit\fP field are set in the \fBpcre_extra\fP block that is
1620 returned by \fBpcre_study()\fP, together with the appropriate flag bits. You
1621 should not set these yourself, but you may add to the block by setting other
1622 fields and their corresponding flag bits.
1623 .P
1624 The \fImatch_limit\fP field provides a means of preventing PCRE from using up a
1625 vast amount of resources when running patterns that are not going to match,
1626 but which have a very large number of possibilities in their search trees. The
1627 classic example is a pattern that uses nested unlimited repeats.
1628 .P
1629 Internally, \fBpcre_exec()\fP uses a function called \fBmatch()\fP, which it
1630 calls repeatedly (sometimes recursively). The limit set by \fImatch_limit\fP is
1631 imposed on the number of times this function is called during a match, which
1632 has the effect of limiting the amount of backtracking that can take place. For
1633 patterns that are not anchored, the count restarts from zero for each position
1634 in the subject string.
1635 .P
1636 When \fBpcre_exec()\fP is called with a pattern that was successfully studied
1637 with a JIT option, the way that the matching is executed is entirely different.
1638 However, there is still the possibility of runaway matching that goes on for a
1639 very long time, and so the \fImatch_limit\fP value is also used in this case
1640 (but in a different way) to limit how long the matching can continue.
1641 .P
1642 The default value for the limit can be set when PCRE is built; the default
1643 default is 10 million, which handles all but the most extreme cases. You can
1644 override the default by suppling \fBpcre_exec()\fP with a \fBpcre_extra\fP
1645 block in which \fImatch_limit\fP is set, and PCRE_EXTRA_MATCH_LIMIT is set in
1646 the \fIflags\fP field. If the limit is exceeded, \fBpcre_exec()\fP returns
1647 PCRE_ERROR_MATCHLIMIT.
1648 .P
1649 The \fImatch_limit_recursion\fP field is similar to \fImatch_limit\fP, but
1650 instead of limiting the total number of times that \fBmatch()\fP is called, it
1651 limits the depth of recursion. The recursion depth is a smaller number than the
1652 total number of calls, because not all calls to \fBmatch()\fP are recursive.
1653 This limit is of use only if it is set smaller than \fImatch_limit\fP.
1654 .P
1655 Limiting the recursion depth limits the amount of machine stack that can be
1656 used, or, when PCRE has been compiled to use memory on the heap instead of the
1657 stack, the amount of heap memory that can be used. This limit is not relevant,
1658 and is ignored, when matching is done using JIT compiled code.
1659 .P
1660 The default value for \fImatch_limit_recursion\fP can be set when PCRE is
1661 built; the default default is the same value as the default for
1662 \fImatch_limit\fP. You can override the default by suppling \fBpcre_exec()\fP
1663 with a \fBpcre_extra\fP block in which \fImatch_limit_recursion\fP is set, and
1664 PCRE_EXTRA_MATCH_LIMIT_RECURSION is set in the \fIflags\fP field. If the limit
1665 is exceeded, \fBpcre_exec()\fP returns PCRE_ERROR_RECURSIONLIMIT.
1666 .P
1667 The \fIcallout_data\fP field is used in conjunction with the "callout" feature,
1668 and is described in the
1669 .\" HREF
1670 \fBpcrecallout\fP
1671 .\"
1672 documentation.
1673 .P
1674 The \fItables\fP field is used to pass a character tables pointer to
1675 \fBpcre_exec()\fP; this overrides the value that is stored with the compiled
1676 pattern. A non-NULL value is stored with the compiled pattern only if custom
1677 tables were supplied to \fBpcre_compile()\fP via its \fItableptr\fP argument.
1678 If NULL is passed to \fBpcre_exec()\fP using this mechanism, it forces PCRE's
1679 internal tables to be used. This facility is helpful when re-using patterns
1680 that have been saved after compiling with an external set of tables, because
1681 the external tables might be at a different address when \fBpcre_exec()\fP is
1682 called. See the
1683 .\" HREF
1684 \fBpcreprecompile\fP
1685 .\"
1686 documentation for a discussion of saving compiled patterns for later use.
1687 .P
1688 If PCRE_EXTRA_MARK is set in the \fIflags\fP field, the \fImark\fP field must
1689 be set to point to a suitable variable. If the pattern contains any
1690 backtracking control verbs such as (*MARK:NAME), and the execution ends up with
1691 a name to pass back, a pointer to the name string (zero terminated) is placed
1692 in the variable pointed to by the \fImark\fP field. The names are within the
1693 compiled pattern; if you wish to retain such a name you must copy it before
1694 freeing the memory of a compiled pattern. If there is no name to pass back, the
1695 variable pointed to by the \fImark\fP field is set to NULL. For details of the
1696 backtracking control verbs, see the section entitled
1697 .\" HTML <a href="pcrepattern#backtrackcontrol">
1698 .\" </a>
1699 "Backtracking control"
1700 .\"
1701 in the
1702 .\" HREF
1703 \fBpcrepattern\fP
1704 .\"
1705 documentation.
1706 .
1707 .
1708 .\" HTML <a name="execoptions"></a>
1709 .SS "Option bits for \fBpcre_exec()\fP"
1710 .rs
1711 .sp
1712 The unused bits of the \fIoptions\fP argument for \fBpcre_exec()\fP must be
1713 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
1714 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
1715 PCRE_NO_START_OPTIMIZE, PCRE_NO_UTF8_CHECK, PCRE_PARTIAL_HARD, and
1716 PCRE_PARTIAL_SOFT.
1717 .P
1718 If the pattern was successfully studied with one of the just-in-time (JIT)
1719 compile options, the only supported options for JIT execution are
1720 PCRE_NO_UTF8_CHECK, PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY,
1721 PCRE_NOTEMPTY_ATSTART, PCRE_PARTIAL_HARD, and PCRE_PARTIAL_SOFT. If an
1722 unsupported option is used, JIT execution is disabled and the normal
1723 interpretive code in \fBpcre_exec()\fP is run.
1724 .sp
1725 PCRE_ANCHORED
1726 .sp
1727 The PCRE_ANCHORED option limits \fBpcre_exec()\fP to matching at the first
1728 matching position. If a pattern was compiled with PCRE_ANCHORED, or turned out
1729 to be anchored by virtue of its contents, it cannot be made unachored at
1730 matching time.
1731 .sp
1732 PCRE_BSR_ANYCRLF
1733 PCRE_BSR_UNICODE
1734 .sp
1735 These options (which are mutually exclusive) control what the \eR escape
1736 sequence matches. The choice is either to match only CR, LF, or CRLF, or to
1737 match any Unicode newline sequence. These options override the choice that was
1738 made or defaulted when the pattern was compiled.
1739 .sp
1740 PCRE_NEWLINE_CR
1741 PCRE_NEWLINE_LF
1742 PCRE_NEWLINE_CRLF
1743 PCRE_NEWLINE_ANYCRLF
1744 PCRE_NEWLINE_ANY
1745 .sp
1746 These options override the newline definition that was chosen or defaulted when
1747 the pattern was compiled. For details, see the description of
1748 \fBpcre_compile()\fP above. During matching, the newline choice affects the
1749 behaviour of the dot, circumflex, and dollar metacharacters. It may also alter
1750 the way the match position is advanced after a match failure for an unanchored
1751 pattern.
1752 .P
1753 When PCRE_NEWLINE_CRLF, PCRE_NEWLINE_ANYCRLF, or PCRE_NEWLINE_ANY is set, and a
1754 match attempt for an unanchored pattern fails when the current position is at a
1755 CRLF sequence, and the pattern contains no explicit matches for CR or LF
1756 characters, the match position is advanced by two characters instead of one, in
1757 other words, to after the CRLF.
1758 .P
1759 The above rule is a compromise that makes the most common cases work as
1760 expected. For example, if the pattern is .+A (and the PCRE_DOTALL option is not
1761 set), it does not match the string "\er\enA" because, after failing at the
1762 start, it skips both the CR and the LF before retrying. However, the pattern
1763 [\er\en]A does match that string, because it contains an explicit CR or LF
1764 reference, and so advances only by one character after the first failure.
1765 .P
1766 An explicit match for CR of LF is either a literal appearance of one of those
1767 characters, or one of the \er or \en escape sequences. Implicit matches such as
1768 [^X] do not count, nor does \es (which includes CR and LF in the characters
1769 that it matches).
1770 .P
1771 Notwithstanding the above, anomalous effects may still occur when CRLF is a
1772 valid newline sequence and explicit \er or \en escapes appear in the pattern.
1773 .sp
1774 PCRE_NOTBOL
1775 .sp
1776 This option specifies that first character of the subject string is not the
1777 beginning of a line, so the circumflex metacharacter should not match before
1778 it. Setting this without PCRE_MULTILINE (at compile time) causes circumflex
1779 never to match. This option affects only the behaviour of the circumflex
1780 metacharacter. It does not affect \eA.
1781 .sp
1782 PCRE_NOTEOL
1783 .sp
1784 This option specifies that the end of the subject string is not the end of a
1785 line, so the dollar metacharacter should not match it nor (except in multiline
1786 mode) a newline immediately before it. Setting this without PCRE_MULTILINE (at
1787 compile time) causes dollar never to match. This option affects only the
1788 behaviour of the dollar metacharacter. It does not affect \eZ or \ez.
1789 .sp
1790 PCRE_NOTEMPTY
1791 .sp
1792 An empty string is not considered to be a valid match if this option is set. If
1793 there are alternatives in the pattern, they are tried. If all the alternatives
1794 match the empty string, the entire match fails. For example, if the pattern
1795 .sp
1796 a?b?
1797 .sp
1798 is applied to a string not beginning with "a" or "b", it matches an empty
1799 string at the start of the subject. With PCRE_NOTEMPTY set, this match is not
1800 valid, so PCRE searches further into the string for occurrences of "a" or "b".
1801 .sp
1802 PCRE_NOTEMPTY_ATSTART
1803 .sp
1804 This is like PCRE_NOTEMPTY, except that an empty string match that is not at
1805 the start of the subject is permitted. If the pattern is anchored, such a match
1806 can occur only if the pattern contains \eK.
1807 .P
1808 Perl has no direct equivalent of PCRE_NOTEMPTY or PCRE_NOTEMPTY_ATSTART, but it
1809 does make a special case of a pattern match of the empty string within its
1810 \fBsplit()\fP function, and when using the /g modifier. It is possible to
1811 emulate Perl's behaviour after matching a null string by first trying the match
1812 again at the same offset with PCRE_NOTEMPTY_ATSTART and PCRE_ANCHORED, and then
1813 if that fails, by advancing the starting offset (see below) and trying an
1814 ordinary match again. There is some code that demonstrates how to do this in
1815 the
1816 .\" HREF
1817 \fBpcredemo\fP
1818 .\"
1819 sample program. In the most general case, you have to check to see if the
1820 newline convention recognizes CRLF as a newline, and if so, and the current
1821 character is CR followed by LF, advance the starting offset by two characters
1822 instead of one.
1823 .sp
1824 PCRE_NO_START_OPTIMIZE
1825 .sp
1826 There are a number of optimizations that \fBpcre_exec()\fP uses at the start of
1827 a match, in order to speed up the process. For example, if it is known that an
1828 unanchored match must start with a specific character, it searches the subject
1829 for that character, and fails immediately if it cannot find it, without
1830 actually running the main matching function. This means that a special item
1831 such as (*COMMIT) at the start of a pattern is not considered until after a
1832 suitable starting point for the match has been found. When callouts or (*MARK)
1833 items are in use, these "start-up" optimizations can cause them to be skipped
1834 if the pattern is never actually used. The start-up optimizations are in effect
1835 a pre-scan of the subject that takes place before the pattern is run.
1836 .P
1837 The PCRE_NO_START_OPTIMIZE option disables the start-up optimizations, possibly
1838 causing performance to suffer, but ensuring that in cases where the result is
1839 "no match", the callouts do occur, and that items such as (*COMMIT) and (*MARK)
1840 are considered at every possible starting position in the subject string. If
1841 PCRE_NO_START_OPTIMIZE is set at compile time, it cannot be unset at matching
1842 time. The use of PCRE_NO_START_OPTIMIZE disables JIT execution; when it is set,
1843 matching is always done using interpretively.
1844 .P
1845 Setting PCRE_NO_START_OPTIMIZE can change the outcome of a matching operation.
1846 Consider the pattern
1847 .sp
1848 (*COMMIT)ABC
1849 .sp
1850 When this is compiled, PCRE records the fact that a match must start with the
1851 character "A". Suppose the subject string is "DEFABC". The start-up
1852 optimization scans along the subject, finds "A" and runs the first match
1853 attempt from there. The (*COMMIT) item means that the pattern must match the
1854 current starting position, which in this case, it does. However, if the same
1855 match is run with PCRE_NO_START_OPTIMIZE set, the initial scan along the
1856 subject string does not happen. The first match attempt is run starting from
1857 "D" and when this fails, (*COMMIT) prevents any further matches being tried, so
1858 the overall result is "no match". If the pattern is studied, more start-up
1859 optimizations may be used. For example, a minimum length for the subject may be
1860 recorded. Consider the pattern
1861 .sp
1862 (*MARK:A)(X|Y)
1863 .sp
1864 The minimum length for a match is one character. If the subject is "ABC", there
1865 will be attempts to match "ABC", "BC", "C", and then finally an empty string.
1866 If the pattern is studied, the final attempt does not take place, because PCRE
1867 knows that the subject is too short, and so the (*MARK) is never encountered.
1868 In this case, studying the pattern does not affect the overall match result,
1869 which is still "no match", but it does affect the auxiliary information that is
1870 returned.
1871 .sp
1872 PCRE_NO_UTF8_CHECK
1873 .sp
1874 When PCRE_UTF8 is set at compile time, the validity of the subject as a UTF-8
1875 string is automatically checked when \fBpcre_exec()\fP is subsequently called.
1876 The entire string is checked before any other processing takes place. The value
1877 of \fIstartoffset\fP is also checked to ensure that it points to the start of a
1878 UTF-8 character. There is a discussion about the
1879 .\" HTML <a href="pcreunicode.html#utf8strings">
1880 .\" </a>
1881 validity of UTF-8 strings
1882 .\"
1883 in the
1884 .\" HREF
1885 \fBpcreunicode\fP
1886 .\"
1887 page. If an invalid sequence of bytes is found, \fBpcre_exec()\fP returns the
1888 error PCRE_ERROR_BADUTF8 or, if PCRE_PARTIAL_HARD is set and the problem is a
1889 truncated character at the end of the subject, PCRE_ERROR_SHORTUTF8. In both
1890 cases, information about the precise nature of the error may also be returned
1891 (see the descriptions of these errors in the section entitled \fIError return
1892 values from\fP \fBpcre_exec()\fP
1893 .\" HTML <a href="#errorlist">
1894 .\" </a>
1895 below).
1896 .\"
1897 If \fIstartoffset\fP contains a value that does not point to the start of a
1898 UTF-8 character (or to the end of the subject), PCRE_ERROR_BADUTF8_OFFSET is
1899 returned.
1900 .P
1901 If you already know that your subject is valid, and you want to skip these
1902 checks for performance reasons, you can set the PCRE_NO_UTF8_CHECK option when
1903 calling \fBpcre_exec()\fP. You might want to do this for the second and
1904 subsequent calls to \fBpcre_exec()\fP if you are making repeated calls to find
1905 all the matches in a single subject string. However, you should be sure that
1906 the value of \fIstartoffset\fP points to the start of a character (or the end
1907 of the subject). When PCRE_NO_UTF8_CHECK is set, the effect of passing an
1908 invalid string as a subject or an invalid value of \fIstartoffset\fP is
1909 undefined. Your program may crash.
1910 .sp
1911 PCRE_PARTIAL_HARD
1912 PCRE_PARTIAL_SOFT
1913 .sp
1914 These options turn on the partial matching feature. For backwards
1915 compatibility, PCRE_PARTIAL is a synonym for PCRE_PARTIAL_SOFT. A partial match
1916 occurs if the end of the subject string is reached successfully, but there are
1917 not enough subject characters to complete the match. If this happens when
1918 PCRE_PARTIAL_SOFT (but not PCRE_PARTIAL_HARD) is set, matching continues by
1919 testing any remaining alternatives. Only if no complete match can be found is
1920 PCRE_ERROR_PARTIAL returned instead of PCRE_ERROR_NOMATCH. In other words,
1921 PCRE_PARTIAL_SOFT says that the caller is prepared to handle a partial match,
1922 but only if no complete match can be found.
1923 .P
1924 If PCRE_PARTIAL_HARD is set, it overrides PCRE_PARTIAL_SOFT. In this case, if a
1925 partial match is found, \fBpcre_exec()\fP immediately returns
1926 PCRE_ERROR_PARTIAL, without considering any other alternatives. In other words,
1927 when PCRE_PARTIAL_HARD is set, a partial match is considered to be more
1928 important that an alternative complete match.
1929 .P
1930 In both cases, the portion of the string that was inspected when the partial
1931 match was found is set as the first matching string. There is a more detailed
1932 discussion of partial and multi-segment matching, with examples, in the
1933 .\" HREF
1934 \fBpcrepartial\fP
1935 .\"
1936 documentation.
1937 .
1938 .
1939 .SS "The string to be matched by \fBpcre_exec()\fP"
1940 .rs
1941 .sp
1942 The subject string is passed to \fBpcre_exec()\fP as a pointer in
1943 \fIsubject\fP, a length in bytes in \fIlength\fP, and a starting byte offset
1944 in \fIstartoffset\fP. If this is negative or greater than the length of the
1945 subject, \fBpcre_exec()\fP returns PCRE_ERROR_BADOFFSET. When the starting
1946 offset is zero, the search for a match starts at the beginning of the subject,
1947 and this is by far the most common case. In UTF-8 mode, the byte offset must
1948 point to the start of a UTF-8 character (or the end of the subject). Unlike the
1949 pattern string, the subject may contain binary zero bytes.
1950 .P
1951 A non-zero starting offset is useful when searching for another match in the
1952 same subject by calling \fBpcre_exec()\fP again after a previous success.
1953 Setting \fIstartoffset\fP differs from just passing over a shortened string and
1954 setting PCRE_NOTBOL in the case of a pattern that begins with any kind of
1955 lookbehind. For example, consider the pattern
1956 .sp
1957 \eBiss\eB
1958 .sp
1959 which finds occurrences of "iss" in the middle of words. (\eB matches only if
1960 the current position in the subject is not a word boundary.) When applied to
1961 the string "Mississipi" the first call to \fBpcre_exec()\fP finds the first
1962 occurrence. If \fBpcre_exec()\fP is called again with just the remainder of the
1963 subject, namely "issipi", it does not match, because \eB is always false at the
1964 start of the subject, which is deemed to be a word boundary. However, if
1965 \fBpcre_exec()\fP is passed the entire string again, but with \fIstartoffset\fP
1966 set to 4, it finds the second occurrence of "iss" because it is able to look
1967 behind the starting point to discover that it is preceded by a letter.
1968 .P
1969 Finding all the matches in a subject is tricky when the pattern can match an
1970 empty string. It is possible to emulate Perl's /g behaviour by first trying the
1971 match again at the same offset, with the PCRE_NOTEMPTY_ATSTART and
1972 PCRE_ANCHORED options, and then if that fails, advancing the starting offset
1973 and trying an ordinary match again. There is some code that demonstrates how to
1974 do this in the
1975 .\" HREF
1976 \fBpcredemo\fP
1977 .\"
1978 sample program. In the most general case, you have to check to see if the
1979 newline convention recognizes CRLF as a newline, and if so, and the current
1980 character is CR followed by LF, advance the starting offset by two characters
1981 instead of one.
1982 .P
1983 If a non-zero starting offset is passed when the pattern is anchored, one
1984 attempt to match at the given offset is made. This can only succeed if the
1985 pattern does not require the match to be at the start of the subject.
1986 .
1987 .
1988 .SS "How \fBpcre_exec()\fP returns captured substrings"
1989 .rs
1990 .sp
1991 In general, a pattern matches a certain portion of the subject, and in
1992 addition, further substrings from the subject may be picked out by parts of the
1993 pattern. Following the usage in Jeffrey Friedl's book, this is called
1994 "capturing" in what follows, and the phrase "capturing subpattern" is used for
1995 a fragment of a pattern that picks out a substring. PCRE supports several other
1996 kinds of parenthesized subpattern that do not cause substrings to be captured.
1997 .P
1998 Captured substrings are returned to the caller via a vector of integers whose
1999 address is passed in \fIovector\fP. The number of elements in the vector is
2000 passed in \fIovecsize\fP, which must be a non-negative number. \fBNote\fP: this
2001 argument is NOT the size of \fIovector\fP in bytes.
2002 .P
2003 The first two-thirds of the vector is used to pass back captured substrings,
2004 each substring using a pair of integers. The remaining third of the vector is
2005 used as workspace by \fBpcre_exec()\fP while matching capturing subpatterns,
2006 and is not available for passing back information. The number passed in
2007 \fIovecsize\fP should always be a multiple of three. If it is not, it is
2008 rounded down.
2009 .P
2010 When a match is successful, information about captured substrings is returned
2011 in pairs of integers, starting at the beginning of \fIovector\fP, and
2012 continuing up to two-thirds of its length at the most. The first element of
2013 each pair is set to the byte offset of the first character in a substring, and
2014 the second is set to the byte offset of the first character after the end of a
2015 substring. \fBNote\fP: these values are always byte offsets, even in UTF-8
2016 mode. They are not character counts.
2017 .P
2018 The first pair of integers, \fIovector[0]\fP and \fIovector[1]\fP, identify the
2019 portion of the subject string matched by the entire pattern. The next pair is
2020 used for the first capturing subpattern, and so on. The value returned by
2021 \fBpcre_exec()\fP is one more than the highest numbered pair that has been set.
2022 For example, if two substrings have been captured, the returned value is 3. If
2023 there are no capturing subpatterns, the return value from a successful match is
2024 1, indicating that just the first pair of offsets has been set.
2025 .P
2026 If a capturing subpattern is matched repeatedly, it is the last portion of the
2027 string that it matched that is returned.
2028 .P
2029 If the vector is too small to hold all the captured substring offsets, it is
2030 used as far as possible (up to two-thirds of its length), and the function
2031 returns a value of zero. If neither the actual string matched nor any captured
2032 substrings are of interest, \fBpcre_exec()\fP may be called with \fIovector\fP
2033 passed as NULL and \fIovecsize\fP as zero. However, if the pattern contains
2034 back references and the \fIovector\fP is not big enough to remember the related
2035 substrings, PCRE has to get additional memory for use during matching. Thus it
2036 is usually advisable to supply an \fIovector\fP of reasonable size.
2037 .P
2038 There are some cases where zero is returned (indicating vector overflow) when
2039 in fact the vector is exactly the right size for the final match. For example,
2040 consider the pattern
2041 .sp
2042 (a)(?:(b)c|bd)
2043 .sp
2044 If a vector of 6 elements (allowing for only 1 captured substring) is given
2045 with subject string "abd", \fBpcre_exec()\fP will try to set the second
2046 captured string, thereby recording a vector overflow, before failing to match
2047 "c" and backing up to try the second alternative. The zero return, however,
2048 does correctly indicate that the maximum number of slots (namely 2) have been
2049 filled. In similar cases where there is temporary overflow, but the final
2050 number of used slots is actually less than the maximum, a non-zero value is
2051 returned.
2052 .P
2053 The \fBpcre_fullinfo()\fP function can be used to find out how many capturing
2054 subpatterns there are in a compiled pattern. The smallest size for
2055 \fIovector\fP that will allow for \fIn\fP captured substrings, in addition to
2056 the offsets of the substring matched by the whole pattern, is (\fIn\fP+1)*3.
2057 .P
2058 It is possible for capturing subpattern number \fIn+1\fP to match some part of
2059 the subject when subpattern \fIn\fP has not been used at all. For example, if
2060 the string "abc" is matched against the pattern (a|(z))(bc) the return from the
2061 function is 4, and subpatterns 1 and 3 are matched, but 2 is not. When this
2062 happens, both values in the offset pairs corresponding to unused subpatterns
2063 are set to -1.
2064 .P
2065 Offset values that correspond to unused subpatterns at the end of the
2066 expression are also set to -1. For example, if the string "abc" is matched
2067 against the pattern (abc)(x(yz)?)? subpatterns 2 and 3 are not matched. The
2068 return from the function is 2, because the highest used capturing subpattern
2069 number is 1, and the offsets for for the second and third capturing subpatterns
2070 (assuming the vector is large enough, of course) are set to -1.
2071 .P
2072 \fBNote\fP: Elements in the first two-thirds of \fIovector\fP that do not
2073 correspond to capturing parentheses in the pattern are never changed. That is,
2074 if a pattern contains \fIn\fP capturing parentheses, no more than
2075 \fIovector[0]\fP to \fIovector[2n+1]\fP are set by \fBpcre_exec()\fP. The other
2076 elements (in the first two-thirds) retain whatever values they previously had.
2077 .P
2078 Some convenience functions are provided for extracting the captured substrings
2079 as separate strings. These are described below.
2080 .
2081 .
2082 .\" HTML <a name="errorlist"></a>
2083 .SS "Error return values from \fBpcre_exec()\fP"
2084 .rs
2085 .sp
2086 If \fBpcre_exec()\fP fails, it returns a negative number. The following are
2087 defined in the header file:
2088 .sp
2089 PCRE_ERROR_NOMATCH (-1)
2090 .sp
2091 The subject string did not match the pattern.
2092 .sp
2093 PCRE_ERROR_NULL (-2)
2094 .sp
2095 Either \fIcode\fP or \fIsubject\fP was passed as NULL, or \fIovector\fP was
2096 NULL and \fIovecsize\fP was not zero.
2097 .sp
2098 PCRE_ERROR_BADOPTION (-3)
2099 .sp
2100 An unrecognized bit was set in the \fIoptions\fP argument.
2101 .sp
2102 PCRE_ERROR_BADMAGIC (-4)
2103 .sp
2104 PCRE stores a 4-byte "magic number" at the start of the compiled code, to catch
2105 the case when it is passed a junk pointer and to detect when a pattern that was
2106 compiled in an environment of one endianness is run in an environment with the
2107 other endianness. This is the error that PCRE gives when the magic number is
2108 not present.
2109 .sp
2110 PCRE_ERROR_UNKNOWN_OPCODE (-5)
2111 .sp
2112 While running the pattern match, an unknown item was encountered in the
2113 compiled pattern. This error could be caused by a bug in PCRE or by overwriting
2114 of the compiled pattern.
2115 .sp
2116 PCRE_ERROR_NOMEMORY (-6)
2117 .sp
2118 If a pattern contains back references, but the \fIovector\fP that is passed to
2119 \fBpcre_exec()\fP is not big enough to remember the referenced substrings, PCRE
2120 gets a block of memory at the start of matching to use for this purpose. If the
2121 call via \fBpcre_malloc()\fP fails, this error is given. The memory is
2122 automatically freed at the end of matching.
2123 .P
2124 This error is also given if \fBpcre_stack_malloc()\fP fails in
2125 \fBpcre_exec()\fP. This can happen only when PCRE has been compiled with
2126 \fB--disable-stack-for-recursion\fP.
2127 .sp
2128 PCRE_ERROR_NOSUBSTRING (-7)
2129 .sp
2130 This error is used by the \fBpcre_copy_substring()\fP,
2131 \fBpcre_get_substring()\fP, and \fBpcre_get_substring_list()\fP functions (see
2132 below). It is never returned by \fBpcre_exec()\fP.
2133 .sp
2134 PCRE_ERROR_MATCHLIMIT (-8)
2135 .sp
2136 The backtracking limit, as specified by the \fImatch_limit\fP field in a
2137 \fBpcre_extra\fP structure (or defaulted) was reached. See the description
2138 above.
2139 .sp
2140 PCRE_ERROR_CALLOUT (-9)
2141 .sp
2142 This error is never generated by \fBpcre_exec()\fP itself. It is provided for
2143 use by callout functions that want to yield a distinctive error code. See the
2144 .\" HREF
2145 \fBpcrecallout\fP
2146 .\"
2147 documentation for details.
2148 .sp
2149 PCRE_ERROR_BADUTF8 (-10)
2150 .sp
2151 A string that contains an invalid UTF-8 byte sequence was passed as a subject,
2152 and the PCRE_NO_UTF8_CHECK option was not set. If the size of the output vector
2153 (\fIovecsize\fP) is at least 2, the byte offset to the start of the the invalid
2154 UTF-8 character is placed in the first element, and a reason code is placed in
2155 the second element. The reason codes are listed in the
2156 .\" HTML <a href="#badutf8reasons">
2157 .\" </a>
2158 following section.
2159 .\"
2160 For backward compatibility, if PCRE_PARTIAL_HARD is set and the problem is a
2161 truncated UTF-8 character at the end of the subject (reason codes 1 to 5),
2162 PCRE_ERROR_SHORTUTF8 is returned instead of PCRE_ERROR_BADUTF8.
2163 .sp
2164 PCRE_ERROR_BADUTF8_OFFSET (-11)
2165 .sp
2166 The UTF-8 byte sequence that was passed as a subject was checked and found to
2167 be valid (the PCRE_NO_UTF8_CHECK option was not set), but the value of
2168 \fIstartoffset\fP did not point to the beginning of a UTF-8 character or the
2169 end of the subject.
2170 .sp
2171 PCRE_ERROR_PARTIAL (-12)
2172 .sp
2173 The subject string did not match, but it did match partially. See the
2174 .\" HREF
2175 \fBpcrepartial\fP
2176 .\"
2177 documentation for details of partial matching.
2178 .sp
2179 PCRE_ERROR_BADPARTIAL (-13)
2180 .sp
2181 This code is no longer in use. It was formerly returned when the PCRE_PARTIAL
2182 option was used with a compiled pattern containing items that were not
2183 supported for partial matching. From release 8.00 onwards, there are no
2184 restrictions on partial matching.
2185 .sp
2186 PCRE_ERROR_INTERNAL (-14)
2187 .sp
2188 An unexpected internal error has occurred. This error could be caused by a bug
2189 in PCRE or by overwriting of the compiled pattern.
2190 .sp
2191 PCRE_ERROR_BADCOUNT (-15)
2192 .sp
2193 This error is given if the value of the \fIovecsize\fP argument is negative.
2194 .sp
2195 PCRE_ERROR_RECURSIONLIMIT (-21)
2196 .sp
2197 The internal recursion limit, as specified by the \fImatch_limit_recursion\fP
2198 field in a \fBpcre_extra\fP structure (or defaulted) was reached. See the
2199 description above.
2200 .sp
2201 PCRE_ERROR_BADNEWLINE (-23)
2202 .sp
2203 An invalid combination of PCRE_NEWLINE_\fIxxx\fP options was given.
2204 .sp
2205 PCRE_ERROR_BADOFFSET (-24)
2206 .sp
2207 The value of \fIstartoffset\fP was negative or greater than the length of the
2208 subject, that is, the value in \fIlength\fP.
2209 .sp
2210 PCRE_ERROR_SHORTUTF8 (-25)
2211 .sp
2212 This error is returned instead of PCRE_ERROR_BADUTF8 when the subject string
2213 ends with a truncated UTF-8 character and the PCRE_PARTIAL_HARD option is set.
2214 Information about the failure is returned as for PCRE_ERROR_BADUTF8. It is in
2215 fact sufficient to detect this case, but this special error code for
2216 PCRE_PARTIAL_HARD precedes the implementation of returned information; it is
2217 retained for backwards compatibility.
2218 .sp
2219 PCRE_ERROR_RECURSELOOP (-26)
2220 .sp
2221 This error is returned when \fBpcre_exec()\fP detects a recursion loop within
2222 the pattern. Specifically, it means that either the whole pattern or a
2223 subpattern has been called recursively for the second time at the same position
2224 in the subject string. Some simple patterns that might do this are detected and
2225 faulted at compile time, but more complicated cases, in particular mutual
2226 recursions between two different subpatterns, cannot be detected until run
2227 time.
2228 .sp
2229 PCRE_ERROR_JIT_STACKLIMIT (-27)
2230 .sp
2231 This error is returned when a pattern that was successfully studied using a
2232 JIT compile option is being matched, but the memory available for the
2233 just-in-time processing stack is not large enough. See the
2234 .\" HREF
2235 \fBpcrejit\fP
2236 .\"
2237 documentation for more details.
2238 .sp
2239 PCRE_ERROR_BADMODE (-28)
2240 .sp
2241 This error is given if a pattern that was compiled by the 8-bit library is
2242 passed to a 16-bit or 32-bit library function, or vice versa.
2243 .sp
2244 PCRE_ERROR_BADENDIANNESS (-29)
2245 .sp
2246 This error is given if a pattern that was compiled and saved is reloaded on a
2247 host with different endianness. The utility function
2248 \fBpcre_pattern_to_host_byte_order()\fP can be used to convert such a pattern
2249 so that it runs on the new host.
2250 .sp
2251 PCRE_ERROR_JIT_BADOPTION
2252 .sp
2253 This error is returned when a pattern that was successfully studied using a JIT
2254 compile option is being matched, but the matching mode (partial or complete
2255 match) does not correspond to any JIT compilation mode. When the JIT fast path
2256 function is used, this error may be also given for invalid options. See the
2257 .\" HREF
2258 \fBpcrejit\fP
2259 .\"
2260 documentation for more details.
2261 .sp
2262 PCRE_ERROR_BADLENGTH (-32)
2263 .sp
2264 This error is given if \fBpcre_exec()\fP is called with a negative value for
2265 the \fIlength\fP argument.
2266 .P
2267 Error numbers -16 to -20, -22, and 30 are not used by \fBpcre_exec()\fP.
2268 .
2269 .
2270 .\" HTML <a name="badutf8reasons"></a>
2271 .SS "Reason codes for invalid UTF-8 strings"
2272 .rs
2273 .sp
2274 This section applies only to the 8-bit library. The corresponding information
2275 for the 16-bit and 32-bit libraries is given in the
2276 .\" HREF
2277 \fBpcre16\fP
2278 .\"
2279 and
2280 .\" HREF
2281 \fBpcre32\fP
2282 .\"
2283 pages.
2284 .P
2285 When \fBpcre_exec()\fP returns either PCRE_ERROR_BADUTF8 or
2286 PCRE_ERROR_SHORTUTF8, and the size of the output vector (\fIovecsize\fP) is at
2287 least 2, the offset of the start of the invalid UTF-8 character is placed in
2288 the first output vector element (\fIovector[0]\fP) and a reason code is placed
2289 in the second element (\fIovector[1]\fP). The reason codes are given names in
2290 the \fBpcre.h\fP header file:
2291 .sp
2292 PCRE_UTF8_ERR1
2293 PCRE_UTF8_ERR2
2294 PCRE_UTF8_ERR3
2295 PCRE_UTF8_ERR4
2296 PCRE_UTF8_ERR5
2297 .sp
2298 The string ends with a truncated UTF-8 character; the code specifies how many
2299 bytes are missing (1 to 5). Although RFC 3629 restricts UTF-8 characters to be
2300 no longer than 4 bytes, the encoding scheme (originally defined by RFC 2279)
2301 allows for up to 6 bytes, and this is checked first; hence the possibility of
2302 4 or 5 missing bytes.
2303 .sp
2304 PCRE_UTF8_ERR6
2305 PCRE_UTF8_ERR7
2306 PCRE_UTF8_ERR8
2307 PCRE_UTF8_ERR9
2308 PCRE_UTF8_ERR10
2309 .sp
2310 The two most significant bits of the 2nd, 3rd, 4th, 5th, or 6th byte of the
2311 character do not have the binary value 0b10 (that is, either the most
2312 significant bit is 0, or the next bit is 1).
2313 .sp
2314 PCRE_UTF8_ERR11
2315 PCRE_UTF8_ERR12
2316 .sp
2317 A character that is valid by the RFC 2279 rules is either 5 or 6 bytes long;
2318 these code points are excluded by RFC 3629.
2319 .sp
2320 PCRE_UTF8_ERR13
2321 .sp
2322 A 4-byte character has a value greater than 0x10fff; these code points are
2323 excluded by RFC 3629.
2324 .sp
2325 PCRE_UTF8_ERR14
2326 .sp
2327 A 3-byte character has a value in the range 0xd800 to 0xdfff; this range of
2328 code points are reserved by RFC 3629 for use with UTF-16, and so are excluded
2329 from UTF-8.
2330 .sp
2331 PCRE_UTF8_ERR15
2332 PCRE_UTF8_ERR16
2333 PCRE_UTF8_ERR17
2334 PCRE_UTF8_ERR18
2335 PCRE_UTF8_ERR19
2336 .sp
2337 A 2-, 3-, 4-, 5-, or 6-byte character is "overlong", that is, it codes for a
2338 value that can be represented by fewer bytes, which is invalid. For example,
2339 the two bytes 0xc0, 0xae give the value 0x2e, whose correct coding uses just
2340 one byte.
2341 .sp
2342 PCRE_UTF8_ERR20
2343 .sp
2344 The two most significant bits of the first byte of a character have the binary
2345 value 0b10 (that is, the most significant bit is 1 and the second is 0). Such a
2346 byte can only validly occur as the second or subsequent byte of a multi-byte
2347 character.
2348 .sp
2349 PCRE_UTF8_ERR21
2350 .sp
2351 The first byte of a character has the value 0xfe or 0xff. These values can
2352 never occur in a valid UTF-8 string.
2353 .sp
2354 PCRE_UTF8_ERR2
2355 .sp
2356 This error code was formerly used when the presence of a so-called
2357 "non-character" caused an error. Unicode corrigendum #9 makes it clear that
2358 such characters should not cause a string to be rejected, and so this code is
2359 no longer in use and is never returned.
2360 .
2361 .
2362 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NUMBER"
2363 .rs
2364 .sp
2365 .B int pcre_copy_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2366 .ti +5n
2367 .B int \fIstringcount\fP, int \fIstringnumber\fP, char *\fIbuffer\fP,
2368 .ti +5n
2369 .B int \fIbuffersize\fP);
2370 .PP
2371 .B int pcre_get_substring(const char *\fIsubject\fP, int *\fIovector\fP,
2372 .ti +5n
2373 .B int \fIstringcount\fP, int \fIstringnumber\fP,
2374 .ti +5n
2375 .B const char **\fIstringptr\fP);
2376 .PP
2377 .B int pcre_get_substring_list(const char *\fIsubject\fP,
2378 .ti +5n
2379 .B int *\fIovector\fP, int \fIstringcount\fP, "const char ***\fIlistptr\fP);"
2380 .PP
2381 Captured substrings can be accessed directly by using the offsets returned by
2382 \fBpcre_exec()\fP in \fIovector\fP. For convenience, the functions
2383 \fBpcre_copy_substring()\fP, \fBpcre_get_substring()\fP, and
2384 \fBpcre_get_substring_list()\fP are provided for extracting captured substrings
2385 as new, separate, zero-terminated strings. These functions identify substrings
2386 by number. The next section describes functions for extracting named
2387 substrings.
2388 .P
2389 A substring that contains a binary zero is correctly extracted and has a
2390 further zero added on the end, but the result is not, of course, a C string.
2391 However, you can process such a string by referring to the length that is
2392 returned by \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP.
2393 Unfortunately, the interface to \fBpcre_get_substring_list()\fP is not adequate
2394 for handling strings containing binary zeros, because the end of the final
2395 string is not independently indicated.
2396 .P
2397 The first three arguments are the same for all three of these functions:
2398 \fIsubject\fP is the subject string that has just been successfully matched,
2399 \fIovector\fP is a pointer to the vector of integer offsets that was passed to
2400 \fBpcre_exec()\fP, and \fIstringcount\fP is the number of substrings that were
2401 captured by the match, including the substring that matched the entire regular
2402 expression. This is the value returned by \fBpcre_exec()\fP if it is greater
2403 than zero. If \fBpcre_exec()\fP returned zero, indicating that it ran out of
2404 space in \fIovector\fP, the value passed as \fIstringcount\fP should be the
2405 number of elements in the vector divided by three.
2406 .P
2407 The functions \fBpcre_copy_substring()\fP and \fBpcre_get_substring()\fP
2408 extract a single substring, whose number is given as \fIstringnumber\fP. A
2409 value of zero extracts the substring that matched the entire pattern, whereas
2410 higher values extract the captured substrings. For \fBpcre_copy_substring()\fP,
2411 the string is placed in \fIbuffer\fP, whose length is given by
2412 \fIbuffersize\fP, while for \fBpcre_get_substring()\fP a new block of memory is
2413 obtained via \fBpcre_malloc\fP, and its address is returned via
2414 \fIstringptr\fP. The yield of the function is the length of the string, not
2415 including the terminating zero, or one of these error codes:
2416 .sp
2417 PCRE_ERROR_NOMEMORY (-6)
2418 .sp
2419 The buffer was too small for \fBpcre_copy_substring()\fP, or the attempt to get
2420 memory failed for \fBpcre_get_substring()\fP.
2421 .sp
2422 PCRE_ERROR_NOSUBSTRING (-7)
2423 .sp
2424 There is no substring whose number is \fIstringnumber\fP.
2425 .P
2426 The \fBpcre_get_substring_list()\fP function extracts all available substrings
2427 and builds a list of pointers to them. All this is done in a single block of
2428 memory that is obtained via \fBpcre_malloc\fP. The address of the memory block
2429 is returned via \fIlistptr\fP, which is also the start of the list of string
2430 pointers. The end of the list is marked by a NULL pointer. The yield of the
2431 function is zero if all went well, or the error code
2432 .sp
2433 PCRE_ERROR_NOMEMORY (-6)
2434 .sp
2435 if the attempt to get the memory block failed.
2436 .P
2437 When any of these functions encounter a substring that is unset, which can
2438 happen when capturing subpattern number \fIn+1\fP matches some part of the
2439 subject, but subpattern \fIn\fP has not been used at all, they return an empty
2440 string. This can be distinguished from a genuine zero-length substring by
2441 inspecting the appropriate offset in \fIovector\fP, which is negative for unset
2442 substrings.
2443 .P
2444 The two convenience functions \fBpcre_free_substring()\fP and
2445 \fBpcre_free_substring_list()\fP can be used to free the memory returned by
2446 a previous call of \fBpcre_get_substring()\fP or
2447 \fBpcre_get_substring_list()\fP, respectively. They do nothing more than call
2448 the function pointed to by \fBpcre_free\fP, which of course could be called
2449 directly from a C program. However, PCRE is used in some situations where it is
2450 linked via a special interface to another programming language that cannot use
2451 \fBpcre_free\fP directly; it is for these cases that the functions are
2452 provided.
2453 .
2454 .
2455 .SH "EXTRACTING CAPTURED SUBSTRINGS BY NAME"
2456 .rs
2457 .sp
2458 .B int pcre_get_stringnumber(const pcre *\fIcode\fP,
2459 .ti +5n
2460 .B const char *\fIname\fP);
2461 .PP
2462 .B int pcre_copy_named_substring(const pcre *\fIcode\fP,
2463 .ti +5n
2464 .B const char *\fIsubject\fP, int *\fIovector\fP,
2465 .ti +5n
2466 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2467 .ti +5n
2468 .B char *\fIbuffer\fP, int \fIbuffersize\fP);
2469 .PP
2470 .B int pcre_get_named_substring(const pcre *\fIcode\fP,
2471 .ti +5n
2472 .B const char *\fIsubject\fP, int *\fIovector\fP,
2473 .ti +5n
2474 .B int \fIstringcount\fP, const char *\fIstringname\fP,
2475 .ti +5n
2476 .B const char **\fIstringptr\fP);
2477 .PP
2478 To extract a substring by name, you first have to find associated number.
2479 For example, for this pattern
2480 .sp
2481 (a+)b(?<xxx>\ed+)...
2482 .sp
2483 the number of the subpattern called "xxx" is 2. If the name is known to be
2484 unique (PCRE_DUPNAMES was not set), you can find the number from the name by
2485 calling \fBpcre_get_stringnumber()\fP. The first argument is the compiled
2486 pattern, and the second is the name. The yield of the function is the
2487 subpattern number, or PCRE_ERROR_NOSUBSTRING (-7) if there is no subpattern of
2488 that name.
2489 .P
2490 Given the number, you can extract the substring directly, or use one of the
2491 functions described in the previous section. For convenience, there are also
2492 two functions that do the whole job.
2493 .P
2494 Most of the arguments of \fBpcre_copy_named_substring()\fP and
2495 \fBpcre_get_named_substring()\fP are the same as those for the similarly named
2496 functions that extract by number. As these are described in the previous
2497 section, they are not re-described here. There are just two differences:
2498 .P
2499 First, instead of a substring number, a substring name is given. Second, there
2500 is an extra argument, given at the start, which is a pointer to the compiled
2501 pattern. This is needed in order to gain access to the name-to-number
2502 translation table.
2503 .P
2504 These functions call \fBpcre_get_stringnumber()\fP, and if it succeeds, they
2505 then call \fBpcre_copy_substring()\fP or \fBpcre_get_substring()\fP, as
2506 appropriate. \fBNOTE:\fP If PCRE_DUPNAMES is set and there are duplicate names,
2507 the behaviour may not be what you want (see the next section).
2508 .P
2509 \fBWarning:\fP If the pattern uses the (?| feature to set up multiple
2510 subpatterns with the same number, as described in the
2511 .\" HTML <a href="pcrepattern.html#dupsubpatternnumber">
2512 .\" </a>
2513 section on duplicate subpattern numbers
2514 .\"
2515 in the
2516 .\" HREF
2517 \fBpcrepattern\fP
2518 .\"
2519 page, you cannot use names to distinguish the different subpatterns, because
2520 names are not included in the compiled code. The matching process uses only
2521 numbers. For this reason, the use of different names for subpatterns of the
2522 same number causes an error at compile time.
2523 .
2524 .
2525 .SH "DUPLICATE SUBPATTERN NAMES"
2526 .rs
2527 .sp
2528 .B int pcre_get_stringtable_entries(const pcre *\fIcode\fP,
2529 .ti +5n
2530 .B const char *\fIname\fP, char **\fIfirst\fP, char **\fIlast\fP);
2531 .PP
2532 When a pattern is compiled with the PCRE_DUPNAMES option, names for subpatterns
2533 are not required to be unique. (Duplicate names are always allowed for
2534 subpatterns with the same number, created by using the (?| feature. Indeed, if
2535 such subpatterns are named, they are required to use the same names.)
2536 .P
2537 Normally, patterns with duplicate names are such that in any one match, only
2538 one of the named subpatterns participates. An example is shown in the
2539 .\" HREF
2540 \fBpcrepattern\fP
2541 .\"
2542 documentation.
2543 .P
2544 When duplicates are present, \fBpcre_copy_named_substring()\fP and
2545 \fBpcre_get_named_substring()\fP return the first substring corresponding to
2546 the given name that is set. If none are set, PCRE_ERROR_NOSUBSTRING (-7) is
2547 returned; no data is returned. The \fBpcre_get_stringnumber()\fP function
2548 returns one of the numbers that are associated with the name, but it is not
2549 defined which it is.
2550 .P
2551 If you want to get full details of all captured substrings for a given name,
2552 you must use the \fBpcre_get_stringtable_entries()\fP function. The first
2553 argument is the compiled pattern, and the second is the name. The third and
2554 fourth are pointers to variables which are updated by the function. After it
2555 has run, they point to the first and last entries in the name-to-number table
2556 for the given name. The function itself returns the length of each entry, or
2557 PCRE_ERROR_NOSUBSTRING (-7) if there are none. The format of the table is
2558 described above in the section entitled \fIInformation about a pattern\fP
2559 .\" HTML <a href="#infoaboutpattern">
2560 .\" </a>
2561 above.
2562 .\"
2563 Given all the relevant entries for the name, you can extract each of their
2564 numbers, and hence the captured data, if any.
2565 .
2566 .
2567 .SH "FINDING ALL POSSIBLE MATCHES"
2568 .rs
2569 .sp
2570 The traditional matching function uses a similar algorithm to Perl, which stops
2571 when it finds the first match, starting at a given point in the subject. If you
2572 want to find all possible matches, or the longest possible match, consider
2573 using the alternative matching function (see below) instead. If you cannot use
2574 the alternative function, but still need to find all possible matches, you
2575 can kludge it up by making use of the callout facility, which is described in
2576 the
2577 .\" HREF
2578 \fBpcrecallout\fP
2579 .\"
2580 documentation.
2581 .P
2582 What you have to do is to insert a callout right at the end of the pattern.
2583 When your callout function is called, extract and save the current matched
2584 substring. Then return 1, which forces \fBpcre_exec()\fP to backtrack and try
2585 other alternatives. Ultimately, when it runs out of matches, \fBpcre_exec()\fP
2586 will yield PCRE_ERROR_NOMATCH.
2587 .
2588 .
2589 .SH "OBTAINING AN ESTIMATE OF STACK USAGE"
2590 .rs
2591 .sp
2592 Matching certain patterns using \fBpcre_exec()\fP can use a lot of process
2593 stack, which in certain environments can be rather limited in size. Some users
2594 find it helpful to have an estimate of the amount of stack that is used by
2595 \fBpcre_exec()\fP, to help them set recursion limits, as described in the
2596 .\" HREF
2597 \fBpcrestack\fP
2598 .\"
2599 documentation. The estimate that is output by \fBpcretest\fP when called with
2600 the \fB-m\fP and \fB-C\fP options is obtained by calling \fBpcre_exec\fP with
2601 the values NULL, NULL, NULL, -999, and -999 for its first five arguments.
2602 .P
2603 Normally, if its first argument is NULL, \fBpcre_exec()\fP immediately returns
2604 the negative error code PCRE_ERROR_NULL, but with this special combination of
2605 arguments, it returns instead a negative number whose absolute value is the
2606 approximate stack frame size in bytes. (A negative number is used so that it is
2607 clear that no match has happened.) The value is approximate because in some
2608 cases, recursive calls to \fBpcre_exec()\fP occur when there are one or two
2609 additional variables on the stack.
2610 .P
2611 If PCRE has been compiled to use the heap instead of the stack for recursion,
2612 the value returned is the size of each block that is obtained from the heap.
2613 .
2614 .
2615 .\" HTML <a name="dfamatch"></a>
2616 .SH "MATCHING A PATTERN: THE ALTERNATIVE FUNCTION"
2617 .rs
2618 .sp
2619 .B int pcre_dfa_exec(const pcre *\fIcode\fP, "const pcre_extra *\fIextra\fP,"
2620 .ti +5n
2621 .B "const char *\fIsubject\fP," int \fIlength\fP, int \fIstartoffset\fP,
2622 .ti +5n
2623 .B int \fIoptions\fP, int *\fIovector\fP, int \fIovecsize\fP,
2624 .ti +5n
2625 .B int *\fIworkspace\fP, int \fIwscount\fP);
2626 .P
2627 The function \fBpcre_dfa_exec()\fP is called to match a subject string against
2628 a compiled pattern, using a matching algorithm that scans the subject string
2629 just once, and does not backtrack. This has different characteristics to the
2630 normal algorithm, and is not compatible with Perl. Some of the features of PCRE
2631 patterns are not supported. Nevertheless, there are times when this kind of
2632 matching can be useful. For a discussion of the two matching algorithms, and a
2633 list of features that \fBpcre_dfa_exec()\fP does not support, see the
2634 .\" HREF
2635 \fBpcrematching\fP
2636 .\"
2637 documentation.
2638 .P
2639 The arguments for the \fBpcre_dfa_exec()\fP function are the same as for
2640 \fBpcre_exec()\fP, plus two extras. The \fIovector\fP argument is used in a
2641 different way, and this is described below. The other common arguments are used
2642 in the same way as for \fBpcre_exec()\fP, so their description is not repeated
2643 here.
2644 .P
2645 The two additional arguments provide workspace for the function. The workspace
2646 vector should contain at least 20 elements. It is used for keeping track of
2647 multiple paths through the pattern tree. More workspace will be needed for
2648 patterns and subjects where there are a lot of potential matches.
2649 .P
2650 Here is an example of a simple call to \fBpcre_dfa_exec()\fP:
2651 .sp
2652 int rc;
2653 int ovector[10];
2654 int wspace[20];
2655 rc = pcre_dfa_exec(
2656 re, /* result of pcre_compile() */
2657 NULL, /* we didn't study the pattern */
2658 "some string", /* the subject string */
2659 11, /* the length of the subject string */
2660 0, /* start at offset 0 in the subject */
2661 0, /* default options */
2662 ovector, /* vector of integers for substring information */
2663 10, /* number of elements (NOT size in bytes) */
2664 wspace, /* working space vector */
2665 20); /* number of elements (NOT size in bytes) */
2666 .
2667 .SS "Option bits for \fBpcre_dfa_exec()\fP"
2668 .rs
2669 .sp
2670 The unused bits of the \fIoptions\fP argument for \fBpcre_dfa_exec()\fP must be
2671 zero. The only bits that may be set are PCRE_ANCHORED, PCRE_NEWLINE_\fIxxx\fP,
2672 PCRE_NOTBOL, PCRE_NOTEOL, PCRE_NOTEMPTY, PCRE_NOTEMPTY_ATSTART,
2673 PCRE_NO_UTF8_CHECK, PCRE_BSR_ANYCRLF, PCRE_BSR_UNICODE, PCRE_NO_START_OPTIMIZE,
2674 PCRE_PARTIAL_HARD, PCRE_PARTIAL_SOFT, PCRE_DFA_SHORTEST, and PCRE_DFA_RESTART.
2675 All but the last four of these are exactly the same as for \fBpcre_exec()\fP,
2676 so their description is not repeated here.
2677 .sp
2678 PCRE_PARTIAL_HARD
2679 PCRE_PARTIAL_SOFT
2680 .sp
2681 These have the same general effect as they do for \fBpcre_exec()\fP, but the
2682 details are slightly different. When PCRE_PARTIAL_HARD is set for
2683 \fBpcre_dfa_exec()\fP, it returns PCRE_ERROR_PARTIAL if the end of the subject
2684 is reached and there is still at least one matching possibility that requires
2685 additional characters. This happens even if some complete matches have also
2686 been found. When PCRE_PARTIAL_SOFT is set, the return code PCRE_ERROR_NOMATCH
2687 is converted into PCRE_ERROR_PARTIAL if the end of the subject is reached,
2688 there have been no complete matches, but there is still at least one matching
2689 possibility. The portion of the string that was inspected when the longest
2690 partial match was found is set as the first matching string in both cases.
2691 There is a more detailed discussion of partial and multi-segment matching, with
2692 examples, in the
2693 .\" HREF
2694 \fBpcrepartial\fP
2695 .\"
2696 documentation.
2697 .sp
2698 PCRE_DFA_SHORTEST
2699 .sp
2700 Setting the PCRE_DFA_SHORTEST option causes the matching algorithm to stop as
2701 soon as it has found one match. Because of the way the alternative algorithm
2702 works, this is necessarily the shortest possible match at the first possible
2703 matching point in the subject string.
2704 .sp
2705 PCRE_DFA_RESTART
2706 .sp
2707 When \fBpcre_dfa_exec()\fP returns a partial match, it is possible to call it
2708 again, with additional subject characters, and have it continue with the same
2709 match. The PCRE_DFA_RESTART option requests this action; when it is set, the
2710 \fIworkspace\fP and \fIwscount\fP options must reference the same vector as
2711 before because data about the match so far is left in them after a partial
2712 match. There is more discussion of this facility in the
2713 .\" HREF
2714 \fBpcrepartial\fP
2715 .\"
2716 documentation.
2717 .
2718 .
2719 .SS "Successful returns from \fBpcre_dfa_exec()\fP"
2720 .rs
2721 .sp
2722 When \fBpcre_dfa_exec()\fP succeeds, it may have matched more than one
2723 substring in the subject. Note, however, that all the matches from one run of
2724 the function start at the same point in the subject. The shorter matches are
2725 all initial substrings of the longer matches. For example, if the pattern
2726 .sp
2727 <.*>
2728 .sp
2729 is matched against the string
2730 .sp
2731 This is <something> <something else> <something further> no more
2732 .sp
2733 the three matched strings are
2734 .sp
2735 <something>
2736 <something> <something else>
2737 <something> <something else> <something further>
2738 .sp
2739 On success, the yield of the function is a number greater than zero, which is
2740 the number of matched substrings. The substrings themselves are returned in
2741 \fIovector\fP. Each string uses two elements; the first is the offset to the
2742 start, and the second is the offset to the end. In fact, all the strings have
2743 the same start offset. (Space could have been saved by giving this only once,
2744 but it was decided to retain some compatibility with the way \fBpcre_exec()\fP
2745 returns data, even though the meaning of the strings is different.)
2746 .P
2747 The strings are returned in reverse order of length; that is, the longest
2748 matching string is given first. If there were too many matches to fit into
2749 \fIovector\fP, the yield of the function is zero, and the vector is filled with
2750 the longest matches. Unlike \fBpcre_exec()\fP, \fBpcre_dfa_exec()\fP can use
2751 the entire \fIovector\fP for returning matched strings.
2752 .
2753 .
2754 .SS "Error returns from \fBpcre_dfa_exec()\fP"
2755 .rs
2756 .sp
2757 The \fBpcre_dfa_exec()\fP function returns a negative number when it fails.
2758 Many of the errors are the same as for \fBpcre_exec()\fP, and these are
2759 described
2760 .\" HTML <a href="#errorlist">
2761 .\" </a>
2762 above.
2763 .\"
2764 There are in addition the following errors that are specific to
2765 \fBpcre_dfa_exec()\fP:
2766 .sp
2767 PCRE_ERROR_DFA_UITEM (-16)
2768 .sp
2769 This return is given if \fBpcre_dfa_exec()\fP encounters an item in the pattern
2770 that it does not support, for instance, the use of \eC or a back reference.
2771 .sp
2772 PCRE_ERROR_DFA_UCOND (-17)
2773 .sp
2774 This return is given if \fBpcre_dfa_exec()\fP encounters a condition item that
2775 uses a back reference for the condition, or a test for recursion in a specific
2776 group. These are not supported.
2777 .sp
2778 PCRE_ERROR_DFA_UMLIMIT (-18)
2779 .sp
2780 This return is given if \fBpcre_dfa_exec()\fP is called with an \fIextra\fP
2781 block that contains a setting of the \fImatch_limit\fP or
2782 \fImatch_limit_recursion\fP fields. This is not supported (these fields are
2783 meaningless for DFA matching).
2784 .sp
2785 PCRE_ERROR_DFA_WSSIZE (-19)
2786 .sp
2787 This return is given if \fBpcre_dfa_exec()\fP runs out of space in the
2788 \fIworkspace\fP vector.
2789 .sp
2790 PCRE_ERROR_DFA_RECURSE (-20)
2791 .sp
2792 When a recursive subpattern is processed, the matching function calls itself
2793 recursively, using private vectors for \fIovector\fP and \fIworkspace\fP. This
2794 error is given if the output vector is not large enough. This should be
2795 extremely rare, as a vector of size 1000 is used.
2796 .sp
2797 PCRE_ERROR_DFA_BADRESTART (-30)
2798 .sp
2799 When \fBpcre_dfa_exec()\fP is called with the \fBPCRE_DFA_RESTART\fP option,
2800 some plausibility checks are made on the contents of the workspace, which
2801 should contain data about the previous partial match. If any of these checks
2802 fail, this error is given.
2803 .
2804 .
2805 .SH "SEE ALSO"
2806 .rs
2807 .sp
2808 \fBpcre16\fP(3), \fBpcre32\fP(3), \fBpcrebuild\fP(3), \fBpcrecallout\fP(3),
2809 \fBpcrecpp(3)\fP(3), \fBpcrematching\fP(3), \fBpcrepartial\fP(3),
2810 \fBpcreposix\fP(3), \fBpcreprecompile\fP(3), \fBpcresample\fP(3),
2811 \fBpcrestack\fP(3).
2812 .
2813 .
2814 .SH AUTHOR
2815 .rs
2816 .sp
2817 .nf
2818 Philip Hazel
2819 University Computing Service
2820 Cambridge CB2 3QH, England.
2821 .fi
2822 .
2823 .
2824 .SH REVISION
2825 .rs
2826 .sp
2827 .nf
2828 Last updated: 27 February 2013
2829 Copyright (c) 1997-2013 University of Cambridge.
2830 .fi

Properties

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

  ViewVC Help
Powered by ViewVC 1.1.5